THE WORLD BANK
Glob al                    Ja25305
January 2003
I =2oonomic
:Prosp ects
and the Developing Countries
4  ..............  -p ,         - -      *<t >........:i a  .   ...
r ;!;, t''r' ~-  . - 2 'L's,           et':s 
;  1p'  a     ,~~~ft." '       '       t    'iZ!' 
L..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~P 
1s'     >         --            4't ,^0-i  
Investing to UrI-ock Global Opportunities vjn n  A
Investing to -Utlock Global Opportunfities203






lo bal
Economic
Prospects
and the Developing Countries
2003



( 2003 The International Bank for Reconstruction and Development /
The World Bank
1818 H Street, NW
Washington, DC, USA 20433
Telephone: 202-473-1000
Internet: www.worldbank.org
E-mail: feedback@worldbank.org
All rights reserved.
1 2 3 4 04 03 02
This volume is a product of the staff of the World Bank. The findings, interpretations, and
conclusions expressed herein do not necessarily reflect the views of the Board of Executive
Directors of the World Bank or the governments they represent.
The World Bank does not guarantee the accuracy of the data included in this work. The
boundaries, colors, denominations, and other information shown on any map in this work
do not imply any judgment on the part of the World Bank concerning the legal status of any
territory or the endorsement or acceptance of such boundaries.
Rights and Permissions
The material in this work is copyrighted. Copying and/or transmitting portions or all of this
work without permission may be a violation of applicable law. The World Bank encourages
dissemination of its work and will normally grant permission promptly.
For permission to photocopy or reprint any part of this work, please send a request with
complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,
MA 01923, USA, telephone 978-750-8400, fax 978-7S0-4470, www.copyright.com
All other queries on rights and licenses, including subsidiary rights, should be addressed to
the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA,
fax 202-522-2422, e-mail pubrights@worldbank.org
ISBN 0-8213-5338-1
ISSN 1014-8906
Library of Congress catalog card number: 91-6-440001 (serial)



Contents
Foreword    ix
Acknowledgments    xi
Summary     xiii
Abbreviations and Data Notes  xxi
Chapter 1 The International Economy and Prospects for Developing Countries
A recovery constrained by major risks  4
Investment cycles in developing countries  23
Growth and poverty to 2015: coming changes in savings and
investment patterns  28
Notes    41
References   42
Chapter 2 Changes in Global Business Organization  45
The surge in trade and FDI  46
The rise in service sector FDI  52
Global production networks  55
Good policies attract FDI  66
Notes    69
References   71
Chapter 3 Domestic Policies to Unlock Global Opportunities  77
Investment climate and investmenit policies 78
Promoting efficient private investment: harnessing competition  85
Public investment in infrastructure and human capital  103
Policies to promote competition  107
Notes    109
References   110



G LO B AL EC ONO MI C PRO SP EC TS 2 0 0 3
Chapter 4 International Agreements to Improve Investment and
Competition for Development  117
International efforts to promote investment  119
International agreements to promote competition and competition policy  133
Conclusions   145
Notes    146
References   147
Appendix 1 Regional Economic Prospects  151
Appendix 2 Global Commodity Price Prospects  175
Appendix 3 Global Economic Indicators  199
Figures
1.1    The recovery was initiated in a typical fashion  6
1.2    A brief rebound in industrial countries was underway  6
1.3    Rebound in industrial countries boosted production in East Asia  6
1.4    Non-oil commodities are recovering but stand well below previous peaks  7
1.5    Private sector creditors have cut debt exposures so far in 2002  9
1.6    FDI flows to emerging Asia are proving to be quite resilient  10
1.7    Investment recovery is still uncertain  10
1.8    World trade rebounds along with GDP, 1998-2004  13
1.9    2002 marks the start of a moderate recovery  15
1.10   LAC and MENA are not experiencing the recovery  15
1.11   Low case: world trade and other indicators will be much lower than
the baseline  20
1.12   G-7 investment falls sharply  21
1.13   Investment growth and net capital flows into Latin America are
strongly correlated  22
1.14   Oil prices spike  23
1.15   Investment is more volatile than GDP in East Asia  23
1.16   Investment is more volatile than GDP in Latin America  23
1.17   Central Europe and Turkey experience greater volatility in
investment than in GDP  24
1.18   Investment volatility declines with income  24
1.19   A better investment climate reduces volatility of investment cycles  25
1.20   Impact of policy climate on investment volatility after correcting for
income remains strong  25
1.21   Growth of working-age population decelerates  33
1.22   Productivity has not been the dominant source of growth in regions  35
1.23   Productivity is expected to be more significant in the longer term  35
1.24   Major structural shifts in investment and savings behavior have occurred  37
1.25   Youth dependency ratio will fall everywhere except Japan  39
1.26   Elderly dependency ratios will rise in some regions  39
2.1    Exports-to-GDP ratios have increased since the 1970s  47
2.2    All regions have benefited from rising FDI flows  47
2.3    FDI is concentrated in large countries, but many small countries receive large amounts
relative to GDP  48
iv



C O N T E N r s
2.4    The service sector dominated the 1990s mergers and acquisitions boom  49
2.5     Cross-border mergers and acquisitions are small compared with
stock market capitalization  49
2.6    U.S. aggregate concentration has held steady  50
2.7     Global concentration has not increased significanitly  51
2.8    Share of FDI in the service sector increased in major industrial countries  52
2.9    U.S. cars are produced in many countries  55
2.10    Cross-border networks capture increasing shares of production and trade  56
2.11   The role of production networks continiued to increase through
most of the 1990s  56
2 12   Tariff rates fell in the last two decades  57
2.13    Reforming countries boosted exports through production networks  58
2.14    Parts and components exports grew rapidly, 1981-2000  61
2.15    Developing countries' share of global parts and component exports
rose between 1981 and 2000   61
2.16    Developing countries' parts and component exports are highly
concentrated, 2000  62
2. 17  Strong rule of law attracts foreign investors  66
2.18    Foreign investors have been shifting away from weaker investmenlt
climate locations  67
2.19    Private iiifrastructure investmiient surged in the 1990s  68
3 1    Domestic capital is the largest source of investmenit in developing countries  78
3.2    Incentives for FDI are varied and numerous  80
3.3    Competition and ease of entry are associated with higher growth  86
3.4     Competition from imports checks markups in concentrated markets  87
3.5     High tariffs are correlated with lower productivity  88
3.6    High entry costs inhibit FDI inflows  90
3.7    Barriers to entry can become barriers to exit  92
3.8     Barriers to entry and exit allow inefficient firms to stay in the market  94
3.9    Difficulties in obtaining licenses and permits discourage FDI  94
3.10    Inefficient customs hurt Indian exports  95
3.11   Inefficient ports raise India's transport costs far above competitors'
transport costs  95
3.12   Privatization revenues soared in the 1990s  96
3.13   Granting monopoly rights brings in revenues  98
3.14   More competition means more phones   98
3.15   Better infrastructure means higher growth  105
3.16    Greater literacy is associated with higher growth  106
3.17   Education raises the productivity of FDI, which leads to higher growth  107
4.1     Countries are increasingly liberalizing their investmiient regimes  123
4.2     South-South FDI is rising  123
4.3     Share of South-South FDI in total FDI is rising  124
4.4     Revealed preferences: governments shield services more often than
manufacturing from the winds of investment competitiol  126
4.5     FDI is growing faster than exports and output  128
4.6     OECD countries spent $230 billion in 2001 to support agrictiltural producers  135
4.7    Imports affected by cartels rose from 1981 to 2000 for both rich and
poor countries  138
v



G LO BA L E CO NO MI C PROS PE CT S 2 0 0 3
Tables
1.1    Global conditions affecting growth in developing countries and
world GDP growth    3
1.2    External environment for developing countries, 1991-2004  14
1.3    Global effects in a low-case scenario, 2003-04  20
1.4    Low case: contributions to global effects in 2003  21
1.5    Relative volatility of investment is high in developing countries  24
1.6    Upturns can be financed abroad and domestically  26
1.7    Capital inflows lead investment in middle-income countries: correlation
between investment ratios and (past or future) capital flows  26
1.8    Long-term prospects are projected to be stronger for most regions  29
1.9    Large poverty reductions in EAP and SAR partially offset by
poverty increases in SSA  30
1.10   Savings fall in high-income countries, but increase in most other regions  38
2.1    FDI inward stocks in services and manufacturing, 1988-99  52
2.2    Growth of exports of parts and components, 1981-2000  55
2.3    Export activity for product groups with the fastest growth in
world exports, 1980-98  62
2.4    Rapid growth and structural change experienced by
network participants  65
3.1    Profitability on equity, concentration, and market share (percent):
Brazil, 1971-78  99
3.2    Cartel enforcement in selected developlng countries  101
4.1    Many antidumping investigations were initiated during the 1995-2001
period   136
4.2    International cartels can be expensive: estimates of sales and overcharge  138
4.3    National exemptions to competition law for exporters  141
4.4    Breaking up floating cartels could help developing countries  142
Boxes
1.1    Is Latin America going against the rising tide?  8
1.2    Integration pays off where policies are supportive  12
1.3    The terrorist attacks of September 11, 2001 had an economic impact  16
1.4    Consumption in low- and middle-income countries is smoothed over
the business cycle  27
1.5    Is the World Bank overestimating global poverty?  31
1.6    Technological progress is an important determinant of growth  34
2.1    Intra-firm trade increases worldwide  59
3.1    Trade restrictions shield MNCs from competitive forces at enormous cost:
the case of Argentina  89
3.2    Competition policy and competition law share similar objectives
across countries  102
3.3    Does public investment "crowd out" or "crowd in"
private investment?  104
VI



C O N r E N T S
4.1    What is a BIT?   120
4.2    The Multilateral Agreement on Investment (MAI)  122
4.3    South-South flows: who invests and who receives?  125
4 4    Do BITs increase investment flows? Only a bit  129
4.5    Disciplines on corporations can also improve the investment climate  132
4.6    The lysine cartel, 1995-2001  139
4.7    International cooperation aids competition policy  144
\'11






Foreword
Productivity increases and efficient investment are essential conditions for rapid growth and
poverty reduction. The key to accelerating technological improvement and increasing in-
P    vestment is improving the "investment climate " In the broadest sense, this term encom-
passes the policy and institutional environment that fosters entrepreneurship, learning, and pro-
ductive investment.
In this report, we argue that the investmlient climate for developing countries has both a global
dimension and a national dimension. The global investment clinate, although less amenable to
policy initiatives of developing countries, nonetheless presents opportunities, risks, and at times
obstacles for developing countries. In this report, we focus on1 two aspects of the global invest-
ment climate: the current state of the world economy as it affects developing countries' financial
outlook, exports, and growth prospects (chapter 1) and the organization of global business,
notably the proliferation of multinational companies and associated production networks (chap-
ter 2). In previous reports we have studied other aspects of the global investment climate, in-
cluding the world trading system (Global Economzic Prospects 2002) and aspects of the global
financial system (Global Development Finance 2002).
The national dimension of the investment climate for developing countries is discussed in
chapter 3. This dimension is composed of the policy and institutionial environment that fosters
entrepreneurship-and that strongly influences the pace of productivity growth and the rate of
investment. Differences in national policies help explain why some countries grow rapidly and
others do not, even though all operate within the same international investment climate In short,
policymakers have considerable scope for choosing policies that influenice the amounit and pro-
ductivity of investment.
For the purposes of this report, we focus on two types of national policies that affect how
countries use globalization to grow. The first type is investmenit policies-for example, tax in-
centives, tariffs, subsidies, and policies to channel investmlient into particular activities, as well as
public iivestmenit. The second type is policies that promote or limit competition-for example,
tariffs, entry restrictions, and state monopolies as well as conventionally defined competition
policy.
We chose these policy areas for three reasons. First, these policy areas directly link the do-
mestic policy dimensions of the investment climate with the global economy. Second-in contrast
to macroeconomic policies, property rights, and other institutional features that primarily affect
the quantity of investment-policies fostering investnment and competition work instead through
microeconomic incentives to influence the quality of investment (as measured by its productivity).
Finally, these policies are at the center of global debate, figuring prominently in discussions of the



G LO B AL EC O NO MI C PROS PE CT S 2 0 0 3
Doha Development Agenda launched at the World Trade Organization (WTO) Ministerial Meet-
ing in November 2001.
To inform that debate, the final chapter of this report asks how the international community
can support developing countries in their quest for better investment climates, both global and
national. The chapter focuses on synergies that can emerge from developing countries' participa-
tion in internationial agreements on investment and competition policies, topics that are not only
central to the WTO Doha Development Agenda but that also figure prominently in many
regional trade negotiations around the world.
Nicholas Stern
Chief Economist and Senior Vice President
The World Bank
x



Acknowledgments
T       his report was prepared by the Development Prospects Group in the World Bank, draw-
ng on resources throughout the Development Economics Vice Presidency and the World
Bank's operational units. Richard Newfarmer was the lead author and manager of the
report, under the direction of Uri Dadush. The principal chapter authors were Hans Timmer
(chapter 1), William Shaw (chapter 2), Jeffrey Lewis and Scott Wallsten (chapter 3), and Richard
Newfarmer and Pierre Sauve (chapter 4). We are grateful for the ideas and insights of Bernard
Hoekman, Michael Klein, and Theodore Moran (Georgetown University). The report was pre-
pared under the general guidance of Nicholas Stern.
Many staff from insidc and outside the World Bank contributed to the report. In chapter 1,
Caroline Farah, Himmat Kalsi, Robert Keyfitz, Aniette I. Dc Kleine, Robert Lynn, Fernando
Martel Garcia, Mick Riordan, and Bert Wolfe contributed to the global trends; Dominique Van
der Mensbrugghe provided the long-term analysis; and Shaohua Chen and Martin Ravallion
contributed to the poverty analysis; Kathlecn Rollins was the staff assistant. Chapter 2 benefited
from background papers and other inputs from Gary Gereffi, Mary Hallward-Driemeier, Shafiq
Islam, Frances Ng, Matthcw Slaughter, Lawrence Whlite, and Yong Zhang. Similarly, in chapter 3,
Claudio Frischtak, Rughvir Khemani, Kamal Saggi, Luis Villela, and Alberto Barreix as well
as Mary Hallward-Driemeier provided background papers that became building blocks for the
sections on incentives, investmenit, competition, and foreign investment, and Denis Medvedev
provided valuable written inputs and rcsearch assistance. Chapter 4 was based on background
papers by Simon Evenett and Benno Ferrarini (competition), Bijit Bora (performance requirc-
ments), and Mary Hallward-Driemeier (bilatcral investment treaties and investor protections) as
well as inputs from Aaditya Mattoo and Shweta Bagai (various). Beata Smarzynska and Jacques
Morisset provided helpful guidance and comments on various aspects of the report. The regional
appendixes benefited from the written inputs of the Regional Chief Economists around the Bank
and their staff. John Baffes, Betty Dow, Donald Mitchell, and Shane Streifel prepared the com-
modity appendix. The staff assistant for the report was Awatif Abuzeid. Mark Felge provided ed-
itorial assistance. Yinne Yu and Yong Zhang as well as Jonathan Koh provided valuable research
assistance at various stages of the report's preparation. Dorota Nowak coordinated publication
and disseminatioii activities
Several experts provided written comments that immeasurably improved the quality of the rc-
port at various stages: Pierre-Richard Agenor, Robert Anderson, Milan Brahmbhatt, Jean-jacques
Dethier, Simeon D;ankov, Richard Eglin, Antonio Estache, Shahrokh Fardoust, Alan Gelb,
Coralic Gevers, Ian Goldin, Rughvir Khemani, Aart Kraay, Ernesto May, Mustapha Nabli, Anne
xi



G LO B AL EC ONO MI C PROS PE CT S 2 O 0 3
McGuirk, Ashoka Mody, Ijaz Nabi, John Panzer, Guillermo Perry, Guy Pfeffermann, Karl
Sauvant, Todd Schneider, Mark Sundberg, and Roberto Zagha.
We are also grateful for the many substantive comments we received from governments
around the world through their Executive Directors on the World Bank's Board.
Melissa Edeburn, Susan Graham, and lIma Kramer managed the production for the World
Bank's Office of the Publisher.
Xii



Summary
The global recovery is fragile, because    and middle-income countries as it is in high-
investment spending is insufficient to    income countries. But the volatility of invest-
underpin continuing growth-                ment is greater in developing countries than in
Strong cyclical dynamics, together with an  rich countries. Countries with sound invest-
easing of macroeconomic policies in the    ment climates experience far less volatility
United States and elsewhere, have boosted  than countries with deficient policies and
large parts of the global economy into the mi-  institutions.
tial phase of a recovery in 2002. The driving  Capital flows to developing countries have
forces behind the initial phase of the recovery  proved to be procyclical. But the direction of
were strong, but they have proved short-lived  causality between investment and capital in-
because inventory and high-tech cycles are  flow appears to differ significantly between
short and appear to have peaked. Though    rich and poor countries. In rich countries, a
consumption spending has held firm, this is  boom in domestic fixed investment tends to
precisely the time when investment demand  attract foreign capital, while in middle-income
should pick up and boost recovery onto a   countries it is the acceleration of capital
higher trajectory. So far it has not. Financial  inflows that typically stimulates domestic
imbalances, evident in  different forms    investment. Similarly, a fall in rich countries'
throughout the world economy, seem to be  investment tends to reduce net capital inflows,
weighing down growth. Wide-ranging uncer-  while for middle-income countrics reduced net
tainty in financial markets may jeopardize the  capital inflows (or increased capital outflows)
needed rebound in fixed investment and may  are the driving forces behind contractions in
thus diminish prospects for projecting the  domestic investment. This dependence on cap-
global recovery into the future. Falling and  ital flows makes the middle-income countries
volatile stock markets, accounting scandals,  especially vulnerable to tensions in global fi-
accumulated debts (domestic and foreign, pri-  nancial markets. Low-income countries, with
vate and public), and reassessments of long-  greater reliance on official aid and with lim-
run profitability keep investors cautious, if  ited access to private capital markets, do not
not jittery, throughout the world. For these  exhibit either of these patterns.
reasons, growth in 2003 seems certain to be
weaker for almost all developing regions than  -but long-term prospects remain
we anticipated as recently as six months ago.  promising
Analysis of long-tern trends indicates that  Over the long run, new opportunities for
the investment cycle as a determinant of over-  technological advances (often driven by
all cyclical behavior is as important in low-  globalization), together with more stable
xii



G LO BA L E CO NO MI C PROS PE CT S 2 0 0 3
macroeconomic policies and an improved     1990s, rising from $331 billion in 1995 to
business climate, have the potential to acceler-  $1.3 trillion in 2000, before falling off to an
ate growth and to increase investment ratios  estimated $725 billion in 2001. Most of these
in developing countries that currently lag  flows are destined to rich countries.
behind. The outlook for reductions in global  FDI flows to developing countries are
poverty, while generally positive and of the  about $160 billion. This amount is still rela-
same order of magnitude as in our previous  tively small compared with all domestic in-
report, is marginally dimmer because of the  vestment in developing countries, now about
absence of a robust recovery today.        $1 trillion. Nonetheless, in virtually every re-
At the same time, demographics are likely  gion, FDI is a driving force of globalization
to alter existing savings and investment pat-  and has risen relative to total capital expen-
terns and will tend to push countries to be-  ditures during the 1990s. It has doubled in
come more interdependent through capital   middle-income countries and has tripled in
flows. Major demographically driven shifts in  low-income countries. However, recently FDI
current account balances-particularly in   flows have fallen. They peaked in 1999 at
Japan, which is moving toward reduced sur-  $184 billion and are experiencing their most
plus, and in middle-income countries, which  sustained fall since the global recession of
are moving toward increased surplus-are    1981-83.
likely to accelerate financial integration.  These trends over the past decade have in-
Underneath large swings in net flows are even  creased competition in most markets around
larger movements of gross capital flows, as  the world. Despite a sharp increase in mergers
foreign direct investment (FDI) expands into  and acquisitions, the share of global economic
growing markets in developing countries and  activity accounted for by the largest compa-
as financial agents in developing countries  nies does not appear to have risen over the
seek to diversify their portfolios in rich coun-  1990s. The profits of the top 50 companies ac-
tries. However, because international financial  counted for 0.8 percent of world GDP in
flows have at times fluctuated widely, they  2001. Although their share of aggregate prof-
have sometimes proved damaging to growth  its amounted to 3.3 percent of global savings
and poverty reduction. The international com-  in 2000, up from 1.8 percent in 1994, this in-
munity and developing countries have to    crease is likely to be the result of the boom in
search for mechanisms to provide greater sta-  the United States and the overstatement of
bility in integration. Developing countries can  earnings of some large U.S. corporations.
do much on their own. Improving the domes-  These factors point to a pattern of stability
tic investment climate, particularly through  rather than a trend of increases. Similar pat-
sound macroeconomic policies and gover-   terns exist for the largest 500 companies.
nance, can reduce the volatility of capital  Four changes in the organization of busi-
flows and attract less-volatile FDI.       ness are particularly important for developing
countries. First, the rise of foreign investment
Global competition is creating new         in services is creating a new source of compe-
opportunities for developing countries     tition-and potential productivity gains-in
Cross-border trade and direct investment have  developing countries, where staid state com-
expanded rapidly over the past three decades.  panies have often monopolized production for
Global exports of goods and services in-   decades. Recent efforts to privatize these com-
creased from 14 percent of output in the early  panies and to open industries to competition
1970s to 23 percent by the late 1990s, while  have allowed some developing countries to
global FDI flows have more than doubled rel-  harness this competition for gains. In many
ative to the gross domestic product (GDP).  developing countries, restrictions on services
The surge in FDI flows accelerated in the late  still remain high, because some countries have
xiv



S U M M A R Y
privatized only slowly and others have priva-  in global risk premiums has reduced investors'
tized badly, creating private monopolies still  appetite for risk and for projects with long ges-
insulated from competition.                 tations. Adversity to such projects is reflected
Second, production networks that span the  not only in the average spreads over U.S. Trea-
globe, once barely a dot on the horizon of in-  sury interest rates that developing countries
ternational business, have now become a cen-  must pay to their bondholders in the Emerging
tral feature. That so many large firms have  Market Bond Index (even excluding country
chosen to outsource production of parts and  "outliers" in crisis) but also more generally in
equipment or to otherwise locate production  spreads of high-risk corporate bonds in the
facilities offshore offers new  opportunities  United States. Both have more than doubled
for developing countries. Firms choosing to  from under 500 basis points to more than
"deverticalize" production through outsourc-  1,000. The recent collapse of the telecommu-
ing create new opportunities for suppliers and  nications sector, as well as difficulties experi-
create a foundation for a steady increase in  enced by major power companies associated
trade for participating developing countries.  with the Enron scandal, has diminished the
The downside is that this production and the  number of players and enthusiasm  among
associated high rates of export growth are  potential long-term financiers. Second, many
highly concentrated geographically, and so  projects have suffered payment problems
this door into a greater share of the global  because of the inability of contracts to weather
economy has, to date, opened only for rela-  sharp  contractions  in  demands.  From
tively few countries. Taking advantage of   Argentina to Indonesia, the string of defaults
networks requires a strong policy environ-  associated with infrastructure projects and re-
ment that fosters private investment and pro-  structurings has left in its wake a severe re-
vides complementary public investmenits (see  trenchment. Thus, governments throughout
below).                                    the developing world will have to do more to
Third, with growing concerns about risk,  offset this risk-principally through better
investors are becoming increasingly sensitive  policies, and perhaps through a slowing of the
to investment climates in developing coun-  retreat from government financing of infra-
tries, and the result is that money is moving to  structure that has occurred under the banner
the countries with large, rapidly growing, and  of privatization.
relatively stable economic environments.
Countries such as China, the Republic of    Harnessing globalization requires
Korea, and Mexico benefited from the largest  reducing barriers to competition-
inflows in 2000. As a share of domestic in-  To raise the productivity of both foreign and
vestment, however, small-market countries are  domestic investment, developing countries
proving they can keep pace-provided that    have to harness the full force of competition
they protect property rights, have stable   inherent in globalization. Too often they have
macroeconomic environments, and have good   not done so. In many countries, policy barri-
institutions. Poor countries that fall short on  ers to competition-whether they are im-
policies and institutions compound the disad-  pediments to trade, restrictions on incoming
vantages they already experience from having  foreign investment, administrative barriers to
small markets. Hence, they may be virtually  competition, or monopolies granted to state
shut out from foreign investment flows in any  enterprises-have channeled domestic as well
sector other than natural resources.        as foreign investment into less-productive
Finally, long-term  private investment fi-  activities that dampen productivity improve-
nancing for infrastructure has fallen off to  ment and hobble growth. Import competition,
levels that may prove persistent. This retrench-  for example, can limit what would otherwise
ment has two origins. First, the post-1997 rise  be the shared monopoly pricing of a few local
xv



G LO B AL EC ONO MI C PROS PE CT S 2 0 0 3
producers. In a wide sample of developing   broader reform packages, either to catalyze
countries, decreasing imports in concentrated  support for emerging opportunities (such as
industries from 25 percent of domestic sales to  clusters) or to create transitional mechanisms
zero is associated with increases of 8 percent  and initial constituencies for reform that can
in oligopolistic markups on sales.          be progressively expanded (such as EPZs). But
Competition-impeding regulations in re-  more broadly, investment incentives will gen-
cently privatized industries have undermined  erally not make up for serious deficiencies
potential benefits from privatization and have  in the investment environment or generate
insulated new  owners-frequently foreign    sustained growth. To encourage productive
companies-from   efficiency-improving com-  investment and benefit from  globalization,
petition; the result has been slow growth and  governments must tackle the challenges of
resource misallocation. In Africa, for example,  promoting competition and entrepreneurship
telephone services in countries with private  and of undertaking complementarily pro-
monopolies have expanded growth only one-   ductive public investment in areas such as
third as fast as telephone services in countries  education.
with competitive networks.
Over time, firms in countries with lower  -and therefore sound public investments
barriers to trade and to investment competi-  are essential
tion tend, as a general rule, to enjoy signifi-  Public investment also plays a crucial role in
cantly higher productivity of investment, both  enhancing growth. Some countries get both
foreign and domestic, and with it more rapid  the levels and the composition of investment
growth. This fact does not imply a single pre-  right, and their growth rates are high. Other
scription for all countries irrespective of their  countries invest too much through the public
stage of development. As the experience of  sector and crowd out private investment.
China-among others-has shown, reforms       Because these effects are also associated with
have to be tailored to country circumstances  investments in state enterprises that enjoy mo-
and integrated into sustainable development  nopoly positions protected from competition,
strategies. The analysis does imply, however,  the composition effects of public investment
that countries wishing to increase their oppor-  are negative. Other countries invest too little
tunities from globalization would do well to  through the public sector. This problem is usu-
look first at the incentive features of their  ally manifested in poor education, poor in-
investment climate, with special attention to  frastructure, and poor public institutions
barriers that impede competition.           generally-all of which reduce profitable in-
vestment opportunities for both domestic and
-and using targeted interventions           foreign companies. Investing in effective pub-
with care-                                  lic institutions has an especially high return.
Governments may hope to make up for an
unfriendly investment environment through in-  International agreements on investment
centive mechanisms. But while there are clearly  and competition policies can provide
examples in which targeted interventions-   benefits through reciprocity-
such as fiscal incentives, export processing  Countries get most of the positive growth
zones (EPZs), or support for economic clus-  stimulus from  domestic unilateral reforms
ters-may indeed lead to higher investment  tailored to local strategy and conditions, and
levels (and the jobs and related spillovers that  these reforms should not be held hostage to
go along with them), there is, unfortunately,  international agreements. Nonetheless, re-
little evidence that such initiatives can be  forming governments may be able to obtain
systematically successful. Instead, they tend to  additional benefits from international agree-
work best when they work in support of      ments. Benefits can take several forms. For



S U M M A R Y
investment policies, participating in interna-  reinforce sound domestic policies and can
tional agreements that are linked to greater  contribute to better performance. Since most
market access may elicit more investment by  of the remaining restrictions are on services,
signaling to investors that changes are perma-  governments around the world can increase
nent. Also, participating in international nego-  market access by using the existing multilat-
tiations may strengthen the hand of domestic  eral framework rather than creating a new
reformers by holding out the prospect of mar-  one. The General Agreement on Trade in Ser-
ket access abroad in exchange for new domes-  vices (GATS) provides an as-yet-underutilized
tic policies; simultaneously, negotiations can  arrangement to negotiate reciprocal market
prompt reciprocal reforms among partners    access in services. To date, the coverage of
that would not otherwise occur. For competi-  commitments for a large number of countries
non policy, international agreements may lead  is limited. About two-thirds of the World
to the removal of restraints that inhibit com-  Trade Organization membership has sched-
petition, thereby unleashing new price compe-  uled 60 or fewer sectors (of the 160 or so spec-
tition that benefits all countries.         ified in the GATS list). Moreover, in many
cases, commitments do not reflect the actual
-but agreements on investment policy are    degree of openness. Finally, in some countries,
likely to have strong development effects   the commitments that have been made serve
only if they deal with the big issues facing  only to protect the privileged position of in-
developing countries-                       cumbents rather than enhance the contestabil-
The purposes of coordinating investnent     ity of markets. To remedy these problems,
policy are to expand the flow of investment  governments must take greater advantage of
around the world, to minimize policy exter-  the opportunity offered by the GATS to lend
nalities that hurt neighbors, and to help   credibility to reform programs by committing
improve economic performance. Agreements    to maintain current levels of openness or by
might contribute to achieving these goals   precommitting to greater levels of future
through three main channels: protecting     openness. To advance the process of services
investors' rights, which increases incentives to  reforms beyond levels undertaken indepen-
invest; liberalizing investment flows, which  dently and to lead to more balanced outcomes
permits enhanced access and competition; and  from the developing-country point of view,
curbing policies that may distort investment  countries could better harness the power of
flows and trade at the expense of neighbors.  reciprocity by devising negotiating formulas
International agreements that focus on es-  that widen the scope for tradeoffs across sec-
tablishing protections for investors cannot be  tors (both in goods and in services) and across
expected to expand markedly the flow of in-  modes of delivery, particularly the temporary
vestment to new signatory countries. This is  movement of workers. While difficult, such ef-
because many protections are already con-   forts may prove easier than designing a whole
tamed in bilateral investment treaties (BITs).  new international investment arrangement.
Even the relatively strong protections in BITs  Similarly, curbing policy externalities that
do not seem to have increased flows of in-  "beggar thy neighbor" can benefit developing
vestment to signatory developing countries.  countries, especially if the countries focus
These facts suggest that expectations for new  on two critical issues. The first is to reduce
flows associated with protections emerging  investment-distorting trade barriers. By de-
from any multilateral agreement should be   priving developing countries of market access
kept low.                                   and by discouraging their exports, many trade
International agreements that allow coun-  barriers also lessen the attractiveness of oppor-
tries to negotiate reciprocal market liberaliza-  tunities to invest in developing countries' ex-
tion and to promote nondiscrimination can   port industries for both foreign and domestic
xvu



G LO BA L E CO NO MI C PROS PE CT S 2 O 0 3
investors. In Canada, the European Union      Restrictions on competition in the global
(EU), Japan, and the United States, average ad  marketplace that most hurt development take
valorem-equivalent tariffs for manufactures  three forms. The first form consists of policy
are roughly twice as high for developing coun-  barriers in markets abroad that limit competi-
tries as they are for members of the Organisa-  tion among developing countries in these
tion for Economic Co-operation and Develop-  markets. These barriers, like those discussed
ment. The ad valorem-equivalent tariffs on  above, discourage investment and create ob-
agriculture (to say nothing of subsidies) in  stacles to competition. Particularly harmful
those countries are also more than three times  are the $311 billion in agricultural subsidies
higher than such tariffs on manufactures. Re-  and textile quotas, as well as the correspond-
ducing trade barriers among developing coun-  ing high border protection, tariff distortions
tries themselves is as important as reducing  (that is, tariff peaks and escalation), and pro-
trade barriers in rich countries. Developing  tectionist use of antidumping. These practices
countries import from each other at average ad  are only too common in all countries, rich and
valorem-equivalent rates comparable to EU   poor alike. All of these trade restrictions limit
rates for imports from developing countries.  the ability of exporters in developing countries
This level of protection dampens investment-  to compete in international markets.
both domestic and foreign-in affected export  Second, private restraints on competition
industries, and removal of these barriers   can adversely affect prices for consumers and
would have significant development effects.  producers in developing countries-much as
The second critical issue is to curb the  they can in industrial countries. For example,
emerging competition among countries to lure  cartel practices among companies based in
foreign investment through investment incen-  high-income countries taxed consumers in
tives. Unfortunately, information on the ex-  developing countries by up to $7 billion in the
tent of investment incentives is inadequate to  1990s. Actions that facilitate prosecution of
assess their effects, and so a high priority for  cartels should be high on the priority list. Such
international collaboration is to systematically  actions can range from developing more sys-
compile this information.                   tematic arrangements to exchange informa-
Finally, participation in international agree-  tion among competition agencies, to granting
ments on investment may also have benefits  standing for developing countries to sue under
over and above unilateral reforms if the agree-  foreign antitrust laws when their trade is ad-
ments include reciprocal market access in   versely affected. Indeed, both developing and
areas of importance to developing countries.  industrial countries would benefit from much
These benefits can become clear only in the  greater efforts to identify and document re-
course of negotiation.                      strictive business practices that adversely af-
fect prices of their trade.
-and thus competition agreements should       Third, many governments in high-income
focus on restraints to competition that     countries officially sanction trade restraints
hurt developing countries                   through antitrust exemptions for their compa-
Greater competition is associated with more  nies in domestic law. For example, many gov-
rapid development, and lowering policy barrn-  ernments permit their companies to cartelize
ers to trade and foreign investment in devel-  exports. Shrouded in the secrecy of government
oping countries, as shown in chapter 3, is a  registries, these national export cartels may
powerful procompetitive force. Beyond unilat-  well raise prices to developing countries. Ef-
eral actions, international agreements on com-  forts should be made to make information on
petition policy might also bring benefits, pro-  national export cartels transparent. Everyone
vided they address the major restrictions that  would benefit from a decrease in cartels that
adversely affect developing countries.      have damaging price effects. Similarly, antitrust
xviii



S U M M A RY
exemptions for ocean transport have given rise  adoption of best practices in developing coun-
to price-fixing arrangements that systematically  tries can pay high dividends.
hurt consumers everywhere, including those in  Unlocking global opportunities begins with
developing countries. These restraints are esti-  the efforts of developing countries to improve
mated to cost developing countries more than  their investment climates. Deployed well, in-
$2 billion per year and entail similar costs to  vestment policies and policies to unleash com-
consumers in industrial economies.          petition can accelerate economic growth and
Finally, competition policies in developing  reduce poverty. This report offers a general
countries themselves can in many cases be   framework and lessons, but each country has
improved through increased transparency,   to formulate its own development strategy.
nondiscrimination, and procedural fairness.  Nonetheless, the international community,
However, international cooperation in this  working together, can help through develop-
complex area of regulation has to recognize  ment assistance, voluntary collaboration, and
that countries have different capacities and  well-conceived international agreements. For
institutional settings, warranting caution in  these efforts to have greatest effect, they have
recommending-much less in mandating-        to tackle the most pressing investment and
across-the-board policies. In this area, volun-  competition problems-and that is the chal-
tary programs that facilitate the learning and  lenge ahead.
xix






Abbreviations
and Data Notes
ADB            Asian Development Bank
ASCM           Agreement on Subsidies and Countervailing Measures
BITs           Bilateral Investment Freaties
EPZ            Export processing zone
EU             European Union
FDI            Foreign direct investmenit
FSAP           Financial Sector Assessment Program
GATS           General Agreemcnt on Trade in Services
GPA            Government Procurement Agreement
HIV/AIDS       Human immunodeficiency virus/acquired immune deficiency syndrome
lCC            International Chamber of Commerce
ICN            International Competition Network
ICSID          Internatioial Center for Settlement of Investment Disputes
IMF            International Monetary Fund
ITO            International Trade Organization
LDC            Least developed countries
M&A            Mergers and acquisitions
MAI            Multilateral Agreement on Investmiient
MERCUSOR       Latin America Southern Cone trade bloc (Argentina, Brazil, Paraguay, and
Uruguay)
MFN            Most favored nation
MNCs           Multinational corporations
NAFTA          North American Free Trade Agreement
NGO            Nongovernmental organization
OAS            Organizationi of American States
OECD           Organisatloni for Economic Co-operation and Devclopment
ROSC           Reports on the Observance of Standards and Codes
SOEs           State-owned enterprises
xxt



G LO B AL EC O NO MI C PROS P E C I 5 2 0 0 3
TNCs           Transnational corporations
TRIMs          Trade-Related Investment Measures
TRIPS          Trade-Related Aspects of Intellectual Property Rights
U.S. BEA       U.S. Bureau of Economic Analysis
UNCITRAL       United Nations Commission on International Trade Law
UNCTAD         United Nations Conference for Trade and Development
UNDP           United Nations Development Programme
USAID          U.S. Agency for International Development
WTO            World Trade Organization
Data notes
The "classification of economies" tables at  "developing countries" as used in this vol-
the end of this volume classify economies by  ume covers all low- and middle-income coun-
income, region, export category, and indebt-  tries, including countries with transition
edness. Unless otherwise indicated, the term  economies. All dollar figures are U.S. dollars.
xxii









N~
The International Economy and
Prospects for Developing Countries
Developments in early 2002 showed a       the second phase is in jeopardy because of ten-
cyclical rebound-                         sions in fiancial markets, which reflect accu-
Macroeconomic stimulus, a rebound from a  mulated financial imbalances and significanit
record trough in the high-tech sectors and  uncertainties. These pressures have made the
a bottoming-out of inventory cycles, brought  recovery in 2002 less uniform, and they are
large parts of the global economy onto a re-  likely to moderate growth in 2003.
covery path at the end of 2001. Lower interest  In Japan, deflation and high and rapidly
rates helped keep consumers' demand for   growing government debt have placed severe
durable goods strong. Together with fiscal eas-  limits on both monetary and fiscal stimuli.
ing, that demand provided support for the  Combined with the fragility of the banking
rebound in the United States and, to a lesser  sector, which is burdened by bad loans and
extent, in some East Asian and European   diminishing capital caused by lower equlty
countries. High-tech markets-in which tech-  values, financial uncertainty prevents the
nologies quickly become obsolete-returned  spread of recovery fromi export sectors to
to strong growth by creating replacements for  those that produce for the domestic market.
old products. Inventory selloffs ceased, thereby  The accounting scandals in the United Statcs
contributing to an acceleration of gross do-  have undermined the trust in reporting sys-
mestic product (GDP) growth in early 2002.  tems. Investors, who have come to rely on
The driving forces behind the initial phase  continuously rising equity prices, nlow find it
of the recovery were strong, but short-lived, as  difficult to assess the profitability of firms.
business confidence remained weak. Inventory  That difficulty sharply pushed up risk premi-
and high-tech cycles typically are short, and  ums in equity markets. European financial in-
both appear to have peaked toward the middle  stitutionis were forced to adjust their balance
of 2002. The effects of fiscal stimulus and  sheets in the wake of large-scale defaults,
monetary easing can also, under current cir-  notably by Argentina and several major
cumstances, dissipate quickly.            U.S. firms, which probably played a role in
suppressing a nascent recovery in European
-but uncertainty in financial markets has  economies. In Europe and elsewhere, tclecom-
sapped momentum                           municationi sectors still suffer from overin-
In the second phase of a typical recovery, the  vestment and high debt burdens, making a
upturn spreads to other sectors and other  speedy recovery of capital spending in those
regions, and the driving force shifts from in-  sectors unlikely.
ventory dynamics to accumulation of fixed    Uncertainty is keeping investors cautious,
investment. In the current upswing, however,  if not skittish, throughout the world. While



G LO BA L EC O NO MI C PROS PE CT S 2 0 0 3
investors in high-income countries take their  -and the outlook for 2003 is
losses and replenish their reserves, they limit  for tepid growth
their exposure to developing countries and  Reflecting financial uncertainty and the dis-
concentrate their assets in investment-grade  appointing recovery of business confidence,
borrowing countries.1 Capital flows into large  prolected growth for 2003 has been marked
parts of Latin America dropped sharply, re-  down for almost all developing regions, be-
flecting the aftermath of Argentina's default  cause a robust rebound in industrial coun-
and the vicious combination of global uncer-  try growth-driven by strong advances in
tainty, domestic problems in some large coun-  investment-has become less likely. In line
tries, and intra-regional contagion. The reversal  with these revisions, inflation, interest rates,
of capital flows-with the accompanying rise  and non-oil commodity prices are also likely
in spreads and depreciation of currencies-  to be lower. The sole exception to this pattern
when combined with vulnerable balance-    is the Middle East and North Africa region,
sheet characteristics triggered a dangerous  where oil exporters have benefited from high
worsening of debt dynamics in some Latin   oil prices during 2002. Several of these coun-
American countries. In such an environment,  tries are seeing increased government expendi-
average per capita income in Latin America  ture, financed by rapidly mounting surpluses
has fallen in 2002 for the second year in  of oil revenues.
succession.
Investment cycles in developing countries
The rebound in 2002 was less uniform       are more volatile than in rich countries
than anticipated-                          With the sharp fall in global investment in
Rapid recovery in the beginning of 2002,   2001 and the uncertainty surrounding a re-
driven in part by sharp increases in the U.S.  bound in capital expenditure, investment
government's expenditure in the aftermath of  behavior has become a key element of the
terrorist attacks, has resulted in upward revi-  outlook. A closer look at investment cycles in
sions of 2002 growth for the United States,  developing countries suggests the following
East Asia, and Japan, relative to forecasts  conclusions:
prepared in February (table 1.1). At the other
extreme, growth in Latin America has been    * Investment behavior in low- and middle-
lowered by 1.6 percentage points for the       income countries is a determinant of
year. This decrease reflects not only the crisis  overall volatility that is even more
in Argentina, but also the major contrac-      important than it is in high-income
tions of GDP in Uruguay and the Repiblica      countries.
Bolivariana de Venezuela, plus slow growth   * Those developing countries with a
in Brazil, Chile, and Mexico. Those events     stronger policy environment exhibit
made the 2001-02 period the worst for the      lower volatility in investment.
region since the debt crisis of the early 1980s.  * Although in rich countries domestic
Consistent  with   higher-than-anticipated     fixed investment tends to drive foreign
global growth, non-oil commodity prices in      capital inflows, in middle-income coun-
2002 have risen more than anticipated.         tries the opposite tends to occur (that is,
Nonetheless, the present rebound in com-       capital inflows typically drive domestic
modity prices is modest froin an historical    investment).
perspective, thus highlighting the continuing
downward pressures on prices tied to strLc-  These conclusions imply that the middle-
tural factors. Higher commodity prices have  income countries are especially vulnerable to
supported modestly improved performance in  the current jitters in financial markets. Such
Sub-Saharan Africa.                        countries are exposed to sudden reversals in
2



r H E   I N T E R N AI I O  N A L  E C O N 0 M  Y   A N D   P R O S P l C T S   I O R  D E V E l O P I N G   C O U N I R I E S
Table 1.1 Global conditions affecting growth in developing countries and world GDP growth
(percenttage cbange front previouis year, except interest rates anid oil price)
Global Developm>ent
Currcnt estinate                Current forecasts            Finance 2002 forecasts
2000         2001         2002        2003         2004          2002        2003
Global conditions
World trade (volume)                            13 1         -0.5          2.9         7.0           8.0           1.8        8 3
Inflation (consumer prices)
G-7 OECD countriesa                            1 9          1 7          0 9          1 2          1 5          0 9         1 6
United States                                  3 4          2 8          1 5         2.1           2 3           1 5        2.4
Commodity prices (nominal $)
Commodity prices, except oil ($)              -13          -91           5 0         5 8           4 4           1 3        7 3
Oil price ($, weighted average), $/bbl        28 2         24 4         25 0        23 0          20 0    1    20 0        21 0
Oilprice(%change)                             562         -137           27         -80          -130         -179          50
Manufactures export unit value ($)'           -21          -1.4          0 5         3 0           2 2         -0 5         3 6
Interest rates
LIBOR, 6 inonths (US$, percent)                6 6          3 6          18           1 5          3 1          2 3         4 0
EURIBOR, 6 months (euro, percenit)             4 5          4 2          3 4         3.2           3 8           3 0        4 0
GDP (growth)d
World                                            3 8           1.1         1.7         2.5           31            1.3        3.6
Menio iteni World GDP (ppp)'                     4 5          2.1          2 8         3 4           4 0           2,4        4 3
High-income countries                          3.5          0.7          1.5         2.1           2 7           0 9        3.3
OECD countriesf                              3 4          0 8          14          2 1           2 6           0 8        31
United States                              3 8          0 3          2 3         2 6           31            1.3        3 7
Japan                                      21          -0 3          00           08            1 3        -15           1 7
Euro Area                                  3 7           1.5         0 8          1 8          2 6           1 2        3.3
Non-OECD countries                           6.8         -0.7          2 3         3 7           5 3          2 7         5 3
Developing countries                           5 2          2.9    I     2.8         3.9           4.7           3.1        4.9
East Asia and Pacifict                       70           S 5          6 3         61            6 4           5.6        7.1
Europe anid Central Asia                     6.6          2.3          3 6         3 4           36           3 2         4 3
Transition countries                       6.4          4 6          3 5          3 3          35            3 4        4 0
Latin America and the Canribbean             3 7          04         -1 1          1 8           3 7          0 5         3.8
ExcIliding Argentina                       4.5           1.2         0 7          1.9          3 6           2 1        4 3
Middle East anid North Africa                4 2          3 2          2 5         3 5           37            2 7        3 3
Oil exporters                              3.6           2.4         2 4         3 7           3 6           2.2        2 8
Diversified economies                      3 7          4.3          2 2         2 7           36     1      31         4 4
South Asia                                   4.8          4.4          4 6          5 4          58            4 9        5 3
Sub-Saharais Africa                          3.2           2.9         2 5          3 2          3 8           2 6        3.6
Memorandums items
Developing countriesI
Excludinig the tiansitioni countries       5.0          2 6          2 7         4 0           4.9           3 1        5 1
Excluding Chitia atid India                4 6           1 7         1 5         2.8           3.8           2 0        4.1
Note OECD = Organization for Economic Co-operation and Developnenit, hbl = barrel, EURIBOR = European interbank offered rate, LIBOR  London
interhank offered rate, ppp = purchasing piower parity
a Caniada, France, Germany. Italy, Japani, the Utmited Kingdom, anid the United States.
b In local cuirrency, aggregated using 1995 GDI' weights
c. Unit value index of matsufactsires exports fioii the G-5 coLuitries to developimg countries, expressed in U S dollars
d GDP itt 1995 constanit dollars 1995 prices and irnirket exch.ange rates
e GDP measured at 1995 puirchasing power parity (itnrernatoiial dollar) weights.
f Republic of Korea iticome classificationi chaniged from middle to sighi in comie (July 2002). Bioth forecassts were adijisted for this revision
Souirce World Bank, November 2002 anid Glrbal Developmenit Fiiance 2002 projectiois of Febiuary 2002



G LO B AL EC O NO MI C  PROS PE CT S 2 0 0 3
capital flows, which can dampen investment  quarter of 2001. Accumulated financial im-
sharply and can undermine growth momen-    balances that had built up during the 1990s
tum. Countries with strong policy environ-  emerged as a critical factor that clouded the
ments are more likely to avoid or smoothly  economic outlook. In the United States, the
absorb potential external financial shocks.  bursting of the equity bubble and cumulated
private sector debt kept investors cautious and
In the long term, faster growth            resulted in a continuous flight to quality,
can be achieved in most                    whlich moved the yield on government bonds
developing regions                         to a 40-year low while hampering the recovery
Market reforms and trade liberalization dur-  in private investment. In Japan, banking prob-
ing the 1990s have opened opportunities for  lems and the lack of scope for monetary eas-
accelerating technological advances through-  ing and fiscal stimulus limited the spillover
out the developing world for the next 15 years.  from an export-driven recovery to a rebound
An exception is emerging East Asia, where  in domestic investment. In Europe, weakness
some moderation of technological progress  was concentrated in the highly indebted
is anticipated, reflecting in part the extra-  telecommunications sector and in financial
ordinarily rapid catching-up that occurred dur-  sectors that had to absorb sharp devaluations
ing the 1980s and 1990s. The acceleration of  of their assets.
growth in many of the other regions is likely  Bankruptcies and reductions in investment
to coincide with increasing savings and invest-  during the global downturn of 2001 and the
ment rates. Demographic transitions are anti-  subsequent first phase of recovery in early
cipated to boost saving rates in developing  2002 had not reduced corporate debt nor re-
countries, while reducing them in high-income  stored profitability sufficiently. In a number of
countries.                                 cases, the downturn has generated new imbal-
On balance, the declining availability of  ances. Throughout the world, fiscal balances
savings in the aging populations of high-ilconme  deteriorated and balance sheets of financial
countries and the increased savings in the  institutions weakened. Continued tension in
developing world-set against investment    financial markets made the recovery less uni-
patterns needed to accommodate potential  form in 2002-as well as probably less robust
growth-imply that more and more develop-  In 2003-than would have been the case
ing countries will move toward surplus on the  under more normal circumstances. Vulnerabil-
current account and that the recent shift from  ity to adverse shocks has increased, and even
debt accumulation to debt reduction is likely  the potential for a "double-dip" recession sce-
to continue. As long as domestic credit mar-  nario in the industrial countries cannot-at
kets continue to mature and public savings  this juncture-be entirely ruled out.
do not deteriorate, domestic savings can be  Three distinct phases characterize recent
expected to rise, and the required reduction in  developments. The first phase portrays the
debt levels will not conflict with the required  driving forces behind the initial phase of the
investment.                                recovery that started in late 2001, a recovery
that was more robust in the United States and
East Asia. These forces range from the end of
A recovery constrained                    inventory adjustment, monetary easing, and
by major risks                             fiscal stimulus to a technical rebound in the
During the summer of 2002, investor risk   high-tech industrial sectors. This picture nor-
D   perceptions increased and market senti-  mally would be characterized as a favorable
ment deteriorated across large parts of the  environment for developing countries. That
world's economy, thereby jeopardizing the  environment includes low inflation and inter-
global recovery that had started in the fourth  est rates, plus a significant recovery in global
4



T HE IN TERN AT IO NA L EC ONO MY A ND PRO SP EC TS F OR D EVE LO PING C O U N F R I E S
trade and commodity prices, albeit a recovery  annual rate, despite a drag of nearly 1.5 per-
from low levels. During the second phase, a  centage points stemming from a deterioration
recovery typically broadens to other regions  of net exports. In contrast, output in Japan in-
and other sectors. Therefore, a recovery of  creased by 2.5 percent, of which foreign trade
profits and strong growth in fixed investment  contributed 1.8 percentage points, while in the
becomes the driving force that sustains or even  Euro Area, GDP growth of 1.6 percent was
accelerates growth. In the absence of such  supported bv almost 1 percentage potnt from
broadening and deepening, driving forces that  positive net exports contributions.
underpin the initial phase would suddenly
appear to become short lived, which is the  -inventory dynamics-
situation today. Finally in the third phase-  Inventory dvnamics played a pivotal role in
typically the shift from "recovery" to eco-  the recovery, thus complementing macro-
nomic expanision-implications of the set of  economic stimulus efforts. The same reduction
opposing forces (cyclical rebound and finan-  in the inventory stock that led to a negative
cial turbulence) for the medium-term global  contribution of stock building to CDP growth
outlook (2003-04) are analyzed. The lack of  in 2001 implied a positive contribution of
uniformity in growth performance during    stock building to GDP growth in 2002. Once
2002, following an almost synchronized slow-  the lower level of desired inventories was
ing of growth across regions in 2001, is par-  achieved, stock building shifted from sharply
ticularly notable. The growth projections for  negative to close to zero. The slowing of inven-
2003 are more uniform across regions, but  tory liquidation significantly shifted the con-
are distinctly weaker than would have been  tributLion to GDP growth from the second half
anticipated in a strong, synchronized global  of 2001 to the first half of 2002t that shift
recovery.                                  added 1.2 percentage points to the accelera-
tion of GDP growth in both Japan and the
The first phase of the global              Euro Area, and a full 2.2 percentage points in
recovery was driven                        the United States (figure 1.1).
by policy stimulus-
In the wake of the terrorist attacks in Septem-  -a high-tech rebound-
ber 2001, forceful monetary easing in the  Recovery in global high-tech markets was an
United States-and to a lesser extent in    equally powerful stimulant. After demand for
Europe-helped prevent a deepening of the   semiconductors and related equipment plum-
global downturn. U.S. consumers benefited  meted during 2001, markets were anticipated
from historically low interest rates to boost  to rebound sharply, but the scope of recovery
their purchases of durable goods. Combined  exceeded expectations. There are several
with double-digit growth in government     natural limits to declines at rates of up to
spending-mainly driven by security, defense,  50 percent. The nature of the product-the
and reconstruction efforts-the stimulus was  technology  of which  becomes obsolete
sufficient to turn U.S. GDP growth positive, to  quickly-warrants a periodic return to high
2.7 percent (annlualized), in the fourth quarter  growth, as old products are replaced by new
of 2001. One quarter later, Japan, which suf-  ones and as the introduction of advanced
fered steep output declines for three quarters  technologies generates new and growing mar-
in succession, and Europe, having experienced  kets. Defensc- and security-related spending in
only a modest fall in GDP, broke away from  the United States also played a role in bolster-
negative growth rates as well.            ing demand (U.S. manufacturing orders for
The importance of U.S. domestic demand  computers and communications equipmenlt
in this recovery is striking. During the first  ratcheted to annual rates of 40 and 90 per-
half of 2002, GDP advanced at a 3 percent  cent, respectively, in early 2002). As the
5



G LO BA L     ECON OM IC          PROS PE CTS           2 0 03
Figure 1.1 The recovery was initiated in a typical fashion
(growth in percent)
5
4       GDP growth
4 
Japan
3
2   |l      *      / ~~~~Stock contribution   
I2
-1
-2
United States                                                               Euro Area
-3 
-4
-5
-6
03     04      01     02             Q3      04     Q1     02             03      04     01      02
2001   2001    2002   2002           2001    2001   2002   2002           2001    2001   2002    2002
Sousrce U S. Departmenit of Commerce, Japan Econiomic Statistics and Research Institute (ESRI) and Eurostat
rebound intensified, a strong boost was given               for a   broader global recovery         through    the
to manufacturing output in industrial coun-                 traditional channels of international transmis-
tries, and especially to production and exports             sion. With world trade increasing, commodity
from   East Asia (figures 1.2 and 1.3).                     prices firming, and interest rates-fostered by
Macroeconomic stimuli, inventory dynam-                  low   inflation-standing       at historically    low
ics, and a powerful turnaround in high-tech
markets in the industrial countries set the stage
Figure 1.3 Rebound in industrial
-                    .* countries boosted production in East
Figure 1.2 A brief rebound in industrial                    Asia
countries was underway                                      Semiconductor dollar sales and Industrial
Manufacturing production                                    productlon
(percent change, 3m/3m saar)                                (percent change, 3m/3m saar)
20                                                         125 -                                        50
100 J.Sem~~~~~~~~~Siconductor sales j
15                                                                             (let axis)               40
10         ,,~~Japan                                   75         A30
10                       EUzE U -1S    tt                  250 -           Industrial production       20
0 ~~~~A25                                                                                     \     10
-5                                           ~~~~~~~~~~~~~~~~~~0                          0
-10                                                         -25                                         -10
United States 
-15                                                .        -0                                          -20
Jan    Jul    Jan     Jul    Jan    Jul
-20                                                             2000   2000   2001   2001   2002    2002
Jan May Sep Jan May Sep Jan. May Sep
2000 2000 2000 2001 2001 2001 2002 2002 2002             Note. Through July 2002 'Republi of Korea, Malaysia,
Singapore, aiid Taiwan, China.
Note 3m/3m saar refers to 3-morith/3-month seasonially      Souirce Semiconductor Iindustry Associataon (SIA) and
adiLsred annualized rate                                    national souirces through Datastream, World Bank staff
Souirce Datastream.                                         estimiates
6



T HE IN T ER N AT IO NA L EC ONO MY A ND PRO SP CC TS F OR D EVE LO PING CO UN r RI ES
levels, developing countries faced a broadly  crisis, such as Argentina and Turkey, have
favorable environment during the early part of  posted inflation peaks of just 20 to 40 percent.
2002.
The second phase of the recovery
-and a recovery of global trade             is on uncertain footing
World trade began to grow at near double-   Notwithstanding the positive environment
digit annual rates, recovering from a fall to  taking shape in early 2002, it became appar-
negative territory during 2001. The World   ent before the middle of the year that financial
Bank's non-oil commodities price index      strains were clouding the outlook. In the
gained 19.2 percent between October 2001    wake of large corporate bankruptcies and
and October 2002 (figure 1.4), while the rele-  the accounting scandals in the United States,
vant index for Sub-Saharan African (SSA)    stock markets still seemed overvalued and
countries rose further-by 30 percent. How-  debt levels seemed underestimated. Falling
ever, commodity prices are still one-third  equity prices further eroded the capital base
below their peak levels, which occurred during  of Japanese commercial banks and other
the summer of 1997, and sevcral exporters,  financial instltutLions. The growth outlook
notably those in Caribbean countries that   for Latin America deteriorated noticeably.
specialize in coffee and sugar, did not benefit  Following Argentina's default, Uruguay and
from the rebound in average prices during the  Paraguay were also hit hard through financial
first half of 2002. Historically low inflation  and trade linkages. Political uncertainty in the
characterized  not only the high-income     Repiiblica Bolivariana de Venezuela triggered
economies, but also those in the majority of  capital flight, and in several other countries
developing countries. The median inflation  debt dynamics worsened as a result of a coIll-
rate in developing countries is presently one-  bination of domestic problems and increased
third of that during the 1990s, despite rela-  risk aversion in international capital markets
tively high oil prices and more widespread  (box 1.1). In Europe, financial institutionis
adoption of flexible exchange rates. Indeed,  were hit hard by defaults in the United States
double-digit inflation rates have become an  and in Argentina, as well as by falling equity
exception, and countries experiencing recent  prices and a weakening of the dollar. And the
Figure 1.4 Non-oil commodities are recovering but stand well below previous peaks
(index, Sept lOct 2001 = 100)
190
180  /\g/ Agriculture
170
160
150 rC\                        /Sub-Saharan African
10    _      commodities
130
120               metals/minerals
110
100
Jan   Jul    Jan    Jul   Jan    Jul   Jan    Jul   Jan     Jul  Jan    Jul Oct
1997   1997  1998   1998  1999   1999  2000   2000   2001  2001  2002  20022002
Souirce World Baunk sraff
7



G LO BA L EC ONO MI C PRO SP EC TS 2 0 0 3
Box 1I1 Hs Latin America going against the
rising tide?
E conomic activity in Latin America and the      Argentina, and of expanding political uncertainties
ECaribbean (LAC) has fallen behind production    in the run-up to the October elections all con-
trends in other developing countries (box figure).  tributed to a weakening in financial market senti-
The region's per capita gross domestic product   ment toward the country. The combination also re-
(GDP) is estimated to have dropped by 2.6 percent  sulted in a sharp reduction in private financing
in 2002, the only developing region where per capita  flows, which, in turn, led to a deterioration of public
output contracted during the year. This decline was  debt dynamics.
the second consecutive year of contraction in per    With financial markets increasingly averse to
capita incomes-the worst performance since the   taking risks through financial flows to the major
beginning of the debt crisis in the early 1980s. Why  LAC countries over the course of the year, other
is LAC going against the tide of rising global income  countries in the region were unable to obtain signif-
growth?                                          icant new financing from international capital
The crisis in Argentina and the spillover    markets at reasonable terms. This lack of financing
effects on Southern Cone Comrnon Market          limited growth in high-debt countries with significant
(MERCOSUR) countries (particularly Uruguay       financing requirements. Political instability in
and Paraguay) clearly contributed the most to the  Rep6blica Bolivariana de Venezuela-an aborted
decline in regional output during 2002. In Brazil,  coup in March that came on top of poor economic
the combination of rising public debt, of declining  management in previous years-led to a large
export revenues tied to the collapse of demand in  contraction in GDP of some 6 percent.
The crisis in Argentina and its fallout is a classic
example of a vicious circle of instability in interna-
Latin America falls.behind                  tional financial markets and domestic vulnerabilities:
(Ldustnalproduction, fdex, 1995=100, 3-month  high levels of debt; large financing requirements;
moving average)                               and, in some countries, fixed exchange rates, politi-
cal uncertainties, and weak banking systems. Once
135                                            a crisis erupts, the vicious circle turns into a brutal
130      A                                     downward spiral in which a depreciation of the
125         countries                          currency, debt burdens, a deterioration of dollar
120                                            returns, a rise in spreads, and a reversal of capital
115            /6c 2                         flows reinforce each other. Argentina's peso lost
11       t_Latin America                   more than two-thirds of its value in the year to
110 {                                         September. In a comparative perspective, stock
105                                           prices increased over the same period by 20-30 per-
100,  , ,               I I   I I C            cent in several countries in East Asia and Central
Feb   Feb   Feb   Feb   Feb   Feb            Europe, currencies were stable and spreads did not
1997  1998  1999  2000  2001  2002          increase substantially.
Source National statistics, World Bank staff
Source- World Bank staff.
plight of European telecommunications sec-    dards, have reduced some of the corporate
tors continued to deepen under the weight of  debt, but they have also led to a deterioration
overinvestment and mounting debt loads.       of the balance sheets of financial institutions.
Bankruptcies and sharp reductions in in-   To improve their reserves and to decrease their
vestment expenditure, driven by the erosion of  risk exposure, these financial institutions, in
equity values and a tightening of credit stan-  turn, sold part of their equity assets. In doing
8



T HE IN T E RN ATI O NA L EC ONO MY A ND PRO SP EC TS F OR DE V EL O PING C O U N T R I E S
so, they further fueled the fall in stock prices  interest rates seems still likely, especially in
and amplified new    imbalances. Similarly,   the Euro Area. The limited scope for macro-
the drop in capital flows has increased debt  economic policy makes the risks surrounding
problems in several vulnerable middle-income  the recovery even more severe. Policy solutions
countries.                                    in the industrial world may best be focused
Another example of new or deteriorating     on eliminating bad debts and restoring in-
imbalances is the public sector deficit across  vestor's sentiment by strengthening institu-
the industrial countries. The U.S. general gov-  tional oversight.
ernment deficit deteriorated from a surplus
position of 2.3 percent of GDP in calendar    Capital flows to emerging markets
year 2000 to a deficit of 2.5 percent in 2002,  are declining
with 2.5 percentage points of that shift attrib-  Also important for middle-income coun-
uted to structural deterioration. In turn, the  tries was a continued decline in international
United States has not taken advantage of the  market-based capital flows, despite low inter-
recession to narrow its deficit on current ac-  national interest rates and the initial phase of
count. Despite increases in the household sav-  recovery in the industrial world. New gross
ings rate and declines in the private investment  capital market flows fell from $228 billion in
rate, the current account deficit widened to  2000 to $175 billion in 2001, falling further
a watershed mark of S percent of GDP as of    to near $140 billion during 2002 (figure 1.5),
the second quarter of 2002.                   with bank lending showing the steepest falloff.
In the Euro Area as well, fiscal deficits have  The latter point highlights the cautious posi-
deteriorated from a 0.9 percent surplus to a like  tion that international banks have adopted
level of deficit, though this deterioration re-  following financial crises in Turkey and
flects mainly the work of automatic stabilizers.  Argentina and following defaults by several
Unlike the U.S. fiscal deficit, it is not a struc-  large U.S. corporations. These capital flow fig-
tural deterioration. France, Germany, Italy,  ures suggest that total external debt in the
and Portugal are now approaching the current  developing world continues to contract.
limits of a 3 percent of GDP deficit, limits     Net foreign direct investment (FDI) inflows
that were imposed by the European Monetary    to developing countries also trended down-
Union Growth and Stability Pact. The original
plan to eliminate deficits by 2004 has been
abandoned and replaced by an agreement to        Figure 1.5 Private sector creditors have
reduce structural deficits by at least O.S per-  cut debt exposures so far in 2002
centage points per annum  over the coming        (gross market-based monthly flows in billions of dollars)
years. Japanese fiscal deficits remain extra-    25 2001 monthly average = $143
ordinarily high, at levels above 7 percent of                     2002 monthly average = $121
GDP. And East Asian emerging economies, on       20
average, continue to run relatively high deficit
levels-above 4 percent of GDP-contrasted         15
with a deficit of 1 percent before the 1997 crisis.
In other parts of the developing world, primary                   1      1
surpluses are increasing, but improvement of      s
the overall deficit remains difficult to achieve,
given the burden of debt service.                 0
Deteriorated government deficits, com-            J F M A M J J A S O N D J F M A M J J A
bined with low nominal interest rates, have                  2001            2002
left little room for further fiscal stimulus or  Souirce Euromoney and World Bank staff estimates
monetary easing, although some lowering of
9



G LO BA L    EC ON OM IC     PRO SP EC TS      2 0 0 3
financial strains currently battering corporate
Figure 1.6 FDI flows to emerging Asia            sectors are likely to have a restraining effect
are proving to be quite resilient                on investment. Falling equity prices, concerns
Net infiows of FDI by region                     about corporate debt, uncertainty about prof-
(billlons of dollars. half yearly rate)          itability, and cautious commercial bank lend-
50                                               ing in high-income countries tend to curtail
45    \   Lat nAmerica                           financing for investment. Moreover, current
40                                               business uncertainty regarding future demand
35                      t                        growth   serves as an additional restraining
30                                               force on capital spending. Reduced capital
25
20                    Emerging Asia              flows to emerging markets place a damper
15    Emerging Europe \                          on investments in the developing world. If
10                                               one looks further ahead, large and increas-
s                                               ing public sector deficits carry the potential
1S       2nd I s        I l                     to "crowd out" private investment in high-
ts in si  2nd   1si  2nd   1si
half  half  hall  half  half  half  half    income and developing countries alike, al-
1999       2000       2001    2002        though   low   interest rates on   government
Source World Banik staff estimates               bonds show that such crowding out is not yet
a problem.
-and driving forces for the initial phase
ward, from   about $170 billion in both 2000        could be short-lived-
and 2001, to an estimated $145 billion during       Without a solid upswing in investment and
2002 (figure 1.6). The decline can be almost        a concomitant broadening of recovery, the
wholly attributed to reduced flows into Latin       forces driving the first phase of the rebound
America, as FDI into Central Europe and East
Asia retained a resilient tone. That resilience
underscored the importance of the course of
integration into global and neighboring mar-           Figure 1.7 Investment recovery is still
kets that China [recently gaining membership           uncertain
in the World Trade Organization (WTO)] and             (4-quarter moving average, percentage change
in the World Trade Organization (WTO)] and qlq, sear)
Central   European    countries   [anticipating         15
European Union (EU) accession] are following                     Developing countries
at present.                                             10
\  /Industra 
Financial strains may inhibit                            s      $      oun
corporate investment-
During the fall months of 2002, investment               o
was still declining in both high-income and
developing countries, although the declines             -5          V
were bottoming out, which indicated the be-
ginning of a turnaround (figure 1.7). In a typ-        -10       01     01     Q1     Q1    Q1
ical recovery, once investment growth returns            1997   1998   1999   2000   2001   2002
to positive territory, the recovery gets new           Note q/q refers to quarter over quarter
impetus, because a virtuous circle of adjust-          *Argentina, Brazil, Chile, Mexico, Czech Republic,
Poland, Turkey, Indonesia, Rep of Korea, Malaysia,
ments in the capital stock and in expected             Philippines, Thailand, and South Africa (4s% of
market growth may easily generate double-              developing cointry total)
Souirce Datastream and WVorld Bank staff estimates
digit advances n capital spending. However,
10



T HE IN T ER NAT IO NA L E CON OM Y A ND PRO SP EC TS F OR DE V EL O PING C O U N T R I E S
could well become short-lived. Widening pub-  appears to be dependent on (and exposed to)
lic sector shortfalls limit the range of options  global demand and financial conditions.
for fiscal policy. As France, Germany, and  Growth was disappointing in the first half of
Italy approach Maastricht limits; as the U.S.  2002, and expectations for only a sluggish
deficit widens; and as Japan remains encum-  advance in output during the second half of the
bered by continuing massive fiscal imbalance,  year have now become more widespread.
opportunities for further fiscal stimulus in the
industrial world have indeed become quite   -implying a much less supportive
scarce. Moreover, it appears that several other  external environment for developing
driving forces for the initial recovery could  countries
quickly run out of momentum   while their   Against this background-particularly the
stimulative properties dissipate. Official inter-  lack of a rebound In fixed investment across
est rates now standing at historically low  industrial countries and the intensification of
levels (particularly in Japan and the United  financial uncertainties-the environment for
States) leave little prominent role for further  developing countries is much less favorable.
monetary easing. And the inventory and      International interest rates may remain low,
high-tech cycles-typically of short duration-  but borrowing costs have risen in step with in-
probably reached peak levels by mid-2002.   creases in interest rate spreads. Trade volumes
Japanese output growth is now   largely  and commodity prices may be on the rise, but
grounded in export growth, although con-    they are still at low levels, and momentum is
sumers have begun to spend at more rapid    weakening. Metal prices started to decline
rates-despite softening labor market condi-  again in the middle of 2002, adding to the
tions. Japan's strong export performance    doubts about the strength of the global recov-
stands at risk and could fade quickly should  ery. The further rise in agricultural prices was
foreign demand conditions worsen, should the  mainly due to specific supply disruptions-as
yen resume its appreciation against the dollar,  civil strife in C6te d'lvoire jeopardized cocoa
or both. In the United States, considerable un-  production and as droughts in Australia,
certainty is attached to the outlook for con-  Canada, and the United States boosted wheat
sumer spending. Following robust purchases  prices-and was not a sign of rising demand.
of durable goods during the third quarter-  Inflation remains low, but the danger of
particularly automotive sales that were in-  outright deflation has emerged in parts of
duced by zero interest incentives-questions  developing and industrial East Asia-China,
arise concerning the tenor of growth into the  Hong Kong (China), Singapore, Taiwan
final quarter of the year. Incentives cannot con-  (China), and Japan. Stable and low inflation is
tinue indefiniitely, and massive equity-based  a prerequisite for solid growth and creates a
wealth losses of the past two years could play  favorable environment for effective monetary
a larger role in households' consumption deci-  policy. However, In several cases, a sharp drop
sions. Though low interest rates have spurred  in inflation has increased real interest rates
mortgage refinancings and "cash outs," which  and has worsened debt problems. Although
are anticipated to place some $100 billion to  deflation limits the options for monetary pol-
$200 billion of additional liquidity into the  icy, it tends to depress investment and demand
hands of consumers during 2002, this trend  for durable goods.
could prove limited if the recovery falters seri-  With market emphasis on financial strains
ously. U.S. business remains cautious in invest-  and risk perceptions, it is important not to
ing or rehiring, in part because of the clouded  lose sight of several brighter spots in the
outlook for growth in final demand. And in  developing world. Market reforms, including
the Euro Area, especially in Germany, domes-  a diminution of trade barriers and an open-
tic demand is lackluster, and near-term growth  ing up to foreign competition achieved in



G LO B AL EC ONO MI C PROS PE CT S 2 O 0 3
many countries during the 1990s, are note-   China and several Central European countries
worthy. These changes have contributed to    are examples of successful reforming econo-
faster growth in trade and in welfare gains, and  mies that are preparing for even further inte-
the process continues across many countries.  gration into foreign markets. This integration
Box 112         EIntegration pays off where po]licies
are supportive
T   he recent global downturn has depressed export  inregration with global markets largely through the
T growth across the developing world, leading to  EU accession process, association agreements, and
a contraction in aggregate volume from the third  lower trade barriers, which will culminate with full
quarter of 2001 through the first quarter of 2002.  menmbership in the EU, which appears now on track
However, several countries-notably China, the   for 2004.
Czech Republic, Hungary, and Poland-have been       Reduction of trade barriers is but one factor
able to record impressive export volume growth  that has made these coLntries successful and able
(box figure). They have done so by increasing their  to take advantage of trade opportunities. Macro
market share, which has allowed them to partially  stability, rapid institutional reforms toward liberal-
offset the dampening effect of the global downturn.  ization of domestic markets, good or adequate
This increase in market share, in turn, reflects a  levels of education, and competitive wages (relative
continuing pattern of their intensifying integrationi  to productivity growth) all contribute to the
into the global economy and of attendant inflows  success.
of FDI.                                             These four countries have benefited from
Greater integration with world markets has   strong and sustained inflows of FDI from Western
been achlieved by China, for example, through its  investors that have been building production
recently gained membership (December 2001) in   capacity and transferring management skills and
the WTO, after years of negotiations and efforts to  technical know-how, in addition to financing, such
comply with WTO rules and standards. The three  that these countries can gain additional market
Central European countries have raised their trade  share as integration deepens. FDI inflows help ex-
pand production capacity and raise productivity
(that is, they help to improve the recipient country's
Several developing countries                 competitiveness). All four countries have posted
show solid export performance                liigh FDI inflows according to various measures.
(volumes, 3-month moving average, percent y/y)  For example, using the ratio of the economy's share
of world FDI inflows to the economy's share of
0       Chia                           world GDP, China, the Czech Republic, Hungary,
30                                         and Poland all posted ratios of above t (or the world
Eastem Europe         average) for 1998-2000-of 1.3, 2.7, 1.2, and 1.5,
20     /                                    respectively. As a share of gross fixed capital forma-
ro 1 0 h JF tlt nt 3;>V  \N, j  V          tion, FDI inflows to these countries have also been
1l l l ll 111 1f l  i   X  -    high, particularly in China, where FDI inflows aver-
o Ti                 "                    4t *  * l msI 1 1 l aged 13 percent of gross fixed capital formation dur-
Oiherdeveloping countnesWoU@World        ing 1990-99. In the Czech Republic, Hungary, and
Jan   Jul   Jan   Jul   Jan   Jul         Poland, FDI inflows averaged 25, 19, and 16 per-
2000  2000  2001  2001  2002  2002        cent, respectively, as a share of gross fixed capital
formation during 1997-99.
Note y/y refers to year over year
Souerce Datastream and World Bank staff estimates  So__rce__World_Bank_staff
12Source: World Bank staff.
12



r HE IN T ER N ATI ON AL EC ONO MY AND P R O S P E C I S F O R D E V E L O P I N G C O U N T R I E S
not only pays off in the long run, but also has  * Growth in 2003-04 is anticipated to be
assisted these countries in absorbing or even    moderate. World GDP reflects a combi-
avoiding short-term shocks and fluctuations,     nation of a more gradual recovery in
by promoting business confidence and by facil-   Latin America, an emergence of modest
itating export-oriented FDI. Perhaps more        growth in Japan, and a generally sub-
important, through relatively large inflows of   dued rebound in other parts of the global
FDI, these countries have become less vulner-    economy. The average outturn does not
able to turmoil in internationial financial mar-  follow the typically strong patterns of
kets (box 1.2).                                 recovery and expansion that is concurrent
across regions and reinforced by accom-
The medium-term outlook calls                    modating macro policies.
for modest growth in the global                * Downside risks to the baseline forecast
economy-third phase                              are substantial. The global recovery
The baseline forecast reflects the interaction   appears vulnerable to additional shocks.
of strong opposing forces: the stimulative      (These points are discussed after we
policies and the intrinsic recovery in stock    review  the external environimilent and
building and high-tech production that work      outlook for developing countries.)
to accelerate global growth on the one hand,
and the financial strains-high and rising debt  The external environment is mixed
levels, falling equity prices, and uncertainty  The driving forces may be found in expecta-
about profitability-on the other hand. These  tions for growth of global trade volumes
driving and restraining forces affect the out-  2.9 percent in 2002 and an average of 7.5 per-
look in three ways:                         cent in the following two years (figure 1.8 and
table 1.2). The restraining forces make a sub-
Global growth in 2002, the initial year  stantial rise in capital flows unlikely over the
of the forecast, shows little resemblance  medium term. The modestly firming trends in
to the uniform recovery that one would  non-oil commodities prices-5 percent in 2002
normally expect after an almost synchro-  and averaging 5 1 percent in the following
nous downturin in 2001. Instead, it dis-  years-reflect the subdued recovery. The 5 per-
plays quite diverse patterns of activity  cent gains in commodity prices remain far
across industrial as well as developing  below historical patterns during booms. In-
countries.                             creases in real commodity prices are expected
Figure 1.8 World trade rebounds along with GDP, 1998-2004
(percentage change)
1 4
10                               /Worldtrade
6                  /World GDP  \I
6
4
2  .     .l__       0_
0
-2 
1998       1999      2000       2001       2002       2003       2004
Soutrce IBF, OECD, World Banik. anid WXorld Banik prolctloris
13



G LO B AL    EC ONO MI C      PRO SP E CT S      2 0 0 3
Table 1.2 External environment for developing countries, 1991-2004
(percenttage change from previous year, except tinterest rates and oil price)
Current
estimate               Currenit forecasts
Growth rates/ratios                 1 1991-2000       1999      2000      2001      2002      2003      2004      2002-04
Industrial country GDP growth       1     2.4          2 9       3 4       0.8        1 4       2 1       2 6        2 0
World trade growth'                       7.2          5.6      13 1      -05        2.9        6 7       7.7        5 7
Industrial country import demand        6 9          8.5      11.6      -1 0        1 3       5.3       6.8        4 4
United States                         9 4          12.4      13.7     -3.6        4 4       8.1       8 0        6 8
Japan                                  5 6          6 6      10.7     -2.8      -2.0        7 7       8 6        4 7
Euro Area                             6.6          6 4       11.2       0 8     -0.5        4 0       6.6        3.3 
Developing-country import demand        8.2         -1 3      16.9        45        56       10.1      100         8.0   1
Market growth for developing countricsb  10.7          5.3      13.1       0 2        2 6       7.0       8.4        6.0
Non-oil commodity prices (nominal)       -1.4        -11 2      -1.3      -9.1        5.0       5 8       4 4        5 1
Agriculture                            -1.3        -13 9      -5 5      -9 1        8 4       8 5       4 8        7.0
Metals and minerals                    -1 8         -2 3       12 6     -9 6      -3.5        5.6       5 7        2.5
Real non-oil commodity pricesc         -1 1        -11.0       0.8      -7 8        4.5       2 8       2.2        3.2
Oil price ($, weighted average), $/bbl   19.1         18 1      28 2       24 4      25.0      23.0      20 0       22 7
Manufactures unit value indexd      1    -0.3         -02       -2.1      -1 4        0.5       3.0       22         1 9
Developinig-country terms of trade        0.1          3.3       2 8       0 3      -3 2      -1.4      -1.3       -2.0
Terms of trade/GDP ('.),                0.1           0.7       0.6       0 1      -0 8      -0 4      -0 4       -0 5
LIBOR (US$, 6 months)                     5.6          5 5       6 6       3.6        1.8       1.5       3 1        2.2
EURIBOR (euro, 6 months)                  5 4          3.1       4.5       4 2        3.4       3 2       3 8        3.5
Note bbl = barrel, LIBOR = London interbank offered rate, EURIBOR = European interbank offered rate
a Goods and nontfactor services
b. Weighted average growth of import demand in export markets
c Deflated by manufactures unit value index
d Dollar-based export prices of manufactures in the G-5 countries
e Chanige in terms of trade, measured as a proportion to GDP (percent)
Source World Bank, November 2002
to be even more moderate, about 2.5 percent           sipates and that increased supply, both from
per year. Most of the increase in commodity           Organization of Petroleum      Exporting Coun-
prices in 2002 was due to a surge in agricul-         tries (OPEC) and non-OPEC sources, can eas-
tural prices, which bounced off cyclical lows:        ily meet moderate increases in demand. This
the agricultural index had dropped 40 percent         forecast would     imply   oil prices of around
below   its peak, which was reached in 1997.          $20 per barrel by 2004. The lack of strength
The recent surge is to a large extent induced          in the recovery is also reflected in the interest
by supply factors: for example, droughts in           rate projections. The London interbank of-
Australia and the United States have boosted           fered rate (LIBOR) and European interbank
grain prices, and supply disruptions in Cote          offered rate (EURIBOR) are anticipated          to
d'Ivoire and Ghana did the same for cocoa             drop modestly further during 2003, but to in-
prices. Conversely, the rally in metal prices that    crease in step with firming economic activity
started in October 2001, which is normally            by 2004, at rates below 4 percent.
strongly correlated   with the business cycle,            The recovery in global markets is shaped
stalled in the second quarter of 2002. The fore-      primarily by developments in industrial coun-
cast for metal prices is one of decline in 2002,       tries. At 1.4 percent in 2002 and an average of
and of a return to 5-6 percent gains thereafter.      2.3 percent in the following years, growth re-
This trend is another indication that the tenor        mains at or below    potential, which is a rare
of global recovery is anticipated to be modest.       phenomenon during a recovery (figure 1.9). In
The base forecast assumes that the current         this forecast, inflation will accelerate little, re-
risk premium    in the oil market gradually dis-      maining below 2000 or 2001 levels.
14



I H E I N T E R N A T I O N A L E C O N O M Y A N D P R OS P E C T O r o R D E V 1: L O P I N G C O U N I R I E S
Figure 1.9 2002 marks the start of a moderate recovery
(GDP growth rate in percent)
6
Developing countries  Forecast
5
S          ~~~~~Industrial countr e\>
4
3
2
0                                                _/_
1998        1999        2000       2001        2002        2003        2004
Source World Banik data and projections
An outlook for moderate growth across           12 percent, and growth in the region slowed
developing regions-                             from  1.2 percent in 2001 to 0.7 percent in
What does this environment imply for output     2002). The region's per capita income fell
growth in the developing countries? In 2002,    2.6 percent after a drop of 1.2 percent in 2001,
a strong recovery in East Asia coincides with a  which was the worst performance across two
disappointinig performance in Latin America,    years since the debt crisis in the early 1980s
where GDP declined by 1.1 percent (excluding    (figure 1.10). Oil exporters, particularly in
Argentina, where GDP plummeted by nearly        the Middle East and North Africa (MENA),
Figure 1.10 LAC and MENA are not experiencing the recovery
(developing countnes GDP growth, percent change)
7
6
5
4~~~~~~~~~~20
3
2~~~~~~~~~20
-1
2002
-2
East Asia    South Asia      Laiin        Eastern     Middle East  Sub-Saharan
and Pacitic                America and   Europe and    and North      Africa
the Caribbean  Central Asia   Africa
Source World Banik projections
15



G LO BA L E CO NO MI C PROS PE CT S 2 0 0 3
follow an independent growth pattern, which,  fit of hindsight, it is quite difficult to disentan-
to some extent, is the same for developing    gle the set of recent shocks and their effects on
countries that are rapidly integrating into for-  developing countries. Yet the downward revi-
eign markets (for example, Central European   sions do not contradict the assessment made
countries, China, and Mexico), where exports  in the fall of 2001 that adverse effects stem-
were recently able to outperform world trade  ming from the terrorist attacks of September
growth as a whole (see box 1.2).              2001 would not be limited to the United
Average growth in 2002 for developing      States, but would spread to developing coun-
countries is anticipated to be 2.8 percent,  tries as well (box 1.3).
0.3 percentage points lower than was expected   Sharply different growth patterns are likely
in the February 2002 forecast and 0.8 per-   to characterize economic activity across coun-
centage points lower than was projected in the  tries and regions in the short run, as jittery
December 2001 forecast. Even with the bene-   financial markets affect the vulnerable and
]3ox 113          The terrorist attacks of September 119 2001
had an economic effect
Shortly after the terrorist attacks of September 11,  of accounting scandals in the United States, compli-
V2001, the World Bank concluded that the eco-    cate the picture.
nomic effects would be most severe in the United     The fiscal stimulus and rapid monetary easing
States but would be significant in developing coun-  in the United States probably prevented a serious
tries. The main transmission mechanisms were     delay in the U.S. recovery and in world trade growth.
thought to be:                                   Non-oil commodity price increases have slightly out-
performed forecasts made in the fall of 2001. How-
• Tourism revenues would decline, especially in South  ever, the steep decline in tourism revenues and the
Asia, the Middle East, and the Caribbean.      increase in risk perceptions did materialize, and the
o Increased risk perceprions in international markets  outlook for developing countries has further deterio-
would make oil prices more volatile, foreign capital  rated, especially for Latin America. This observation
less readily available, and transportation more costly.  suggests that 9/11 did exert strong ifluence in shap-
o There would be delayed recovery in the United States,  ing ensuing economic trends in developing countries,
where immediately after the attacks air traffic was  albeit reinforced by other factors.
constrained, equity prices had declined sharply, and  Following continuous and robust growth over
consumer confidence had plummered. That delay  the past decade, global tourism arrivals declined
would also hamper the recovery In world trade, com-  by 0.6 percent in 2001, tied in large measure to the
modity prices, and financial flows,            effect of the terrorist attacks. The most-affected
Of the developing world, Latin America was   developing regions were South Asia (down 6 per-
thought to be the hardest hit because of its proximity  cent), Latin America (down 4 percent), and the Mid-
to the United States, its dependence on tourism rev-  dle East (down 3 percent).
enues and commodity prices, and its vulnerability     Business and consumer confidence across the
to financial shocks. Countries in Sub-Saharan Africa  industrial centers fell abruptly in the aftermath of
were also vulnerable because they have limited   the attacks. Although the losses evaporated within
options to absorb adverse shocks.                2 months as the recovery took shape, the tenor of
Even with the hindsight of a year, it still remains  business and financial market confidence has contin-
quite difficult to assess the independent effect of 9/11  ued to be exceptionally fragile (box figure). This fac-
on the global economic environment. Not only are  tor was one underlying the decline of capital market
the counterfactuals unkinown but new shocks, such  flows to emerging market recipients in 2002, as well
as the financial crisis in Argentina or the emergence  as the rise in bond markets.
16



T HE IN r ER NAT IO NA L EC ONO MY AND PRO SP EC TS F OR Dr v EL O PING CO UN TR IE S
Box      1.3      (continued)
U.S. confidence plummets
(ISM and Conference Board business and consumer sentiment indices)
60                                                                              120
:0 -              \              / Consumer confidence                          110
ISM index \              /                                100
50  -
45 ~~~~~~~~~~~~~~~~~~~~~~~~~~90
80
Jan       Apr       Jul        Oct       Jan       Apr       Jul        Oct
2001      2001      2001       2001      2002      2002      2002      2002
Soiurce Iinstitute for Suppl) Managenient and Coniferenice Bloard, throig/i Datastream
Developing countries' GDP growth in 2002 has  thesc developments took form under the influence of
been limited to 2.8 percent-about 0.8 percentage  a global environment that was highly unsettled by
point below that expected onc year ago (notably,  the destruction of the New York World Trade
figures that included an assessment of the 9/11 ef-  Center.
fects). Although not all barriers to stronger growth
are linked directly to the terrorist attacks, most of  Soutrce World Bank staff
highly indebted developing countries more     together with the pursuit of policies geared to-
severely than countries with lower debt ratios.  ward reducing financial strains. Mexico and
Though the forecasts for 2003-and espe-       sorine Central American and Caribbean coun-
cially for 2004-display    accelerationi of   tries are in position to benefit most from the
growth that tends toward more uniformity      expected upswing in the United States, and
across regions, that acceleration is weaker   Mexico's growth in particular is anticipated to
than one would expect in a strong, synchro-   exceed that of most Latin American countries.
nized global recovery.2
The forecast for Latin America and the      -has the strongest growth evident in
Caribbean (LAC) assumes some rebound in       East Asia-
Argentina, where output has fallen some       Prospects for developing East Asia and Pacific
20 percent below 1998 levels, but the rebound  (EAP) appear more buoyanit than those for
is insufficient to return to carlier prevailing  other regions, as growth is expected to reach
levels within the time horizon of this forecast.  6.4 percent by 2004. Continiued solid expan-
Modest growth rates are anticipated for Brazil  sion in China and recovery in most other
and most other countries of the region. That  countries-albeit with growth rates that re-
growth is grounded in a recovery in global    main below the robust performance of 2000-
trade and an end to the freefall in Argentiiia,  underpin this view. Favorable prospects do
17



G LO B AL EC O NO MI C PRO SP E CT S 2 0 0 3
not imply that risks are negligible, however.  infrastructure and other development activities,
East Asia remains vulnerable to oil price   especially in Algeria, the Islamic Republic of
spikes, to uncertain demand conditions in the  Iran, and Saudi Arabia. Growth among the di-
United States, and to the fragile state of the  versified exporters should increase to an aver-
Japanese commercial banking system   and    age of 3.2 percent, as drought conditions ease
growth prospects there. Moreover, the dy-   in Morocco and Tunisia and as fiscal deficits
namics in high-tech markets remain volatile.  are brought under tighter control and business
Options for domestic stimulus are more lim-  confidence returns in Egypt-as the govern-
ited than in previous years because in most  ment there sets an appropriate interest rate and
countries fiscal deficits have widened and in-  pushes ahead with policies, such as privatiza-
terest rates stand at low levels.           tion, that will increase international investor
In Europe and Central Asia (ECA), growth  confidence. Risks to this outlook are substan-
is expected to remain strong, but it will be  tial, however, with political tensions mounting
grounded in a highly differentiated outlook  during apparent preparations for military
between the Central and Eastern European    action in Iraq. At this juncture, the baseline
(CEE) group of countries and the hydrocarbon  does not explore these potential developments,
exporters that dominate growth trends in the  but rather focuses on the country-specific and
Commonwealth of Independent States (CIS-    region-specific economic fundamentals, as well
Russian Federation, Kazakhstan, and several  as global factors that contribute to shape the
smaller states). For the former group, output  outlook.
growth is projected to accelerate from 2.3 per-
cent in 2002 to 3.1 percent and 4.3 percent  -and South Asia
in 2003 and 2004, respectively. Activity is  A forecast of consistent growth in the South
expected to be driven by increased import   Asia region (SAR) of well above 5 percent
demand from the EU and by intensification   over 2003-04 comes after a significant cycli-
of the EU's accession process. For Turkey   cal downturn in 2001, when manufacturing
(included in this group), assuming that there  output in India and Pakistan stopped growing
is relative political stability and that the new  and when GDP growth mainly reflected con-
government continues to pursue the current re-  tinued expansion in the service sectors. The
form path, recovery is expected to strengthen  main challenge for the subcontinent remains
in 2003. In contrast, growth is anticipated to  fiscal reforms to curb over-large government
ease in the CIS subregion in the years through  deficits and to promote further trade liberal-
2004 (through fiscal and trade linkages to the  ization. With almost-balanced current ac-
hydrocarbon exporters in particular), assum-  counts and with substantial capital flows into
ing a significant medium-term decline in the  Pakistan, external financial tensions remain
oil price. CIS GDP is anticipated to decelerate  limited at present. But, it is expected that the
from 4.4 percent in 2002 to 3.5 percent and  effects of accumulated fiscal debt will, at some
3 percent in 2003 and 2004. These divergent  future point become an obstacle to achieving
trends combine to shape the path of growth  the acceleration in growth required for sub-
for the broader region, from 3.6 percent in  stantial alleviation of poverty levels.
2002 to an average of 3.5 percent in the years  Growth in Sub-Saharan Africa (SSA) re-
following.                                  mains restrained by unfavorable domestic con-
Growth in the Middle East and North Africa  ditions, ranging from civil strife, to droughts,
(MENA) region is expected to revive in     to macroeconomic imbalances, and to the AIDS
2003-04 to average 3.6 percent, as hydrocar-  epidemic. Elements of the external environ-
bon output increases in line with global energy  ment should, however, provide some support
demand, and as accumulated oil-surplus funds  for a modest acceleration of growth over the
are progressively committed and expended on  next years. Despite a relatively sluggish pickup
18



T I1 E IN FE RN Al IO NA 1. EC ONO MY A ND    PRO SP EC TS F OR I) E V EL O PING C O U N T R I E S
in world GDP growth in 2003-04, a robust    markets, thereby heightening tensions in sev-
recovery is anticipated for African trade vol-  eral vulnerable middle-inconie countries; and
umes, ratcheting from growth of 3 percent in  (c) the risk of higher oil prices, which are as-
2001 toward 6 percent by 2004. That recov-  sociated with prospective developments in the
ery should be accompanied by generally      Middle East.
firmer non-oil commodity prices (exceptions   The base case presents a moderate but
are cocoa and gold, where prices have surged  steady recovery in investment; it effectively
to unsustainable levels). The resulting terms of  rules out finaicial crises in middle-income
trade gains should support relatively buoyant  countries and foresees a gradual decline In oil
external performances by African non-oil cx-  prices. If downside risks materialize, adverse
porters. For oil exporters of the region, the  outturns in these domains could easily occur
price of crude is expected to weaken in the  at the same time or could, in sequence, rein-
medium term. Even so, oil sectors will remain  force cumulative effects on the economy. To
profitable, and production and export vol-  gauge the sensitivity of economic recovery to
times are anticipated to rise-from i'ligeria  these risks, we have traced the possible effects
and other producers in the Gulf of Guinea, as  on the economic outlook of these elements.
well as from Angola's offshore sector.      The results underscore the set of tensionis em-
In the domestic sphere, agricultural pro-  bedded in the base-case forecast and can illu-
duction will benefit from a return to more  minmate the magnitude of potential downside
normal weathier patterns in southerin Africa,  risk to the projections-with particular focus
thus contributing to a recovery of domestic  on the implications for developing countries.3
output and expenditure. On balance, GDP
growth for the region is expected to rise from  Global recovery could be delayed
2.5 percent in 2002 to 3.2 percent in 2003 and  until 2004
3.8 by 2004. The overall acceleration reflects  Table 1.3 outlines the global effects of a low-
gains by non-oil exporters, which will more  case scenario in which the risks highlighted
than offset modest retrenchment by oil pro-  above occur essentially at the same time, but
ducers. The current projection for the region  each of the adverse shocks is fairly short lived.
represents a sliglt deterioration of prospects  The scenario reflects the joint effects of a
compared with the spring 2002 forecast,     temporary relapse to negative growth in the
which is consistent with the overall down-  industrial country's investimient cycle, of short-
grading of expectations for world output and  lived financial disruptions in several middle-
trade growth. Nevertheless, though perfor-  income countries, and of a momentary spike
mance will continue to lag behind other devel-  to $45 per barrel (bbl) in world oil prices The
opmig regions, per capita incomes are set to  scenario suggests that, rather than an acceler-
resume positive growtlh following several years  ation of global growth in 2003 to 2.5 percent
of stagnation.                              as in the base case, a continuation of sluggish
output advance in a range of 1.9 percent could
Risks to the base case are substantial      characterize the year. Contrasted with base-
The world recovery is clouded by substantial  case forecasts, cumulative differences over
uncertainties in the immediate to near term.  2003-04 in world trade growth, OECD infla-
These uncertainties carry with them implica-  tion, and interest rates are fairly substantial.
tions for medium-term developments in global  The latter element reflects a strong monietary
growth and financial flows. Among critical  policy response to the financial and real dis-
factors in the outlook are (a) continued finan-  turbances of the scenario. OECD  output
cial turbulence in high-income countries that  growth is dampened by 1 percentage point,
could jeopardize a rebouncl in investment;  and for developing countries, it is dampened
(b) a reversal in capital flows to emerging  by 0.8 point over the period (figure 1.11).



G LO B AL    EC ONO MI C      PRO SP EC TS       2 0 0 3
Table 1.3 Global effects in a low-case scenario, 2003-04
2003                         2004                2003-04
Scenario      Diff. lbase)   Scenario      Diff (base)    Cum. diff
GDP growth (%)
World                               1 9           -0.6           2 7           -0 4           -1.0
industrial countries (OECD)       1 5           -0 6           2.2           -0 4           -1 0
Developiiig countries             3.0           -0 9           4 8            0 1           -08
Consumer price index inflation (%)
Industrial counitries               2 2            0 1           1 2           -0 6           -0.5
Developing counitries (median)      5 0            0 6           4.4            0.0            0.6
Short-term interest rates (%)
Industrial counitries               2.9           -0.3           3.3           -0 8           -1.1
Trade volumes (%)
OECD imports                        5 5           -1 3           7 5           -0 4          -1 7
Developing-country exports          9 3           -1 0           9.2           -0 3          -1 3
Source World Bank, Noveinber 2002
Among industrial countries, growth pro-            falling substantially   below   baseline), but it
files in the United States and Japan are more         will rebound wsth some vigor as conditions
adversely affected. This is linked to the relapse     equilibrate in 2004. East Asia-with        strong
of fixed   investment spending     during  2003,      links to export markets in all three industrial
with less room for monetary easing than exists        centers-will also     suffer a sharp    falloff in
in the Euro Area. Among developing regions,           growth, but less so than Latin America.
Latin America will feel the initial brunt of             Initially, inflation increases slightly in reac-
diminished capital inflow during 2003 (growth         tion to the rise in oil price. However, when oil
prices start falling again and the effect of
lower growth becomes noticeable, inflation
Figure 1.11 Low case: world trade and              will drop in high-income countries, thereby
other indicators will be much lower than           triggering substantial monetary easing com-
the baseline                                       pared with the baseline. In developing coun-
(cumulative difterences,2003-04, ow-case scenano   tries, inflation will remain on average above
versiusbase-case scenano, percent)                 its base-case level, as higher oil prices are com-
Developing-       plemented by devaluation of several curren-
World Interest OECD   OECD   country
trade  rates  inflation growth  growth       cles in reaction to financial tensions.
o0        *       *                                   Table 1.4 breaks out for 2003 the contri-
butions of the individual risk scenarios (out-
-0            5                                     lined below) to the overall low-case simula-
I         -oUs            -             tion. A relapse of investment in the industrial
-1C0                         -       -08            countries carries the largest downside poten-
-1 1           -0                     tial to the global outlook, which affects output
-1 5                                                growth, world trade, and interest rates most
acutely. Developing-country growth is more
-1 7                                         affected under the restraint of capital flow
-20                                                 scenario, but is equally diminished by devel-
Soutrce World Bqi,k                                 opments under the G-7 investment scenario
and by higher oil prices. The latter scenario
20



T H E  I NT ER NAT IO NA L     EC ONO MY AN D       P R O S P IiC r S  F O R D E V r L o P I N G  C O U N I R I E S
Table 1.4 Low case: contributions to global effects in 2003
- -|       Diar  kiu-j   1--  -      L.  I.i Flu.."  uJ p;.J
GDP growth                          -                                                           .
World                            19           -0.6         -0 3          -0 1          4 2
Inidustrila countries (OECD)    1.5          -0 6        -0 3          -0 1          -0 2
l)eNelopinig counitries         3 (   .      -0.9        -0 .2         -0 5          -0.2
Consumer price index inflation (%)
Inidustrial countries            2.2           0 1         -0 I           0 0           0 2
Devcloping countries (mediani)   5.0           0 6         -0 1           0 1           0 6
Short-term interest rates          2.9   -      -0.3         -0.3         -0.1            0 1
World trade                        5.8          -1.3         -0.9          -0,1         .0.3
OECD imports                     5.5          -1.3         -1 0           0 0          -03
Developinig exports              9 3          -1 0         -0 6          -0 2          -0 2
Source World Banik, November 2002
prospectively could hamper a fuller easing of
monetary policy across the industrial centers,       Figure 1.12 G-7 investment falls sharply
with growth suffering commensurately.                (base-case scenario and low-case sceonanos, voliume
growth, percent saar)
Financial market turbulence in high-                    s
income countries could trigger a relapse
in the investment cycle-                                          Base-case
\ /scearraio
Though investment growth momentum is now                o
building in selected sectors and countries, the
potential for relapse looms large as imbal-
ances and uncertainties remain acute through-                                      Low-case
out the industrial world. The low-case sce-                                        socnsrio
nario assumes that the growth of real business       -10
investmiient in industrial countries drops to             04 02 04 Q2 04      Q2 Q4 02 Q4
negative territory (-0.7 percent) during 2003,           2000 2001 2001 2002 2002 2003 2003 2004 2004
which   would   be 3.5 percentage points of          Souirce Natnoiial statistics. World Basik proiectio,,s
growth below    the baseline path. Spillovers
into 2004 carry cumulative growth differences
in capital spending to 6.5 percentage points
(figure 1.12).                                    0.4 percent in 2003 and by a further 0.3 per-
If one examines the scenario environment        cent during 2004, as multiplier effects place
among the industrial countries, cumulative        pressure on    household  consumption. With
inflation over 2003-04 is reduced by 0.5 per-     the compression of imports, adverse growth
cent. At the same time, the profile of short-     effects in the major industrial countries are
term interest rates reflects reductions that are  transmitted to smaller advanced economies,
more than the improvement in inflation per-       while weakened     demand    for commodities
formance, which     represents a substantial      places downward      pressure  on1 nonenergy
easing of monetary policy in the wake of de-      prices, thus affecting developing countries'
velopments. The rate of unemployment rises        terms of trade adversely. On balance, GDP
by 0.4 percentage points in OECD countries.       growth in developing countries will decline
Among prominent growth effects, industrial        by 0.3 percentage points relative to the base
country   output gains    are  dampened     by     figures during 2003-04.
21



G LO BA L EC ONO M IC PRO SP EC TS      2 0 0 3
-while financial tensions pose difficulties
for emerging markets-                           Figure 1.13 Investment growth and net
Under a scenario of substantially lower in-     capital flows* into Latin America are
flows of private capital into countries with   strongly correlated
weak r deciningcredi ratigs an withlarge (4-quarter movwng  (4-quarter moving
weak or declining credit ratings and with large  average. percent GOP)  average, percent saar)
financing requirements, interest rates rise, ex-                                 25
change rates depreciate, or both, which raises          Investment               20
the cost of servicing debt and results in diffi-  6  A                           20
culties in meeting debt payments for both the                                    10 1  - 5
public and private sectors. To understand the  4 -
possible implications of such development, we       \
performed a simulation in which capital flows                                   -5
to selected emerging market countries were     2
assumed to drop by 15 percent below the         1          Netcapdalflows      -
baseline forecast in 2003, while spreads on in-            as percent of GOP
I I   I     2
ternational debt would increase substantially.    1 01 01 Q1   01 01 Q1   01 01
With these assumptions, the Latin America     1994 1995 1996 1997 1998 1999 2000 2001 2002
region will suffer an average falloff in output  'Net capital flows defined as CAD-addition to
growth of 1.9 percentage points compared       reserves-lMrF funds
with the baseline in 2003  . Rebound in re-  Source Datastream and World Bank staff estimates
with the baseline In 2003. Rebound In re-
sponse to eventual equilibration in exchange
and interest rates, as well as to falling risk
spreads, should be grounded in stronger ex-  prices. A so-called war premium on prices has
ports, and output growth rises some 0.9 per-  been estimated at up to $8-$9/bbl. The base
centage point above the base in 2004. The    case assumes that the oil market normalizes
cumulative fall in regional growth amounts   with further increases in non-OPEC supply,
to 1.1 percentage points during 2003-04. This  with a modest rise in OPEC quota, and with
fall is a reflection of the strong dependence  a dissipating of the war premium. These as-
of Latin America on capital inflows, whereas  sumptions might turn out to be too optimistic.
reversals in capital flows tend to have far-    With continued tensions in the Middle East
reaching consequences for domestic economies  and  vith the possibility that, for example,
in that region (figure 1.13). Central Europe is  some 2 million barrels per day (mb/d) of Iraqi
another region that is affected by the reversal  oil exports would be temporarily lost to the
in international capital flows, although the  market, oil prices might rise well above
greater diversity within the region makes the  $30/bbl-partly because of low stocks and
overall effect smaller than for Latin America.  tight market conditions-and might peak at
$4S/bbl during the height of the disruption.
-and higher oil prices could temporarily     However, it is likely that any loss in supply
(yet moderately) dampen recovery             will eventually be replaced by other OPEC
Crude oil prices have risen to more than     producers. They currently have about 5 mil-
$29/bbl because of expectations of a supply  lion barrels per day (mb/d) of spare capacity-
disruption in Iraq and of increasingly tighter  of which Saudi Arabia alone has some 3mb/d.
market fundamental conditions. Crude oil     Or supply will otherwise be replenished. Prices
stocks fell sharply in the third quarter of 2002,  could fall relatively quickly below $20/bbl by
particularly in the United States. The drop was  2004 before OPEC begins to restrain output as
due to high runs of refined products, a decline  world demand increases, as the organization
in Iraqi crude exports, and continued restraint  attempts to bring prices back into its target
by OPEC to limit exports in support of higher  price band of $22-$28/bbl (figure 1.14).
22



T II r  I N i E R N A T I O N A L  E C O N O M Y  A N D  P R O S P E C I S  F O R  D)  V E L. 0 P I N G  C O U N I R I E S
Figure 1.14 Oil prices spike                          Figure 1.15 Investment is more volatile
(base-case and low-case scenanos, oil pnce $/bbl)     than GDP in East Asia
36 0                                                  (4-month moving average, percentage change qlq, saar)
A                              30
31 5                   /                                                     ,A    ,-Investment
Base-case            Low-case                20
scenarqo             scenano                           GDP
270                                                     10
18 0                                                    0 7|
04   02  04   02  04   Q2  04   02  Q4            -20
2000 2001 2001 2002 2002 2003 2003 2004 2004       -30
Soturce World Bank dati, World Bank prolectionis      -40
Q1     01      01     01      01     01
1997   1998    1999   2000   2001    2002
Note Indonesia, Rep of Korea, Malaysia, Phiippiiies,
On a cumulative basis, a $2.501bbl incrcase           aid Thliailaid
in oil price above baseline levels (but with dy-         Sozurce Datastrearn and World Banik ctaff estimrtes
namics as outlined above) will yield a moder-
ate fall in global growth of 0.1 percent during
2003-04. Effects during 2003 will be some-
what more pronounced, however: a drop in
world output of 0.2 percentage points. Infla-            than GDP in Latin America
tion and interest rates in the industrial coun-
(4-month moving average, percentage change q/q, saar)
tries  will be   boosted   modestly, by     some
0.2 and 0.1 percentage points, respectively.               20
15/
10       a    / Investment
Investment cycles in developing                             5              _4GDP
countries
For developing countries, the risk         of an
F   untimely interruption to the recovery in               -s
global investment comes after a period of sharp           -10
swings in investment in the past five years.              -1     _
During the East Asian crisis, investimient fell at           GI     Q1      Q1     01      Q1     01
an annual rate of 30 percent, three times the               1997    1998   1999   2000    2001   2002
fall in output (figure 1.15). Crises in Argentina        Note Argentiia, Brazil, Chile, aid Mexico
and Turkey dominated recent developments in              Source Dattstrearn aid World Banik sra ff estimates
Latin America and Central Europe, respec-
tively, with regional investment declining at
annual rates of 15-20 percent (figures 1.16 and
1.17). The dynamics of itivestment are much                 isons of developing     with   high-income
more forceful than that of output.4                         countries.
This section    looks at three     important           * The cffects of improvements in the in-
patterns:                                                   vestment climate     on  the volatility   of
investment and output.
. The volatility of investment, relative to           * The role of capital flows that influence the
the volatility  of output, and     compar-            investment cycle in developing countries.
23



G LO BA L EC ONO MI C       PROS PE CT S 2 0 0 3
for some 160 countries over the 1990-2000
Figure 1.17 Central Europe and Turkey          period.6 The volatility of the cyclical compo-
experience greater volatility in               nents of investment declines steeply with
investment than in GDP                         htgher income per capita.
(4-month moving average, percentage change qlq, saar)  Explanations for the high volatility of in-
45                                            vestment in low- and middle-income countries
30  A                                  will vary from large external shocks relative to
30    \     Investment                    the size of the country to a poor investment
1 5            GDP                           climate. Properly functioning domestic finan-
cial institutions may smooth cycles by allow-
o,   ,   1  , >    ?  ,   | t;4-   /sing additional savings to be channeled to
investors during downturns.
-15                                               Poor countries, on average, tend to be rela-
tively small economies. Thus, the GDP of an
-30       Q     Q      Q      Q                average low-income country during the 1990s
1997   1998  1999   2000   2001   2002      was $25.6 billioon, barely 2.5 percent of the
Note Czech Republic, Poland, and Turkey        average high-income country. Baxter and
Source Datastream and World Bank staff estimates  Crucini (1993) and Crucini (1997) argue that,
,_________________ _  as a result, external shocks are relatively large
in proportion to GDP, which explains reason-
Table 1.5  Relative volatility of investment  ably well the patterns observed in figure 1.18.
Table high indevelopingcountriesFor example, international capital flows can
easily be much larger from the standpoint of
1971-80  1981-90  1991-2000    a small country, and reversals in capital flows
Low ncome           4.6      5 2      7.6         can have a relatively large effect. The decision
MiddleIncomne       4.9      3.6      4.5         by a French multinational to invest $100 mil-
Hlgh-income OECD I  2.9  3.2   ~  lion in Senegal instead of at home would re-
Note This table presents tinweighted averages of country-  duce France's investment spending in 2000 by
specific standard deviations of investment growth as a ratio
to the unweighted average of standard deviations of GDP
growth
Source World Bank,
Figure 1.18 Investment volatility declines
Investment cycles are more pronounced                with income
in lower-income countries than in higher-           (mean and standard deviation of volatitty of investment)
income countries                                    020
The volatility of investment growth relative to
the volatility of output growth is twice as large    0 15              ard deviation
in low-income countries as in high-income                                      Mean
economies-and volatility has increased over          0 10
time (table 1.5). An understanding of the in-
vestment cycle is pivotal to the explanation         005
of overall cyclical behavior in developing
countries.                                             o
Low     Middle  High income High inoome
A similar picture emerges if one examines a             income  income  non-OECD   OECD
different measure of the cyclical component of
Note lnvestnienr volatility is defined as the srandard
investment, namely the percentage deviation         deviation of the deviation from (HP filtered) trend
from  trend.5 Figure 1.18 displays the mean         Source World Bank staff estimates
and standard deviation of that measure
24



T HE IN T ER NAT IO NA L EC ONO MY AND            PRO SP EC TS F OR DE V EL O PING         CO UN TR I ES
just 0.04 percent, but it would raise Senegal's
investment by 11.5 percent. The effects on         Figure 1.19 Abetter investment climate
the two economies of such a decision would         reduces volatility of investment cycles
be disproportionate. Similarly, idiosyncratic      (standard deviation of cyclical component of investment)
shocks-economic, weather-related, or the           0 4
reflection of civil strife-have a relatively large             Regulation
effect on smaller countries. Low levels of de-     03
velopment and small size will tend to imply                           i . "     ar.z;Z
less diversification in the output and export      02
mix but stronger dependence on commodity
prices, so that poor countries' terms of trade     0 1
will tend to be more volatile than those of
OECD countries.                                        ist quarOle 2nd quartile 3rd quartile 4th quarbile
Governance
Improvements in the investment climate                Bad                           > Good
can reduce volatility                              Source World Bank staff estimates using governance
The qualityof the invstment climte, exten- indicators developed by Kaufmann, Kraay, and
The quality of the investment climate, exten-      Zoido-Lobar6n (2000)
sively discussed in the following chapters of     _
this report, appears to be highly correlated with
investment volatility. If one examines the
investment climate, several candidates are
available to proxy for this environment, all of    Figure 1.20 Impact of policy climate on
which correlate highly with one another. Fig-      investment volatility after correcting for
ure 1.19 is based on the quality of governance     income remains strong
indicators compiled by Ka ufmann, Kraay,           (standard deviation of cyclical component of investment)
and Zoido-Lobat6n (2002), which have the           0 25
advantage of comprehensiveness. Six sub-           0 20-
indexes attempt to capture various dimensions                      _
of policy and institutional quality. An un-
weighted average of all six is identified in       0 10-
the figure as "All." Meanwhile, two of the        0o05 -
subcomponents seem    especially pertinent to                  *        *       h
measuring the investment climate: "regulatory      oo2nd                    3rd     4th
burden," which reflects the incidence of market          quartile  quartile  quartile  quartile
friendly policies, and "rule of law," which mea-                    Governance
sures respect for society's rules. "Regulation"        Bad                           o Good
consists of the first of these, while "Regulation  Sozurce World Bank staffestimates using governance
and Law" is an average of the two. All three       indicators developed by Kaufmann, Kraay, and
indexes show  a similar pattern of declining       Zoido-Lobat6n (2000)
volatility as the policy environment improves,
suggesting the finding is robust.
One concern about the evidence presented        The independent effect of governance on
in figure 1.19 is that all three indexes correlate  volatility can be shown in a different way.
very highly with income.7 However, in regres-   Figure 1.20 repeats the analysis, but it corrects
sions that explain volatility with both income  for the possible influence of income by first re-
and governance as explanatory variables, the    gressing the governance index on income and
coefficient for governance tends to be more     then examining the relationship between the
significant than is the coefficient for income.  volatility of investment cycles and the residuals
25



G LO B AL EC ONO MI C PRO SP EC TS 2 0 0 3
from this regression (that is, in relation to that  -but capital flows tend to trigger
part of the governance indicator that is not    domestic cycles in middle-income
correlated with income). Investment volatil-    countries
ity continues to fall with a better investment  Both "push" and "pull" factors are respon-
climate.                                        sible for the procyclical nature of capital
inflows. In a downturn, demand for invest-
International net capital inflows are pro-      ment financing is reduced as firms postpone
cyclical, both in developing and high-          investment plans and reduce capital stocks
income countries-                               (the pull factor). At the same time, financial
Because investment cycles are more pro-         investors look for less risky or risk-free invest-
nounced than output cycles, investment as a     ments and show    little appetite to invest in
share of GDP tends to rise during a boom and    countries and sectors that suffer from declin-
to decline during a downturn. For countries at  ing growth and profit rates (the push factor).
all income levels, financing of the increase in  For developing countries, the push factors
the investment ratio during a boom    comes     have often been emphasized as a major chal-
from both domestic and foreign sources. Dur-    lenge. Sharp increases in external finance fre-
ing an upturn, domestic savings rates normally  quently preceded severe crises (such as in
build, while current accounts deteriorate       Mexico, East Asia, and Turkey), which were
(table 1.6). In other words, foreign investors  triggered by sudden reversals of these flows.
turn away during downturns, while domestic      The dynamics of net capital inflows and the
consumers reduce their savings and increase     changes of official reserves over the cycle do
their consumption as a share of income.         indeed indicate that the push factor is more
Clearly, procyclical capital flows do not pre-  important for middle-income countries, while
vent consumers from    absorbing shocks by      the pull factor dominates in high-income
smoothing consumption over time (box 1.4).      countries. Net foreign capital inflows actually
lead the domestic investmenit cycle in middle-
income countries, while they lag the cycle in
Table 1.6  Upturns can be financed abroad       high-income countries. For example, one-year-
and domestically                                lagged capital inflows are correlated with in-
Coreeiaulon of cyciicai                High |   vestment by 0.27 in middle-income countries,
mwstvmentcomponenis  Low.    Middlc   inconic.  compared with a correlation of only 0.08 for
with changes mn    income   income    OEC   ,   high-income   countries. In   middle-income
Current account                                 countries, one-year-lagged capital nflows are
(as % of GDP)      -021     -0 39    -0.43    ias strongly correlated with the domestic m-
Domestic savings                                a
(as % of GDP)       0.16     0 16      0.45    vestment cycle as they are with contempora-
neous capital inflows (table 1.7). In high-
Note The table shows unweighted averages of correlation
coefficients of variables in individual countries  income countries, a one-year lead in capital
Source World Bank                               inflows is, by contrast, as strongly correlated
Table 1.7 Capital inflows lead investment in middle-income countries: correlation between
investment ratios and (past or future) capital flows
2-year lag    I.ycar lag   No lag or lcad   1-year Iead   2-scar lead
Net capital inflows
Low income              -0.09         009            021             007          -001 
Middle income            0.13         0.27           035             0 10         -0 11
High income (OECD)      -0 20         0.08           0 32            0 35           0.25
Source World Bank staff estimates.
26



T HE IN T ER NAT IO NA L EC ONO MY AND PRO SP EC TS F OR DE V EL O PING CO UN TR I ES
lBox 1.4          Consumption in low- and middle-income
countries is smoothed over the business cycle
n all country groupings, the savings rate is posi-  Relative and increasing vulnerability of
ltively (or the consumption rate is negatively) corre-  low-, middle-, and high-income countries to
lated with GDP growth in the short run (see box    terms-of-trade shocks
figure). The correlation is relatively weak for
developing countries, but that weakness is mainly                   1973-80  1981-90  1991-2000
because GDP is an inaccurate measure of income in  Low income        2.1     2 2       2.8
the presence of terms-of-trade shocks. Real growth  MAddle income    1 9      1.8     2.2
of gross national income (GNI), which includes     High incom2e               1        15
(non-OECD)        2.8              I 
terms-of-trade gains and losses, is much more      High Income
volatile than growth of real GDP in developing     (OECD)            1.4     1.3       1 3
countries. In industrial, high-income countries, the
terms of trade have a much smaller effect (see box      deviation of GDP growth/
table). The correlation between the savings rate and  Source: World Bank data.
real growth of GNI is decisively more similar across
countries.
The strong evidence of consumption smoothing
in low- and middle-income countries is remarkable,  ing fluctuations in income are less favorable than in
because the conditions in such countries for absorb-  high-income countries. First, domestic credit markets
tend to function less smoothly, and access to interna-
tional capital markets is more difficult than in high-
income countries. Second, as far as fluctuations in
Consumption smoothing is strongest in        income are caused by terms-of-trade shocks, the
high-income countries                        smoothing of consumption over time is less attractive
(correlation of change in average propensily to consume  than in the case of volume shocks, which prevail in
Iand GDP growthi)
andIGDP growth)                              high-income countries. If, for example, import prices
Low    Middle  Hgh income High income  fall, the decline in price amounts to a rise in income,
income  mnoome  non-OECD  OECD
0 1       1       1 _          _  J        which could trigger an increase in the savings rate.
However, a temporary price fall implies a future
-o1                                          price rise, making future spending of current savings
-O 2                                         less attractive. In the case of a temporary rise in the
volume of income, it is more appealing to save now
-0 3                                         and spend later, when income is back at a lower
level,
| -0 4   T   _  _           _ v                   )Despite these impediments, developing countries'
-0 5                                         consumption is being smoothed over the business
| -0 6  Income growth    _   GP growth       cycle, providing relief for consumers and supporting
domestic financing of procyclical investment ratios.
Sourcc VWorld Bank staff eshiinares
Soturce World Bank staff.
with the domestic investment cycle as with    official reserves. Some of the foreign capital
contemporaneous capital inflows.              inflow that precedes a domestic investment
The picture of capital inflows being a push  boom in middle-income countries is temporar-
factor for middle-income countries is strength-  ily accumulated as foreign reserves, while cap-
ened when one considers the behavior of       ital outflows that precede a domestic bust are
27



G LO B AL EC ONO MI C PRO SP E CT S 2 0 0 3
temporarily absorbed by reductions in official  significant reduction of poverty. Thus, the mil-
foreign exchange reserves.                 lennium development goal of halving poverty
In low-income countries, capital inflows  by 2015 could be reached on a global level,
are also procyclical, but the correlation with  although growth will be insufficient to achieve
the domestic cycle is less significant. The  poverty targets in all regions. At the same
weaker correlation reflects the specific charac-  time, financial imbalances and volatility in
ter of these flows-official aid and FDI are less  international capital flows continue to jeopar-
cyclical-and also the dependence of those  dize uninterrupted growth. Vulnerable coun-
countries on commodity prices. Part of the fi-  tries will benefit from further debt reduction
nancing for investment booms comes not from  in their pursuit of sustained high growth.
foreign borrowing, but from increased export  The acceleration of growth in developing
revenues as a result of terms-of-trade gains.  countries is expected to coincide with in-
Similarly, investment busts are not necessarily  creases in investment ratios. Saving rates are
driven by a reversal of capital flows, but they  also expected to increase, driven particularly
can originate from terms-of-trade losses.  by a declining proportion of youths and by
On the basis of these relationships, it ap-  the need for adults to save for retirement. Op-
pears that cycles are still a prominent feature  posite movements are expected in industrial
of macroeconomic developments, which is    countries, wvhere aging is bound to reduce
even more important in developing countries  savings rates and where a sharp decline in
than in industrial countries, and cycles were  population growth will suppress investment
more pronounced during the 1990s than dur-  ratios.
ing earlier decades. Investment swings are   The long-term forecast suggests that, on bal-
financed both domestically and abroad, which  ance, net inward capital flows toward develop-
makes current account deficits and capital  ing countries could well decline, though gross
inflows strongly procyclical. A major differ-  flows will continue to play an important role in
ence between developing countries and high-  enhancing growth potential. These changes in
income countries is that middle-income     global savings and investment behavior raise
countries are more exposed to independent  questions about the critical role of financial
reversals in capital flows, while capital flows  integration and the need for improvements in
are more accommodating in high-income      international financial intermediation.
countries, and the cyclical dynamics in low-
income countries are to a significant extent  Long-run per capita growth is expected
influenced by terms-of-trade shocks and    to accelerate-
idiosyncratic disturbances.                Developing-country growth, on a per capita
basis, is projected to more than double during
the 10-year period from 2005 to 2015 when
Growth and poverty to 2015:                compared with the performance of the 1990s
coming changes in savings and              (table 1.8). This projection reflects substan-
investment patterns                        tially improved growth prospects for Europe
A  fter an impressive wave of market re-  and Central Asia-leaving behind sharp con-
Aforms and increased openness in develop-  tractions that characterized the transition to
ing countries during the 1990s-both of     market economies during the 1990s-and for
which prompted acceleration of technological  Sub-Saharan Africa. For Africa, the scenario is
progress and brought about a more stable   predicated as a continuation of broad trends
macroeconomic environment-long-term eco-  toward better governance and economic poli-
nomic growth prospects for developing coun-  cies, of progress toward resolving conflicts and
tries are relatively optimistic. If the projections  diversification away from agriculture, and of
come to pass, growth patterns could lead to a  export dependence on primary commodities.
28



TIf E  I N T E R N A T I O N A L  E C O N O M Y  A N D  P R O S P E C T S  F O R  D E V I 1 O P I N G  C O U N T R I E S
Table 1.8 Long-term prospects are projected to be stronger for most regions
(real CDI' per capita, annuitial average percentiage change)
Forecast sceniario
Medium terni        Long term
19805          19905            2001-05            2006-is
World total                             1.3            1.2              1.1                2.1
High-income countrics                   2.5            1.8              1.5                2.4
OECD                                  2.5            1.8               1.5                2.4
United States                       2.2             2.2              1.6                2.4
Japan                                3.5            1.2              0.4                2.0
European Uttion                     2.1             1.7              1.9                2.4
Non-OECI) counlitries                 3.3            3.6               1.7                3.3
Developing countries                    0.7            1.6              2.4                3.5
East Asia anid the Ilacific           5.6            6.4              5.1                 5.4
Europe anid Cet,tral Asiai            0.7           -1.9              3.3                 3.4
lIatini America anld the Caribbeani  -0.9             1.6              0.3                2.6
,M0iddle East anid North Africa      -0.6            1.0               1.3                1.3
South Asia                            3.4            3.3               3.5                4.0
Sub-Saharan Africa                   -1.2           -0.4               0.8                1.6
Note: Aggregationis are inovinig averages, reweighted annuially after calcUlations of growth in constant prices.
Sotirce: World Bank.
At the same time, a lack of humani capital,          development goal of halving extreme poverty
poor infrastructure, and the AIDS epidemic           by 2015 from the 1990 poverty level should be
remaini pressing problems.                           achieved on a global level, though with wide
Per capita   growth   in Latin   America   is     regional disparities. The revised poverty projec-
expected to accelerate by I percentage point         tions indicate a poverty rate of some 13.3 per-
under this scenario, but as a result of slowing      cent in 2015 compared with 29.6 percent in
population growth, the acceleration of real         :1990. The actual number of poor would
GDP growth is small. Latin American coun-            decline to around 809 million from   1.3 billion
tries are expected    to  have benefited   from      in 1990 and 1.1 billion in 1999. Asia should
reform  efforts durinig recent years and from        readily achieve the target, but the MENA
sustained   improvemiielits in  macroeconomic        and SSA regions will make little progress in
stability. The East Asia and Pacific Region          imiproving poverty incidence (table 1.9).
should witness a declininig per capita growth           Though the central message remains the
rate from 6.4 percent in the 1990s to 5.4 per-       same, the long-term    outlook   reflects rather
cent in the longer term, as economies mature         significant changes from   last year's forecasts.
and as options for rapid catchinig up become         These changes are a combination        of three
less abundant. Per capita growth in the rest of      factors:
the world, including South Asia, the Middle
East and North Africa, and highi-income couni-          * The economic projections reflect recent
trics, is projected to accelerate moderately.             trends and a downgrading of the medium-
term  forecast, as detailed earlier in the
-leading toward significant poverty                       chapter, page 5. The long-term     forecast
reduction                                                 has remainied relatively unchanged, but
As projected   in previous Global Econotnic               lower growth-actual and         forecast-
Prospects (GEP), achieving the millennilum                between   2000   and   2005   has slightly
29



G LO B AL    EC ONO MI C       PRO SP E CT S       2 0 0 3
worsened the poverty forecast, all else                (that is, for the same growth rate, the
being equal.                                          rate of poverty reduction has declined).
* New surveys and methodology have sig-
nificantly altered the 1999 estimate of             The reader should bear in mind that these
poverty incidenice. For developing coun-         numbers are sensitive to the poverty line cho-
tries, this change has led to a 1.6 percent      sen and underlying assumptions and data (see
rise in the estimate of the number of            box 1.5).
poor living on    less than   $1  per day.          The relation between growth and poverty
However, the revisions are not uniform           may not have changed in a fundamental way,
across regions. There is a significant rise      but the change may be a consequence of past
in East Asia and in Europe and Central           trends at a more disaggregated level. Recent
Asia, while the     estimated   number of        studies9 of poverty trends in India indicate
poor has dropped in Latin America.8             that poverty has been successfully reduced in
* The thlird   factor is the change in the         a number of states in which growth rates are
relation between economic growtlh and            high  and   in  which   the  responsiveness    of
poverty reduction. This rclation has been        povcrty reduction to growth is likcwise highcr.
re-estimated using the new survey data.          Moreover, the evidence suggests that these re-
Overall, the relationship has weakened          gions had significantly better initial conditions
Table 1.9 Large poverty reductions in EAP and SAR partially offset by poverty
increases in SSA
Number of people livng on less than 51 per day (millions)
CEI' 2002                            GEP 2003
Region                              1990         1999       2015         1990         1999       2015
East Asia and Pacific               452          260          59         486          279          80   |
Excludinig Chinia                   92          46           6          110          57           7
Europe and Central Asia               7           17           4           6           24           7
Latin America anld the Caribbean     74           77          60          48           57          47
Middle East and North Africa          6            7           6           5            6           S
South Asia                          495          490         279         506          488         264
Stib-Saharan Africa                 242          300         345         241          315         404
Total                              1,276        1,151        753        1,292        1,169        809
Excluding China                   916          937         700         917          945         735
SI pcr dav hcadcount indcx (percent)
GEP 2002                             GEP 2003
Region                              1990         1999       2015         1990         19')9      2015
East Asia and Pacific               27.6         14 2        2 8         30.5         15 6        3 9
E\cludingChina                    185           79         09          242          106          1 1
Europe and Central Asia              1 6          3 6        0 8          1 4          51          1 4
Latin America atid the Caribbcats   168          151         97          11.0         I I1        75
Middle East and North AfriLa         2 4          2 3         I S         2 1          2 2        2 1
South Asia                          44 0         36 9       16 7         45 0         36 6        15 7
Sub-Saharan Africa                  47 7         467        39 3         47.4         49 0       46 0
Total                               29 0         22 7        12 3        29 6         23 2        13 3
E\cluding Chinia                  28 1         24 5       14 8         28.5         25 0        15 7
(continued oni page 31)
30



I 11 L  I N I E R N A T I O N A l  E C O N OM Y  A N D  P R OS P 1: C I S  F O R  D E V E L OP I N G   C O U N T R I E S
Table 1.9 (continued)
Numbe,r Of p-ople Iving oi, le,, th-n 52 pCr dai% (milli-ns
C'kl 2002             r               GEP 2003
Region                               1990        1999        lots         1990         1999        2015
Last Asia anid lacific             1,084          849         284    |   '1,114         897        339
ExclUding Chinsa                   285          236          93          295          269         120
hirope anid Ccntral Asia              44           91          42           31           97         45
Latin Amiicma and the Caribbean      167          168         146          121          132         117
,Middle East anid North Africa        59           87          65           50           68          62
South Asia                           976        1,098       1,098        1,110        1,128       1.139
Sub-Saharan Aftica                   388          484         597    I     386          480        618
btoal                               2,718        2,777      2,232        2.712        2,802       2,320
lxcltiding Chila                  1,919       2,164       2,041        1,892        2,173       2,101
S2 per ciay head-uni midex (per,enit
GCLV 2002                             CEP 2003
Region                              1990(         1999       201   i      199U         1999        21015
Fasr Asia anid P'acific             66 1         46 2        11 5         697          50 i        166
Excltiduig China                   57 3         40 4        13 3        61.9         50 2        18 4
Eulropc and Central Asia              9 6         19 3        8 7          6 8         20 3         9 3
Latin America and the Caribbean      38 1         33 1       23 4         27 6         26 0        18 9
Minddle East and North Africa        24 8         29 9        167         210          23 3        16 0
South Asia                           86 8         82 6       655          89 8         84 S        6S.0
SLub-SahAran Africa                  76 4         75 3       68 0         76 0         74.7        70 4
Total                                61 7         54 7       363          62 1         55 6        38 1
Excluding Chinia                   58 8         56 5        41 0        58 7         57 .5       44 7
Noie The GEP 2002 figures IClUde the RepuiblI of Kiirea, which has been relaLssified ieito the high-incoiine gtruLp
Soiti cc World Baiik stati estirrlit
Box 1.5 Is the World Bank overestimating
global poverty?
A    recent study (Bhalla 2002) conclLIdCs that the       line, his use of seconidary data sources rather than
A    World Bank has overestimatecl the number of            primary surveys, and consumption adjtistnlents.
poor in developing countries, and that the miLlen-         These differences highlight the complexity in count-
nium devctopment goal ot halving extrcme poverty           ing the number of poor and are described here. A
by 2015 (from its 1990 level) was already achieved         more cotnplete critique of the Bhalla study can be
In 2000. The study estimates that the percentage           found in a separate paper (Ravallion 2002).
of poor in developmig countries in 2000 was only                The r World Bank has chosen to use $1 per day
13.1 percent and that the World Bank's estimate is         and $2 per day poverty lines for global estinations,
10 percentage points higher (see table 1.9). Three         roughly spanning the ranige of national poverty lines
differences between the Bhalla estimates and the            in developinig countries. Bhalla uses $1.50 per day.
World Bank's explain Bhalla's different conclusion.         Because the cost of purchasing 2,200 calories differs
Thesc differences Hiclude the choice of the poverty         from country to country, each country estimates its
31



G L O B AL EC ONO MI C PROS PE CT S      2 0 0 3
Box     IL o5    (continuted)
own national poverty line. This approach also re-  countries-and the calculations are properly
flects the fact that the nature of poverty varies signif-  weighted to reflect survey design and differences in
icantly across and within countries. Moreover,  household size. Consumption is deemed to be the
poverty has many dimensions: inadequate consump-  preferred measure to income, but income is used
tion of essential commodities. as well as low life ex-  when consumption is not available.
pectancy, high child mortality, and low school enroll-  Finally, Bhalla's study makes consumption
ment rates, among other attributes related to the  adjustments, which may not be warranted and could
quality of life. Poverty is also a r elative and sublec-  lead to a biased poverty estimiate. The headcount
tive concept. What is deemed a necessity in some  index (that is, the percentage of the population living
countries (for example, indoor plumbing in rich  at or below the poverty line) is calculated using an
countries) may be a luxury in others. In Latin Amer-  estimate of the per capita consumption level relative
ica, the regional estimate for the percentagc of peo-  to the poverty line. There can be significant discrep-
ple living in extreme poverty is 17.8 percent in 1998  ancies between the survey-based mean consumption
(VWodon and others 20021, compared with 11 I per-  and the mean consumption as measured by the na-
cent in 1999 usmig the Sl per day poverty line. The  tional accounts. Part of this discrepancy is explained
foIrImner is based oni national poverty lines and is from  by the way consumption is estimated in the national
a regional perspective. This higher number-ar-  accounts (which typically includes consumption of
guably a more accurate reftection of the incidence of  non-household private agents such as nonprofit
poverty from a social point of view-is more rele-  organizations). Other discrepancies can arise from
vant in determining policies for reducing poverty.  measurement error (for example, misreporting of  i
However, for purposes of global comparisons, the  consumption in surveys). Adjustments for these dis-
World Bank has tried to select a level of real con-  crepancies can lead to different estimates of poverty.
sumption that best measures the same level of con-  If, for example, the misreporting is biased toward
sumption across countries so it can make aggregate  high incomes-that is, if only rich households under-
judgments independent of where the poor live.    report consumption in the surveys-and if an tup-
Unlike the Bhalla study-which relics on aggre-  ward adjustment is applied to all households (includ-
gate secondary data sources-the World Bank's    ing the poorest), then the number of poor will clearly  |
poverty estimates rely exclusively on primary data  be underestimated (see Ravallion 2002). The World
from comprehensive household surveys. Since the  Bank's poverty estimates rely on the survey-based
1980s, the World Bank and developing-country gov-  consumption levels.                            i
ernments have been actively involved in undertaking  Transparency in the methodology and data used
national household surveys to get an accurate picture  to assess the level of poverty is critical in this debate
of the distribution of consumption across individo-  regarding how to count the number of poor The
als. To date, more than 300 comprehensive surveys  ability of different researchers to easily assess and
have been collated and used to estimate the number  compare results will lead to improved estimates and
of poor. Currently, the surveys cover more than  to a better understanding ot the nature of poverty
90 countries, with surveys available for various years  and policies that accelerate poverty reduction
for most counties. The surveys used all have national
coverage They include consumption from own-
production-a key feattire in many developing     Source: World Bank staff.
than did states with lower growth rates. This  initial conditions in the laggard states and
evidence implies that progress in the future  should not rely exclusively on raising the na-
will be harder to achieve with the same na-  tional growth rate.
tional growth rate. Poverty-reducing policies   The number of poor in last year's report
will, therefore, have to focus on raising the  was projected to be 753 million in 2015,
32



T H E  IN T E RN Al 1IO NA L EC oNO MY A N D P R O S P E C r S F O R D E V 1: L 0 P I N G C O U N F R I E S
or a headcount rate of 12.3 percent. The cur-  Population growth will ease
rent forecast shows the number of poor will   in all regions-
decline to only 809 million by 2015, or a     In virtually all countries, growtth of the
headcount rate of 13.3 percent This change    working-age population is slated to decline
represents a 7.4 percent increase when com-   over the next 15-year period, thereby affect-
pared with last year's figure. The percentagc  ing labor supply and thus contributing Icss to
increase can be decomposed into two factors.  GDP growth in the long run (figure 1.21).
The higher initial level of the number of poor  High-income countries and those in ELirope
in 1999 would lcad to an increase of 2.3 per-  and Central Asia are likely to see an absolute
cent in the 2015 forecast, all else being equal.  drop in the working-age population by 2015.
The remaining 5.1 percent of the increase In   Developing regions will see a slower pace of
the forecast is attributed to a weakening of the  decline, although East Asia is expected to see
relation between growth and poverty.          its growth rate halved to 0.7 perccnt per year
by the end of the period.10 A slower growth
Population, savings, and investment            rate in the labor force means that achieving
are factors underlying the long-term          the same ratc of per capita growth will requirc
forecast                                      an acceleration of investniciit, a highcr level of
Some developing regions will need to see a    productivity, or a combination of both.
reversal in performance of underlying growth
factors-particularly in productivity and in   -but technological progress should
savings and investment-from the last decade.  accelerate
That reversal should show either changes to   The role of total factor productivity (TFP)
policies or persistence with ongoing reforms.  growth in determininiig growth ratcs has bcen
Assessing the factors underlying the long-    the subject of significant research over the past
term forecast-population and labor supply,    decade (see box 1 6), but there is a growing
technological progress, and   savings and     consensus that technological advances and
investment-will elucidate some of the under-  efficiency improvements are pivotal determi-
lying dynamics in the long-term growth fore-   nants of growth patterns. If one looks for-
cast and its policy implications.              ward, many countries are expected to reap the
Figure 1.21 Growth of working-age population decelerates
(annual growth of population for ages 15 to 65)
40
35               0              XMNA
30    ____________SSA__SAS_
25                                     LAC
20
1 0             \ ~~~~EAP=
10
05                              ', EGA
0 0 1  I      I     I    I   I    I   I   I    1                 1 I
-05                                                                  HIY
-1 0
1998     2000    2002     2004     2006     2008    2010     2012     2014
Note HIY refers to high-income counitries, SSA refers to Sub-Saharan Africa, EAP refers to East Asia and lPacific, SAS refers to
South Asia, ECA refers to Eastern Furope and Central Asia, MNA refers to Middle East and North Africa, LAC refers to Latin
Amiierica and the Caribbean
Source World Baitk staff demographic projectiois
33



G LO B AL EC ONO MI C PROS PE CT S       2 0 0 3
Box 1.6          Technological progress is an important
determinant of growth
A growing conscnsus in the economic literature is  sign for developing countries, because constraints on
L   that TFP accounts for the bulk of cross-country  domestic resources and access to external financing
differences in the level of income and the rate of  severely limit a country's ability to raise growth rates
GD1' growth (Easterly and Levine 2001).11 Whether  by increasing the volume or imnproving the quality
TFP or capital and labor accounted for the bulk of  of physical and human capital. To the extent that
income differences among countries has been an   differences In TFP growth will reflect differences in
issue of dispute since seminal articles by Denison  technology, then developing countries (whicih are
(1972) and Jorgenson and Griliches (1967, 1972).  well below the technological frontier) can potentially
The debate sharpened in the 1990s when a number  achieve high "catch-up" rates of growth hy import-
of comparative growth studies found that the success  ing technology. Parente and Prescott (1994) see the
of the East Asian Tigers was driven mostly by in-  main source of cross-country productivity differences
creases In capital and labor rather than by increases  as stemminlg from policy-induced barriers to adopt-
in TFP (Young 1992, 1995; Collins and Bosworth  ing advanced technology. TFP growth also reflects
1996). Because capital is subject to diminishing re-  other aspects of economic efficiency that are
turnis, such studies implied that the highi rates of  amenable to change through improving policies. For
growth achieved in East Asia were not sustainable  example, policies that increase competition may raise
(Krugman 1994).                                  TFP by improving the allocation of labor and capital
More recently, the weight of evidence appears  and by increasing the ability of the economy to re-
to be moving toward the conclusion that TFP is the  spond to changes In the economic environment
main driver of growth. The East Asian studies have  (Easterly and Levine 2001, Solow 2001; Hulten
been criticized for not accounting for the role that  2000).
technological progress plays in encouraging greater  Critiques of using this accounting approach
capital accunmulation (Hulten 2000; Barro and Sala-i-  inclide the fact that it typically relies on various
Martin 1992).12 Nelson and Pack (1999) emplhasize  restrictive assumptions (for example, constanit
that learning, technology absorption, and forceful en-  returns to scale and competitive markets) that may
trepreneurship were critical to the success of large in-  not hold in reality (although models that incorpor-ate
vestments in physical and human capital. Klenow and
Rodriquez-Clare (1997) estimate that TFP made an
important contribution to growth in all of the East   TFP is a major contributor to growth
Asian "mniacle" economies, excepr Singapore. East-    (percent)
erly and Levine (2001) summarize studies that show     60
TFP accounts for more than 40 percent of output        50
growth in most of the industrial countries, 30 percent  40
or more in most Latin American countries, and a
wide range (from -5 percent to 30 percent) in East     30
20--
Asia (box figure). Some writers attempt to refine esti-
mates of the contribution of TFP to growth by imcor-   10
porating some measuremenit of the quality of in-        0 __
puts-for example, adjusting data on labor input for   -10
the degree of education or training (see Easterly and            rj          q, CO    o Z
Levine 2001 and Parente and Prescott 2000 for recent                  F     V    4 b.
contributions). In general, efforts have not been suc-           b
cessful In greatly reducing the amount of growth or                        ,°
income differences that are accounited for by TFP.
The evidence that growth in TFP is the maini      Source Easterly anid Levinie (2001)
driver of economic growth is essentially an optimistic
34



T ` E I N I E R N A T I O N A L  E C O N O M1 Y  A N D  P R O S P E C T S  F O R D E V E L OP I N G  C O U N T R I E S
Box      1.6       'continued)
market Imperfections also confront clifficult econo-  covered. Finally, growtlh accounting generally does
metric prolblems; see Brock and Durlauf 2001).     not reflect eithel improvements in the quality of
Often, key parameters, such as the share of capital in  goods or the introductioni of new products, which
output, are assumed to be based on linited empirical  are also Important for welfare (Hulten 2000). Those
work (Senhadii 2000). Solow (2001) notes that the  criticisms underscore the substantial methodological
growtlh-accounting framework assumes that the      and measurement difficulties involved in quantify-
economy is moving along the potential output fron-  ing the contiibution of inpuits and productivity to
tier. In developing countries, the volatility of the  growth rates. Nevertheless, as Hulten (2000) stresses,
business cycle means that actual output may be con-  this approach has provided a simple and internally
siderably different from potential at any point in  consistenit intelleCtual framework that has been used
time. Thus, growth-accounting estimates using time  to gain vital insights iito the process of economic
series data may be biased by differences between   growth.
potential and actual output levels at the beginning
and end points, even if a considerable time period is  Source: Woild Bank ntr.ff
changes in its servicc sectors, which will have
Figure 1.22 Productivity has not been        long-term  payoffs. Both Europe and Japan,
the dominant source of growth in             although lagging somewhat behind the United
regions                                      States, have the opportumity to reap gains from
(average percent per annum 1990-2000)        improved use of information technologies.
8                                              The East Asia region has been the leader
Productivity       among developilg regions in terms of acceler-
6        _                  \               ating productivity over the past two decades
4                            ,              It has built on compositional shifts (from
Capi al                        agriculture to mallufacturing and services),
o ra_ r _ iiii-r
Figure 1.23 Productivity is expected to
-2                                              be more significant in the longer term
(average percent per annum 2005-15)
-4
HIY  EAP ECA   LAC MNA SAS    SSA            8  Productivily
HIY  = high-inconie coitintriesi6
Souerce Wo Id taik staff estimates                                   Capital
4
benefit of reforms undertaken during the past       2
10 years. These benefits will likely show up as 
an acceleration of technological progress (fig-                                   Labor
ures 1.22 and 1.23). Among industrial coun-       -2
tries, those countries in Europe will see accel-  -4
erating benefits from the single currency plus         HIY  EAP ECA   LAC MNA SAS SSA
greater capital and labor mobility. Japan,        i-lY = high-iicome cotiiitries
though still burdened with significant prob-      Soure VWorld Bank staff stimates
lems in its financial sector, is witnessilg
35



G LO BA L E CO NO MI C PRO SP EC TS 2 0 0 3
educational improvements, and productivity-    Figures 1.22 and 1.23 summarize the de-
enhancing policies (for example, increasing  composition of the various sources of GDP
openness). The region will continue to benefit  growth into three broad components: the
from good policies and compositional shifts.  labor supply, capital accumulation and pro-
After all, the largest country in the region,  ductivity for the historical period 1990-2000,
China, still has more than 60 percent of its  and the long-term  projection for 2005-15.
work force in agriculture. However, as the gap  The decomposition profiles projected for the
with technologically more advanced countries  period 2005-15 are rather similar to the
closes, the opportunities for extreme advances  1990-2000 period, with the significant excep-
in productivity diminish.                   tion of the contribution from technological
Europe and Central Asia is benefiting from  progress. As argued on page 38, we have pro-
the substanitial reforms of the 1990s, accom-  jected that most regions will see an accelera-
panied by large FDI flows. Moreover, many of  tion of technological progress, which will
the accession countries will accelerate the re-  drive the improvements in GDP growth. That
form process in preparation for joining the EU  progress will trigger further capital accumula-
early in the forecast period.               tion to accommodate and to further enhance
In Latin America, progress has been visible  growth prospects-hence the need for pro-
regionwide. As has been positively demon-   viding an enabling investment environment.
strated by Chile and Mexico, openness and   Capital accumulation implies an investment
stability are key elements in providing sustaiin-  profile-discussed in more detail below-
able growth. A recent study of Latin American  linked, of course, to behavioral assumptions
growth over the past three decades concludes  regarding savings during the coming decade.
that structural reforms and stabilization poli-
cies accounted for a large contributioni to the  Convergence of investment ratios is likely
overall acceleration in the rate of growth in  to continue
the 1990s compared with the "lost decade" of  After a distinct divergence of investment ratios
the 1980s.13 These trends are expected to con-  across regions during the past decades, some
tinue in the next decade, with increasing in-  convergence is expected  during  coming
vestments in infrastructure and education and  decades, though disparities will remain signifi-
with greater openness to trade underpinning  cant. Investment rates in the developing East
solid productivity growth. The key downside  Asia and the Pacific region gradually increased
risk includes the vulnerability of the region to  from 15 percent of GDP during the 1960s to
external shocks, particularly given its sizable  almost 30 percent during the 1980s (fig-
external debt burden, and the effects that this  ure l.24a). This increase coincided with rapid
debt could have on the stability of the domes-  growth of the economies in the region, major
tic financial sector.                       sectoral shifts as a result of diversification,
Sub-Saharan Africa has benefited from sim-  and regional and global integration, which all
ilar policy reforms and stabilization. FDI has  required expansion or replacement of capital
been increasing, and the resolution of some  stocks. During the first half of the 1990s, the
long-term  civil conflicts should provide a  average investment rate jumped further to
more enabling environment for sustained     35 percent, of which a substantial part was
growth. South Asia and the Middle East and  used for real estate development when foreign
North Africa regions have maintained some of  capital streamed in. In 1997, investors realized
the highest trade barriers in the world. These  that their collective behavior was based on
barriers will slowly be removed under the im-  overly optimistic expectations. The resulting
petus of multilateral and regional trade agree-  financial crisis sharply lowered investment
ments, with ensuing efficiency gains.       rates back to levels near 25 percent.
36



T HE I NT E RN AT IO NA L EC ONO MY A ND         PRO SP EC TS F OR DE V EL O PING C O U N T R I E S
The investment rate in South Asia was ex-    changes and privatization programs. One rea-
tremely low until the mid-1970s, but then it   soIn for the lack of a strong rebound in invest-
started to rise. To a large extent, that rise is a  meint was the continued low domestic savings
reflection of the green revolution and industri-  rate In both regions (figure 1 24b). Even with
alization programs in India. Although not as    low  investment rates, the current account
strong by far as in East Asia, the average rate  showed large deficits during the 1 990s.
continued to increase during the 1980s and        Future investmiient trends will be influenced
1990s, and the current level is now close to    by expected GDP growth, by real domestic
OECD levels of around 22-23 percent. That       interest rates, and by expected domestic rates
level is well above rates in Latin America and  of return to capital compared with the average
Sub-Saharan Africa.                             global rate of return. If one assumes a stable
Structural developments in the latter two    risk environment, the last effect should tend
regions (LAC and SSA) were quite different      to benefit developing countries, where rates
from the Asian experience. A gradual rise in    of return are higher than In rich counties.
investment rates during the 1970s, whiclh was  Table 1.10 summarizes the changes in invest-
partly financed abroad, suddenly turned into    mcnt behavior betwccn the averagc of the
a sharp decline when the debt crisis hit In the  1997-2001 period and the final year (2015)
early I 980s. Investment rates dropped around   of the basehine scenario. On average, the high-
5 percentage points. Macroeconomic imbal-      income countries will see a drop in the invest-
ances and hyperinflation In Latin America, as   ment rate of about 2.9 percentage points (rel-
well as sharp terms-of-trade losses in Africa,  ative to GDP). This figure is derived largely
led to extremely low growth during the 1980s.   from lower projected GDP growvth rates and
In this environment, investment rates contin-   thus changes in the optimal capital to output
tied to fluctuate around historical lows. Dur-  ratio. The average investment rate in develop-
ing the 1990s, investment rates showed only    ing countries increases sligltly by 0.2 percent-
a slight recovcry, although the composition     age points, with higher investment rates in
shifted significantly away from public to pri-  many regions offset by lower investment in
vate investmenit, following major structural    East Asia. The latter region is still suffering
Figure 1.24 Major structural shifts in investment and savings behavior have occurred
a Investment-to-GDP ratio                    b. Savings-to-GDP ratio
(current dollar-weighted average)            (current dollar-weighted average)
35                                          35
1980s
30                 1                   -    30
197\SO  /1990s                              1970s 1980s
25                                          25
1960s                                        1960s190
20                                     _     20
15                                          15
510                                         100                            II
South     Labn   Sub-Saharan East Asia       South     Latin  Sub-Saharan East Asia
Asia   Amenca and  Atrica  and Pacitic       Asia   America and  Africa  and Pacitic
the Caribbean                                the Caribbean
Souirce World Batik staff estrmites
37



G LO B AL    EC ONO MI C      PROS PE CT S       2 0 0 3
Table 1.10 Savings fall in high-income countries, but increase in most other regions
(as a percentage of CDP)
1997-2001                               2015
Capital                               Capital
Savings     Investment     nflows     Savngs      Investment    inflowsv
(S)          (I)          (KA)         (S)         (I)          (KA)
Total                           22.4         22.5          0 1        20.5         20.5           0.0
High income                     22 0         21 9         -0.1        18.6         19.0           0.4
Low and middle income           24 2         24 7          0 5        26 0         24.9         -1.1
European Union                  208          204          -04         17.5         174          -0.1
Japan                           29 8         27 5         -2 2        26 0          25.4        -0 6
United States                   17 5         19 5          2.0        14.5         16.3           1 9
Rest of high income             30 8         26 1         -4.7        24 9         23 2         -1.7
East Asia and Pacific           36.9         33 9         -3.0        35 0         29.3         -5.8
South Asia                      21.3         22.2          0 9        23 7         23.5         -0.2
Middle East and North Africa    26.2         21 4         -4 8        23.1         22.2         -0.9
Sub-Saharan Africa              13.9         17 6          3.7        19 0         20 1           1 1
Europe and Central Asia         21 9         22 6          0 7        22 0         271            51.
Latin America and the Caribbean  17.9        21.7          3.7        20 5         20.6           0.0
Note The columils (S), (I), and (KA) represent, respectively, the nationial saving rate, the national investmenit rate, and the capital
account, all as a share of GDP The values for 2015 are simulated values from the global general equilibrium model (maintained
by the Development Economics Prospects Group) The values for the 1997-2001 period represent the average observed values
from the World Bank's statistical databases For thie high-income countries, these values are the 1997-99 or 1997-2000 averages,
depending on data availability For the totals, the averages cover only the years 1997-99
Source. World Bank model simulations
from   past overinvestment, particularly        in    lennium. Rich countries are seeing an exten-
some sectors, and will adjust its investment          sion   of life spans and     a rapid   decline  in
needs to a slight deceleration in growth. Most        birthrates, leading to a sharply aging popula-
of the other developing regions will see accel-       tion. Developing countries are witnessing a
eration in investment in anticipation of higher       relatively  sharp  drop   in the percentage of
growth. Investment in those regions will also         youths as well as modest increases in the
provide relatively higher returns than in the         number of elderly. Recent economic evidence
more mature economies.                                suggests that these trends could have sigmfi-
Future changes in investment rates do not          cant implications on national saving rates-
necessarily lead to corresponding changes in          lowering them     in rich countries because of
current accounts. Despite      capital mobility,      aging, but raising them     in developing coun-
investment rates tend not to correlate with           tries as workers save for future retirement.
current account deficits in the long run, the         Lower birthrates also lead to a reduction in
so-called  Feldstein-Horioka     puzzle   (1980).     resource demand for the young.
Also in the coming 15-year period, the savings            The demographic transition with respect to
rate is expected to increase in those countries       the proportion of youths is largely over in the
that are anticipated to enjoy an acceleration of      high-income countries (figure 1.25). The over-
growth, partly because of rapidly changing            all average was 27.4 youths per 100 workers14
demographics.                                          in 2000 and is expected to drop to 24.4 by
2015. Developing regions are likely to witness
Changing demographics will raise                      much greater changes. First of all, the propor-
savings in developing countries                       tion of youths is starting from    a significantly
and will lower savings in industrial                  higher base, wsth the dependency ratio aver-
countries                                             aging 51.0, which is nearly double the high-
Global population      dynamics    are  evolving      income average. The ratio is expected to drop
fairly rapidly at the beginning of the new mil-       to 40.7 by 2015, twice the percentage of the
38



T HE IN I ERN ATI ON AL EC ONO MY AN D            P R O S P E C I S F O R D E V E L O P I N G  C O U N T R I ES 
Figure 1.25 Youth dependency ratio will         Figure 1.26 Elderly dependency ratios
fall everywhere except Japan                    will rise in some regions
(number of youths, age 15 and under, per 100    (number of elderly age over 60, per 100 working-age
worklng-age population)                         population)
90             2000 -T             2000         50
80
2                      ~~~~~~~~~40                         2015
70                                 2015
60
50                                             230                                 200
40                          T    ------215 
20            _20
30                                                              2000 
20                                              1
10  ~ ~ ~     ~     ~     ~     ~     ~    ~     ~    ~    1*~~~~1
0 F77T                                                  1  t    I 0  I I 
RHY = rest of high inLome, TOT = global average,  RHY = rest of high riicoime, TOT = global aserage,
HIY = high-linctine average, LMY = low- and miiddle-  I-IIY high-iiicomie average, L.\Y = lows- aiid miiddlc-
inconile dvrrage                                iiicoine average
Note Working-age popllatiotn is defined ,is those people  Note \Working-age population is defined as those people
betwseen the ages of 15 to 65                   hetxveen the ages of 15 to 65
Soutrce World Banik staff                       Souerce World Batik staff
fall expected iD industrial countries. This anti-  alone could have major macroecononmic impli-
cipated decline is the consequence of two fac-  cations for Japan and have conscquences for
tors: a boom   in births over the past two      the rest of the world, because Japan has long
decades, leading to a rapid rise in the working-  had largc excess savings that have been recy-
age population, and, more recently, declining   cled abroad. Outside the industrial countries,
birthrates, which are caused by a combination   the average elderly dependency ratio is not
of economic growth     and   family-planning    expected to change significantly. The average
programs.                                       was 8.9 in 2000, increasinlg to only 9.6 in
The elderly dependency ratio, defined as      2015.
number of members of the population who            Several recent studies of private savings
are older than 65 per 100 workers, is rising in  behavior have linked the private savings rate
almost every region (figure 1.26). The transi-  to a number of different factors: demograph-
tion is occurring rapidly in industrial countries  ics, income levels and growtlh, interest and in-
as the Baby Boom    generation ages and as      flation rates, and degree of finaicial mterme-
life expectancy improves, but there will only   diation.16 The results discussed next focus on
be modest changes in developing countries       only three channels affecting savings: the rate
through 2015 because relatively more recent     of per capita GDP growth, the youth depen-
improvements in health and life expectancy      dency ratio, and the elderly dependency ratio.
will affect elderly demographics only further   Other chaninlels may prove to be equally im-
in the future.15 The elderly dependency ratio   portant, however, such as improved financial
average in 2000 for industrial countries was    intermediation and a stable macroeconomic
21.2 per 100 workers. It is expected to rise to  environment. Combining these three effects
26.9 by 2015, surpassing the youth depen-       suggests that global savings may decline by
dency ratio. Japan will witness the most dra-   around 1 percentage point over the longer
matic iiicrease, from 25 1 to 39.8. This change  term, a figure that takes into accounlt a
39



G LO BA L ECO NO MI C PRO SP E CT S 2 0 0 3
2.7 percentage point decline in high-income  investment ratio. Those figures would reflect
countries and a 3.7 percentage point rise in  no change in overall growth compared with
developing countries.                      the 1990s because declining population
With relatively large swings in domestic  growth will counteract the acceleration in per
savings rates, the scenario suggests a rather  capita growth.19 These trends in savings and
significant change in net capital flows. In  investment rates will require ad]ustments to
1997, the base year of model simulations,  relieve the strains on budget and other vari-
industrial countries were exporting around  ables. Open capital and goods markets would
$67 billion in net investment to low- and  facilitate the potential strains from  demo-
middle-income countries, some 1.1 percent of  graphic transition.
low- and middle-income countries' GDP. East
Asia was the only developing region that was  Major policy challenges are likely
a net exporter of capital in the base year.17 By  to emerge
2015, under these savings and investment as-  While there is evidence in support of the eco-
sumptions, low- and middle-income countries  nomic growth projection in this long-run sce-
would be significant net exporters (some   nario, the degree of uncertainty regarding the
1.1 percent of their GDP), while industrial  projections is high. There is recognition that
countries would be net importers. Among in-  these trends will require good policies, be-
dustrial countries, Europe would be in ap-  cause many developing countries are still sub-
proximate investment-to-savings balance, but  ject to major shocks.
the United States could continue to be a sig-  These scenarios raise important long-term
nificant net importer of capital. The United  policy issues. First, the changing savings and
States would actually see little change in its  investment patterns will have consequences
capital account surplus from the base year (as  for net capital flows. Several developing coun-
a share of GDP), though it would experience  tries that in past decades have relied on net
a reversal from its present high level. Japan's  capital inflows will find it harder to do so in
capital account deficit would diminish sharply  the coming 15 years. Indeed, the recent shift
with a rapidly increasing elderly population  from debt accumulation to debt reduction in
leading to a decrease in savings. 8        many of these countries is likely to herald
The actual magnitude of these changes   a new long-term trend of further declines
in savings and investment rates remains   in debt. If the expected increase in private
speculative-even though they are grounded in  domestic savings is accompanied by further
empirically validated economic theory within  maturing of domestic credit markets and is
a consistent accounting framework-because  not thwarted by deterioration of public sav-
some of the underlying behavioral relations  ings, then debt reduction will not conflict with
could change the savings rates, the investment  the investment patterns needed to underpin
rates, or both to some extent and could have  growth.
a noticeable effect on the balance (that is, the  Underlying large swings in net capital flows
capital account). Nevertheless, there is little  are even larger movements of gross capital
doubt regarding the broad changes in the  flows, because FDI expands into growing
pattern in these projections. A clear trend  markets in developing countries and because
emerges of further reduction in developing  financial agents in developing countries seek
countries' foreign debt. East Asia could con-  to diversify their portfolios in rich countries.
tinue to show significant surpluses on the cur-  These capital movements will require further
rent account, while Latin America could con-  international financial integration. Because
verge toward a balanced current account. The  the history of financial integration has not
latter would be the result of, and represent a  been entirely felicitous-and at times has been
rise in, the savings rate and a relatively stable  damaging to growth and poverty reduction-
40



T HE   IN rE RN A     1IONA L   EC ONO MY      A ND   PRO SP EC TS      F OR  DE V EL O PING      C OU N T R I E S
the international community and developing          have a lower magnitude because average per capita in-
countries have had to search for mechanisms         come in Argentina ($12,100 in purchasing power par-
to provide for greater stability in integration    ity terms in 2001, and $7,750 in 1995 terms) is much
Developing countries can further facilitate      higher thani the $2 per day level.
9 See, for example Datt and Ravallion (2002)
potential growth by improving their invest-            10. The growth rates are weighted by labor value
ment climates. A sound policy environment           added, which may bias the regional estimates down-
will trigger investment flows and, more im-         ward if high-wage countries iave slower growth rates
portant, will ensure that these flows go into       Among other things, this method would affect a world
internationially competitive activities. A sound    total because industrial couLntries have sigilficantly
investment climate in developinig countries         higher wages than developing countries.
can           11ta FIwcialII TFP is nor the same as technical progress, but
instead includes all contributions to growth that are
volatile capital inflow.                            inot captured by data on capital and labor However,
the integration of more efficient techinologies into pro-
duction can raise the level of TFP
Notes                                                  12 Technically, only the fraction of capital accu-
1 Examples of investment-grade borrowing coun-   mulation that arises from the underlying propensity to
tries are Chile, China, the Republic of Korea,      invest at a constant rate of TFP growth should be
Malaysia, Mexico, Thailand, and several Central     viewed as capital's independenit contribution to output
European countries                                  growth Hulten (2000) found that correcting for the
2 See Appendix 1, "Regional Economic Prospects,"  induced capital accumulationi caused by highier TFP
for detail on recent developments, policy, and prospects  almost doubled estimates of the contribution of TFP
for developing regions                              to growth for the United States
3. Simulation used the world model by Oxford        13 See Loayza, Fajnzylber, and Calder6n (2002)
Economic Forecasting Inc                               14. We will use the term "workers" as shorthand
4 Investment is the most cyclical component of   for the working population between ages 15 and 65
GDP and is the key drivinig force underlyling the emer-  15. With the significanit exception of SSA, where
gence of turning points in the economy The flow of  AIDS has dramatically reversed life expectancy, and to
investment expenditures is volatile because investment,  a lesser extent ECA, whiere health systems deteriorated
unilike consumption, represents the desired change of  during the transition toward market economies.
a stock. As the capital stock tends to move with income  16. See, for example, Loayza, Fajnzylber, and
and consumption, the change in the srock shows      Calder6n (2002) and Masson, Bayoumi, and Samici
sharper fluctuations than the change in income. Fur-  (1998).
thermore, a downturn in in vestmiienit is inherently tem-  17. Model simulations start in the base year 1997,
porary and bears the seeds of subsequent recovery   and net capital flows for that year are derived from the
Once the lower desired capital stock has been achieved,  Global Trade Analysis Project (GTAP), release 5.0,
the flow of investment stops falling and starts increas-  database. Note that the GTAP data, though based on
ing again to keep the capital stock stable at the new  official statistics, are adjusted to ensure global ac-
level.                                              counting consistency (that is, the sum of the capital
5 The trends are computed witlh Hendrick-Prescott  account across all regions is identically equal to zero).
filters                                             The world capital account has had a significant and
6. To reduce the potential risk of bad data contam-  increasing residual over the past few years. The values
mating the results, we have excluded outliers (invest-  in the first three columns of table 1.10 (that is, the
ment volatility more than three standard deviations  average savings to investment ratios for the period
above the mean).                                    1997-2001) are largely consistenit with the initial 1997
7. The correlation coefficient exceeds 0 75      base levels from the GTAP dataset The only region
8 The large drop in the poverty incidence in Latin  where the signi of the capital account balance differs be-
America comes from new surveys and from revisions to  tween 1997 and the 1997-2001 average is the MENA,
consumption levels The sur eys predate the recent tur-  which is subject to considerable volatility because of
moil in the region, particularly in Argentina, where the  the price of oil, the region's main export.
incidence of poverty has increased substantially after  For all other regions, the sign of the observed early
three years of recession. The large recent rise in poverty  period capital account balance is consistent with the
in Argentina reflects the national poverty line The rise  1997 base year, though the magnitude may differ The
using the World Bank's $1 per day or $2 per day may  difference in maginitude is easily magnified because the
41



G LO B AL EC ONO MI C       PROS PE CT S     2 0 0 3
capital account balance is a residual item. Thus, even  Assimilation." Brookings Papers on Economic
if there is little volatility in the savings and investment  Activity 2: 135-91.
components individually, there could be significantly  Crucini, Mario J. 1997. "Country Size and Economic
more variation in the capital account balance.         Fluctuations " Review of International Econom-
Finally, as already highlighted in the table nore, the  ics 5(2): 204-20.
capital account imbalance at the global level has been  Datt, Gaurav, and Martin Ravallion. 2002 "Is India's
large of late-and is increasing In the model simu-     Economic Growth Leaving the Poor Behind>"
lations, the base data are adjusted to remove the     Journal of Economic Perspectives 16(3). 89-108.
global imbalance, and the model itself ensures global  Denison, Edward E 1972. "Some Major Issues in
accounting consistency in each year of any simulation.  Productivity Analysis: An Examination of the
(The model could track perfectly the observed saving   Estimates by Jorgenson and Griliches." Survey of
and investment ratios for each country except one. Or  Current Business 49(5, part II): 1-27.
there would have to be a residual country in the model  Easterly, William, and Ross Levine. 2001. "It's Not
that would absorb any adjustments to ensure global     Factor Accumulation: Stylized Facts and Growth
accounting consistency) Even if the global capital     Models." The World Bank Economic Review 15:
account imbalance is small as a share of global GDP,   177-219.
squeezing out the $100 billion to 200 billion residual  Feldstein, Martin, and Charles Horioka. 1980.
error is bound to have significant effects on the capital  "Domestic Savings and International Capital
account of individual countries, even a large one such  Flows." Economic Journal 90(358): 314-29
as the United States. If the U.S. imbalance is $400 bil-  Hulten, Charles R 2000 "Total Factor Productivity. A
lion and the entire adjustment is forced on it, the U S  Short Biography" National Bureau of Economic
capital account imbalance could change by as much as   Research (NBER) Working Paper no. 7471 Boston.
50 percent                                        Jorgenson, Dale W., and Zvi Griliches. 1967. "The
18. Note that this concept is flow based. All indus-  Explanation of Productivity Change." Review of
trial countries have significant assets in other industrial  Economic Studies 34(3) 249-83
countries as well as in developing countries. One        1972 "Issues in Growth Accounting: A Reply
would anticipate, as in the case of Japan, that as the  to Edward F Denison " Survey of Current Busi-
population ages, the retired elderly will draw down    ness 52(5, part 11): 65-94.
their accumulated savings, including those funds in-  Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-
vested abroad                                          Lobat6n. 2002. "Governance Matters II' Up-
19. Table 1.10 compares the observed investment     datedLndicatorsfor2000-01 "WorkingPaperno.
ratios over the 1997-2001 period with the end-of-      2772 World Bank, Washington, D.C. February.
period investment ratio generated by the model The  Klenow, Peter, and Andres Rodriguez-Clare 1997
starting point of the model, 1997, has an investment   "Economic Growth: A Review Essay." Journzal of
ratio of 20.5 for Latin America. Thus, the scenario is  Monetary Economics 40(3)- 597-617
forecasting virtually no change in the ratio      Krugman, Paul 1994 "The Myth of Asia's Miracle."
Foreign Affairs 73(6): 62-77 (November/
December).
References                                        Loayza, Norman, Pablo Falnzylber, and    Cesar
Barro, Robert, and Xavier Sala-i-Marnn. 1992. "Con-    Calder6n 2002. "Economic Growth in Latin
vergence." Journal of Political Economy 100(2).   America and the Caribbean' Stylized Facts,
223-51                                            Explanations, and Forecasts " World Bank,
Baxter, Marianne, and Mario J. Crucini. 1993.          Washington, D C. June Processed
"Explaining Saving-Investment Correlations."  Masson, Paul R., Tamim   Bayoumi, and Hossein
American Economiic Review 83(3). 416-36           Samiei. 1998. "International Evidence on the
Bhalla, Surlit. 2002. Imagine There's No Country:      Determinants of Private Saving." World Bank
Poverty, Iiequality, and Growvth in the Era of    Econonmic Review 12(3): 483-501
Globalization Washington, D.C . Institute for  Nelson, Richard R, and Howard Pack. 1999. "The
International Economics                           Asian Miracle and Modern Growth Theory."
Brock, William A., and Steven N. Durlauf 2001          EconomictJournal 109(July). 416-36.
"Growth Empirics and Reality." The WVorld Bank  Parente, Stephen L, and Edward C. Prescott. 1994.
Economic Revietv 15(2): 229-72                    "Barriers to Technology Adoption, Learning-by-
Collins, Susan M, and Barry P. Bosworth 1996 "Eco-     Doing, and Economic Growth." Journal of Polit-
nomic Growth in East Asia: Accumulation versus    ical Economy 102(2): 298-321.
42



T HE   IN T ER NAT IO NA L EC ONO MY          AND    PRO SP EC TS F OR       DE V EL O PING      C O U N T R I E S
-   2000 Barriers to Riches. Cambridge, Mass      Carlos Sobrado, and Jean-Philippe Tre 2002
MIT Press.                                         "Poverty in Latin America: Trends (1986-1998)
Ravallion, Martin 2002. "Have We Already Met the        and Determinants." World Bank, Washington,
Millennium  Development Goal for Poverty;"         D C February. Processed
World Bank, Washington, D.C. Processed.        Young, Alwyn. 1992. "A Tale of Two Cities Factor
Senhadli, Abdelhak. 2000  "Sources of Economic          Accumulation and Technical Change in Hong
Growth   An Extensive Growth Accounting            Kong and Singapore." In 0. J. Blanchard and
Exercise." International Monetary Fund Staff       S. Fisher, eds., NBER Macroeconomics Annual
Papers 47, no. 1 Washington, D.C.                  1992 Cambridge, Mass.: MIT Press, pp. 13-53
Solow, Robert. 2001. "Applying Growth Theory             . 1995 "The Tyranny of Numbers: Confronting
across Countries." Th7e World Bank Economic        the Statistical Realities of the East Asian Experi-
Review 15(2). 283-88.                              ence." Quarterly Jouzrnal of Economics 110(3)
Wodon, Quentin, Rodrigo Castro-Fernandez, Kihoon        641-80
Lee, Gladys Lopez-Acevedo, Corinne Siaens,
43






Changes in Global Business
Organization
T       he organization of global business is  networks, and the growinig importance of
rapidly changing in ways that affect  strong investment climates for the allocation
the competitive opportunities open to  of foreign investment.
developing countries. A principal feature of
business organizationi is the steady expansion  Developing countries have benefited
of multinational corporations and their related  from the boonm in global trade and
trade and investment activities. Multinational  investment-
companies, including many based in other  Cross-border trade and direct investment have
developing countries, are altering the compet-  expanded rapidly over the past three decades.
itive landscape by providing for developing  Global exports of goods and services increased
countries a new source of entry into markets.  from 14 percent of output in the early 1970s to
Moreover, by taking advantage of falling com-  23 percent by the late 1990s, whiile global FDI
munication and transport costs, multiniation-  flows have more than doubled relative to gross
als have learned to manage different stages  domestic product (GDP). The surge in FDI
of production in multiple, distant locations,  flows, with a large boost from cross-border
thereby creating opportunities for developing  mergers and acquisitions (M&A), accelerated
countries to produce during those stages of  in the late 1990s. FDI rose from $331 billioni
production-often labor-intensive stages-  in 1995 to $1.3 trillion in 2000 before falling
that correspond to their comparative advan-  to an estimated $725 billion in 2001. Despite
tage. But tapping into this potential source  the sharp increase in M&A, the share of global
of competition is not automatic, and not all  economic activity accounted for by the top
countries have benefited. Moreover, some  50 companies does not appear to have risen
observers have openly worried that the re-  significanitly during the 1990s. The top 50 com-
cent surge in global mergers among leading  panies accounted for 0.8 percent of world
multinationals might be dampening competi-  GDIP, and their aggregate profits amounted to
tion and creating obstacles for developing  3.3 percent of global savings in 2000.
countries.
This chapter reviews four recent trends in  -the rise in service sector FDI-
the organization of global business that affect  A second change in global business organiza-
developing countries' ability to harness for-  tion creates an opportunity for developing
eign investment for greater competition:  countries to expand productivity-enihancinig
changes in global business concentration, the  competition. Foreign investimienit in services-
rise in service sector foreign direct investment  financial, wholesaling and retailing, real
(FDI), the growth of global production    estate, and business services-is accelerating.
45



G LO BA L EC O NO MI C PRO SP EC TS 2 0 0 3
Today, services account for more than half  -but a strong investment climate is
of the FDI stock in most major industrial   critical
countries. The rise in service sector FDI helps  The policy and institutional framework is an
developing countries to introduce new tech-  important determinant of whether countries
nology, to boost competition in services, and  have participated in the rise in FDI. During
to increase the availability and quality of ser-  the 1990s, countries with strong investment
vices. Because many services are essential in-  climates captured an increasing share of rapidly
puts to production, with multiple linkages to  expanding global FDI flows. The removal of
virtually every dynamic part of the economy,  restrictions on private investment in services
increasing their efficiency directly boosts  (particularly infrastructure services) has in-
economy-wide productivity. However, many    creased private investment and has improved
countries still maintain impediments to this  the quality of services available to firms in de-
new source of competition and technology    veloping countries. The lowering of trade bar-
and, as a result, are at risk of being left  tiers and reduction in restrictions on FDI has
behind.                                     facilitated developing countries' participation
in cross-border production networks. External
-and the growth of cross-border             factors also play a role in determining access to
production networks-                        FDI. For example, the recent deterioration of
Technological progress in transport, commu-  the global business environment has led to a
nications, and data processing-coupled with  reduction in investment in high-risk projects,
policy reforms-has fueled the growth of     and foreign investment in infrastructure has
cross-border production networks, in which  dropped all over the world. Still, those coun-
multinational corporations break down the   tries with macroeconomic stability, sound gov-
production of final goods into stages that  ernance, and healthy institutions will attract an
vary in the intensity of capital, skilled labor,  increasing share of available funds.
unskilled labor, and other requirements, and
multinationals produce each stage where it
can be done at lower cost. In part, production  The surge in trade and FDI
through networks is accomplished by greater
outsourcing of production, as multinationals  Trade and FDI have grown rapidly since
become less vertically integrated. In part,  the 1970s-
networks are established through foreign    Cross-border trade and direct investment have
subsidiaries.                               expanded rapidly over the past few decades.
Developing countries' increased participa-  Global exports of goods and services increased
tion in production networks is seen in the  by 5.5 percent per year in real terms from
rapid growth in their exports of parts and  1978 to 2001, rising from just over 14 percent
components, as well as in their increasing  of output in the 1970s to almost 25 percent
importance in intra-firm trade by multination-  of output in 2001. High-income countries
als. Participation in networks has generated  account for the bulk of world trade and hence
substantial gains for developing countries  the largest increment to trade flows. Develop-
through improving access to technology, thus  ing countries' exports rose by just under 6 per-
increasing the demand and supply of skilled  cent per year in real terms from 1978 to 2001,
labor, as well as providing the opportunity for  and their aggregate exports-to-GDP ratio in-
moving up the value chain to produce more   creased by more than half over this period
sophisticated products. However, production  (figure 2.1).
for networks is highly concentrated in coun-   Global FDI flows have also expanded
tries with strong policy regimes, skilled work-  rapidly. The surge in FDI flows accelerated
forces, and adequate infrastructure.        in the late 1990s, rising from $331 billion in
46



C H A N G E S   I N  G L O B A L  B U S I N E SS   O R G A N I Z A II 0 N
sharp rise in the average ratio of FDI to invest-
Figure 2.1 Exports-to-GDP ratios have                ment during the 1990s (figure 2.2), with the
increased since the 1970s                            largest increase in the industrial countries dur-
(percent)                                            ing the last years of the decadc. Low-income
35                                                   countries have seen a five-fold rise in FDI rela-
30                                                   tive to investment, to almost the same ratio as
in lower-middle-incomile countries. FDI flows
25          High-incomecountnes                      to developing countries equal about $160 hil-
20          H   -                                    lion, while domestic investimlent in developinig
1 5                                                  countries equals about $1 5 trillion.
Developing countries                        The rise in trade and FDI has played an
10                                                   important role in boosting the productivity of
s                                                   firms in developing countries. In part, devel-
oping countries may become more productive
0        o         I Q,   I I C  I Iz ,  I  (b I  53  because trade inmproves the allocation of re-
'A          '6 g  'C' ', ' ' Ksources a                  nd because multinational sLubsidiaries
Sorurce World Banik data                             Inay be more productivc than domestic firms.
In addition, domestic firms may increase their
productivity through participation in trade
1995 to $1.3 trillion in 2000 before falling to         and contacts with local subsidiaries of foreign
an estimated $725 billion in 2001 (UNCTAD               firms, although the extent and channels are a
2002a). All income       groups    experienced    a     matter of considerable debate in the economic
Figure 2.2 All regions have benefited from rising FDI flows
(FDI-to-investment ratios, 1990-2000, in percent)
40
35
30
25
20
1 5
105                                                il1
0
Top 10 developing     Low-income      Lower-middle-ncome  Upper-middle-income    High-income
country recipients     countries          countnes            countnes            countries
Note Each set of bars repi eseists the period 1 990-2000, each bar represents onie year dtiring that period lThe data on FDI itcliUde
both greenfield and M&A transactiols. wbhereas national iiiconiie account data oni itivestmient representi oiilv new inivestimienits .Top
10 developisig counitry recipienlts" refers to the 10 deselopiig counitries that received the largest inflows of FDI
Souirce World Banik data
47



G LO B AL EC O NO MI C PRO SP E CT S       2 0 0 3
literature. Much empirical work has focused     -but not all countries have participated
on the potential for technological spillovers   equally in the rise in FDI
through importing, exporting, and FDI (see      Among industrial countries, the top five recip-
chapter 3 for a full discussion). On balance,   ients of net FDI flows accounted for 74 per-
the evidence for technological spillovers       cent of total FDI. However, a few    of the
through imports is strong, while the evidence   smaller countries (for example, Ireland and
that exporting promotes technology diffusion    Denmark) have the highest ratio of FDI to
is less robust. Evidence for the existence of   GDP. The same pattern can be seen in devel-
technology spillovers from   FDI is mixed.      oping countries, where the top 12 recipients
Many industry-level studies (for example,       captured 80 percent of total FDI flows, but
Blomstrom 1986) have documented a positive      some smaller countries had FDI-to-GDP ratios
correlation between FDI inflows and produc-     that were several times the average ratio. Fig-
tivity, although the causal direction is unclear.  ure 2.3 compares each developing country's
Some firm-level studies have failed to find evi-  share of total FDI with its ratio of FDI to
dence of technological spillovers in developing  GDP (the countries are ordered by the share
countries. The effect of FDI will depend, in    of total FDI). Almost half of the 12 largest
part, on the form that FDI takes. FDI directed  recipients of FDI (at the far right of the distri-
to heavily protected industries or attracted by  bution in figure 2.3) have FDI-to-GDP ratios
very costly incentives may have a low, or even  that are lower than the average. According to
negative, effect on growth and productivity.    a more comprehensive measure developed by
But FDI used to integrate domestic sub-         the United Nations Conference on Trade and
sidiaries in production networks could have     Development (UNCTAD), FDI to developing
substantial spillover effects (Moran 2001).     countries is mildly concentrated: only 30 out of
Figure 2.3 FDI is concentrated in large countries, but many small countries receive large
amounts relative to GDP
(percent share of total FDI)                                         (FDI as a share of GDP)
25 -                                                                            45
- 40
20 -X-35
20 -                              Some countries with a small share of total
30
15 -        Top 10 receive almost        FDI have high FDI-to-GDP ratios (nght axis)
80% of total FDI                                                    25
-flows (left axis)                                                    20
10 
15
5                                                          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 0
5            ,                                       «f                           1 ; n n nn n n   + * *   * | I- 50
Mos 0D es D
0         iHnno'nnnn                       0,6 0, .a.
tQ                '  i- Q4. 
Note Countries sorted by share of total FDI, l1999
Soiirce World Baiik data
48



C h AN GE S IN G LO B AL B US IN ES S OR GAN I Z AT IO N
Figure 2.4 The service sector dominated the 1990s mergers and acquisitions boom
Mergers and acquisitions boomed in the 1990s  Services dominated the mergers and acquisitions
boom from 1987 to 2000
(billions of dollars)                      (billions of dollars)
1200                                       900
1000                                       800
700
800                                       600
600                             /         500                       Services
600                            ~ ~~~~~~~~~~~~400  _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
400         Developing countries          300                   Primary
200   High-income countries '~200                     / Manuatrn          '
100
O                                         O
Soiirce UNCTAD (2001)
102 developing countries had shares of FDI   Global concentration of ownership
that equaled or excecded their average shares of  does not appear to be increasing
world GDP, employment, and exports (see box  Contrary to popular perceptions, the boom In
2 2 in World Bank 2002b). Obviously, many    cross-border M&A does not appear to have
factors other than market size, particularly the  had a major effect on the global concentra-
policy and institutional framework, are impor-  tnon of ownership. Cross-border M&A trans-
tant In determining a country's attractiveness  actions in the late 1990s represented only a
to FDI.i                                     small fraction of industrial countries' stock
The takeoff in M&A transactionis among     market capitalization. The dollar value of
industrial countries-driven in part by extra-  cross-border M&A transactions equaled less
ordinarily rapid increases in the stock prices  thani 3 percent of stock markets in most of
of some major corporations and in part by    the top seven industr-ial countries (figure 2.5).'
expectations (during the boomn) that continu-
ing productivity increases would fuel contin_-
Lied rises in stock prices-was a driving force
behid graterFDI see igur  2.4. Glbal Figure 2.5 Cross-border mergers and
behind greater FDI (see figure 2.4). Global     acquisitions are small compared with
M&A rose more than five-fold between 1995       stock market capitalization
and 2000 (after increasing by only 24 percent   (average value 1997-99, in percent)
In the first half of the 1990s) to a peak of    4_0
$1 1 trillionii n 2000, before dropping by some  3 5
45 percent in 2001 with the decline in stock    30 -
markets and the global economic slowdown.2      2 5
This experience was not unprecedented:          25     *   *    *
through the 1980s and 1990s, the global econ-   1 0    U   *    *
omy experienced major waves of corporate        05     C             U    I   ___
mergers.3 The bulk of the cross-border M&A       0    !    !     1   1   !      M
transactions was in service sectors (more than  c    -        & ,
half in finance, transport, storage, and com-                    a
niunications alone), which accelerated rapidly
beginning in 1998 (see also the discussion of   Souirce Evenert (21112)
FDI in service sectors, page 10).
49



G LO BA L E CO NO MI C PROS PE CT S 2 0 0 3
Of course, the M&A boom coincided with the
sharp rise in stock market valuations in the  Figure 2.6 U.S. aggregate concentration
late 1990s, particularly in the United States.  has held steady
Thus it is useful to keep in mind that both the  Share of top 500 companies of U.S private sector
numerator and the denominator in figure 2.5   (percent)
are rising rapidly.                           80
A related concern is whether the concentra-  70     ,Profits
tion of global economic activity, either in mdi-  60
vidual sectors or in total, has increased for  50                           _
reasons other than cross-border M&A. For      40
social and political reasons, high concentra-  30
non in an economy may be a matter of con-     20
cern. For example, 5.5 million corporations    t0
are in the United States, with the largest 100  0 l,,,    ,,
companies accounting for about l1 percent        .z  W$z            IR, IC' Ie@$'I'.'IC
of employment and payroll.5 The fabric of
the U.S. economic (and socio-political) land-  Source White
scape would surely be different if there were
no small enterprises, no start-ups, and no
alternative places (beyond a few mammoth    because each firm may participate in multiple
corporations) where someone with a new en-  markets.9
trepreneurial idea might go to obtain financial  Concentration at the global level appears
support and institutional encouragement.    to remain low, although one confronts enor-
Similar concerns would hold for a high level of  mous data problems and difficult tradeoffs in
concentration in the global economy. There  making such estimates. In 2001, total employ-
may be an extra element of concern for devel-  ment by the largest 50 global companies (as
oping countries in this regard. Few of the  identified by Forbes) accounted for 0.3 percent
largest companies in the global economy are  of the world labor force, or 1.6 percent of
headquartered in and identified with a devel-  employment in Organisation for Economic
oping economy.f6 A global economy that is   Co-operation  and  Development (OECD)
dominated by a relative handful of giant com-  countries. Those companies' profits amounted
panies (if that were the case), which are head-  to 0.8 percent of world GDP and 3.3 percent
quartered in a relatively few industrial coun-  of world gross domestic savings.10 Although it
tries, may raise even greater socio-political  is difficult to say what level of concentration
concerns in developing countries that feel that  should be viewed as a cause for concern, at
they can exert little effective control over these  least these aggregate data do not reflect a dom-
enterprises.                                ination of the global or OECD economies by a
Although the measurement issues involved  small number of firms.
are enormous, it does not appear that global   Global concentration does not appear to
concentration is high or has been rising signif-  have risen significantly during the 1990s. The
icantly during the 1990s.7 White (2001, 2002)  share of the top 50 companies (as measured by
reports declining or stable aggregate concen-  Forbes) in the world labor force and in OECD
tration in the U.S. economy from the 1980s  employment has fallen slightly since 1994 (fig-
through the late 1990s, depending on whether  ure 2.7).11 The declining share of the large
employment, payroll, or profit data are used  companies' employment levels is consistent
(figure 2.6).8 Note that this measure of aggre-  with the trend for the United States reported
gate concentration does not provide an indica-  by White (2001, 2002). Despite the merger
tor of market power in individual markets,  wave of the 1990s, very large companies have
50



CH AN GE S IN G LO B AL B US IN ES S OR GAN I ZAT IO N
Figure 2.7 Global concentration has not increased significantly
Employment of the top 50 global companies, 1994-2001  Profits of the top 50 global companies, 1994-2001
(percent)                                   (percent)
18      Share of OECD em6loymet             6
1 6     S   o8  E  employen                                 Share of OECD savings
T                     ~~~~~~~~~~~~5
14 
1 0
08   \   200d labor1force
06                                           2 m   iii 
040
1994 1995 1996 1997 1998 1999 2000 2001    1994 1995 1996 1997 1998 1999 2000 2001
Souirce Wloite 20021,
not experienced a significant expansion of       A different approach to the calculation of
employment relative to other companies. This  global concentrationl is reported by DeGrauwe
has been partly due to interinal rationaliza-  and Camerman (2002), with similar conclu-
tions and cost-cutting by those companies and  sions. They find that sales of the top 50 in-
partly due to significanit numbers of spinoffs  dustrial corporations from  the Fortune 500
and divestitures.                              list have grown slightly less rapidly than world
The share of the top 50 companies' profits  GDP from 1980 to 2000. Thus the 2000 sales
in global savings and OECD savings has risen  of the 50 largest industrial corporations were
sincc the rnid-1990s (figure 2.7). The rise in  slightly smaller in relation to world GDP than
profits among the largest companies is also   was true for the 50 largest corporations in
consistent with the U.S. trend reported by     1980.12
White (2001, 2002a). However, in the United      These indicators of global concenitiationi
States, economy-wide profits were rising      reveal nothing about the concentration of
rapidly during the late 1990s; thus the ratio of  market power in individual sectors. Risinig
the largest companies' profits to total profits  concentrationi at the sectoral level may reduce
was relatively constant. Unfortunately, tinme se-  competition, thereby increasing prices faced
ries data on global profits, or even OECD prof-  by consumers and suppliers and shifting
its, are not available. Nevertheless, it is likely  wealth from  consumers and suppliers in
that OECD profits were rising more rapidly    competitive industries to producers in more
than nominal GDP or savings; hence the share  concentrated incustries. Unfortunately, com-
of the top companies' profits may not have in-  prehensive data do not exist on global sectoral
creased as much as indicated by the ratios given  concentration.
in figure 2.7. Moreover, the recent accounting   Sectoral data are available for some major
scandals affecting telecommunications, energy,  countries, and they do indicate a rise in coil-
and other high-tech companies indicate a sig-  centration ratios. The average concentration
nificant overstatement of profits in many of the  of industries at the Standard Industrial Classi-
largest companies during the late 1990s. Thus  fication (SIC) 4 level increased from 1947 to
the rise in profits of the top 50 companies rela-  1992 in the United States, while concentration
tive to global savings may be overstated.     declined slightly from 1983 to 1992 in Japan



G LO B AL EC O NO MI C     PRO SP EC TS    2 0 0 3
before increasing   sharply in  1992-98.13     Table 2.1  FDI inward stocks in services
However, sectoral concentration ratios at the   and manufacturing, 1988-99
country level provide little information on the  (groth rate and shares in dollars)
competitiveness of markets, because most of                         Growth rate.
1988-99      Share,
those companies face competition from   im-                         I 1   9 change  1999
ports. Indeed, the rapid rise in world trade                          pr vecar)  (percent)
over the past two decades, coupled with the
emergence of developing-country exporters,       orlDi
Total FDl             123
indicates that competitive pressures may have    Manufacturing      12.2           41 6
increased in many industries.14                  Services              13.8        503
Despite the difficulties in measuring global
Industrial countries, I
sectoral concentration ratios and in determin-   Total FDI             9 9
ing the implications for competition, anticom-   Nlanufacturing         9 1        36 4
petitive practices have clearly affected some    Services              i 6         555
industries. The 1990s saw the uncovering of a  Developing countries,
large number of international cartels, in which  Total FDI             21 5
firms from   more than one country made          ManufaCtUring         19 6        54 5
explicit agreements to fix prices, divide up mar- _Services_28_2               _   37_3
kets, or rig bids for contracts (see chapter 4).  Note Seconid column data for France are from 1998, and
seconid columin data for Japan are from 1994
Souirce UNCTFAD (2001)
The rise in service sector FDI
lines the importance of an effective regulatory
FDI flows into services have overtaken          regime, because designing and enforcing an
those in manufacturing-                         appropriate regulatory framework is more
Service sector FDI has grown rapidly over the   difficult in many service sectors (such as
past few decades, and services are now the     natural monopolies in infrastructure) than in
dominant sector for foreign investment.1s       manufacturing.
The stock of FDI in services was only about
one-fifth of the total in the 1950s (United
Nations 1989), but by the mid-1970s the
share of services in the stock of outward FDI     Figure 2.8 Share of FDI in the service
of major industrial countries ranged mostly       sector increased in major industrial
between 30 and 40 percent.i6 By 1990, this        countries
share rose to between 45 and 60 percent, and      (percent)
over the past decade, FDI in services has con-    -0
tinued to rise more rapidly than FDI in man-      80
ufacturing in both developing and industrial      70                     1998
countries (table 2.1). By the end of the 1 990s,  60
services accounted for more than half of the      50       _
stock of inward FDI in most major industrial      40 i
countries (figure 2.8). Despite the rapid         20                                   _
increase in service sector FDI, the global ratio  10                                   _
of FDI to value added in services remains less     0
than half the ratio of FDI to value added in          France Germany italy Japan Unted United
manufacturing, thus indicating the potential                                Kingdom States
for further increases in service sector FDI.      Sou,rce OECD (2001) and OECI) onliie database
The dominance of service sector FDI under-
52



C It A N G E S I N G L O B A L B U S I N E SS  0 R GANIZATION
-reflecting the rising role of services in  intensive services-research and development
the global economy-                         (R&D), engineering, design, computing and
FDI in services has increased relative to man-  data processing, inventory management, qual-
ufacturing, in part because of the growing  ity control, design, accounting, legal services,
importance of the service sector in economic  personnel services, and so on-have become a
activity. By the late 1 990s, the service sector  critical part of the production process in the
had increased from half of global output in the  manufacturing sector. With moderin manufac-
early 1970s to 64 percent. Income growth has  turing production and distribution becoming
been the driving force behind the rise in ser-  increasingly dependenit on the processing and
vices: cross-country comparisons show that  dissemination of information, the demanld for
the richest countries have the greatest share  those producer services is rising rapidly.
of services. Services account for 70 percent of  Moreover, the growing sophistication and
output in industrial countries, 55 percent  variety of services, coupled with specialization
in middle-income countries, and 44 percent in  emerginig from economies of scale, have led
low-intcomie countries. The correlation coeffi-  manufacturinig firms to rely more on out-
cient between income level and the share of  sourcing than o1 in-house departments to
services is 0.6. The relationship between   provide the services necessary for production.
higher income and a greater share of services  The immediate consequence is a statistical
in part reflects consunmer demand, because  effect: the size of the service sector rises wheni
luxuries such as travel and entertainmrent  services that were previously classified as
often have a large services component. Also,  manufacturing output are suddenly counted
higher incomes permit an allocation of more  as services. Typical examples of these types of
resources toward protecting assets (insurance  services are accounting, computer services
and legal services), richer and complex soci-  (data processing and software), warehousing,
eties require more resources devoted to educa-  public relations, information technology, and
tion and advisory services, and technological  management information systems.
advaiices associated with higher incomie widen  Technological progress has greatly reduced
the scope for the protection of health. Finally,  the cost of some services, thus increasing the
the higher labor intensity of services and rising  scope of services that are feasible to supply
real wages have increased the nominal value  (for example, mobile telephones, complex
of services relative to manufacturing.      financial transactionis suclh as derivatives, and
a host of other services facilitated through
-technological changes that have            advanced data processing). Technological
increased the demand and supply             progress has also generated new means of
of services-                                deliverinig services (for example, the dissemi-
Technological progress has tended to increase  nation of research over the Internet). This
the demanid for services connected with the  process is similar to what occurred during the
production of goods and to facilitate the sepa-  industrial revolution, when technological
ration of goods production from services pro-  progress and income growth greatly increased
duction.17 The larger scale of production, the  the importance of manufactures when com-
greater technological sophistication of goods,  pared with the primiary sector.
and the increased trade in goods and imianage-  Both the reduced cost of some services and
ment of enterprises across large distances have  the increased scope of services have increased
all contributed to the greater demand for ser-  the tradability of services, a trend that has,
vices. The importance of management, market-  in part, beeii exploited through increased FDI.
ing, distribution, and after-sale maintenance  For example, software can be produced in
has risen relative to the value of manufactured  low-cost locations such as India and sold
products. Many information-and-knowledge-   directly to firms and consumers in the United
53



G LO B AL EC O NO MI C PROS PE CT S       2 0 0 3
States over the Internet. Many multinationals  provide services through subsidiaries, rather
have established centers in developing coun-  than other kinds of relationships, include
tries, where wages are low, to handle contacts  (a) financial institutions, in which much of
with consumers in industrial countries. For   proprietary knowledge is tacit, is expensive to
example, call centers in the Caribbean manage  produce, and is complex and idiosyncratic;
phone calls to multinationals from U.S. clients  (b) firms that require control over production
(this is an example of participation of service  to maximize efficiency and to protect the
firms in production networks; see page 13 for  quality of the end product (and thus customer
an elaboration of this concept in manufactur-  goodwill) for trademarks (for example, in
ing). In some cases, the rising tradability of  advertising, market research, construction,
services may have reduced FDI by enabling     business consulting, consumer-oriented ser-
firms to provide services at a distance rather  vices, and goods-related personal services such
than establishing a subsidiary. However, in   as motor vehicle maintenance and repair); and
general, increased tradability has created new  (c) trade-related service affiliates set up by
opportunities for multinationals' subsidiaries  non-service multinationals to obtain inputs
to export services to home markets and, in    for domestic activities or to supply markets.
some cases, to operate as international hubs
to supply services to firms in other countries.  -and policy changes that encourage
Because industrial countries are the lead con-  the private provision of services
sumers of tradable services, developing coun-  The removal of restrictions on FDI and regula-
tries have benefited from  the establishment  tory reforms that have improved competition
of subsidiaries to service the richer markets.  in service sectors have contributed to the rise in
Technological advances that increase services  service sector FDI. Until recently, many coun-
tradability have also imparted an advantage   tries (including many industrial countries) pro-
in service delivery to multinationals relative to  hibited foreign investment in sectors such as
domestic firms, thus enabling the former to   transport, communications, banking, finance,
overcome the natural advantages of proximity  utilities, and media. Since the mid-1980s,
and knowledge of the market (Sauvant and      governments in both industrial and develop-
Mallampally 1996).                            ing countries have been gradually opening up
Income growth and technological progress    those service sectors to foreign investment.18
have boosted the provision of services through   Multinationals can enhance the efficiency
various forms of cross-border relationships   of services industries in developing countries
in several sectors: (a) management and fran-  by providing services that developing-country
chise contracts in hotels, restaurants, and car  suppliers cannot provide, as well as by intensi-
rentals (in which performance requirements    fying competition. In particular, providing
can often be adequately codified, local man-  producer services (for example, managerial
agerial input is desirable, and the synergistic  services, engineering, finance, and marketing)
advantages of global reservations and referral  that are often subject to economies of scale and
systems can be obtained without the risks and  that have a much higher cost from a distance
costs involved in an equity stake); (b) joint  can generate important benefits to developing-
ventures in some business services, in recre-  country firms. Availability of producer services
ational activities, in some accounting and legal  may be an important reason to form industrial
services, and in civil engineering in turnkey  complexes and may explain a significant share
projects (in which individual customization   of the differences in economic performance
and specialized knowledge of local practices  among regions. Producer services are likely to
are required); and (c) services in which a local  be provided through FDI (rather, for example,
partner is required for marketing and dist-   than through training unaffiliated firms)
ribution (Dunning 1981). Firms that tend to   because they involve knowledge-based assets
54



C H AN GE S IN G LO B AL B US IN ES S OR GAN I Z AI IO N
that are easily copied if firms lose control
over the knowledge (Markusen, Rutherford,    Figure 2.9 U.S. cars are produced in
and Tarr 2000). FDI has surged in developing  many countries
countries' banking sectors, in many cases re-  (percentshareinvalueadded)
ducing the costs of financial intermediation,  40
increasing the scope of financial services avail-
able to local firms, and transferring skills to  30
workers in developiiig countries.19          20
20    ,   _ _ _ _ __ _ _ _ _ _ __ _ _ _ _ _
Global production networks                   10                  -
The globalization of production has           ol   l             _             l
le,  31 ~ C',~9
helped fuel the growth in global trade                   10  :'c-  4o       --I'7 
Rapid growth in trade and in FDI flows has    ', .
reflected, in part, the expansion of production
networks.20 The production of many fmial
goods, which formerly took place in one loca-  Souirce WTO (1998)
tion, has been broken down into discrete steps,
with each step moved to locations where it  United Kingdom, although it fell in Japan
canl be performed at the lowest cost (Venables  (figure 2.10). Using iIput-output tables,
1999; Kimura 2001). Thus a significant pot-  Hummels, Rappaport, and Yi (1997) calculate
tion of international trade and FDI has shifted'' 
from the exchange and production of final  that the fraction of the total value of trade
consumer goods to the exchange and produc-  accounted for by inputs that are both
consofmart anods  componthe nts.hae Ts glob-  imported and then embodied in exports rose
tilon of parts and componentrs. This global- I
ization of producing individual goods has  in France, the United Kingdom, and the
United States from- 1970 to 1990, while drop-
progressed to the point that it can become dif-  Unitd tae fo  197022 rt 1990 whe drop
ficult to identify the nationality of some prod-  ping slghtly In Japan.   Data from the U N.
ucts. For example, the World Trade Organiza-  portrad  databaenshow tha   portscof
tion (WTO 11998]) gives figures for the share  parts and components-a proxy for partlcipa-
ofvalue added in producing a U.S. automobile,  tion in global networks-Increased by almost
of vaue       grouped by       atomobile,- 2 percentage points faster than exports of
with countries grouped by category of pro-  total manufactured goods from 1981 to 2000
duction. The United States accounts for only  (table 2.2).23
37 percent of value added (figure 2.9).(te.)-
Thpercent is consderabled evidee  that th  The rise in the share of trade accounted for
share of global trade accounted for by net-  by global networks in part reflects the increas-
works is inicreasing, although the results  ing importance in global production of goods
vary amonig countries and studies. Baldonie,  Table 2.2  Growth of exports of parts and
Sdogati, and Tajoh (2002) estimate that the  components,1981-2000
share of intermediate products in total trade  (averagenas,198al-perce 00tage0chatgiidollars)
within the European Union (EU) rose only
sligltly in the 1990s, from 17 percent in 1990                     1981-90  1990-2000
to 19 percent in 1999.21 One measure of inter-  ianufactured exports  10.6   7 2
national outsourcing-the ratio of imported to  Parts and components exports  12 1  9 6
total intermcdiate inputs in manufacturing-  Merno itern Share of parts
doubled in the United States from  1974 to  and colnponents         13.2    18 5
1993 and increased in Canada and the       Source UN Conmtrade database
55



G LO B AL     EC ONO MI C       PROS PE CT S        2 O 0 3
Figure 2.10 Cross-border networks capture increasing shares of production and trade
Imports of intermediate inputs increased, 1974-93      Production through networks increased, 1974-93
(ratio of imported to total intermediate inputs In     (percent share of trade accounted for by imported inputs
manufactunng, in percent)                              that are embodied in exports)
25                                                     20
2016
1 4
1c 5                                                   12 2                                        199
the~~~~~~~~~~~~~~19 1990
1 0                                                     8
6
5                                                      4
2
0                                        ~~~~~~~~~~~~~~~0
Canada      Japan      United      United               France     Japan       United     United
Kingdom     States                                      Kingdom     States
Soerce Feenstra (1998), Campa and Goldberg (1997), Himmets, Rappaport, aid Yi (1997)
such as electronics, chemicals, and transport             product     classes.  Hanson,      Mataloni,     and
equipment and       machinery, where       trade   in     Slaughter (2001) report that the share of U.S.
components is most important. The share of                multinational affiliates' imports of intermedi-
those sectors in world trade rose from       27 per-      ate inputs in their total sales rose significantly
cent in 1986 to 43 percent in 1997 (Schive and            from   1982 to 1994 in electronics, transporta-
Chyn    2001). However, the          increase   also      tion equipment, and industrial machinery and
reflects a rise in components trade within the            equipment (figure 2.11).
Figure 2.11 The role of production networks continued to increase through most of
the 1990s
Use of Intermediate inputs rose, 1982-94                Exports by MNCs abroad rose, 1982-98
(percent of U S affiliates' imports of intermediate inputs as  (percent of U S affiliates' exports as a share of total affiliate
a share ot sales)                                       sales)
25                                                      70        19892
20                                                      60
1982                                ~~~~~~~~~~~50 
1 5                                                    40 
10'                                                    30 
20 
5                                                      1 0
Souirce U S Bureau of Economic Analysis, as reported in Hanson, Maratoiii, and Slaughter (2001)
56



C If A N G E S I N G L O B A L B U S I N E SS  O R G A N I Z A T I O N
The establishment of global networks has       systems greatly reduce the costs of procure-
been facilitated by technology-                 ment and improve the coordinationi of pro-
Technological progress in transport, commu-     ductron across dispersed factories by automat-
nications, and data processing has fueled      ing the processing of routine transactions
increased FDI flows and the establishment       (Chen 1996).
of cross-border production networks. A nearly
70 percent decline in sea freight unit costs   -policy improvements-
between the early 1980s and the mid-1990s       Improvements in economic policies, notably
(in part caused b) a rise in the share of cargo  the decline in barriers to trade, have also con-
carried in containers; see World Bank 1997)24   tributed to forming cross-border production
and an increased reliance on air shipments,     networks. Successive rounds of multilateral
plus the growth of express services (such as    negotiations reduced average tariffs on manu-
overnight and two-day delivery)25 have facili-  factured products in industrial countries from
tated the shipment of com1ponents for pro-      10 percent in 1980 to 5 percelnt by 1998. The
cessing in different locations. The low cost of  average tariff rate in developing countries fell
long-distance telephotne rates, the develop-    from between 25 and 30 percent in the early
ment of fax machines, and, most recently, the   1980s to 13 percent by 1998 (figure 2.12).
advent of the Internet have made it easier for  Even relatively low  tariff rates can have a
multinationals to closely coordinate produc-    significant role in deterring the formation
tion at dispersed locations. Those changes      of cross-border networks, because goods often
have also greatly reduced the costs of finding  pass through borders several timies in the
and evaluating potential suppliers for more     course of prodLuction (Navaretti, Haaland,
arm's-length  transactions  (Grossman   and     and Venables 2000), and the gross margins
Helpman 2002). Finally, an increased ability    of manufacturing companies are often lower
to process and analyze vast amounts of data     than  5 percent.26 Hanson, Mataloni, and
has facilitated  the management of global       Slaughter (2002) find that tariffs are an im-
networks. Electronic data interchange (EDI)     portant detcrminant of the size of intermediate
Figure 2.12 Tariff rates fell in the last two decades
Average tariff raies, 1980-98                 Industrial countries' tariffs on parts and components,
1988-2001
(average tanff on nmports trom industrial countries,
(percent)                                    in percent)
35                                           6
/ Developing countries
30  _ _ _ _ __ _ _ _ _ _ __ _ _ _ _ __ _ _ _ _ _ 5
25
20
15          I3d                                                 E U      _-___
10 ~~~inusnaocuntr es
7710 -                                       0d2| 
OI       I    I       I  I      I  I   I  I   IIIII O r' 
1980 1984 1988 1990 1992 1994 1996 1998        1988-90  1991-93   1995-97  1998-2001
Note Uniweighied average of ad salorern, applied, or NtFN rates, whiichever data are available ftor a loniger period
Soitrce U N Cmrntrade database
57



G LO B AL EC O NO MI C PROS PE CT S       2 0 0 3
inputs from parent companies relative to the  shifted to negative lists (that is, lists specifying
total sales of U.S. affiliates (a direct measure  a limited number of sectors from which for-
of activity within production   networks).    eign investors are excluded or are subject to a
Higher tariffs are significantly correlated with  ceiling on the share of the firm that foreigners
less production sharing, with estimated elas-  may own).27 Often, reforms in trade, FDI,
ticities in the range of 2 to 4. Multinationals  and other areas work together to encourage
may even lobby for reduced tariffs on their in-  greater participation in global networks. The
puts so they can reduce the costs of networks.  export-to-sales ratio of U.S. multinational
The average tariff rate that industrial coun-  affiliates rose dramatically from 1982 to 1998
tries impose on imports of parts and compo-   in Mexico (following trade and investment
nents declined during the 1990s and was well  reforms in the mid-1980s), in China (after
below the overall average tariff rate by the end  reforms in the early 1990s), and in Canada
of the decade.                                (after investment reforms of the mid-1980s
Steps toward greater integration between    and the coming into effect of U.S.-Canada free
geographically close neighbors with signifl-  trade agreement in 1989) (figure 2.13).
cantly different wage rates have had a particu-
larly important role in stimulating the growth  -and incentives
of regional production networks. Before 1990,  Countries may affect their attractiveness to
the export of processed goods from Eastern    global production networks by specific require-
Europe to the EU was minimal. By 1996, such   ments or incentives affecting foreign firms.
exports were almost 20 percent of Poland's    Moran (2001) examines case studies on the
exports to the EU, 40 percent of Romania's,   industries cited above as being most heavily
and well over 10 percent in most other Eastern  involved in global production networks (elec-
European countries (Baldone, Sdogati, and     tronics, machinery, and transportation). He
Tajoli 2002). Kaminski and Ng (2001) report   finds that affiliates in countries that impose
that the value of Central Europe's total trade  relatively stringent or widespread performance
in parts grew almost three-fold from  1993    standards on multinationals (for example,
to 1997. The maqtiiladora industry in Mexico  limits on foreign ownership, domestic-content
has grown spectacularly since introduction of  requirements, and various technology-sharing
the North American Free Trade Agreement
(NAFTA).
Networks have been boosted by special         Figure 2.13 Reforming countries
arrangements. U.S. and European tariff provi-    boosted exports through production
sions encourage production by the subsidiaries   networks
of multinationals, because the tariff on a sub-  (exports of US affiliates asashare of total sales,
sidiaries' import is imposed only on the value  in percent)
added in the assembly country, not on the total  45
value of the good (Ng and Yeats 2001).           40             1998
35
Reduced restrictions on FDI in developing     30
countries have increased the participation in    25
20
international production networks. Of the        15                               1982
1 0
numerous regulatory and policy changes that       s
have affected FDI and that werc introduced        0
by developing-country governments during the       All countries Canada Mexico  China OtherAsia
1990s, 95 percent were aimed at creating a       Note Exports from US affiliates in all countries show
more open environment for FDI (UNCTAD            iso rise in the share of sales due to declines in the export
shares in primary production
2001). Many developing countries have elim-      Souerce Hanson, Mataloii, and Slaughter (2001)
mated broad restrictions on FDI and have
58



CHANGES IN G LO BA L B US IN ES S ORGAN I ZAT ION
mandates) are much less productive, use older  particular products (Arndt 2001). A spectrum
technology, and take longer to introduce new  of choices exists, and each involves some
processes and products than affiliates in coun-  form  of relationship between supplier and
tries that do not impose such requirements.  purchaser.28 The major advantage that multi-
Thus, FDI to countries with strict require-  nationals have over local firms is typically tech-
ments is more likely to be directed at local  nology, and protecting that advantage is a key
markets, because participation in networks   consideration in determining the structure of a
often requires the latest technology.        global network (Ethier and Markusen 1996).29
Because it can be difficult to maintain control
Global networks can be structured            over technology in arm's-length arrangements,
in many ways                                 FDI is often the preferred choice (Hoon and
Global networks are achieved through a range  Ho 2001).30 This preference is reflected in the
of ownership structures, from  conducting    rise in the share of intra-firm trade in multina-
arm's-length transactions (for example, trade  tionals' exports, at least as far as developing
in standardized parts sold in organized mar-  countries are concerned (box 2.1).
kets) to establishing a subsidiary for produc-  There are disadvantages, however, to es-
ing components that are custom-made for      tablishing a network through FDI. Securing
Box 2.1 Intra-firm trade increases worldwide
The boom in FDI flows during the 1990s was       actions rather than greenfield investments. The
associated with only a small rise in the share of  transfers of ownership involved in such transactions
intra-firm trade in U.S. exports during the 1990s (see  would not necessarily have a significant effect on
box figure 1).31 By contrast, the share of intra-firm  trade flows. If one looks at longer time series, there
exports in Japanese trade almost doubled during this  is some evidence that intra-firm trade has become
period. The failure of U.S. outward FDI flows (the  more important for U.S. multinationals, particularly
stock of which nearly tripled in the 1990s) to result  in services (box figure 2).
in a sharper increase in the share of intra-firm trade
probably reflects the dominance of M&A trans-
Intra-tirm trade in services is becoming
Intra-firm trade has increased                   more important
(percent share of intra-firm exports in total exports)  (percent share of intra-firm exports in total U.S MNC
40                       1999                   exports)
1999_______________________
35250
1 5                                 20~~~~~~~~~~~~4  auacue   SrIe
3 0                                             40    -
25                   199
15~~~~~~~~~~~~~~~~~~~
10                                              20
United Slates      Japan                       1977    1983     1993    1999
Souorce: OECD (2001).                           Source: OECD (2001).
59



G LO B A L EC ONO MI C PRO SP EC TS       2 0 0 3
B ox 2.1          (continued)
Intra-firm trade includes production that is  by U.S. multinationials have increased only slightly
shared among locations in global networks, as well  more rapidly than have total U.S. exports. However,
as trade in finished products tor marketing and dis-  a growing share of this trade is devoted to production  l
tribuLIoll in foreign countries There is some evidence  networks, which increasmigly involve developing  i
that production through networ ks has become more  countries.
important over time. The share of exports of inter-
nmediate goods to overseas manufacturing affiliates in
I   total Japanese exports rose from 20 percent in 1994
I   to 29 percent in 1999. Products intended for further
processing increased from 57 percent of U.S. multi-
nationals' exports to foreigni-ownied affiliates in 1989  Trade among U.S. affiliates is rising in
to 68 percent In 1999 (Mataloni and Yorgason 2002).   the developing world
Trade among foreign affiliates of U.S. multinationals  (percent share of intra-firm exports in total U S MNC
has also expanded, which probably indicates that net-  exports)
works have beconme more complex over time. The         70
share of intra-firm exports of foreign subsidiaries
accounted for by exports to other subsidiaries (rather  6019
than to the parent company) rose from 53 percent In    50s-
1983 to 66 pcrcent in 1 999 (box figure 3). This rise  40-
is almost totally due to an increase in foreign affiliate                      1983
trade among developing countries, from 30 percent      30-
of U.S. multinationals' intra-firm trade in 1983 to    20-
51 percent in 1999. Production networks appear to      10
be less important in intra-firm trade among industrial  0
countries, given their more similar labor costs. For  o  _   . . . .
example, 90 percent of intra-firm exports from                All      Industrial  Developing
countries   countries  countries
foreign multinationals to U.S. affiliates are finished
goods for direct distribution to the U.S. market. The  Soturce OECD (2001)
picture that emerges is that total intra-firm exports
intermediate inputs through contracts with    ing the 1990s and now account for 20-30 per-
local firms often entails lower administrative  cent of total electronics production.
costs than establishinig a subsidiary. The multi-
national is free to specialize in providing tech-  Developing countries have increased
nology, marketing, and distribution services,  their participation in global
while a local partner may be better situated  networks
to handle the personnel and regulatory issues  Developing couLntries have been increasingly
involved in establishing a company. Moreover,  involved in the international networks that
some multinationals outsource a substantial   manage the production and trade of interme-
share of manufacturing, because contract      diate goods. Differences in wage levels have led
manufacturers may be better placed than       firms to locate in developing countries those
multinationals to absorb the risk from rapid  portions of the production chain that are in-
product obsolescence (Ernst 2002). Contract   tensive in manual labor, while locating at home
manufacturers that produce components for     the technically skilled labor (such as that in-
well-known multinationals grew rapidly dur-   volved in R&D, management, and marketing)
60



C I1 AN GE S IN   G LO B AL B US IN ES S OR GAN I Z A r I ON
(Hanson, Mataloni, and Slaughter 2001;
Filipe, Fontoura, and Saucier 2002). Multina-     Figure 2.14 Parts and components
tionals operating in developing countries arc     exports grew rapidly, 1981-2000
morc likely to be part of a network (as opposed   (percent per annum)
to supplying the host market) than are multi-     25
nationals in industrial countries. The share of
U.S. affiliate production that is sold back to    20
the United States is more than twice as high      15
for developing countries as it is for industrial  10
countries (Shatz and Vcnables 2000).
Data on parts and components exports,           5
which are a proxy for participation in nct-        0
works, confirm the growing participation of           High-income countries  Developing countries
developing countries.32 Their exports of parts    Source U N Cointrade .ind World Bd.ik
and componients increased by almost 1 8 per-
cenit per year in the 1 980s and by 22.5 percent
in the 1990s (in U.S. dollar terms), almost    Access to networks among developing
three timies more rapidly thaii such exports    countries is highly concentrated
of high-incomile countries In the latter period  Developing countries' participation In global
(figure 2.14). As a result, the share of develop-  networks is highly concentrated, particularly
ing countries in global parts and components    in East Asia. The top five developing-country
exports increased from 4 percent in 1981 to     exporters of parts and components (China,
21 percent In 2000 (figure 2.15). By contrast,  Mexico, Republic of Korea, Malaysia, and
the share of developing countries in exports    Thailand) accounted for 78 percent of devel-
of world   manufactures rose much     morc      oping countries' exports of parts and coplipo-
slowly, from 16 percent in 1 981 to 22 percent  nents, and the next five largest developing
in 2000, while developing countries' share of   countries accounted for about 14 percent (fig-
total trade fell slightly (largely caused by the  ure 2.16). Developing countries outside the top
fall in commodity prices).                      10 made up only about 8 percent. By contrast,
Figure 2.15 Developing countries' share of global parts and component exports rose
between 1981 and 2000
Percent share of parts and components trade, 1981  Percent share of parts and components trade, 2000
Other developing          Developing East Asia  Other developing        Developing East Asia
counltries        rU                          i 3  countries                7
Souirce U N Cointrade and World Banik
61



G LO B A l EC ONO MI C PRO SP EC TS 2 0 0 3
Networks help improve the allocation
Figure 2.16 Developing countries' parts  of resources
and component exports are highly          Global production networks break the pro-
concentrated, 2000                       duction of a given final good into a set of coIn-
(percent share of total)                 stituent activities that vary in the intensity of
30                                       capital, skilled labor, unskilled labor, and
25-                                      other production requirements. Instead of
20                                        making entire products, developing countries
15                                       can be involved in just those stages of products
10                        M L            (for example, labor-intensive stages) that best
5    I    I    I                         suit their mix of endowments. This approach
0o    !   !    !    !   !    !           enables developing countries to shift more
cS   , &   a  9  Zbx   resources to activities in which they have a
o        e} <;   tg  >  c,  soa,4comparative advantage, particularly the fast-
,   ^  be Co9  growing segments that require large labor
inputs in one or more stages of the manufac-
turing chain. Developing countries' participa-
Sourrce U N Cointrade aid World Banik    tion in global networks has enabled those
countries to increase their share of the world's
fastest-growing export products (transistors
and semiconductors, computers, and computer
the top 10 countries accounted for 63 percent  and office machine parts) from 2.4 percent in
of developing countries' total exports ancl  1980 (about the same as the share of those
75 percent of their manufactured exports.   products in global exports) to 16.3 percent by
Thus trade of parts and components is much  1998 (almost 7 percentage points higher than
more concentrated than total or manufac-    the share of such products in global exports)
tures trade. All top 10 countries (except Brazil)  (table 2.3).
either are from East Asia or are participating  Participating in networks may help dampen
in regional arrangements-with the United    the effect of adverse shocks. Multinationals
States or the EU-that provide for low trade  may have an interest in maintaining the oper-
barriers and long-term arrangements to in-  ations of firms with which they have close ties,
crease trade integration. By contrast, countries  either in the form of investment or long-term
that have limited ties to major industrial coun-  contracting relationships. Some authors have
try markets, that lack adequate infrastructure  argued that intra-firm trade is less respon-
(particularly transport facilities) or a suffi-  sive to changes in relative prices than is trade
ciently educated work force, that are subject  between firms, because multinationals will be
to high risks as a result of poor governance or  concerned with the effect of their production
weak institutions, or that have pursued poli-
cies that erode incentives for private sector  Table 2.3  Export activity for product
investment have minimal participation in    groups with the fastest growth in world
global networks. South Asia, Sub-Saharan    exports, 1980-98
Africa, and the Middle East and North Africa  (percent)
together account for only 2 percent of devel-                         1980    1998
oping countries' parts and components
exports (and two-thirds of that amount is   Share in world exports    2 6      9 7
from South Africa and India), compared with  Share in developing-country
11 percent of developing countries' total man-  exports               2.4     16.3
ufactured exports.                          Soturce UNCTAD (2002)
62



C HANG ES IN G LO B AL B US IN ES S OR G A N I Z A I I ( N
decisions on the survival of foreign affiliates  a large share of technical advances. Using in-
(Cho 1990; Helleiner 1978).33 Thus niulti-   ternational patent data, katon and Kortum
nationals may lend funds to subsidiaries suf-  (1999) find that even the major industrial
fering temporary shocks or may provide the   coulitries (the United States, Japan, Germaniy,
backing necessary for them to access credit  the United Kingdomil, aiid France) generally
markets,                                     adopt from one-half to three-fourths of their
Long-term contracts that have been entered  innovations from abroad, and that the United
through networ-ks may help firms survive     States is the only country that clerives most of
severe shocks. For example, exports from the  its growth from its own innovationis (see also
Philippinies maintained double-digit growth  Keller 2001). Because developing countries
rates in 1998, while other countries in the  spend a lot less than industrial coLintries on
region saw outright declines in exports be-  basic researchi, they are even more dependent
cause of the crisis This performance was prin-  on foreign sources of techniology. Thus the
cipally due to the high growth In electronics  potential for increasinig access to techinology
exports (while exports of consunier goods    as a result of participation in trade and FDI
languishcd), and almost all of the Philippines'  may be great (Keller 2002).
electronic exports come from  affiliates of     In part, benefits from the tranisfer of tech-
multinationals. The arrangements in place    nology are directly capturcd by the local firm
meant that a substantial share     of the    or subsidiary participating in a network.
Philippines' production was booked well in   Technology is transferred from the parent to
advance (typically one year), which helped the  a subsidiary, or a local exporter may pLirchase
country maintain output growthi during the   technology as a conditioni of participating
downturn in demand (World Bank 1999).        in a network. One piece of microeconiomilic
evidence consistent with rising intra-firimi
Networks may boost access to                 knowledge transfer is the rising share of
technology-                                  multiniationals' R&D  performed by foreign
Participating in global networks may improve  affiliates. The U.S. Bureau of Economilic
developing countries' access to technology.  Analysis (BEA) reports that, in 1982, affiliates
Multinationials typically possess knowlecige  of U.S. multinationals performed 6.4 percent
assets such as patents, proprietary technology,  of worldwide R&D for these firms. By 1994,
trademarks, and so forth that can be deployed  that share had nearly doubied, to 11.5 per-
in plants outside the parent couLitry (see   cent. This form of technology transfer may
Dunning 1981). Blomstrom and Kokko (1998)    increase domestic productivity, but the benefit
describe how  multinationals typically have  is fully reflected in market prices: the local
proprietary technology that enables them to  subsidiary or indepenidenit firm pays for the
compete against local firms, which presumilably  techniology through profit repatriation or
have superior kniowleclge of local markets and  expenditures on technology. In addition, local
business practices. Approximately 90 percent  firms may absorb technology from networks
of the world's R&D is carried out in five coun-  in ways that are not enti-ely reflected in
tries (the United States, the United Kingdomii,  market transactions (referred to as spillovers;
France, Germany, and Japan) that are among   see discussion in chapter 3). Rodriguez-Clare
the largest source countries for world FDI   (1996) Illustrates how multinational spillovers
flows (Keller 2001).                         from participation in global production net-
works may work: affiliates increase a host
-which may be an important source of         country's access to specialized varieties of
growth potential                             intermediate inputs, the improved knowledge
Access to techinology is particularly important  of which raises the productivity of domestic
for developing countries, which tend to import  producers.
63



G LO BA L EC O NO MI C PRO SP E CT S 2 0 0 3
Networks may help increase the supply       by inward FDI, resulted in a surge in labor
and demand for skilled labor                supply driven largely by the reverse migration
Networks help improve access to technology,  of young Irish people back to Ireland.
which tends to stimulate demand for more-      If the presence of multinationals raises the
skilled workers relative to less-skilled workers.  demand for skilled labor more than the supply,
Increased capital available through FDI may  then wage rates for skilled labor may increase
also increase the demand for skilled workers  relative to those for unskilled labor. This
(see the survey in Hamermesh 1993). Feenstra  change implies a widening of income inequal-
and Hanson (1996) show that the transfer of  ity in countries with a large pool of unskilled
technology and capital accumulation associ-  labor. However, multinationals' demand for
ated with global networks can raise the     labor is likely to raise the level of income of all
demand for more-skilled labor in both indus-  workers, regardless of the effect on relative
trial and developing countries, and Feenstra  wages. Several studies have found that multi-
and Hanson (1997) estimate that FDI into    nationals pay higher wages than do domesti-
Mexico's maquziladoras has contributed to ris-  cally owned establishments, even when con-
ing demand for skilled labor. Slaughter (2002)  trolling for a wide range of observable worker
finds a robust and positive correlation be-  or plant characteristics such as industry,
tween skill upgrading and the presence of U.S.  region, and overall size. The magnitudes in-
foreign affiliates. This correlation is more than  volved are significant. Doms and Jensen
twice as large for the subsample of developing  (1998) document that for U.S. manufacturing
countries when compared with the subsample  plants in 1987, wages in foreign affiliates ex-
of industrial countries.                    ceeded wages in domestically owned firms by
Participation in global production net-  a range of 5 to 15 percent, with larger differ-
works may raise the supply of skilled labor  entials being enjoyed by production workers
in developing countries. One channel can be  than by nonproduction workers.34 The pre-
the short-term activities by which individual  mium  could be accounted for by higher
firms interact with host-country labor markets  worker productivity as a result of multination-
through on-the-job training or support for  als' superior technology or capital. It could
local educational institutions. Multinationals  also be a result of other factors, such as higher
might directly affect labor supplies, because  worker productivity caused by unobserv-
their transferred knowledge might boost the  able worker qualities, or of multinationals
skills of their employees (and, with labor  being more profitable and therefore more able
mobility, the skills of the employees of domes-  to share more rents with workers. Whatever
tic firms as well). They might indirectly affect  the case, the bottom line is that global pro-
labor supplies (for example, by influencing the  duction networks are likely to present high-
educational infrastructure of host countries in  wage opportunities for both more-skilled and
terms of curriculum choices and vocational  less-skilled workers.
training). As Hanson (2001) reports, Intel
recently chose to establish a large assembly  The benefits from networks can contribute
and testing facility in Costa Rica, in part  to growth and structural transformation
thanks to Costa Rica's agreement to expand  Improved allocation of resources, access to
high school training in electronics and English  technology, and increases in skilled labor can,
(see also Moran 2001). Also, to the extent that  in principle, make important contributions
FDI inflows and trade increase the supply of  to raising productivity and to facilitating the
attractive employment opportunities, they   transition from primary commodities to pro-
may inhibit the emigration of more-educated  ducing products with higher value added and
workers to industrial countries. For example,  greater potential for growth. However, the ex-
the 1 990s boom in Ireland, caused in large part  tent of benefits from participation in networks
64



C H A N G E S I N G L O B A L B U S I N E SS  O R G A N I Z A I I O N
is an empirical question. Local firms may not  participation in networks.37 The failure of
capture the benefit from the transfer of tech-  manufacturing value added to rise as a share
nology and increased productivity through    of GDP reflects the rapid rise in services as in-
networks if multinationals have a wide choice  come rises, particularly in these fast-growing
of production locations and a monopsonist    economies. That network participants did
position in the purchase of supplies. In this  achieve significanit structural change is indi-
situation, competition among suppliers may   cated by the rapid fall in the share of agricul-
drive prices down, and the benefits of local  tural value added in GDP, from 17 percent in
firms' productivity improvemenits will accrue  1980 to 10 percent in 2000. These data do not
to the multinational.35                      demonstrate that network participation was a
Some observers have argued that the bene-  major cause of growth and structural change
fits from network participation have been lim-  in these economies, but they do indicate that
ired for most countries, with the important  participation in networks has been consistent
exception of a few of the most successful East  with such progress.
Asian countries (see UNCTAD 2002b). Rising      Sectoral studies indicate that networks have
manufactured exports thtough networks may    enabled countries to move from low-value to
not be accompanied by increased value added  relatively high-value activities. For example,
in manufactures, and network participation   the global apparel industry contains many ex-
may simply mean the continued use of un-     amples of industrial upgrading by developing
skilled labor in low value added activities  countries.38 Several countries have shifted
rather than the development of the manufac-  from assembling apparel from Imported in-
turing sector. For the 20 developing countries  puts (whlich requires only low-wage labor) to
with the largest exports of parts and compo-  filling orders from global buyers. This latter
nents (a proxy for network participation), the  role requires the ability to make samples; to
average share of GDP devoted to manufac-     purchase or manufacture the needed inputs for
tures has shown no increase over the past    garments; to meet international standards in
20 years (table 2.4).36 However, average main-  terms of price, quality, and delivery; and to as-
ufacturing value added (at constant prices)  sume responsibility for packing and shipping
has increased by more than 5 percent per year  the finished items. A few East Asian countries
in these countries, a very respectable perfor-  made this transition in the 1970s, then began
mance over a 20-year period, aiid 2.5 per-   to set up their own international production
centage points more rapidly than the average  networks in the   1980s using   low-wage
of developing coun1tries with limited or no  countries in Asia and elsewhere. Since then,
several countries (for example, India, Mexico,
Romaniia, Turkey, and Vietnani) have devel-
Table 2.4  Rapid growth and structural       oped expertise in managing apparel produc-
change experienced by network                tion chains. Their role is likely to expand
participants                                 greatly as the apparel quotas under the Multi-
"    -. ,  ,,fibre Arrangement are phased out in 2005.
Global production networks have been a
Share of manu ,. GD ,., .                    central feature in the development and upgrad-
added inl GD'             2.      - .
Share of agricuii.wf         -               mig of Asia's large, dynamic electronics sec-
added in GDIP           - -       10 3      tor. While the East Asian newly industrializing
Memotern average annual       - Eeconomies-Honig                Kong   (China), Korea,
growth rate of tmianiutactrLring
valLue added, 1980-2000   5 3               Singapore, and Taiwan (China)-were the first
participants, the major Southeast Asian couI1-
Note Data represent the 20 developinig economies (includinig
Taiwan ICIniial) with largest exports of parts and componeints  tries of Indonesia, Malaysia, the Philippinies,
Souirce World Bank staff                     and Thailand have taken places directly below
65



G LO B AL EC O NO MI C PRO SP EC TS 2 0 0 3
them in the production chain, including work-
ing in design and setting up their own1 produc-  Figure 2.17 Strong rule of law attracts
tion networks. More recently, China has been  foreign investors
cvolving from a provider of low-wage, assem-  (FDI)
bly operations to the leading producer of elect-  3
ronics across a wide range of industries       2
(Borrus, Ernst, and Haggard 2000). Finally,     1               S       **
exporters of fresh vegetables in Kenya and     o,
Zimbabwe have benefited from their relation-  -1* 
ship with U.K. supermarkets, first through    -2*              *   y=03996x + 00117
assistance in meeting production standards,   _3                       (t = 1 90)
and more recently in taking on higher value    -1 5  -1   -05    0   05    1    1 5
added activities within the production chain.                 Rule of law
These activities have included packaging and  Note Partial correlation for developing coLntries
applying barcodes; investing in state-of-the-art  controll iig for size, inicome, openness, inflation, education
applymg Drcoues; nvestmg n state-o-tne-art Source World Bank staff
methods for cold storage; adopting just-in-   Source_World_Bank_staff
time management techniques (including infor-
mation technology) to reduce the time between  law, and effectiveness of the regulatory regime
harvesting, packing, and delivery; and expand-  have been shown to be significant determinants
ing to joint ventures with freight forwarders to  of the location of foreign investment, after con-
gain more control over the distribution process  trolling for other variables (Stein and Daude
(Dolan and Humphrey 2000).                  2002). For example, figure 2.17 shows that a
Of course, participating in networks has  ranking of countries according to the rule of
not always been accompanied by progress to  law (see Kaufman, Kraay, and Zoido-Lobaton
higher value added activities. Survey evidence  2000) is significantly related to the level of FDI
indicates that East Asian firms participating  inflows, after controlling for size, income,
in networks have experienced an increased   openness to trade, inflation, and educational
propensity to innovate as they draw on for-  attainment. By this measure, increasing the rule
eign expertise (as well as increased export  of law by one standard deviation (for example,
growth), compared with firms in the same sec-  from the level of Bangladesh to that of Turkey,
tors that did not participate in networks   or from the level of Turkey to that of Chile)
(World Bank 2002a). However, networked      would raise FDI inflows by 40 percent.
firms did not show faster growth in employ-   Time series analysis underlines the impor-
ment or value added, on average, than non-  tance of governance and institutional quality
networked firms. The World Bank (2002a)     for the allocation of FDI. Countries with better
also found that few East Asian firms were able  investment climates-as indicated by the level
to move up the value chain through participa-  of corruption, voice (political openness), rule
tion in networks. However, these observations  of law, quality of the regulatory regime, gov-
are consistent with countries benefiting from  eniment effectiveness, and political stability-
network participation through spillovers and  tended to receive an increasing share of total
production by multinational subsidiaries.   FDI over the 1990s (figure 2.18).39 The impor-
tance of each dimension of the investment
climate used in figure 2.18 varies considerably.
Good policies attract FDI                   Countries that have strong rankings for regu-
T   he quality of the policy regime is an   latory quality, government effectiveness, or
T  important determinant of the allocation  political instability consistently received more
of FDI flows among developing countries.    than half of all the FDI to developing coun-
Macroeconomic stability, corruption, rule of  tries, with little change in their share of FDI
66



C II A N G E S  I N  G L O B A L  B U S I N  SS   o R G A N I Z A T I o N
Figure 2.18 Foreign investors have been shifting away from weaker investment climate
locations
(percent annual shares of total FDI to developing countnes)
Anticorruption                                       Votce
100                                                  100
75                                                   75
50                                                   50
11110 Il 
o)    )   00 io    0)   00) Q0         00 0)                                 0        ) 00  () 00  0  00,  0  0
0)  0)  00)    o,1   00),    0)                               ,    ,.   i,  '.0- 
Weakest     Below average  Above average             Weakest     Below average  Above average
Rule of law                                          Regulatory quality
100                                                  100
75                                                   75
50                                                   50
25                                                   25
0)       0
Weakest     Below average  Above average             Weakest     Below average  Above average
Government effectiveness                             Political stability
100                                                  100
75                                                   75
50                                                   50
25                       rii                         25
0 
0    4a    00 4o         00 50        0°            0)0   4)   400      4)   4)4      4)°  4°)
Weakest     Below average  Above average             Weakest     Below average  Above average
Soiirce World Bank staff
by the late 1990s when compared with earlier           shares of FDI, the shares tended to decline in
tn the decade. In contrast, the extent of politi-      the latter half of the 1990s. For example, coun-
cal openness has not been strongly associated          tries with below average anticorruption efforts
with the share of FDI received. And although           received 70 percent of developing-country FDI
countries with relatively poor rankings for rule       in 1994, but only 50 percent in 2000, while
of law and anticorruption received substantial         those with above-average ratings doubled their
67



G LO BA L EC O NO MI C     PRO SP EC TS    2 0 0 3
share from 20 percent to just over 40 percent.  centives. Another hypothesis is that investors
If one holds other determinants of FDI alloca-  became more concerned about risk in reaction
tion (including market size, openness, inflation,  to the crises of the late 1990s, and that coun-
and education) constant, corruption is signifi-  tries with weak governance were viewed as
cantly related to the share of FDI in the late  relatively risky. Indeed, risk premiums on junk
1990s, but not in the early 1990s.              bonds and on emerging market debt jumped
Why should the share of FDI going to         sharply toward the end of 1998, indicating
countries with better investment climates have  a shift toward increasing risk in the global
increased during the 1990s? Remember that       environment.
total FDI flows to developing countries in-        The boom in private infrastructure invest-
creased very rapidly during this period. One    ment during the 1990s highlights the impor-
possibility is that countries with high levels  tance of a policy for attracting foreign invest-
of corruption or weak rule of law had other     ment. Private infrastructure investment in
attractions (such as high tariff barriers or in-  developing countries surged during the 1990s,
centives programs) that made them desirable     rising from $14 billion in 1990 to a peak of
locations for investment In particular sectors,  $117 billion in 1997, before easing to $89 bil-
while they remained relatively undesirable      lion by the end of the decade (figure 2.19).40
locations for FDI in general. Countries that    Foreign  investors were involved in some
offered such attractions for FDI were unlikely   80 percent of recorded private infrastructure
to share equally in the FDI boom during the     transactions from  1990 to 1998, although
1990s, which generally responded to the liber-  foreigners accounted for only about 30 per-
alization of economic policies and improve-     cent of the dollar value of total private infra-
ments in macroeconomic stability in several     structure financing (Sader 2000). The boom
countries. That is, countries with poor gover-  in private infrastructure investment responded
nance  may   have attracted   a substantial     to improvements in the investment climate in
amount of FDI during the 1990s because of       several of the largest developing countries.
costly incentives. However, they would be       Privatization programs opened infrastructure
unlikely to attract increasing amounts of FDI   sectors to private investment, and total priva-
unless they were able to continually raise in-  tization  proceeds in infrastructure jumped
Figure 2.19 Private infrastructure investment surged in the 1990s
By region                                    By sector
(billions of dollars)                        (billions of dollars)
140 -                                        140 -
120 -                            ~~~~~~~~~~~~~~120 -
100- LAC,X, 100°~ Telecoms
100  EA                          ~~~~~~~~~~~~~100 - 
80 -         ECA                             8 0 \  -   80 Gas & electric
SA                            ~~~~~~~~~~~~~Transpari
60-          SS      A,    ,                60 -
40-                                          40-
2!0                                         20
0 -
19901991 199219931994199519961997199819992000  19901991 199219931994199519961997199819992000
Note MENA is Middle East and North Africa, ECA is Europe and Cenitral Asia, SSA is Sub-Saharan Africa, EAI' is East Asia and
Pacific, SAS is South Asia, and LAC is Latin America and the Caribbean
Souirce World Bank staff
68



C H A N G E S  I N  G L O B A L  B U S I N E SS O R GANIZATION
from  $10 billion in 1990 to $40 biliton in         Notes
1998, before faliing off sharply to $12 billion        1 A country's location may have an important rolc
in 1999 (World Bank 2001). More gener ally,         in attracting FDI flows For example, Caribbean cooin-
efforts in  several countries to    open   their   tries benefit from their proximity to the United States
economics to trade and investmenit and to              2. Data are from UNCTAD 2001. The data on
establish more stable macroeconomic environ-        cross-border IMf&A already introduced are nor compa-
rable to the data on FDI Fot examilple, NMl&A is re-
ments encouraged the surge in inifrastructuri-e     ported on a transactions basis, while actual payments
investment.                                        that are reported as FDI may be spread ovei several
The decrease in foreign investment in infra-     years Also, the local financing share will be reported as
structure projects among developing countries       part of an M&A transactioni but will not be reported
since 1997 largelv reflects a reduced demand        as FDI. Thus It Is not useful to compare the magnitude
for infrastructure services, owing to the crises    of M&A flows with FDI
in East Asia, Russia, Brazil, and Argentina.           3 Documentation and discussion of those merger
in East                                waves can be found in Golbe and White (1988, 1993),
For example, in East Asia and the Pacific, pri-     Black (2000), Holmstrom and Kaplan (2001), Wlhite
vate infrastructure investmenit collapsed from      (2001), and Pryor (200 Ia)
a peak of $38 billion in 1997 to an average of         4. Analysis based on Evenett (2002)
less than $15 billion per year from    1998 to         5 These data are from White (2001, 2002) The
2000. 1n Latin America, p-ivate infrastructure      number of coiporations refers to 2000 The share of
investment in 1999-2000 was halved from             the top 100 iefers to 1999
the 1998 peak of $71 billion. The drop-off             6 In the Fortune list of the largest 500 global com-
panies in 2000 (as measured by ievenues), 12 were
affected all sectors, with peak-to-trough de-       headquartered in Chlin (iicliding Hong Koig), 11 in
cines of 63 percenit in gas and electricity,        Korea, 3 in Brazil; 2 each In Mexico, Russia, and South
57 percent in transport, and 24 percent in          Africa, anid I eacch in Iidia, Malaysia, and the Repliblica
telecommunications.                                 Bolivariana de Venezuela The iemaininig companiles
Although comprehenisive data are unavail-        were headquarteied in Japan, North America, or
able, foreign investment in developing-country      Westerni Eurcpe
inifrastruIcture projects has likely conitinlued       7 Choosing the appropriate indicator of conceni-
tration is difficult Value added is clearly the superior
to declinie in the past two years. The overall      all-around measure of aggregate concentration, but It
deterioration  in the interinationial economic     is not regularly reported by companiies in their public
enivironimilenit has driven a sharp decline in coil-  financial statemenits or in governmenit data AccouLnt-
mercial bank lending to developing countries       ing profits will depend on depreciation and amortiza-
(net long-term  lendinig from  the banks fell to    noni rates that vary across firms, and on corporate
a negative $32 billion in 2001), anid funids for    income tax rLules that vaiy by country Data oni sales
poetfiniance have dried up. Also, the key in-   will signiificantly distort thie relative imiportanice of e
prolect tiainfrast  ried secrso, u   ties k ey       til firmis (with large ratios of sales to valuLe added) ver-
sus manufacturing firms. Measurinig concentration in
Europe and United States), equipment manu-          terms of assets would imply double counting for finan-
facturers, and specialized venture capitalists      cial intermediaries Moreover, reported asset values
have seen their profits collapse, and in some       would depeid oi alternative accouniting treatments for
cases the firms have gone bankrupt. Most Of         \e&A, changes over time In aCCounlting and tax t[eat-
those firms are under pressure to recapitalize      ment of asset depreciation, amortization, and write-
and are reluctant to devote their limited re-       offs, and changes in the treatment of expensing verSuS
write-off for various categories of costs. To avoid these
sources to high-risk ventures in developing         inconisistenicies and definitional problems, we tise em-
countries  Finally, the scandals involving en-      ployment data to analyze global conicenitration, but we
ergy deregulation and the spectacular losses of     also look at Indicators that are based on profits
privatized telecommunications firms may have           8. White (2001) also reports a rise in aggiegate
reduced support for the deregulation of service     concentration in manufacturing alone from the 1 940s
sectors, a key step toward providing infra-         to the 1980s, and then a decline in the 1990s, based
,  ,                    ~~~~~~~on value added measures, a decchnc in economy-wide
structure services by the private sector. 
aggregate concentration in the 1970s, as showni by
69



G LO B AL   EC O NO MI C     PROS PE CT S 2 0 0 3
employment and profit data; and a decline in aggregate  and retail trade; travel, transportation; storage and
concentration from  the 1950s through the 1980s,    warehousing; telecommunications; banking, finance,
based on assets Somewhat similar conclusions are    and insurance; entertainment; real estate; accounting
reached by Pryor (2001b).                           and auditing; data processing; research and develop-
9. The only antitrust concern that might be raised  ment; law; health; education; public relations; personal
would be that of multimarket contacts among the     assistance (such as auto and house repair, haircutting,
largest companies. For discussions of the potential and  and laundry); and public administration.
actual influence of multimarket contacts, see Feinberg  16. See UNCTAD (1992), table 1.3, p. 18. For
(1985), Rhoades and Heggestad (1985), Bernheim and  Germany and Japan, the initial year is 1976
Whinston (1990), and Evans and Kessides (1994).        17. Services are becoming increasingly interlinked
10. The Forbes "Super 50" list is based on a com-  with goods, especially in high-tech products in which
posite calculation of sales, profits, employment, and  the use of hardware requires various software and
market value.                                       maintenance service contracts.
11 The time series analysis is based on the Forbes  18. Williamson and Mahar (1998) detail moves
list, which provides comparable data from 1994 to   toward liberalization of banking sectors.
2001. The Fortune 500 list was not used because it     19. The importance of foreign bank participation in
included several government-owned businesses. In par-  developing countries has been discussed by several
ticular, it extended in the latter years to a few state-  authors (see Roldos 2002).
owned Chinese companies, thereby distorting the        20 See World Bank (1997) for a discussion of the
comparison with the mid-1990s.                      globalization of production and the developing
12. As stated above, calculations of aggregate con-  countries.
centration should not be based on sales data because of  21. The share declined from 1996 to 1999 because of
the wide range of ratios of sales to value added found  the abolition of tariffs under the EU's Association Agree-
in different corporations. But this calculation is based  ments, which resulted in companies switching from EU
on changes over time, and presumably differences in  to Eastern European firms for intermediate inputs.
the growth rates of sales and value added are not as   22. This is a narrow definition of the share of trade
disparate as the levels.                            conducted through networks, it excludes imported
13. World Bank computations are based on data   inputs that are processed and sold as a final good in
from the U S Bureau of Census and the Japan Ministry  the domestic market.
of International Trade and Industry.                   23. Hummels, Ishii, and Yi (2001) attributed one-
14. Even if data were available, global trends in the  third of the growth in world trade over the past
number of companies in major oligopolistic industries  25 years to trade in parts and components, rather than
would provide only limited information concerning   trade in final goods.
changes in the degree of competition. On the one hand,  24. Containerized shipment allows for better track-
declining numbers of firms may be consistent with ris-  ing of cargo, more efficient and reliable port services,
ing competition, as lower transportation and commu-  and greater ease of switching to land transport.
nication costs enable formerly regional firms to enter  25. The volume of cargo shipped by airlines
global markets. On the other hand, little change in the  increased by 6 percent per year from 1978 to 1998,
number of firms may be consistent with reduced com-  and the share of revenue from international cargo in
petition (for example, resulting from strategic alliances  total air shipments rose from about 40 percent to well
with the goal of coordinating prices or sharing out  over half (Air Transport Association 1999).
markets) (OECD 2001).                                  26. Deardorff (1998) points out that tariffs can
15 Services are products that are to a large extent  either deter or stimulate participation in global net-
intangible, nonstorable, and nontransportable. Intangi-  works, depending on where they are imposed and
bility implies that the quality of services is uncertain  whether they are on final or intermediate goods.
because of their high and variable human content and   27. The easing of restrictions on FDI flows in
"one-off" nature of production Therefore, services  developing countries has been discussed in various
generally require proximity and close interaction   editions of Global Development Finance
between the producer and the consumer to ensure a      28. Chen (1996) lists alternative forms of prod-
satisfactory level of quality. Nonstorability and non-  uction relationships, including wholly owned affili-
transportability imply that services must be produced  ates, joint ventures, foreign minority holdings, fading-
and consumed at the same time and at the same loca-  out agreements, licensing, franchising, management
non However, some services can be embodied in goods  contracts, turnkey ventures, contractual joint ventures,
or stored and transmitted through electronic means.  and subcontracting  See also Grosse (1996) for an
Services include such economic activities as wholesale  alternative categorization.
70



C 11 A N G E S  I N  G l O B A L  B U S I N E SS  O R G A N I Z A T I O N
29. FDI requires some advantage by the multina-     37  Excludinig Korea, Malaysia, and Taiwan
tional over home production to compenisate for domes-  (China) from  this set reduces the group's average
tic firms' superior knowledge of local markets, con-  growvth rate of manufacturing value added to 4.3 pei-
sumer preferences, and business practices (Blomstrom  cent per year, still much higher than other developing
and Sioholm 1998)                                   countiies.
30. The example is given of a Singaporeani com-     38  General information on apparel is mainly
pany in which some technology was transferred to a  draviv from Geieffi (1999) and the apparel chapters
local subcontractor, but critical components were  in Gereffi and Korzeniewicz (1994) and Gereffi and
unlikely to be outsourced.                          Kaplinsky (2001)
31 One concern is whether these data are distorted  39. The countries are classified into three cate-
because multinationals may not report "true" prices of  gories the worst group (more than half a standard
goods traded amonig affiliates, but they will instead  deviation from the average), below average (half a stan-
increase the price (and therefore profits) of goods from  dard deviation oi less from the average), or above
low-tax locations and will reduce the price of goods  average
from high-tax locations. There is some evidence that   40. Technological inn1ovation also helped boost
U S firms have followed this practice; however, the  investment in iifrastructure over the I 990s For exam-
oveiall patterns of prices are similar to the pricing of  ple, flows to the telecommunications sector rose with
goods tiaded between firms (Clausing 2001) Thus     the dramatic reductions in the price of long-distance
these data may provide a reasonably accurate picture  service.
of trends in intra-firm trade
32 These data refer to product categories identified
as parts and components in the Standard International
Trade Classification (SITC) Revised 2 svstem. This  References
trade is a proxy for, but is not identical to, production  Air Transport Association. 1999. "Aviation Economic
through networks On the one hand, such trade may         Impact "  http:/fwww air-transport org/public/
also reflect the export of relatively undifferentiated   industry
inputs to arm's-length purchasers Conversely, many  Aitken, Brian, Ann Harrison, and Robert E Lipsey
goods that are parts to consumer products are not        1996 "NWages and Foreign Ownership: A Coni-
identified as such in the SITC system. On balance, the  parative Study of Mexico, Venezuela, and the
data probably understate the extent of trade through     Umted States " Jottntal of International Econom-
networks (see Kaminski and Ng 2001)                      irs 40()May) 345-71
33 Howevei, Rangan and Lawi ence (1999) argue,   Arndt, Sven W 2001 "Productioni Networks in an Eco-
on the one hand, that the costs of search and assess-    nomically Integrated Region " ASEAN Econzomiiic
ment of reliability involved in choosing suppliers and   Bulletin, Singapore, 18(1): 24-34
outlets will mean that even arm's-lenigth relationshlips  Baldone, S , F Sdogati, and L. Tajoh 2002 "Interna-
can be relatively insensitive to changes in relative prices  tional Fragmentation of Production, Trade Flows,
in the short term  On the orher hand, multinationials   and Growvth " 53rd Conference Paris March
face smaller search and assessment costs because of      14-17
gleater international experience, so they are moIC  Bernheim, B. Douglas, and Michael D   Whinston
likely to switch pioduction iapidly in response to iela-  1990. "Multimarket Contact and Collusive
tive exchange rate changes They provide some empir-     Behavior,"   Ranid  Journal  of  Economics
ical support for this view                              21 (Spring) 1-25.
34 For additional U.S. evidence, see Howenstinie  Black, Bernard S. 2000. "The First International
and Zeile (1994) Griffith (1999) piesents similar       Merger Wave (and the Fifth and Last U S.
evidence for the United Kingdom, Globerman, Ries,       Wave) " Unziversitt of Miami Lawv Revierv 54
and Vertinsky (1994), for Canada, Aitken, Harrison,     799-818
and Lipsey (1996), for Mexico and the Repiblica     Blomstrom, Magnus 1986 "Foreign Investment and
Bolivariana de Venezuela, and Te Velde and Miorissey    Productive Efficiency  The Case of Mexico"
(2001), for five African countries                      Journal of Indistrial Economics 35(1) 97-110.
35. Conceivably, all sectors may be peifectly coin-  Blomstrom, Magnus, and An Kokko 1998. "Multina-
petitive, and the benefits of increased productivity will  tional Corporations and Spillovers " Journal of
accrue to consumers                                      Economic Surveys 12(2) 1-31
36 This calculation excludes a few network partic-  Blomstrom, Magnus, and Fredrik Sloholm  1998.
ipants that lack adequate time series data on maniufac-  "Technology, Transfer, and Spillovers Does Local
turing value added                                      Participation  vith iMtultinationals  Mlatter?"
71



G LO B AL EC ONO MI C       PRO SP E CT S    2 0 0 3
Centre for Economic Policy Research. Discussion  Ernst, Dieter. 2002. Global Production Networks in
Paper Series no. 2048(December): 1-16. London.    East Asia's Electronics Industry and Upgrading
Borrus, Michael, Dieter Ernst, and Stephan Haggard,   Perspectives in Malaysia. Honolulu: East-West
eds. 2000. Iiiternational Production Networks in  Center. Processed.
Asia. London: Routledge.                     Ethier, Wilfred J , and James R. Markusen. 1996. "Multi-
Campa, J., and L. Goldberg. 1997. "The Evolving Ex-   national Firms, Technology Diffusion, and Trade."
ternal Orientation of Manufacturing Industries:  Journal of International Economics 41 1-28
Evidence from Four Countries." National Bureau  Evans, William N., and loannis N  Kessides 1994
of Economic Research (NBER) Working Paper         "Living by the 'Golden Rule': Multi-Market
no 5919. Cambridge, Mass.                         Contact in the U S. Airline Industry." Quarterly
Chen, Edward K. Y. 1996. "Transnational Corpora-      Journal of Economics 109(May) 341-66.
tions and Technology Transfer to Developing  Evenett, Simon J. 2002 The Cross-Border Mergers
Countries." In Transnational Corporations and     and Acquisitions Wave of the Late 1990s. Bern,
World Development London    Routledge (on        Switzerland: World Trade Institute Processed.
behalf of the UNCTAD Division on Transnational  Feenstra, Robert C. 1998. "Integration of Trade and
Corporations and Investment), International       Disintegration of Production in the Global
Thompson Business Press                           Economy." Journal of Econoniic Perspectives
Cho, Kang Rae 1990 "The Role of Product-Specific      12(4). 31-50
Factors in Intra-Firm Trade of U S Manufacturing  Feenstra, Robert C., and Gordon H. Hanson. 1996.
Multinational Corporations " Journal of Interna-  "Foreign Investment, Outsourcing, and Relative
tzonal Business Studies 21(2): 319-30.            Wages " In Robert C. Feenstra, Gene M. Grossman,
Clausing, Kimberly A. 2001. "The Behavior of Intra-   and Douglas A Ihwin, eds., Political Economy of
Firm Trade Prices in the U.S International Price  Trade Policy. Essays in Honor of Jagdish Bhagwati.
Data." Bureau of Labor Statistics Working Paper   Cambridge, Mass.: MIT Press, pp. 89-127.
no. 333. Office of Prices and Living Conditions,     1997. "Productivity Measurement and the
Washington, D C. January.                         Impact of Trade and Technology on Wages Esti-
Deardorff, Alan V. 1998. "Fragmentation in Simple     mates for the U S., 1972-1990 " National Bureau
Trade Models." School of Public Policy. Research  of Economic Research (NBER) Working Paper
Seminar on International Economics. Discussion    no 6052 Cambridge, Mass
Paper no 422. University of Michigan, Ann Arbor.  Feinberg, Robert. 1985. "Sales-at-Risk  A Test of
DeGrauwe, Paul, and Filip Camerman. 2002 "How         Mutual Forbearance Theory of Conglomerate
Big Are the Big Multinational Companies>"         Behavior." Journal of Business 58(April): 225-41.
United Nations Conference on Trade and Devel-  Filipe, Joao Paulo, Maria Paula Fontoura, and Philippe
opment Geneva Processed.                          Saucier. 2002 "U.S. Intra-Firm Trade: Sectoral,
Dolan, Catherine, and John Humphrey. 2000. "Gover-    Country, and Location Determinants in the 90s."
nance and Trade in Fresh Vegetables: The Impact   International Atlantic Economic Conference
of U.K. Supermarkets on the African Horticulture  Paris. March 13-17.
Industry " Journal of Development Studies 37(2).  Gereffi, Gary. 1999. "International Trade and Indus-
147-75 (December).                                trial Upgrading in the Apparel Commodity
Doms, Mark E, and J. Bradford Jensen 1998. "Com-      Chain " Journal of International Economics
paring Wages, Skills, and Productivity Betveen    48(1). 37-70 (June).
Domestically and Foreign-Owned Manufacturing  Gereffi, Gary, and Raphael Kaplinsky, eds. 2001 The
Establishments in the United States." In Robert   Value of Value Chains: Spreading the Gains from
Baldwin, Robert Lipsey, and J. David Richardson,  Globalisation Special issue of the IDS Bulletin
eds., Geography and Ownership as Bases for        32(3).
Econoniic Accounting. Chicago: University of  Gereffi, Gary, and Miguel Korzeniewicz, eds. 1994.
Chicago Press, pp. 235-55.                        Commodity Chains and Global Capitalism.
Dunning, John H. 1981. International Production and   Westport, Conn Praeger
the Multinational Enterprise. London- George,  Globerman, Steven, John C. Ries, and Ilan Vertinsky.
Allen, and Unwin.                                 1994. "The Economic Performance of Foreign
Eaton, Jonathan, and Samuel Kortum. 1999 "Interna-    Affiliates in Canada." Canadian Journal of Eco-
tional Technology Diffusion: Theory and Mea-      nomics 27(1): 143-56
surement " InternationalEconomic Review 40(3):  Golbe, Devra L., and Lawrence J. White. 1988 "A
537-70.                                           Time Series Analysis of Mergers and Acquisitions
72



C H A N G E S I N  G L O   B A  B U u S I N E SS  O R C. A N I Z A 1I 0 N
in the U.S Economy." In Alan J. Auerbach, ed ,    facturing Establishments." Survey of Current
Corporate Takeovers: Causes and Consequences      Business 74(1). 34-59 (January)
Chicago University of Chicago Press, pp 265-  Hummels, D, J Ishii, and K.-M. Yi. 2001   "The
302.                                              Nature and Growth of Vertical Specialization in
1993 "Catch a Wavc The Time Seties Behav-       World Trade." Journal of International Econom-
ior of Mergers " Review of Econonics and Statis-  ics 54 75-96.
tics 75(August). 493-99.                      Hummels, D., D   Rappaport, and K.-M. Yi 1997
Griffith, Rachel 1999 "Using the ARD Establishmenit    "Verical Specialization and the Changing Nature
Level Data to Look at Foreign Ownership and        of World Trade." Federal Reserve Bank of New
lroductivity in the U K " Econonic journal         York Economic Policy Reviewv (June): 79-99.
109(June). F416-F442.                         Kamioiski, Bartlomiel, and Francis Ng 2001. "Trade
Grosse, Robert. 1996  "International Technology        and Production Fragmentation Central European
Transfer in Services." Journal of International    Economies in EU Networks of Production anid
Business Stidies Fourth Quarter 781-800           Marketing." World Bank Policy Research Work-
Grossman, Gene M, and Elhanana Helpinan. 2002.        ing Paper no 2611. Washingtoni, DC
"Outsourcing in a Global Econonmy." National  Kaufman, Daniel, Art Kraay, and Pablo Zoido-
Bureau of Economic Research (NBER) Working         Lobaton  2000  "Governance Matters, from
Paper 8728. Cambridge, Mass. January.              Measuremcnt to Action." Finance and Develop-
Hamernicsh, Daniel S. 1993 Labor Demand. Princeton,    nient 37(2) (June)
N J Princeton University Press.               Keller, Wolfgang  2001. "The    Geography  and
Hanson, Gordon H   1996. "Localization Economies,      Channels of Diffusion at the World's Technology
Vertical Organization, and Trade." Amierican       Frontier" National Bureau of Economic Research
Economic Review 86(5): 1266-78 (December)         (NBER) Workinig Paper 8150. Cambridge, Mass
2001. "Should Countries Promote Foreign        March
Direct Investment?" United Nations Conference       . 2002. "International Technology Diffusion."
on Trade and Development G24 Discussion Paper      University of Texas, Austin Processed.
Series No. 9. Geneva.                         Kimura, Fukuniari. 2001 "Fragmentation, Internaliza-
Hanson, Gordon H., Raymond J. Mataloni Jr., and        tion, and Interfirm Linkages. Evidence from the
M4atthew J Slaughter 2001 "Expansion Strate-       Micro Data of Japanese Manufacturing Firms In
gies of U.S. Multinational Firms " In Dani Rodrik  Leonard K  Cheng, and Henryk Kierzkowski,
and Susan Collins, eds , Brookings Trade Forum    eds , Global Production and Trade in East Asia
2001. Washington, D C. Brookings Institution,      Boston, Dordrecht, and London: Kluwer Acade-
pp 245-94                                          mic Publishers
2002   "What Drives Production Sharing     Markusen, James, and Anthony Venables    1998
Within Multinational Firms?" Processed.            "Multinationial Firms and the New   Trade
Helleiner, Gerald K. 1978. "Transnational Corpora-     Theory." Journal of International Economics 46
tions and Tradc Structure: The Role of Intra-Firm  183-203.
Trade." In Herbert Giersch, ed., Oni the Economn-  Markusen, James, Thomas F Rutherford, and David
ics of Intra-lndustry Trade Tubingen, Germany     Tart. 2000. "Foreign  Direct Investmenit in
J C B. Mohr, Paul Siebeck, pp. 159-81              Services and the Domestic Market for Exper-
Holmstrom, Bengt, and Steven N. Kaplan 2001. "Cor-     rise." National Bureau of Economic Research
porate Governance and Merger Activity in the      (NBER) Working Paper no 7700. Cambridge,
United States Making Sense of the 1980s and       Mass. May
1990s " Journal of Economic Perspectives      Mataloni, Raymond J., and Daniel R. Yorgason. 2002
15(Spring). 121-44                                "Operations of U.S. Multinational Companies
Hoon, Hian Teck, and Kong Weng Ho. 2001 "Arm's-        Preliminary Results from the 1999 Benchmark
Length Transactions versus Affiliates: A Study of  Survey." Survey of Current Btusiness. Washington,
Two Electronic Component Firms in Singapore."     D C: U S Department of Commerce March.
In Leonard K. Cheng and Henryk Kierzkowski,   Moran, Theodore 2001. Parental Supervision' The
eds., Global Production anid Trade in East Asia.  Netv Paradignm for Foreign Direct Investment
Boston, Dordrecht, and London' Kluwer Acade-      anid Development. Washington, D.C.. Institute for
mic Publishers                                    International Economics August
Howenstine, Ned G., and William J. Zeile. 1994.   Navaretti, Giorgio Barba, Jai 1. Haaland, and Anthony
"Characteristics of Forcign-Owned U.S. Manu-      Venables 2000. Multinationzal Corporations and
73



G LO BA L EC ONO MI C       PRO SP EC TS 2 0 0 3
Global Production Networks. The Implications  Slaughter, Matthew. 2002. U.S. Trade and Investment
for Trade Policy. Report prepared for the Euro-   Policy and the Growth of Infornation Technol-
pean Commission Directorate General for Trade.    ogy. Washington, D.C.. Emergency Committee
London: Centre for Economic Policy Research.      for American Trade
Ng, Francis, and Alexander Yeats 2001. "Production  Stein, E., and C. Daude. 2002. Institutions, Integra-
Sharing in East Asia Who Does What for Whom       tion, and the Location of Foreign Direct Invest-
and Why?" In Leonard K. Cheng, and Henryk         ment. Washington, D.C.: Inter-American Devel-
Kierzkowski, eds., Global Production and Trade    opment Bank Research Department. Processed.
in East Asia. Boston, Dordrecht, and London:  Te Velde, Dirk Willem, and Oliver Morrissey. 2001
Kluwer Academic Publishers                         "Foreign Ownership and Wages: Evidence from
OECD (Organisation for Economic Co-operation and       Five African Countries." Centre for Research in
Development). 2001. New Patterns of Industrial    Economic Development and International Trade
Globalisation: Cross-Border Mergers and Acqui-    Discussion Paper University of Nottingham,
sitions and Strategic Alliances Paris.            United Kingdom.
Pryor, Frederick L. 2001a. "Dimensions of the World-  UNCTAD (United Nations Conference on Trade and
wide Merger Wave." Jour,zal of Economic Issues    Development). 1992. World Investnient Report.
35(4). 825-40.                                    New York.
-   2001h. "Will Most of Us Be Working for         2001 World Investmnent Report. New York.
Giant Enterprises by 2028?" Journal of Economic     . 2002a  "FDI Downturn in 2001 Touches
Behavior and Organization 44(April). 363-82.      Almost All Regions " January 21. Press Release.
Rangan, Subramanian, and Robert Z. Lawrence.            . 2002b. Trade and Development Reports.
1999. "Search and Deliberation in International   Geneva.
Exchange: Learning from  Multinational Trade  United Nations. 1989. Foreign Direct Investment and
about Lags, Distance Effects, and Home Bias "     Transnational Corporations in Services. New
National Bureau of Economic Research (NBER)       York, p. 8.
Working Paper 7012. Cambridge, Mass.          Venables, Anthony 1999. "Fragmentation and Multi-
Rhoades, Stephen A., and Arnold Heggestad. 1985        national Production " European Economic Re-
"Multimarket Interdependence and Performance      view 43(5): 935-45.
in Banking- Two Tests." Antitrust Bulletin 30:  White, Lawrence J. 2001. "What's Been Happening
975-95.                                           to Aggregate Concentration? (And Should We
Rodriguez-Clare, Andres. 1996  "Multinationals,        Care?)" Working Paper no. EC-02-03. Stern School
Linkages, and Economic Development " American     of Business, New York University. December 3
Economic Review 86(4): 852-73 (September)           . 2002a. "Trends in Aggregate Concentration in
Roldos, Jorge. 2002  "FDI in Emerging Market           the United States " Journral of Economic Perspec-
Banking Systems." International Monetary Fund,    tives 16(Summer)
Washington, D.C. Processed                          . 2002b. "Trends in Global Aggregate Concen-
Sader, Frank. 2000 "Attracting Foreign Direct Invest-  tration." Background paper for Global Econom-
menlt Inlto Infrastructure Why Is It so Difficult?"  ics Prospects 2002. Processed
Foreign Investment Advisory Service Occasional  Williamson, John, and Molly Mahar. 1998. "A Survey
Paper no 12. World Bank, Washington, D.C.         of Financial Liberalization." Princeton Essays in
Sauvant, Karl P., and Padma Mallampally. 1996.         International Finance no 211. Princetoni Univer-
"Transnational Corporations in Services." In      sity, Princeton, N.J.
Transnational Corporations and World Develop-  World Bank. 1997. Global Economic Prospects. Wash-
ment Geneva: Routiedge (on behalf of the UNC-     ington, D C
TAD Division on Transnational Corporations and      . 1999. Philippines- The Challenge of Econoniic
Investment), International Thompson Business      Recovery. East Asia and Pacific Regional Office.
Press                                             Washington, D.C.
Schive, Chi, and Regina Yeu-Shyang Chyn 2001.           . 2001  Global Developmeent Finanice 2001
"Taiwan's High-Tech Industries." In Leonard K.    Washington, D.C.
Cheng, and Henryk Kierzkowski, eds., Global Pro-  -    2002a. East Asia's Future   Economy
duction and Trade in East Asia. Boston, Dordrecht,  Washington, D C.
and London: Kluwer Academic Publishers              . 2002b. Global Developmieiit Finance 2002.
Shatz, Howard, and Anthony J. Venables. 2000. "The     Washington, D.C.
Geography of International Investment." World  WTO  (World Trade Organization). 1998. Annual
Bank Research Paper. Washington, D C.             Report. The World Trade Organization. Geneva.
74









Domestic Policies to Unlock
Global Opportunities
Globalization makes it increasingly        spected property rights-encourages invest-
important to get the "investment           ment and promotes economic growth. Many
climate" right-                            countries try to use specific investment poli-
Expanding global scrvice and production net-  cies, such as tax incentives, to attract invest-
works can accelerate growth in developing  ment or to channel it in particular directions.
countries that successfully harness competition  Such schemes are often poorly designed, mad-
to encourage efficient investment. Efficient in-  equately implemented, and costly, and may
vestment does not simply mean more invest-  largely benefit investors who would have in-
ment. In fact, recent research demonstrates  vested anyway.
surprisingly little short-run correlation be-
tween investment levels and growth (Easterly  -and promoting competition that will
1999). Instead, investment and its productivity  increase the productivity of private
are inextricably linked to domestic policies  investment
that, taken together, broadly make up the local  In many countries, policy and private barriers
investment climate.                        either have discouraged private investment or
Sound enabling policies-including good   have channeled it into less-productive activi-
governance, institutions, and property rights-  ties that reduce economic growth. Promoting
can help attract more private investment, both  a positive investment climate, however, does
domestic and foreign. Policies that promote  not imply a laissez-faire approach to the econ-
competition and entrepreneurship increase the  omy. Rather, it requires active government
efficiency of that investment. Complementary  efforts to reduce barriers that stifle entrepre-
public investment, mcanwhile, further con-  neurship and competition. Four policy barri-
tributes to overall productivity growth. Taken  ers to competition are especially common:
together, sound policies in these three areas  barriers to trade, restrictions on foreign
contribute to a positive investment climate,  investmenits, administrative barriers to entry
which is essential to accelerating growth and  and exit, and monopoly positions granted to
reducing poverty (Stern 2001).             state-owned enterprises (SOEs) and newly
privatized firms. While pnivatization has usu-
-including having an enabling policy       ally improved the performance of divested
framework-                                 firms, shortcomings abound in subsequent
A stable macroeconomic environment is essen-  industrial performance. Those shortcomings
tial for a country to realize its investment  may be associated with regulations that
potential. Good public governance-including  reduce competition and grant exclusivity be-
transparent rules, low corruption, and re-  fore sale of the enterprise. In addition, private
77



G LO BA L E CO NO MI C       PRO SP EC TS 2 0 0 3
barriers  to  competition-including      price-     Investment climate and investment
fixing and other collusive practices-can in-        policies
duce resource misallocation. After establishing     11V,hlie foreign     direct investment (FDI)
an adequate macro policy framework, coun-                  flows to developing countries receive
tries that lower both policy barriers and pri-      much attention and have special characteris-
vate barriers to competition can usually mini-      tics that can benefit recipients, most invest-
mize investment distortions. They can also see      ment in these economies remains domestic in
more capital inflows, more rapid growth in          origin (figure 3.1).1 This fact highlights the
trade, and superior overall performance.            importance of policies likely to affect the level
and productivity of all investment, not just
Public investment plays a critical role             foreign. Since the mid-1980s, the share attrib-
in increasing productivity                          utable to   public investment has remained
The level and composition of public invest-         fairly constant, while private domestic invest-
ment has changed over the past two decades.         ment has declined slightly as FDI has grown.
The wave of privatizations has reduced the
level and scope of public investment through        Governance, corruption, and property
state enterprises, and    many    sectors once      rights matter-
thought to be natural monopolies can now            One critical dimension of the domestic policy
be exposed to competition. Public resources         environment is whether the         government
formerly used to subsidize loss-making SOEs         operates with transparency, credibility, and
can potentially be used where the private sec-      stability. Good governance-including inde-
tor is unlikely to invest enough: education,        pendent agencies, mechanisms for citizens to
rural roads, and expanded access to under-          monitor public behavior, and rules that con-
served areas in many networks. While always         strain corruption-is essential to development
a challenge, investment in effective infrastruc-    (World Bank 2002b). Barro (1991) finds a pos-
ture and human capital projects has an espe-        itive relationship between growth and mea-
cially high return.                                 sures of political stability for 98 countries from
Figure 3.1 Domestic capital is the largest source of investment in developing countries
(gross fixed capital formation, percent of GOP)
30
25 _
20 
Private investment
10
6 _                                                                    Public Investment
0 -                 .   .   .    .    .   . .                            I    I l
1975      1978       1981      1984       1987      1990       1993      1996       1999
Note GDP is gross domestic product. These are the aninlual averages for I 1 J developing countries Private investment is calculated
as the difference berween gross fixed capital formation and the sum of public investment and FDI Public investment data measure
total public investment, including SOEs
Souirce World Bank anid International Monetary Fund data, and Everharr and Sumlinski (2001)
78



D O M E S T I C P O L I C I E S  r o U N L 0 C K G l O r A l O P P o R T U N I I I 1: S
1960 to 1985. For example, as discussed in  licenises and permits. Recenit empirical re-
chapter 2, countries with stronger rule of law  search confirms that measures of corruption
see imore FDI (figure 2.17).                are significanitly and negatively related to
Transparency is among the most important  FDI inflows (Smarzyniska and Wei 2000; Wei
components of the domestic enabling environ-  2000). Lipsey ( 1999) observes a strong nega-
ment. Transparenicy relates to both the actions  tive correlation between corruption and the
taken by authorities and the broader business  location choice of U.S. affiliates across Asian
environmient of the host country. A nontrans-  countries.2 Hausinann and Fernanidez-Arias
parcnt business environment increases the   (2000) find positive, albeit weak, evidence
cost of inforniationi, diverts corporate energies  that FDI as a share of gross domestic product
toward rent-seeking activities, and can be con-  (GDP) increases with institutional quality.3
ducive to corruption. Case studies suggest that  Corruption and poor governanice often go
companiies may, for example, be willing to  hand in hand with lack of investor protections
invest in countries with legal and regulatory  and with poorly functioning institutions,
frameworks that would not otherwise be      thereby deterring competitioni and investment
considered "investor friendly," provided the  No investor-domestic or foreign-is likely
investors can obtain a reasonable degree of  to risk assets if there is a highi probability that
clarity about the environment in which they  those assets will be arbitrarily seized. Security
will be operating. Conversely, extremely    of private property helps ameliorate asymmet-
opaque business conditions can deter virtu-  ric ilforimiatioi between inivestors and the guy-
ally all private investimient, regardless of the  ernment and reduces investor uncertainties,
extent of the incenitives.                  thus reducing risk premiums and the overall
While these factors affect all participants in  cost of doing business. Empirical literature
the host country's business sector, they are  provides unamilbiguous support for this basic
arguably more discouraging to outsiders who  point, finding that the institutions protecting
are not privy to locally available information  property rights are among the most critical
and who have other choices about where to   for growth (Knack and Keefer 1995), that
invest. As with earlier relations, causality can  productivity and economic growth will in-
run both ways, because FDI may coiitribute to  prove whien governments impartially protect
creating a more transparent environmcnt.    and define property rights (Claugue and
There are cases in which a foreign corpo-   others 1999), and that countries without
rate presence encouraged more open govern-  adequate property rights are likely to grow
inent practices, raised corporate transparency,  more slowly (Zak 2001). Moreover, historical
and energized the fight against corruption.  evidence from  industrial countries suggests
More generally, by observing commonly       that when ilivestors face the threat of asset
agreed standards such as those in the Conven-  expropriation, they are likely to charge much
tion on Combating Bribery of Foreign Public  higher prices to recoup investments quickly-
Officials, implemented by the Organisationi  if they choose to invest at all (Keefer 1996;
for Economic Co-operation and Develop-      Wallsten 2001c).
ment (OECD), multinational firms can con-
tribute to raising standards for corporate  -but policies to channel private
social responsibility in host countries.    investment warrant caution-
Corruption can deter foreign investors by  Building a strong and stable investment cli-
increasing transaction costs and by raising  mate is neither easy nor quick. Governments
uncertainty regarding the enforcement of    may hope to jump-start the process or to coni-
contracts, the predictability of operating  pensate for a poor investment climate through
costs, and the likelihood of obtaining needed  targeted policies intended to draw investors
79



G LO BA L    E CO NO MI C     PRO SP EC TS       2 0 0 3
(usually foreign). Similarly, governments may         tries each try to give the biggest incentives to
compete    for foreign    investment in    higher      investors. In this section, we will consider
value added industries as a way of moving up          three common policies: tax incentives to en-
the technology hierarchy of international trade       courage FDI, subsidies to promote industrial
and production. The lure of targeted policies          "clusters," and measures to encourage indus-
is clear: incentives can be legislated quickly,       trial development through export processing
and investment that occurs after the incentives       zones (EPZs).
are in place can be touted as a success. While
actual success stories exist, they tend to be the     Tax incentives for FDL      Given the perceived
exception rather than the rule because a com-         benefits of FDI, many countries have explicit
bination of design flaws and implementation           policies to attract it. One recent study esti-
failures could limit the hoped-for response.          mated that 1 16 countries take a proactive
Moreover, such     schemes can     be expensive,      approach to FDI and offer incentives to for-
with the risk that costs will outweigh        any     eign investors (Moran 1998). Figure 3.2 illus-
benefits, that incentives will merely transfer        trates the variety and frequency of fiscal and
money to private investors who would have             financial incenitives for FDI that developing
invested anyway, and that incentives can lead         countries offer. Typically, these policies focus
to a "race to the bottom" as developing coun-         on attracting particular types of investment-
Figure 3.2 Incentives for FDI are varied and numerous
Subsidized loans    }
VAT exemption on exported Inputs
Reduction in local, municipal taxes/duties
Loss write-off
Export income treated preferentially _
Duty drawback
VAT exemption for raw materials
Raw material import duties exempted d
Accelerated depreciation  -            -ECD countries
VAT exemption for capital goods  -Developing countries
Lower tax rate
Investment/reinvestment allowance
Tax exemption/holiday
Capital goods import duties exempted
0            10          20           30           40           50
Number of countnes
Note VAT is value added tax Data on fiscal and financial incentives wvere compiled for 71 developing and 20 OECD countries The
most common incentives (used in at least 18 percent of developing countries) are shown in the chart
Source Bora (2002)
80



D O M E S T I C PO LI C I ES TO UN LO C K G LO B AL O P PO R r IUN I I ES
or changing investors' conduct-rather than  what is meant by an incentive. It is imiportant
on improving the general invcstment climate.  to distinguish betwccn the fiscal and financial
Incentives designed to lure FDI can take the  incentives (which are usually firm specific)
forn of up-front subsidies that are designed to  and the more general policies that promote
help multinationals defray some of their fixed  business activity. Evidence is uncontested that
costs (financial incenitives), tax holidays (fiscal  general policies matter a lot in attracting In-
incentives), and other grants. The main goal of  vestment. In a recent empirical analysis of the
such policies is to alter either the magnitude or  effect of U.S. state-level policies on the loca-
the location of inward FDI.                 tion of manufacturing investment, Holmes
There are three main categories of FDI    (1998) found that the manufacturing share of
incentives: fiscal-policies that are designed  employment in states with a pro-business reg-
to reduce the tax burden of a firm (including  ulatory environment is one-third greater than
loss writeoffs and accelerated depreciation);  that in a bordering state without that environi-
financial-direct contributions to the firm  ment. Policies that encourage the adoption
from the government (including direct capi-  and adaptation of know-how-and other gen-
tal subsidies or subsidized loans); and others  eral incentives that apply across the hoard-
that do not fall easily into either category. In  are important and help foster a sound enabl-
contrast to the industrial world, where the in-  ing environm1ent. Examples include effective
centives offered are usually financial, the over-  enforcement of contracts, absence of red tape,
whelming majority of developing-country     adequate inifrastructure, and efficient traininilg
incentives are fiscal (see figure 3.2). In a recent  and education programs.
study that included 71 developing countries,   Under special circumstances, targeted FDI
Bora (2002) concludes that fiscal incenitives are  incentives may have positive effects. Many
the most popular, accounting for 19 of the 29  government officials seem to think that such
most frequently used incentives. Furthermore,  incenitives work, as illustrated by statements
the five most common incentives are all fiscal.  from a number of representatives in the Work-
Despite the popularity of FDI incentives in  ing Group on Trade and Investment of the
developing countries, the evidence of their  World Trade Organization (WTO   11998i).
effectiveness remains ambiguous. The United  Several studies find that fiscal incentives do
Nations Conference on Trade and Develop-    affect location decisions, especially for export-
ment (UNCTAD 1996) reports that incentives  oriented FDI, although incentives seem to play
can have an effect on attracting FDI at the  a secondary role (see Devereux and Griffith
margin, especially when one considers the type  1998; Guisinger and others 1985; Hines
of incentive and the type of project. Con-  1996). However, fiscal incentives appear
versely, Caves (1996) and Villela and Barrelx  unimportant for FDI that is geared primarily
(2002) conclude that incentives are generally  toward the domestic market; instead, such FDI
ineffective once the role of fundamental deter-  appears more sensitive to the extent to which
minants of FDI is taken into account. Further-  it will benefit from import protection. Thus, a
more, in a recent review of the literature on tax  more nuanced view of the efficacy of incen-
incentives and FDI, Morisset and Pirnia (2000)  tives may be in order. Although useful for at-
conclude that such instruments rarely make up  tracting certain types of FDI, incentives do not
for deficiencies in a host country's overall eco-  seem to work when applied at an economy-
noiTnic environment, and they fail to generate  wide level (see Hoekman and Saggi 2000).
the desired externalities. Overall, recent evi-  Moreover, even when targeted, FDI incen-
dence provides little support for those who  tives may impose excessive costs on govern-
believe that incentives will bring in extra FDI.  ments, especially when fiscal incentives are
To some extent, the ambivalent perspec-  provided through special tax provisions.
tivcs may reflect differences in views regarding  Because benefits (a new manufacturing plant,
8]



G LO BA L EC O NO MI C PRO SP EC TS 2 0 0 3
jobs created) are visible, whereas costs are hid-  some scope for international action to prevent
den (tax revenues are forgone), governments  suboptimal outcomes (see chapter 4). Second,
may offer too much. Also, the existence of ex-  there is the possibility that incentives offered
cessive FDI incentives is not just a developing-  by high-income countries will end up retain-
country phenomenon-in fact, such incenitives  ing or attracting FDI that would be more
are far larger in industrial countries. For exam-  efficiently  used  in  developing  countries
ple, in 1996, Mercedes-Benz received a sub-  (Hoekman and Saggi 2000). For example,
sidy of $300 million, which amounts to a     labor unions and local interest groups may
subsidy of $200,000 per employee, from the   oppose plant closures by offering excessive
U.S. state of Alabama for establishing an auto  incentives for firms to remain. Similar motiva-
plant (Moran 1998). Similarly, following reuni-  tions underlie the use of trade policy instru-
fication, Germany paid a subsidy of $6.8 bil-  ments such as antidumping. It is imlportant,
lion to Dow Chemical, which amounts to an    therefore, to distinguish between the loca-
astounding $3.4 million per employee (Moran  tional competition that may enhance efficiency
1998).                                       and the use of investment and trade policies
Additional concerns about the use of in-  (such as antidumping) that alter the incentives
centives emerge from their effect on the dis-  for outward FDI. The latter policies are inher-
tribution of rents between governments,      ently inefficient because they protect industries
host-country firms, and large multinationals.  that are no longer competitive, and they in-
Developing countries may be tempted to offer  duce various related distortions that are well
investment incentives to multinationals in   documented in the literature (Finger 1993).
part because of an expectation of technology
spillovers to local firms. Yet, investment in-  Clusters. In the past decade or so, the concept
centives to multinationals can put local firms  of industrial clusters has received a great deal
at a competitive disadvantage, at least initially,  of attention (see, for example, Porter 1990).
The net effect is hard to estimate: perhaps  While there is no standard definition of a clus-
incentives impose a short-run cost on local  ter, it is usually characterized as a regional
firms, which may gain from foreign invest-   agglomeration of firms in related industries
ment in the long run.                        (along with complementary infrastructure and
A selective use of investment incentives  support services such as business, financial,
can have strategic consequences among for-   and legal) that all work together in a virtu-
eign firms, especially when multinationials are  ous cycle to attract new firms and to help
pervasive in markets with a high level of con-  existing ones grow. California's Silicon Valley
centration. For example, an exporting foreign  typifies the high-technology cluster, with its
firm from a developing country (or a local host  concentration of high-tech firms, premier
firm) may find itself at a disadvantage with re-  universities that actively interact with local
spect to another foreign firm that experiences  businesses, and venture capitalists. Cluster-
a decline in costs resulting from an investment  ing, however, occurs in many other industries
subsidy. Thus, incentives can alter the distrib-  as well and is quite widespread (Ellison
ution of rents among multinationals.         and Glaeser 1997; Krugman 1991, 1998). In
Finally, the use of investment incentives by  the United States, evidence of knowledge
developing countries poses a possible interna-  spillovers within regions (Jaffe 1989; Jaffe,
tional coordination problem in two respects.  Trachtenberg, and Henderson 1993) and very
First, as noted earlier, the possibility of exces-  small areas (Wallsten 2001b) is consistent
sive incentive "competition" among develop-  with the idea that similar firms may benefit
ing countries may increase the likelihood that  from proximity with one another.
the "winning" country will have given away      Although the policy interest may be rela-
far more than it receives. This area allows  tively new, clusters have been recognized for a
82



D O M E S T I C P O LI C I ES  r o UN LO C K G LO B AL O P P OR T UN III ES
long time. In 1920, Alfred Marshall (as cited  In sum, while much evidence shows that
in Davenport 1935) hypothesized three rea-  clusters of firms are beneficial and occur nat-
sons for the existence of clusters: the benefits  urally over time, there is little understanding
from a pooled labor supply, access to special-  of how to create them from scratch and no
ized resources, and information flows among  experience to suggest that governments have
market participants. Today, these main bene-  any expertise in selecting activities where clus-
fits are still associated with clusters. In a suc-  ters might flourish. Bigger payoffs are likely
cessful cluster, these factors generate positive  to come from interventions to improve the
feedback loops because the concentration of  broader business environment. If governments
people and firms will attract more people and  are obliged to provide incentives to stimulate
firms (Arthur 1994; Krugman 1991).          cluster development, they may do better by
With these potential benefits, it seems nat-  encouraging expansion of existing clusters,
ural that policymakers would want to start  rather than by trying to pick winners and
clusters close to home. Unfortunately, there is  ending up simply transferring resources to the
little evidence that active efforts to create clus-  private sector without generating any positive
ters tend to be successful. This result is in part  externalities.
related to the difficulty that governments
everywhere have in "picking winners." With-  Export Processing Zones. EPZs have become
out any clear market signals about what activ-  a prominent feature of many developing and
ities or clusters might be viable, governments  transition economies, increasing from 1 75 in
have a fairly poor track record. Bergman and  53 countries in 1987 to 500 in 73 countries
Feser (2001) argue that "in less developed  by 1995 (Kreye and others 1987 and OECD
regions a policy decision to concentrate re-  1996, both cited in Schrank 2001). Along
sources on key industries, instead of more gen-  with this increased prevalence, it is not
eral infrastructure needs or other strategies  surprising that EPZs now account for fairly
that would serve best a broad array of indus-  high shares of total employment in many
tries, brings with it significant risks against  countries-for example, as much as 6 percent
which the gains remain unverified." In indus-  in the Dominican Republic (de Ferranti and
trial countries, research suggests that efforts  others 2002). Despite EPZs' ubiquity in the
to promote cluster developmenit througlh    developing world, there is little agreement on
science parks and public venture capital tend  whether EPZs are an effective development
to be unsucccssful (Braun and McHone 1992;  tool. While some view EPZs as the first step
Felsenstein 1994; Wallsten 2001d).          down a virtuous path of liberalizing domestic
Of course, this cautionary conclusioni does  markets (Rodrik 1999), others believe that, by
not mean that emerging clusters should be   creating a special "property right" of value to
ignored. Indeed, it may be that governments  those who participate, EPZs represent an es-
can draw on the problems such clusters face  cape valve that curtails broader reform efforts
when prioritizing where to undertake reforms.  and that hampers overall liberalization and
In other words, cluster promotioni may be   developmncit.
more successful when directed toward areas in  The inimediate benefit of EPZs to the host
whicih significant activity is already ongoing,  economy lies in job creation, greater foreign
as weil as areas where additional efforts on the  exchange earnings, and, possibly, higher real
margin by government may be the catalyst    wages. In many instanices, workers seem to
needed for further expansion. This type of  perceive EPZ employment as an attractive
selective interventioni may underlie the success  opportunity. For example, Brown (2001, cited
stories that do exist, such as Hsinclu Science  in de Ferranti and others 2002) finds that
Park in Taiwan, China (Saxenian and Hsu     men and women employed in Mexico's
2000).                                      maquila (maniufacturing EPZ) sector earn 31
83



G LO BA L EC O NO MI C PRO SP EC TS 2 0 0 3
and 38 percent more, respectively, than their  Schrank suggests that smaller countries may
peers in non-EPZ sectors. Similarly, in a sur-  be unable to "transform feeble manufactur-
vey described by Sargent and Matthews      ers into world market-oriented firms" and are
(1999, cited in de Ferranti and others 2002),  less likely to draw themselves onto a "large-
73 percent of Mexican maquila workers in-  country growth trajectory."
terviewed reported their current job to be at  Research does show that, in some in-
least as good as their previous employment.  stances, EPZs can be successful and can act
Furthermore, worker welfare in EPZs is also  as a catalyst for the rest of the economy (for
improved through employer practices of     example, Jayanthakumaran and Weiss 1997;
providing worker benefits (such as medical  Johansson and Nilsson 1997). Moran (2002)
insurance), stable work schedules, and week-  argues that EPZs will have only a limited
ends off. Moran (2002) evaluates worker con-  effect unless they are supported by efforts to
ditions in EPZs in a number of countries and  integrate them more fully into existing com-
concludes that there is "extensive evidence  mercial and industrial hubs and unless they
that wages and working conditions in foreign-  are located near existing or potential pools
owned or foreign-controlled factories com-  of better-educated labor. In particular, this
pare favorably with those of alternative occu-  argument implies that government efforts to
pations." Moran (2002) further notes that the  use EPZs to encourage development of
demand for jobs is high and that workers   "backward" regions that are far from existing
tend to return to existing jobs following a  industrial centers (where the infrastructure is
holiday. English and de Wulf (2002) credit  limited and skilled labor is scarce) are unlikely
EPZs with creating more job opportunities  to be successful. The more successful EPZ
for women in Bangladesh, with reducing fe-  experiments that Moran considers are in
male poverty in the Dominican Republic, and  Costa Rica, the Dominican Republic, and the
with raising wages for EPZ workers above   Philippines. Those examples show how EPZs
wages for workers in the rest of the economy.  have facilitated a shift in foreign investment
Beyond the direct effect of EPZs on job  away from  lowest-skill operations that are
creation, a comprehensive evaluation of them  limited to export enclaves toward higher-skill
should look at two additional criteria. First,  operations that are better linked to the rest of
do EPZs actually encourage firms to export  the economy and that provide both employ-
(or to increase exports), rather than causing  ment opportunities for higher productivity
firms that already export to relocate into the  (and higher wages) and better worker con-
EPZ so they can take advantage of financial  ditions. Without such complementary efforts,
incentives? Second, do EPZs produce spillover  EPZs risk becoming another entrenched
effects by drawing local manufacturers into  interest that simply maintains trade barriers
the world markets, thereby indirectly bring-  and delays broader market reforms.
ing reform and enhanced competitiveness to a  Another view of EPZs focuses on their role
greater segment of the nation's producers?  as "transition property rights." It highlights
Schrank  (2001) compares EPZs in the      their function in helping the country steadily
Dominican Republic and in the Republic of  improve its investment climate. That is, EPZs
Korea, arguing that market size is a major  may act as a catalyst for the host economy, thus
determinant of EPZ success. Despite the good  sparking a sequence of beneficial changes in the
performance within the Dominican Republic's  economy. The experience of Mexico is highly
EPZ sector, few benefits appear to spill over  illustrative in this case: the transition began
into the rest of the economy. Korean EPZs,  with establishing maquilas in a 2-mile border
however, are increasingly integrated with local  zone, which was next expanded to 12 miles,
suppliers, thereby helping to transform much  then to entire states, and eventually to the whole
of the economy into world-level competitors.  country. In this case, EPZs were able to help
84



DO ME ST IC PO LI C I ES TO UN LO C K G LO B AL 0     P 0 R I1 U N I T I E S
improve the investment' climate by acting as  large effects on productivity and welfare. In-
a bridge between the old and the new systems.  dustries generally function better when they
operate in a competitive environment, and
-and incentives cannot offset a poor       richer and faster-growing countries tend to
policy environment                         have more competition and fewer barriers to
Governments may hope to make up for an     entry. Changes in technology, global business
unfriendly investment environment through  organization, and regulation have created new
incenitive mechanisms. But while there are  opportunities for competition in areas that
clearly examples in which targeted interven-  had formerly been seen as natural monopolies
tions (such as fiscal incentives, EPZs, or sup-  (infrastructure) or that were considered neces-
port for clusters) may indeed lead to higher  sary to preserve domestic sovereignty (services,
investment levels-and the jobs and related  real estate, and the financial sector). Countries
spillovers that go along with such levels-there  that do not change their investment policies
is unfortunately little evidence that such initia-  and do not exercise well the powers and re-
tives can be systematically successful. Instead,  sponsibilities of the state-such as regulating
the impression is that these interventions work  privatized industries, providing education, or
best when they work in support of broader re-  enforcing conditions of competition-will
form packages, either to catalyze support for  forgo poverty-reducing growth opportunities.
emerging opportunities (such as clusters) or to  At the broadest level, competition and ease
create an initial constituency for reform that  of entry are both positively correlated with
can be progressively expanded (such as EPZs).  economic growth (figure 3.3). A host of policy
But more broadly, as Wells and others (2001)  and private barriers in developing countries
note, "Incentives will generally neither make  work to restrict competition. Restrictions on
up for serious deficiencies in the investment  trade and FDI rob an economy not only of
environment nor generate the desired long-run  potential sources of investment, but also of
strategies." To encourage productive invest-  one incentive for firms to improve productiv-
ment and benefit from globalization, govern-  ity. While causality goes both ways, both trade
ments must tackle the challenges of promoting  and FDI are correlated with higher produc-
competition and entrepreneurship and of un-  tivity of firms in an economy. But potential
dertaking complementarily productive public  competition does not come solely from inter-
investment in areas such as education. We now  actions with the global economy. Many devel-
turn to these issues.                      oping countries still protect incumbent firms-
whether state-owned or private-by giving
them monopoly power even when there is lit-
Promoting efficient private                tle rationale for doing so. While such actions
investment: harnessing                     may protect particular firms, they almost al-
competition                                ways impose net costs on everyone else in the
W 7Tthile a stable macro environment and   country. Finally, other private barriers-such
W   good governance are important to at-  as collusion, price-fixing, and cartels-block
tracting investment, policies that promote con-  competition and reduce welfare. This section
testable markets and that protect against  of the chapter reviews some of these barriers
abuses of market power are required to ensure  to competition, and details how they can harm
that new investment is both productive and  developing countries' economies.
efficient. Of particular importance in this re-
gard are investment and competition policies,  Policy barriers to competition are a drag
which are important elements of the invest-  on productive investment
ment climate and also are basic pillars of the  Barriers to competition stemming from gov-
economy's micro foundations that can have  ernment policies can emerge either through
85



G LO B AL     EC ONO MI C      PROS PE CT S       2 0 0 3
Figure 3.3 Competition and ease of entry are associated with higher growth
(GDP growth rate, percent)                          (GDP growth rate, percent)
14                                                  14
12                                                  12
10                                                  10                               *
8              ^   *,*+                             8                 * 
6             . *    *t6 *                          6         ,-
4   .-*S - -t - - -t                                4
0  ----                                             0 
-2                                                  -2
-4                                                  -4
-6  |    |    |     |    |     |    t               -6                        I      I      I
3    35    4    45    5    55    6    65    7      35      4      45      5      55      6     65
Competition                                            Entry
Note "Competition" is the average response in each country to the question "in most industries, competition in the local markets is
(I = limited and price-cutting is rare, 7 = intense and market leadership changes over time) " "Entry" is the average response to the
question"Entry of new competitors (1 = almost never occurs in the local market, 7 = is common in the local market)." Although
competition and entry rankings suffer from methodological problems related in part to averaging of responses across respondents
(see, for example, Lall 2001, Recanatini, Wallsten, and Xu 2000), those rankings can, nevertheless, provide a useful starting point
for more rigorous investigations. One important question that these figures cannot answer is that of causality do entry and
competition make countries richer, do richer countries have more competition, or does something else altogether drive both growth
and competition' What does emerge clearly is that poorer and more slowly growing countries seem to have less entry and
competition
Source World Economic Forum (2002), World Bank SIMA indicators
dtrect channels (such as when governments               dispersion), research     strongly  suggests that
create state monopolies) or through tndtrect            greater openness     is associated    with  higher
channels (such as when policy choices made in           growth    in  both   industrial and    developing
pursuit of other objectives end       up   limiting     nations. Sachs and Warner (1995) find that
competition). In this section, we will focus on         openness is a highly significant determinant of
four channels through which competition is              growth and, combined with property rights,
affected by policy choices:                             may even represent sufficient conditions for
growth in poor economies. Kang and Sawada
e Import competition                                 (2000) find a similar effect of openness on
• FDI competition                                    growth. They argue that, combined with finan-
* Administrative barriers                            cial development, openness increases growth
* State monopolies and private barriers to           rates in developing economies by decreasing
competition.                                      the cost of human capital investment. Maloney
(2001) offers regional support for the above
Import competition can enhance                          result, citing  evidence that Latin      American
productivity                                            economies that are more open and that possess
The important role that trade plays in pro-             a more developed       knowledge infrastructure
moting    productive   investment and      growth       will grow   faster. Consistent with this result,
has long been recognized. Using different mea-          Cuadros, Orts, and Alguacil (2001) find that
sures of openness to trade, including         both      openness positively     affects Latin    American
its relative size (as measured by import and            growth and trade through increasing FDI.4
export shares) and      its degree of distortion           Such aggregate results fail to answer the
(as  measured    by   average   tariff rates   and      question of exactly how       increased openness
86



D O M E S T I C P O L I C I E S 'I 0  U N L O C K G L O B A L O P P O R T U N I I I E S
(however measured) is translated into faster  markets. In figure 3.4, markups are lowest
growth. One approach emphasizes the learn-    (measured on the vertical scale) when import
ing and productivity gains that occur as do-  competition is highest and when there are more
mestic firms confront more competitive world  firms (the front corner), and markups are high-
market conditions, become more efficient,     est when import competition is low and when
and begin exporting. Another more compell-    the market is more oligopolistic (back corner).
ing approach emphasizes the rise in import       Import competition   pressures domestic
propensities that often comes with trade      firms to be more productive. A recent study of
liberalization. Increased imports place domes-  Brazilian manufacturinig firms, for example,
tic firms under direct competitive pressurcs  finds that foreign competition induces quick,
and indirectly induce technological innovationi  marked improvements in domestic productiv-
or cost-cutting restructuring that further en-  ity and, over time, forces inefficient firms to
hances competitiveness and productivity.      shut down (Muendler 2002). Cross-country
Research finds that price-cost margins      data are consistent with these findings, sug-
(markups above cost) tend to fall with import  gestng that higher tariff rates (which make
competition, though the direction of causallty  imports more costly and thus less competi-
is not clear, and that foreign competition tends  tive), are correlated with lower productivity
to improve manufacturers' efficiency (Tybout  (figure 3.5).
2001). Hoekmnan, Kee, and Olarreaga (2001)       In addition to the direct competition af-
found that import competition (defined as the  forded by greater openness to imports, higher
ratio of import volumc to domestic consump-   trade prevalence can create spillover opportu-
tion in an industry) reduces industry markup.  nities through which domestic firms can gain
The effect of import competition is particularly  access to (improved) technology without pay-
powerful when a few oligopolists dominate     ing full cost. In general, imports from industrial
Figure 3.4 Competition from imports checks markups in concentrated markets
1 800
Markup 1 400
1 200 ~   mpo/d      Cfltn,10
Ntote Import penletrationl is dcftnted as the ratio of Iiitport volulme to the dontlesoc riurpulr of asn induistry
Sousrce Hoekmlan,w Kcc, and Olarrc.igi (2001i)
87



G LO BA L EC ON OM IC PRO SP E CT S 2 O 0 3
manufacturing sectors in which productivity
Figure 3.5 High tariffs are correlated with  increased moderately. Imports did not seem to
lower productivity                       affect productivity in sectors with low or high
(percent)                                 productivity growth. Using industry-level
40                                       data, Keller (2000) finds that imports may
30 5                                      boost technology diffusion if countries receive
25            Average productivity  _    a relatively high share of total imports from a
20      Average tariff rate               high-technology trading partner. Hakura and
15                                       Jaumotte (1999) find that the share of imports
10             [                  -      from industrial countries has a positive effect
s- t   _    1[1<      _    1*            on total factor productivity. Finally, Xu and
0                 I          I _,        Wang (2000) find that the share of imports
of capital goods from high-technology coun-
tries is significantly related to productivity
increases.
Source World Batik staff, based on survey of more than
5,000 firimis                             Competitive effects of FDI depend
on policy-
FDI can be a potential vehicle for increas-
countries are positively related to technology  ing competition. Multinational corporations
diffusion and productivity growth (Eaton and  (MNCs) tend to be more efficient and produc-
Kortum 1996; Lumenga Neso, Olarreaga, and   tive than smaller, purely domestic firms. While
Schiff 2001). Sjoholm (1996) finds a positive  MNCs' entry into the domestic market can
relationship between bilateral import shares  put competitive pressures on local producers,
and patent citations for Sweden.5 And Coe   the mere presence of MNCs does not neces-
and Helpman (1995) find that industrial     sarily increase competition. Because they often
countries that receive a larger share of imports  possess significant intangible assets (brand
from countries with a high level of research  names, technology, managerial skills, and so
and development (R&D) expenditures will ex-  forth), MNCs often supply different markets
perience faster productivity growth.6 Despite  directly (through domestic production activi-
agreement that imports are an important     ties) rather than through exports. Such assets
channel for technology diffusion, studies reach  may permit MNCs to wield considerable mar-
somewhat different conclusions on the condi-  ket power. Openness to trade, low barriers
tions under which such diffusion is most likely  to exit and entry, and other regulatory condi-
to occur. Coe, Helpman, and Hoffmaister     tions can in turn help limit the capacity of
(1997) extend the results of Coe and Helpman  MNCs to abuse market power in the domestic
(1995) to developing countries and find that  market.
developing countries' total factor productivity  While the relationship between competi-
is positively related to their openness to trade  tion and FDI remains complex, over time the
with the industrial countries. Furthermore,  competition-increasing association has become
productivity in developing countries increases  more prominent. Historically, FDI was often
as imports' share of GDP increases.         attracted to regions that were protected by
Some research finds that manufacturing   high tariffs, as firms calculated that it was
productivity in developing countries depends  easier to set up a subsidiary than to pay the
on the complexity of imported machines      tariffs required to serve the market through
(Navaretti and Soloaga 2001). Choudhri and  exports. Such tariff-jumping investment was
Hakura (1999) show that imports are signifi-  also motivated by the opportunity to service
cantly related to productivity growth only in  the domestic market behind the tariff barriers
88



DO ME S II C P O L I C I E S T O  U N L O C K G L 0 B A L O P P O R1 U N I T I E S
while shielded from competition from abroad.  allowed high-cost firms to produce for the
This type of FDI has a long history: in the  local market at very high prices, cven though
post-World War 11 period, many developing   they could have imported much more cheaply.
countries encouraged both domestic and for-  An even higher share of domestic projects that
eign firms to invest in high-priority industrial  they reviewed had negative value added.
sectors by imposing high tariffs, quantitative  Encarnation and Wells (1986) found that
restrictions, and other nontariff barriers,  25-45 percent of 50 projects studied (depend-
along with providing various additional incen-  ing on analytical assumptions) reduccd natio-
tives (Caves 1996).                         nal income; again the main culprit was high
Investment induced in such a way, however,  protection.
is unlikely to be efficient and, therefore, is less
capable of providing a basis for sustained  -and benefits are higher when trade
growth. First, the empirical evidence suggests  barriers are lower
that tariff-Jumping FDI is "likely to be tran-  One clear implicationi is that If countries are
sient, lasting as long as the artificial policy-  open to foreign investment, trade barriers can
induced incentives" (Balasubramanyam 2001).  and should be kept low. Such openness to in-
Second, it can harm welfare by increasing con-  ternational competition will keep MNCs from
sumer prices. In an era of much higher tariffs  using high protective tariffs to exert market
than generally exist today, Lall and Streeten  power domestically and will discourage thcm
(1977) found that more than one-third of the  from joining domestic vested interests that are
90 foreign investments they studied actually  lobbying for policies that perpetuate costly
reduced national income. This reduction was  rent-seeking activities. The cost of not doing
mainly from excessive tariff protection that  so can bc enormous, as illustrated in box 3.1.
Box 3.1 Trade restrictions shield MNCs from
competitve forces at enormous cost: the case
of Argentina
T   rade and tax policy often interact in ways that  subsidies were so lucrative that foreign (and some
T  magnify their competition-restricting effects.  domestic) firms bought finished products in Japan;
Newfarmer (2001) illustrates the importance of  shipped them to Panama, where they were broken
policy in determining thc net contribution of multi-  down; and then shipped them to Tierra del Fuego
national corporations (and domestic firms) by using  for subsequent reassembly and resale in the main-
the example of Argentina in the 1980s. In an effort  land of Argentina. By 1990, estimates of the cost
to encourage settlement of Tierra del Fuego, the  to the (then-bankrupt) Argentine treasury ranged
southernmost tip of the country (partly in response  from 0.5 to 1 percent of gross domestic product.
to territorial disputes with Chile), the government  The winners in this scheme were the produciig
set up a special production zone for assembling  companies and a few thousand workers in Tierra
electronic products with generous tariff protection  del Foego; the losers were Argentine consumers and
and tax subsidies. Firms were encouraged to    businesses that had to pay high prices, thousands of
assemble many types of electronic goods for resale  workers who would have otherwise gotten jobs in
to the highly protected Argentine market at    more internationally competitive new activities on
enormous markups. As a result, televisions in  the mainland, and the Argentine poor, who, among
Argentina routinely exceeded international prices  others, had to pay the tax of high inflationi to close
by 150-400 percent. The regime protection and  the fiscal accounts.
89



G LO B AL EC ONO MI C      PRO SP EC TS 2 O 0 3
In recent years, the incentives for tariff-  this outcome is based on advantages associated
jumping FDI have declined. Barriers to trade    with greater efficiency, and if the resulting
have come down considerably. As the impor-      market structure remains reasonably compe-
tance of production networks has risen, for-    titive, these effects are generally positive. Fur-
eign investors have found barriers to entry and  thermore, MNCs could spark the entry of
less-competitive environments less appealing.   productive suppliers, encourage greater inno-
In more recent studies, foreign investmenit is  vation, increase the variety of available prod-
deterred by high taxes or nontariff barriers    ucts, and drive down prices. However, if a
on imported inputs and is attracted to more-    domestic firm's exit is driven more by the mar-
open economies. In reviewing cross-country      ket power of the MNC and if the exit results
regressions on the determinants of FDI,         in greater market concentration, then the long-
Charkrabarti (2001) argues that, after market   run result may be less competition.
size, openness to trade has been the most reli-    The case study literature provides both pos-
able indicator of the attractiveness of a loca-  itive and negative examples. After reviewing
tion for FDI (see Kolstad and Tondel 2002).     the evidence, UNCTAD (1997) concludes that
As figure 3.6 illustrates, there is now a signif-  although there is substantial evidence that the
icant negative relationship between high entry  entry of MNCs yields new products and im-
costs7 and the attractiveness of a market to    provements in existing products, there is no
foreign investors (controlling for other factors  systematic evidence on whether it ultimately
such as market size, macroeconomic stability,   reduces consumer prices. The overall effect
and human capital).                             should not be judged at one moment in time. In
MNCs can have an indirect effect on com-     the short run, some less-efficient producers will
petition by affecting ownership and market      likely be driven out of the market, while over
structure. For example, with a blend of deeper  time, more productive entrants will emerge.
financial pockets, marketing skill, and supe-   There is evidence that domestic suppliers to
rior product or process technology, MNCs        MNCs enjoy higher productivity, both in levels
may drive a significant number of domestic      and growth (see Blalock 2001; Smarzynska
competitors out of business. To the extent that  2002). Thus, the net effect of FDI on competi-
Figure 3.6 High entry costs inhibit FDI inflows
(FDI)
More
Less
Low                                  Entry costs                               High
Note Partial correlations conitrol for market size, humati capital, and macroecoitomic stability The entry cost measure used in the
figure refers to the costs of obtaining the iiecessary permits and licenses and other procedures required to set up a new establishment
See Djankov and others (2002)
Source World Bank staff
90



D O M E S T I C PO LI C I ES TO UN LO C K G LO B AL O P P O R T U N I T I E S
tion, per se, depends on the level of interna-  of reform and "have usually been the fastest-
tional competition in the industry and on the  growing segment in transition countries"
ability of domestic firms to increase their pro-  (McMillan and Woodruff 2002). The scale of
ductivity in response to increased competition.  entry that occurs when reforms promote com-
Perhaps because the channels through    petition can be impressive. Deng Xiaoping
which FDI affects competition will vary de-  expressed his surprise that "all sorts of en-
pending on the institutional environment-  terprises boomed in the countryside, as if a
tariff structure, market size, competition  strange  army  appeared  suddenly from
policy-the empirical findings about the   nowhere" less than a decade after the first re-
effect of FDI on growth are also mixed. FDI  forms in China in 1978 (Zhou 1996 as quoted
should contribute positively to growth, be-  in McMillan and Woodruff 2002). Key to
cause it can bring capital, technology, skilled  promoting entrepreneurship and to improv-
management, and technical staffs, plus busi-  ing productivity is an environment that facili-
ness practices that are usually more modern.  tates entry and exit of firms (see, for example,
Indeed, several econometric studies have  Lansbury and Mayes 1996). Through this
shown that, controlling for other factors, FDI  process, poorly performing firms leave the
flows are positively associated with economic  market and dynamic new ones enter. Unfor-
growth (for example, see UNCTAD 1998 and  tunately, many developing and transition
World Bank 2001 for all developing coun-  governments fall to recognize that firm
tries; Van Ryckeghem 1994 for Latin America;  births and deaths are an inevitable corollary
and Chunlai 1997 for China).              of entrepreneurial risk-taking. Instead, those
However, the direction of causation is not  governments erect a maze of administrative
clear: does FDI cause more rapid growth   obstacles to starting, operating, and closing
because of its associated characteristics, or is  firms.
FDI simply attracted to more rapidly expand-  A growing body of literature documents
ing markets to exploit growth opportunities?  the difficulty that entrepreneurs face in estab-
The answer is probably both. Theory does not  lishing firms in developing countries (for ex-
provide a simple answer because the institu-  ample, Djankov and others 2002; Emery and
tional settings and endowments are quite  others 2000; Friedman and others 2000).
varied and complex (see Cooper 2001). One  Djankov and others (2002) compiled data on
problem, for example, is that those elements  entry regulations in 85 countries and discov-
in the investment climate that are conducive  ered enormous variation in the number of
to FDI are also conducive to more domestic  procedures required to start firms across
investment and to greater growth in pro-  countries, ranging from a low of 2 in Canada
ductivity. Many of the methodological cri-  to as many as 21 in the Dominican Republic
tiques that Rodriguez and Rodrik (1999) and  (with Bolivia and Russia close seconds at 20
Cooper (2001) apply to cross-sectional studies  each). The time required to establish a firm
of trade openness and growth also apply to  ranged from 2 business days in Canada to 152
the somewhat less abundant literature on the  in Madagascar. These procedures can be ex-
relationship between FDI and growth.      tremely costly to the economy. The cost of of-
ficial procedures (that is, not including bribes)
Administrative barriers are usually high  for setting up a new business was 266 percent
in developing countries-                  of per capita income in Bolivia. Djankov and
Entrepreneurship is an important contributor  others (2002) found that stricter regulation
to economic growth and welfare improve-   of entry is correlated with more corruption
ments in transition and developing countries.  and a larger informal economy. Moreover,
For example, new firms created 10 million  "countries with more open access to political
new jobs in Vietnam in the first seven years  power, greater constraints on the executive,
91



G LO BA L E CO NO MI C      PROS PE CT S     2 0 0 3
and greater political rights have fewer re-       ings in countries where It takes longer to
quired procedures for entry regulation-even       start a firm, thus suggesting that keeping new-
controlling for per-capita income-than      do    comers out of the market protects inefficient
the countries with less representative, less lim-  incumbents.
ited, and less free governments" (Djankov and        While exit barriers can be harmful, dealing
others 2002). In a study of such obstacles in     with a firm's exit is not simple. Ideally, bank-
Africa, Emery and others (2000) discovered        ruptcy and insolvency procedures rehabilitate
that "when added together, this whole maze        viable but financially distressed   firms and
of often   duplicative, complex, and     non-     liquidate unviable firms. In practice, deciding
transparent procedures can mean delays of up      which firms are viable is difficult. Djankov,
to two years to get investmenits approved and     Hart, and Nenova (2002) note that many
operational."                                     countries have crude insolvency laws that
Although policymakers and advisers tend        push financially distressed companies directly
to emphasize market entry, exit is important      into liquidation, while other countries allow
as well because it releases resources that can    completely unviable companies to enter reha-
be used in more productive ways. Healthy          bihtation procedures. In the latter case, such
economies have a high "churn rate" of firms,      companies are often liquidated only after a
and research demonstrates a strong positive       long and expensive period of rehabilitation. In
link  between entry and exit (Love 1996).         recent years, there is a growing movement in
Moreover, as Caves (1996) has pointed out,        insolvency reforms to introduce rehabilitation
barriers to exit can be barriers to entry both    procedures in countries that do not have them,
by absorbing the scarce resources necessary       but to allow creditors to replace management
to start new enterprises and by making it dif-    during the rehabilitation   process (Djankov,
ficult for new firms to compete. Entry barriers,  Hart, and Nenova 2002).
moreover, can become exit barriers (see fig-         Barriers that limit firms' operating flexibi-
ure 3.7). Claessens and Klapper (2002) find a     lity exist even when entry and exit is not at
smaller share of firms in bankruptcy proceed-     stake. Friedman and others (2000) compile
Figure 3.7 Barriers to entry can become barriers to exit
(ratio of bankruptcies to number of firms)
8
7
6
5
4
0                                     o
0                    50                   100                   150                   200
Days required to start a new business
Note Averages for 35 industrial and developing countries, 1990-99 The data panel used for cailculationi of averages is unbalanced
that is, the enitire range of observationis (1990-99) was not available for some countries lThe measure of days required to start a new
business is taken froin Djankov and others (2002)
Source Claessens and Klapper (2002)
92



D O M E S T I C P O LI C I ES TO UN LO C K G LO B AL O P PO RI UN III LS
indices of taxation levels and overregulation  see how such costs could quickly undo other
(essentially, indices of the business environ-  advantages that these firms might have when
ment) of firms in 69 countries. Although    competing in world markets.
Friedman and others (2000) find no evidence    A telling indicator of whether markets are
that higher tax rates drive firms under-    competitive in a country is the productivity
ground, ". . . every available measure of over-  dispersion of firms within an industry. In a
regulation is significantly correlated with the  competitive market with reasonably free entry
share of the unofficial economy and the sign  and exit, dispersion should be low because
of the relationship is unambiguous: more    unproductive firms either become more pro-
over-regulation is correlated with a larger un-  ductive or leave the market. Higher dispersion
official economy." The important result here  indicates that less-efficient producers are not
is that higher tax rates do not seem to drive  being forced to improve their productivity or
away investors, but the myriad and often    to exit the market. Firm-level studies in a
arbitrary array of obstacles to starting and  number of countries bear this out.8 Subsidies
running a business do.                      or strict regulations that impede entry or exit
can ultimately bolster high-cost producers.
-and have real costs                        When such firms remain in the market, more
The administrative obstacles have real costs to  productive firms may not have either the ade-
the economy, which means that even poten-   quate incentives or the ability to increase pro-
tially competitive firms often cannot compete  ductivity or to grow. However, as competition
because any efficiency advantages they may  increases, firms face greater incentives to inno-
have are consumed by the costs of administra-  vate and greater penalties for failure to do so.
tive hassles. Indian firms, for example, are  Loss of protection and greater competition
potentially competitive in a range of labor-  from foreign firms can drive inefficient do-
intensive industries; the combination of their  mestic producers to better exploit scale eco-
labor productivity and their wages makes    nomies, eliminate waste, reduce managerial
them low-cost producers at the plant level.  slack, adopt better technologies, or shut down.
The value added per unit of labor cost is lower  As a result, productivity dispersion should
in India than in East Asian competitors such  shrink as productivity levels rise in the face of
as Malaysia, the Philippines, and Thailand.  greater competition.
However, in practice this potential competi-   Productivity dispersion-a  measure of
tiveness is often offset by investment climate  inefficiency-tends to be associated with bar-
bottlenecks, resulting in lower Indian exports.  riers to competition, such as the administra-
Several dimensions are of particular relevance.  tive barriers to start a business for India,
The regulation of factor markets, particularly  China, the Philippines, Thailand, Malaysia,
of labor and land, severely restricts the entry  and Korea (figure 3.8). In Indian textiles, gar-
and exit of firms. For example, firms with  ments, and electronics, the higher performers
more than 100 employees have not been       have value added per worker that is five times
allowed to retrench workers without govern-  that of lower performers. The dispersion of
ment permission. Meanwhile, the lack of cred-  productivity is lower in four East Asian coun-
itor rights and the severe backlog in judicial  tries where the World Bank has conducted
cases mean that India has one of the lowest  similar surveys. In Thailand and Malaysia,
levels of bankruptcies internationally. The  the productivity dispersion ratios are just
Confederation of Indian Industry estimates  below 3, and in Korea not much more than 2.
that proceedings can easily take more than  Thus, more competitive countries in the group
two years, and more than 60 percent of Iiqui-  (as proxied by weeks to start a business) have
dation cases before the High Courts have been  lower levels of productivity dispersion than do
in process for more than 10 years. It is easy to  the less-competitive countries.
93



G LO B AL EC ON OM IC PROS PE CT S 2 O 0 3
countries. These administrative procedures
Figure 3.8 Barriers to entry and exit       vary across countries, with especially severe
allow inefficient firms to stay in the      delays in obtaining land and building permits.
market                                      They have found evidence that increased ad-
(for productivity dlspersion, percent, for weeks, number)  ministrative barriers deter foreign investment.
14                                            These findings are supported by a World
12 -             Weekstostartbusiness       Bank survey study that finds a similar result
10                             r            in a larger sample of 69 developing countries:
8                  Produciviy             there is a significant negative correlation be-
6  ~       ~      / r_    1           |    tween the amount of management time spent
4               VA;              L-        on obtaining the necessary paperwork and
2    **     **l        *                   the levels of FDI (figure 3.9).
0               **    Eu         1 i         Another obstacle to competition is mani-
0a Ss >s ,  F    os 8fested in product delivery costs that go be-
1:P     |   yond producers' control and yet can have an
enormous effect on their overall competitive
positions. The effect of the quality of trans-
Note Produictivitv dispersion serves as a measure of  portation, as well as the performance of gov-
inefficieiicy-more productivity dispersion means iltore
iiieffiLient firms are allowed to stav in business  ernment agencies such as customs admin-
Souirce World Bank staff, based on suirvey of more than  istration, can more than offset the cost
5,000 firms 
_____________________firms______________    advantage  that producers  enjoy  at the  factory
gate. Indian textiles provide one such exam-
ple. India's value added per unit of labor cost
These obstacles can deter foreign investors.  is lower than almost all its East Asian neigh-
Morisset and Lumenga Neso (2002) have         bors. However, if one takes into account the
compiled data on the permits and procedures   longer delays in clearing customs and the
required for entry, access to land and infra-  higher shipping costs, Indian textiles are much
structure, and operation in 32 developing     less competitive on international markets.
Figure 3.9 Difficulties in obtaining licenses and permits discourage FDI
(FDI)
More
y -0 5937x T 01851
0
v        0     ~0°   0 
0~~~~
Less
Less                      Time managers spend obtaining permits            More
Noie Partial correlations control for market si7e, human capital, anid macroeconomic stability
Source World Bank staff
94



D O MI E S I I C  ro      I c  I  s  1   u N l O  C K  C I, o B A l  o P P o R I U N I T I E S
Figure 3.10 Inefficient customs hurt Indian exports
(average number of days to clear customs)       (average number of days in India to clear customs)
12                                              30
f 02                                           20
6                                              1 5                     Lsie
4                                              10
2                                               5
India     China  Korea, Rep of Thailand         Garmnents     Textiles  Pharmaceuticals
Souirce World Baiik staff, figures based oit firm survey data
Furthermore, World Bank surveys rcport infor-
mation on the numilber of days needed to clear       Figure 3.11 Inefficient ports raise India's
customs (see figure 3 10). Here, India scores        transport costs far above competitors'
poorly relative to Korea and Thailanid, with the     transport costs
time about 50 percent longer in lindia (anid         (percent cost advantage compared with India of shipping
triple what many OECD      countries report) 9
But the issue is not only the average time,           50 Wesl coast, U S             All U S
but also the variances in clearance time. Fig-        40                 East_C     l
ure 3.10 shows the longest delay In the past          30                 u s
year for a typical firm in thrce sectors in India.    20
Although the average clearance timie is I I days,      10
the longest delays averaged almost 28 days for °
garmenats and 25 days for pharmaceuticals.'0         -10
The transportation costs associated with                Brazil  China Indonesia Korea, Thailand
shipping a container of textiles to the United                                  Rep of
States from  India are more than 20 percent          Source World ltliik stitf
higher than shipping costs from Thailand and
35 percent higher than shipping costs from
China (figure 3.1 I). Variations in maritimie
distances explain only a small part of the gap.   economics whilc producilIg just more than
Delays and inefficienicies in the ports account    10 percent of GDP (World Bank 1995). But
for a higher share of the difference In port      state ownership on such a scale was not sus-
productivity. Together, inefficient customiis and  tamable. Many SOEs rcquired large subsi-
ports can hurt the investmiienit climate and can  dies from cash-strapped governments to stay
erode comparative advantage.                      afloat, thus constraining government spendinig
on other priorities. For cxample, it was esti-
SOEs use resources inefficiently-                 mated that "divcrtinig SOE opcrating subsi-
Another way in which states make competi-         dies to   basic educatioin  . . would increase
tion and entreprencurship difficult is by their   central governmeint education expcnditures by
direct ownerslhip of many firms and industries.   50 percent In Mexico, 74 percent in Tanzania,
By 1990, SOEs consumed nearly 20 percent           160 percent in Tunisia, and 550 percent in
of gross domestic investmiienit in developing     India" (World Banik 1995)
95



G LOB AL EC ONO MI C PROS PE CT S 2 0 0 3
both public and private superiority). The
Figure 3.12 Privatization revenues       dominance of studies finding superior pri-
soared in the 1990s                      vate performance is robust across all sub-
(in nominal billions of dollars)         categories" (Shirley and Walsh 2000).
70                                          Privatization usually improved financial
60                                       and operating performance in privatized firms
50                                       (see Megginson and Netter 2001 for a com-
40                                        prehensive review of the literature). This result
30                                       holds in industrial and developing countries
alike (Boubraki and Cosset 1998a, 1998b;
20     1      11 4 4       l             Megginson and Netter 2001). The finding is
.10                       -robust across case studies, cross-sections of
firms from different industries within a given
g t P t .R. # #? ts$ $ WP              country, cross-sections of firms from differ-
ent countries, and performance of firms be-
Sosrce World Bank privatization databaseencotrs,adpfrmceffisb-
fore and after privatization (Sheshinski and
Lopez-Calva 2000). Moreover, other research
suggests that privatizations tend, overall, to
Driven in part by the high and unsustain-  increase welfare (Galal and others 1994). In
able fiscal costs of state ownership, countries  other words, privatization tends not only to
around the world embarked on a massive      improve the performance of privatized firms
privatization wave. Privatization revenues in  and to benefit investors, but also to make the
developing and transition countries increased  country better off.
from almost nothing in the early 1980s to
more than $60 billion in 1997, before de-   -and is more successful when combined
creasing somewhat to $50 billion in 1998    with competition
(figure 3.12). It was estimated that by 2000,  Simply pointing out the overwhelming evi-
cumulative privatization revenues worldwide  dence demonstrating improvements in pri-
had exceeded $1 trillion (Megginson and     vatized firms, however, masks important
Netter 2001). The bulk of privatization in  differences across industries in the challenges
developing countries occurred in services, par-  and pitfalls of privatization, especially with
ticularly infrastructure.                   regard to introducing competition. Some sec-
tors, such as manufacturing, generally lack
-and privatization improves firm            any economic lustification for state ownership
performance-                                from the outset. SOEs that have been priva-
Overall, privatization has dramatically im-  tized into such competitive markets-while
proved the performance of former SOEs. State  being freed from  unprofitable government
enterprises were substantially less efficient  controls or social "mandates"-tend to per-
than private firms. Shirley and Walsh (2000)  form quite strongly. Indeed, studies show that
reported that most of the extensive literature  the most robust results occur from privatiza-
finds private firms superior to state firms.  tion in competitive sectors (Kikeri and Nellis
Of 52 empirical studies, 32 found that the per-  2001).
formance of private and privatized firms is
"significantly superior to that of public   Infrastructure industries present special
firms," and 15 studies found "either that there  challenges
is no significant relationship between owner-  However, in infrastructure sectors such as
ship and performance, or that the relationship  telecommunications, electricity, gas, and
is ambiguous (different evidence supports   transport, existing SOEs traditionally were
96



D O M E S T I C PO LI C I ES TO UN 1.0 C K G LO B AL O P PO RI UN III ES
considered "natural monopolies." It was     distinguish-and to varying degrees separate-
almost an article of faith that, in these indus-  the true, natural monopoly clements of a sys-
tries, a single firm could provide services at  tem from the competitive segments. Second,
the lowest cost. In most of the world outside  even when competition is feasible, a dominant
of North America, such natural monopolies   incumbent in a network industry often has
translated into state-owned monopolies from  both the incentive and the means to thwart
the 1920s through the 1980s. But by the late  competition.
1980s, the combination of technological       With privatization more likely to be suc-
change, a clearer understanding of the costs of  cessful in competitive sectors and with infra-
state ownership and monopolies, and a wide-  structure sectors, in gencral, less amenable
spread failure of SOEs in developing countries  to compctition, it is not surprising that expe-
to deliver reliable services to consumers in  riences in infrastructure privarization offer
natural monopolies made privatization and   more mixed outcomes. Perhaps not unexpect-
competition both technically feasible and   edly one key determinant of privatization suc-
politically desirable. The benefits from this  cess has been the degree of competition intro-
process are clear: studies suggest that privati-  duced in the regulatory regime. As Ambrose,
zation or contracting out of public services,  Hennemeyer, and  Chapoii (1990) note,
including many infrastructure services-if   "[Slimply moving a monopoly from the pub-
done right-can yield efficiency gains equiv-  lic to the private sphere will not rcsult in com-
alent to 10 to 30 percent of previous cost  petitive behavior." Another factor affecting
(Bartone and others 1991; Carnaghan and     success relates to the sequencing of sector
Bracewell-Milnes  1993; Domberger and       reforms (including privatization) and the cre-
Piggott 1994). When real competition is not  ation of the regulatory institutions that are
or cannot be introduced, it is more likely that  necessary to achieve the broader objectives,
privatizationi will be less effective, and well-  including promoting competition. Policy
run public firms may do as well as private  reforms such as privatization often have
ones (Kwoka 1996). But even in these cir-   proceeded faster than the necessary support-
cumstances, many private projects have      ing institutions manage (see, for example,
outperformed public eiiterprises. Examples  Wellenius 1992). This outcome is hardly sur-
include the water sector in Argentina, C6te  prising because privatizing a firm, complicated
d'lvoire, and Guinea (Clarke, Menard, and   though it may be, is a relatively straight-
Zuluaga 2000; Noll, Shirley, and Cowan      forward and discrete task when compared
2000).                                      with building a regulatory agency where none
At least two broad difficulties exist in pro-  existed. Nonetheless, varied expcriences with
moting competition when privatizing infra-  privatization in the infrastructure sector cau-
structure utilities. First, not all components of  tion developing countries to develop a system
infrastructure industries are equally amenable  of checks and balances before privatizing sec-
to competition; therefore, privatization might  tors in which competition has until recently
not be appropriate for all activities in a sector.  been a foreign phenomenon.
For example, relatively low-cost wireless tech-  For several reasons, governments may sell
nologies make most elements of telecommuni-  off state monopolies and may grant whole or
cations potentially competitive, whereas gen-  partial monopoly privileges to new private
eration of electricity is more likely to support  incumbenits. The government may face sub-
competition than is transmission of electricity.  stantial pressure to maximize privatization
The key to successful reform in any sector is,  revenues, and the first metric by which the
therefore, an adequate reform  of market    success of the sale is likely to be judged is the
structure to maximize the potential for real  sales price. Privatizations tend to be contro-
competition. Market structure reform tries to  versial, and the government may be wary of
97



G LO B AL EC O NO MI C PRO SP EC TS 2 0 0 3
being accused of giving away the crown jewels  tive environment and the monopolists face the
if the sale price is too low. This wariness, plus  same cost of capital, and neither will invest
a need to build support for privatization, may  unless the expected revenues make the invest-
create an incentive to generate a high sales  ment worthwhile. The monopolist's market
price, even at the expense of future improve-  power makes it less, not more, likely to un-
ments in the network. These pressures may     dertake a given investment because monopoly
have been especially intense during the first  profits are typically obtained by providing
privatizations when there was little evidence  lower quantities of the good or service at
that privatizations could be successful or that  higher prices. A firm with a guaranteed mo-
failing state-owned firms could attract private  nopoly is also likely to invest less because it
investors.                                    does not have to worry about more efficient
Consider the growth rate of networks in     competitors stealing market share. Even the
telecommunications when     investors were    threat of entry-which is typically the situa-
given "exclusivitv"-temporary    monopoly     tion when reforms are introduced-can be
rights-compared with when they were not.      enough to induce the incumbent to invest.
In a sample of about 20 countries that pri-      Indeed, in telecommunications, empirical
vatized their telecommunications firms, one   work consistently demonstrates that competi-
study found that although private investors   tion, typically in the form of mobile providers
were willing to pay more for an exclusivity   (which have much lower fixed costs than
period (figure 3.13), telecommunications in-  wire line firms) is extremely successful in im-
vestment was substantially lower in countries  proving telephone penetration (for example,
that gave exclusivity periods than in countries  Fink, Mattoo, and Rathindran 2002; Galal
that did not (Wallsten 2000). In other words,  and Nauriyal 1995; Li and Xu 2001; Ros
investors wcre likely paying for the expected  1999; Wallsten 2001a, 2001c). Figure 3.14
streaimn of monopoly profits, not for the right  illustrates how the penetration of the mobile
to invest.                                    telephone market in Africa is influenced by
Another reason for granting monopolies is   competitive versus monopolistic regimes.
the mistaken belief that restricting competi-    However, introducing competition-even
tion can stimulate investment. As Noll (2000)  when technically feasible-can be difficult.
notes, both the firms operating in a competi-  Incumbent firms can use their considerable
Figure 3.13 Granting monopoly rights          Figure 3.14 More competition means
brings in revenues                            more phones
(dollars per line)                            (mobile subscnbers per 100 Inhabiltants)
3,000                                         4 0
2,5003 
2,500        ,                              ~~~~~~~~~~~~~~Competitive market S  r_
2,000                     ,2;_                              _  __;_
1,500                                         20
500                              _ _
0 5
Noexclusivity    Exclusivity             1996 1998 2000 2001 1996 1998 2000 2001
Niotr  Average price paid in releconii privatiations  Souirce African Telecommtniiication Research Project
Souirce World Batik sraff                     Database, DECRG, World Bank
98



D O M E S T I C  P O L I C I E S T O  U N L O C K GC  OIi B A lO  P o R T U N I I I E S
market power to ensure that competitioni never  Table 3.1  Profitability on equity,
succeeds. Taking full advantage of the compet-  concentration, and market share (percent):
itive forces in the global economy requires     Brazil, 1971-78
introducing a regulatory    framework   that    Four-firm
maximizes competition. Establishing a clear     concentratio-              -
regulatory framework in advance of privatiz-     amo (CR4)!        I      L
ing companies is key to achieving a competitive
outcome. Wallsten (2002) studied 200 coun-          20       l              -               |
40             i 4 - 
tries from 1985 to 1999 and has found that,         60       l             _
60       l*.   1 4   -| .
in telecommunications, creating a regulatory        80       14 I   I   1-  I    I- I
capacity before privatization is significantly     100       14 -   I       -    i -   i - 
and positively correlated with subsequent       a CR4 is the ratio of foor largcst plants to total uicdUstry
performance (using measures of capacity and     sales, weightLcc by the fotir-digit grotip product grouip to sales
of firni
investment). Moreover, earlier existence of a   b RMS is the ratio of firmn's sales to indLuistr) sile,, weighted
regulator seemed to increase the price received  by four-digit produLct sales of the firm
for privatized telecommunications flrms by      Source Newfarimicr and Marsh (1994) FigUres are hased o
regressioni coefficicents holding other structural vai iahles (for
reducing uncertainty over the future stream of  example, size, leverage, capital iitenisity) at their 1eacis
earnings.11 Regulatory agencies are discussed
in more detail on page 27 and page 33.
found a statistically significant relationship
Private barriers to competition are             between concenitration and firm  profitability
often difficult to identify and can             in Brazilian manufacturing (table 3.1). Similar
be pernicious                                   results were reported by Connor (1977) for
Even if policy barriers to competition are re-  Brazil and Mexico.
moved, private firms-usually in concentrated       Interpreting these results, however, re-
industries-can raise barriers to competition.   quires care. Both concentration and profits
In particular, dominant firms can exercise their  could be high because firms exercise market
market power to prevent entry by competitors    power and block entry, or because better,
in order to keep prices and profits high. Such  more efficient firms are morc likely to succeed,
anticompetitive behavior may be especially      to capture higher market shares, and to be
prevalent among newly privatized firms in       more profitable    (Bresnahan  1989; Feeny
industries that are traditionally dominated     and Rogers 2000). Nonetheless, empirical
by a single firm, such as telecommunications.   research-primarily in industrial countries-
Another form of private barriers is collusive   demonstrated that there is a great deal of
behavior-often in the form of cartels-to fix    market power in some industries and that
prices and discourage entry.                    anticompetitive conduct can lead to high
Early research explored links first between   price-cost margins (Bresnahan 1989). And, as
concentration and profitability and then be-    Weiss (1989) noted, "[lUn smaller lands and/or
tween concentration and prices. The underly-    in nations with less enthusiasm  for antitrust
ing hypothesis in this line of research was that  [than in industrial countries], the problem
firms in highly concentrated markets would      must surely be greater."
earn higher profits (implying monopoly prof-       Many believe that markets in general are
its) and would be able to charge higher prices.  less competitive in developing countries. With
In general, empirical work supported this       the exception of Brazil, China, India, and
view, finding that firms in highly concentrated  Indonesia, domestic markets tend to be small,
markets were more profitable and charged        with low human capital, poor infrastructure,
higher prices (for example, Weiss 1989).        volatile economies, and few    manufactured
In addition, Newfarmer and Marsh (1994)         inputs produced domestically. Surprisingly,
99



G LO BA L ECO NO MI C PRO SP EC TS 2 0 0 3
though, some evidence suggests that manufac-  horizontal and vertical restraints, of abuse of
turing sectors, on average, are not less com-  dominant market position, and of mergers
petitive than elsewhere. As Tybout (2000)   and acquisitions [M&A1) are similar, their
notes, "[B]ecause of institutional barriers,  scope, institutional design, budgets, staffing,
labor market regulations, poorly function-  and other resources vary widely. Competition
ing financial markets, and limited domestic  laws generally complement and buttress other
demand, the industrial sectors of developing  policies, such as policies on deregulation, pri-
countries are often described as insulated,  vatization, and trade and investment liberal-
inefficient oligopolies. To date, however, there  ization, that enhance competition. However,
is little empirical support for this characteriza-  the overzealous application or misapplication
tion. Turnover is substantial in developing  of competition law in the context of weak
countries that have been studied, unexploited  administrative capacity can also have serious
scale economies are modest, and evidence of  negative consequences. Effectively implement-
widespread monopoly rents is lacking."      ing competition law requires an adequately
Nonetheless, he notes, "[lit would be foolish  funded agency with well-trained, knowledge-
to conclude that market power is a non-issue  able, and experienced staff members. This is
in developing countries."                   a challenge in industrial countries and even
more so in the developing world.13 In this re-
Collusive behavior and domestic cartels     gard, some developing countries have made
limit competition                           noteworthy progress, but it is still too early to
A single firm  abusing a dominant market    form an overall view of the effectiveness of
position is not the only way firms can engage  their competition agencies. International in-
in anticompetitive practices. Vertical restraints  vestors have raised the issue that the prolifer-
between manufacturers or suppliers and      ation of competition laws has led to higher
downstream distributors in the form of exclu-  costs for M&A transactions-a primary vehi-
sive dealing and geographic market restric-  cle for FDI. And in some cases, the decisions
tions can also raise barriers. In addition, firms  arrived at by the competition authorities are
that would be price-takers individually-and  highly questionable.
unable alone to control any significant part of  The remedy for anticompetitive conduct of
the market-can work together to control     firms necessarily depends on a country's ca-
the market, thus increasing prices and dis-  pabilities. As the first order of business, all
couraging entry. Collusive behavior is not  countries are well advised to look for ways to
uncommon, and competition authorities in    reduce policy barriers to competition. Small-
developing countries have prosecuted several  market countries in particular can look to
cases of price-fixinig, as the illustrative list in  trade to discipline domestic pricing. Govern-
table 3.2 suggests. In one colorful example of  ments in countries with weak regulatory
a bid-rigging conspiracy in the electrical equip-  capacity, high levels of corruption, and poor
ment industry (high-voltage switchgears),   accountability would be better advised to do
participants used the phases of the moon to  the following: first, limit the powers of a com-
determine which firm's turn it was to submit  petition agency to review of government poli-
the "low" bid.12                            cies for their competitive consequences and,
During the past decade, a number of devel-  second, concentrate on improving information
oping and transition market economies have  and reporting requirements of firms so that
adopted or strengthened existing competition  increased transparency will attract entry.
laws (see box 3.2). More than 90 countries  Trying to establish more comprehensive com-
have such legislation; more than half the laws  petition authorities in countries without an
were enacted since 1990. Although the core  appropriate legal-economic framework may
provisions of these laws (addressing issues of  simply create another avenue for corruption
100



DO ME ST IC       PO l IC I ES      IO   UN l O C K     G l O B AL    O P PO R r UN IIII S
Table 3.2 Cartel enforcement in selected developing countries
Country                Year                        Market                                 Acuons
Bulgaria            2000               Intenrediate transportation            Price-fixing
2000               I'honc cards sales                    Price-fixinig
2002               Gasification                           Contracts with non-compete clauses
China               1998               School building                        Bid-rigging conspiracy
1998               Engineering construction               Bid-rigging conispiracy
1999               Brickyard                             Bid-rigging conspiracy
Estonia             1999               Taxi services                          l'rice-fixing
1999               Road transport                        Price-fixing
2000               Milk produicts                         Price-fixing
Indonesia           2000               Pipe andl pipe-processilng services    Bid-riggiiig
Latvia              1998-99            Aviation                               Cooperatioii in organization
of passenger flights
1959               Courier post                          Agrecissent betweeti two
postservice coinpaniies
l'eru               1995-96            Poultry market                         Price-fixing, voluime control, and
cinspiracy to establish entry barrier
1997               Building and constructioii             Bid-rigging
1999               Taxi tours                            Price-fixing
Romania             1997               Mineral water                          Price-fimng
1997-2000          Drugs                                 Conspiracy in market-sharinig in
pharrmaceutical distribution
Slovenia            2000                Electricity                          lPrice-fixing
2000               Organization of cultural events        Cooperation anid establishment of
enitry bartiers
South Africa        1999               Citrus fruits                          Conispiracy relatinlg tri the purchase,
packaging, anid sale of citrus fruits
Taiwan, Chinia      1997-98            Wheat                                  Buyer's cartel imposing quatitry
control and quota system
1998               Mobile craines                        Bid-rigginig
Liquefied petroleum gas               Pu ce-fixing
Ukraine             1999               Electronlic cash nsachines             Price-fixing
2000               Kaolin                                 Noncompete contract
Zambia              Not available      Ploultry                               Agreement foreclosing competition
1997               Oil                                   Price-fixisig
Souirce OECD (2001)
and   rent seeking. Governments In countries               Regulatory agencies may help promote
with stronger regulatory capacity have many                competition, but one size does not fit all
options    that go    beyond    policy   review    for     One way that regulatory authorities can play
competitive consequences and          for improved         a positive    role  in  encouraging     competition
disclosure. They     may    be able    to  prosecute       and investment has to do with bringing com-
price-fixing and other horizontal restraints, as           petition to   industries that are dominated        by
well as prosecute      restrictive  marketing     and      a small number of firms or to          industries in
other vertical restraints that hobble entry.               which    cartels have developed. For example,
101



G LO BA L EC ONO MI C      PROS PE CT S 2 0 0 3
Box 3.2            ComperHirdin pohcy and compeaion Iaw
share sinilar objectives across ccmun%11i
n recent years, many countries have enacted spe-    predatory pricing) is more difficult to enforce
Lcific legislation to safeguard and encourage com-    because authorities must focus not on the firm's size
petition. Enforcing competition (or antitrust and     or dominance itself (which is not illegal) but rather
antimonopoly) laws can increase welfare and can       on the "abuse" of a dominant position. Competition
improve efficiency by combating the negative exter-   agencies must have the expertise to distinguish
nalities generated by anticompetitive firm behavior,  between dominance resulting from superior business
The focus and objectives of competition agencies      practices and dominance from erecting anticompeti-
entrusted with this task vary across countries:       tive barriers.
some, such as Canada, New Zealand, and the          o Vertical restraints between manufacturers, suppliers,
United States, place the emphasis on consumer wel-    and distributors (such as resale price maintenance,
fare and economic efficiency, while others, such as   exclusive dealing, and geographic market restriction)
Brazil and the European Union (EU) member coun-       can be tricky because such measures can improve
tries, look to serve the broader "public interest." But  efficiency just as easily as discourage competition.
even with these differences in scope, the underlying  The emerging consensus is that adverse effects are
principles are similar.                               more likely to dominate if the participant firms enjoy
The conduct provisions of competition law         a certain degree of market power. Therefore, vertical
relate primarily to the following:                    restraints should be evaluated within the context of
AOD market competition laws.
o Horizontal agreemnents are entered into by firms to
fix prices (and agree to similar practices such as    The structural provisions relate primarily to the
bid-rigging; restricting output; and allocating market  following:
shares, geographic markets, or customers). Such
agreements represent an unambiguous welfare loss to  o Mergers and acquisitions, where the principal con-
consumers in terms of reduction in price or output  cems arise in horizontal transactions, and joint
competition. Firms that enter into these agreements  ventures compose two structural approaches. Two
are severely prosecuted and, in some countries      different views are prevalent. First, when transactions
(Canada and the United States), such conduct is     significantly reduce firm numbers or increase concen-
treated as a criminal offense, with CEOs liable for  tration, competition may substantially decrease.
imprisonment. However, such anticompetitve behav-   Second, transactions may be strongly motivated by
ior is often difficult to investigate because managers  efficiency goals, and substantial anticompetitive out-
generally avoid written communication. As a conse-  comes are likely only if there are barriers to entry or
quence, some countries have adopted amnesty or      to new competition. Because most horizontal M&A
leniency programs for cartel members who are the    activity will lessen competition but may also increase
first to "blow the whistle" against other members.  efficiency, a cost-benefit approach is often pursued in
Encouraging new entry by removing both private      which mergers are exempted or are permitted to
and policy barriers may be the best policy to combat  proceed on a restructured basis if the efficiency gains
horizontal agreements because collusive behavior    are likely to be greater than the competition losses.
drops as the number of firms rises.
o Abuse of dominant (AOD) market position (that is,
monopolistic practices such as market foreclosure and  Source: Khernani (2002).
Galal and Nauriyal (1995) compare the per-      balance regulatory objectives: commitment,
formance of the telecommunications sector in    information asymmetry, and pricing issues. In
several countries before and after reforms as   their sample, they find that the country
they explore how well countries were able to    (Chile) that resolved all three issues achieved
102



D OM N E S f I C  P o L I C I E S T O  U N L O C K G L O B A L O P P O R T U N I T I E S
the greatest improvement, while the country  tion "iS generally associated with greater cor-
(the Philippines) that did not experienced the  ruption and a larger unofficial economy, but
worst performance. Countries that resolved   not with better quality of private or public
some issues but not others experienced mixed  goods ... The principal beneficiaries rof reg-
success. A more recent study of competition,  ulationJ appear to be politicians and bureau-
privatization, and regulation hints at the im-  crats themselves."
portance of effective regulatory institutions   This observation does not mean that devel-
(Wallsten 2001 a). Like other research (for ex-  oping nations are doomed to failure when
ample, Petrazzini 1996; Ros 1999), the study  building effective regulatory institutions. It
finds that competition resulting from priva-  also does not detract from the general point
tization positively affects network growth,  that introducing competition in potentially
but it also concludes that privatization brings  competitive sectors that are dominated by a
greater benefits in the presence of an indepen-  single firm requires competent regulation that
dent regulator.                              both protects consumers and assures investors
Given the potential importance of regula-  that their assets will not be expropriated. In-
tory institutions in promoting competition, It  stead, as already discussed, it suggesrs that
rnay seem surprising that regulation has been  such agencies should focus on promoting entry,
given relatively little emphasis in developing  not regulating it, and that they themselves
countries. Three factors may have worked to  should operate in an especially transparent
diminish the focus on regulation. First, the pri-  fashion to gain credibility. This feat is not easy
vatization wave was picking up strength lust  to achieve, and such agencies must find the del-
as the United States and other industrial coun-  icate balance between accountability and inde-
tries were engaged in a process of deregula-  pendence from short-term political pressures.
tion, which often meant removing government
controls to allow the industry to compete or
to encourage new entry. Second, privatization  Public investment in infrastructure
in developing countries often faced competing  and human capital
objectives because governments want not only  11//hle the government plays a crucial role
to maximize revenues from privatizing state-  VW   in providing a general framework to
owned firms but also to improve the delivery  encourage investmiient and in establishing the
of service by firms in the industry. The easiest  conditions that use competition to create effi-
(and most common) means to increase the      ciency, its role as a direct investor is pivotal
firm's value for private investors is to include  in shaping investment climate. There is some
monopoly rights in service provision, but, Ull-  question as to what effect public investment
fortLnately, precluding competition is likely to  has on private investmiient (see box 3.3). More-
retard investment.                           over, all governments make public investments
Third, the challenge of building effective  that work through several channels: Govern-
regulatory agencies is enormous and will not  ments can invest directly in physical and
automatically lead to better outcomes. These  hutnan infrastructure provision. In addition,
agencies are costly, require tremendous capac-  their involvement in less-tangible areas (pro-
ity in terms of human rcsources, and probably  viding policy stability, setting standards, and
work best in the presence of complemenitary  establishing legal and regulatory frameworks)
organizations such as competition agencies.  affects opportunities even in areas in which
Moreover, there is little evidence that, in gen-  direct government involvenment is minimal. In
eral, regulatory agencies in developing coun-  this section, we wvill evaluate the scope and
tries have been successful. Regulation often  rationale for governmenit engagement in the
takes the form of regulating entry, and, as  areas of infrastructure and human capital
Djankov and others (2002) document, regula-  provision.
103



G LO B AL EC ONO MI C PRO SP EC TS 2 0 0 3
Box 3.3          Does pubhc investmet "6crowd out" or
6crowd in" private investment?
W      hile direct government investment in the econ-  Recent results from the developing world are ambigu-
omy is necessary, direct expenditures should  ous and show little consistency. For example, Ahmed
be targeted carefully, because resources are scarce. In  and Miller (2000) find general evidence of crowding
particular, governments should not invest where the  our, but note that public infrastructure expenditures,
private sector is willing and able to go. A government  such as spending oin transport and communication,
can crowd out private investment in two general  seem to crowd in private investment.
ways. First, it can invest in areas where the private  C hura and Goodwin (2000) report a crowd-in
returns are likely to be high. Such investmenits  result for a sample of 31 countries, but find more
may tempt politicians because they can then tout  variation on the regional level: Asian and Latin
"successful" investments on the basis of their high  American countries exhibit crowding out, while Sub-
returns. But such successes are illusory because  Saharan Africa shows that public and private invest-
private investors would have undertaken the invest-  ment are complements. In a slightly smaller sample,
ments anyway, and there is an opportunity cost from  Herrera and Garcia (2000) find crowding out in
the suboptimal use of scarce government resources.  both Latin America and East Asia, with the effect
Second, using high governmlenit deficits and borrow-  much stronger for Latin American countries.
ing to fund public investments can indirectly crowd  Everhart and Sumlinski (2001) add SOEs to the
out private investment by means of macro channels  definition of public investment and, with a sample
such as pushing up interest rates-and thereby raising  of 62 developing countries, find that corrupnon
borrowing costs-or creating credit constraints for  exacerbates crowding out. In fact, the effect of the
private investors. Conversely, some government   corruption interaction variable is so significant that
spending can also crowd in investment by attracting  in thc long run there may well be no crowding out
additional private investment that would not other-  if no corruption is present
wise occur. The government may need to build         Finally, Wang (2002), using ainual data for
certain parts of the infrastructure to attract private  East Asian economies (developing as well as indus-
investors, to build roads that connect rural areas to  trial), finds evidence of substantial gains from positive
markets, and to make education universally available.  externalities generated by the public sector, therefore
In any particular country, it is likely that some  hinting at crowding in. However, he cautions that in
public investment choices will crowd out private in-  the long run the influence of private production on
vestment, while other choices will have the opposite  public capital expansion is stronger than the reverse,
effect. It is, therefore, not surprising that the empiri-  which might indicate a causal chain in which higher
cal research on which effect dominates is mixed.  private demand leads to greater public investment.
Aschauer (1989) argues that while public investment  There is thus no consensus within the literature
reduces private investment almost one-to-one by  on whether crowding in or crowding out dominates.
encouraging the private scctor to take advantage of  Empirical results are often sensitive to sample choice
public capacity instead of building its own, it is also  and vary with regard to individual countries and
true that public capital (and infrastructure capital  time periods. Hence, whether crowding in or crowd-
in particular) complements private capital in the  ing out dominates may depend on complex interac-
production of goods and service. Hence, public   tions betweeni private and public investment. It is
investment raises the marginal product of private  possible that, up to a certain level, higher public
investment so that-despite the direct negative   investment may encourage private investment and
effect-the long-term consequence of an increase in  growth. However, when undertaken in excess (or
public investment on private investment is positive.  when existing capacity is used inefficiently), any
These results, however, were based exclusively  additional increases in the level of public capital
on the United States, and a subsequent wave of cross-  may crowd out private investment.
country tests of the crowd-out or crowd-in hypothe-
sis has thus far failed to yield any clear conclusions.  Source: World Bank staff
104



D O M E S T I C PO LI C I ES  TO  UN LO C K G LO B AL O P P O R T U N I T I E S
Infrastructure affects opportunities          with stronger climates, there is significant
for growth                                    room for improving the climate in particular
The quality and availability of infrastructure  areas: moving to the best Indian practice
has a major effect on investnment opportunities  would add 1.5 percentage points to the
in the private sector. World Bank (2002) notes  growth rate for these states. Note that in many
that "improvements in infrastructure services  ways this is a conscrvative counterfactual
can help promote competition in other mar-    scenario because it would raise states to the
kets, and there is evidence that infrastructure  levels of regulation and infrastructure quality
has a positive impact on growth and poverty   that are already observed in India. If India
reduction." In a sample of 100 countries,     could achieve Chinese or Thai levels in various
Easterly and Rebelo (1993) attach an impor-   investment climate areas, its potential growth
tant role to infrastructure capital-particularly  acceleration would be even more dramatic
transportation and communications-in eco-     (World Bank 2002a).
nonmic growth. Elemenits of infrastructure such  The efficiency of infrastructure capacity
as paved roads, telephone density per worker,  utilizationi is just as important as (if not more
and adequate generation of electricity have    important than) the capital stock itself.
been found to have a strong effect on growth  Easterly and Levine (2001 ) propose that "cre-
(see Easterly and Levine 1997; Canning 1999;  ating conditions for productive capital accu-
and Caniing and Bennathan 2000). Even in      mulation is Imlore importanlt thanl accuminula-
industries that have very low requirements    tion per se and policynmakers should focus on
for energy and transportation services, such as  encouraging TFP Itotal factor productivityl
the software industry, the quality and avail-  growth." Hulten (1996) notes that those low-
ability of infrastructure play a key role in se-  and middle-income countries that use infra-
lecting firm locations because firmns rely on  structure inefficiently pay a growth penalty in
satellite facilities to export their products  the form of a much smaller benefit from infra-
(Balasubramanyanm 2001). In addition to pro-  structure investments. More than 40 percent
moting economic growth, greater coverage of   of the growth differential is due to the
infrastructure services is also a key detcrini-  efficiency effect, making it the single most
nant of FDI (Balasubramanyam 2001; Stein
and Daude 2001).
Infrastructure is a key determinant of the     Figure 3.15 Better infrastructure means
quality of a nation's investment climate. A      higher growth
recent survey study that Iilked quantitative     (annualaverage GDP growth rate 1992-98, percent)
measures of the investnment climate to firm
investment and growth experiences demon-         9
8
strates the potential for improvements in in-    7
frastructure. The study, which is based on       6
more than 1,000 firms in 10 Indian states,       5-                                 _
finds that if each state could attain the "best  4-
practice" in India in terms of regulation and    2       hi             .
infrastructure, the national economy could       1-
grow about 2 percentage points faster (see fig- o
ure 3.15). The gains would be particularly          Good and best  Poor climate  All India
large in the states with weaker investmenit         climate states  states
climates (an extra 3.2 percentage points of      Note Sirvey of more than i,000 firmns in 10 Itidiani states
growth), thus reflecting the fact that the move  Indtia in terms of infrasnent of r the best nd ractgceaiD
from current to best practice in India would     Sotirce World Bank staff
be a large improvement. But even in the states
105



G LO BA L EC O NO MI C PRO SP EC TS 2 0 0 3
important explanatory variable in differential  that for a sample of 98 countries, the growth
growth performance. Similarly, Aschauer     rate of real per capita GDP during 1960-85
(2000) attributes an important role to the  was positively related to initial human capital
efficiency variable, although he cannot reject  (proxied by 1960 school-enrollment rates).
the hypothesis of parallel importance of the  Figure 3.16 illustrates this concept by showing
quantity and effectiveness of public capital at  a clear positive relationship between the 1970
conventional levels. Aschauer (2000) also cal-  literacy rate and the growth in GDP per capita
culates the growth-maximizing level of public  between 1970 and 2000 for 75 developing
capital, which is vastly exceeded by the actual  countries.
sample average of 46 developing countries.    Easterly and Levine (2001) caution that
Thus, it would seem that the average country  economic growth differences across countries
in his sample has overspent on capital expen-  cannot be easily explained by factor (including
ditures, thereby lowering the productivity of  human capital) accumulation and should
its public investment program.              focus instead on technology and productivity
growth. However, the success of dissemina-
Investments in human resources are          tion of more advanced technologies in devel-
critical                                    oping economies is largely determined by the
Human capital is widely recognized as an    absorptive capacity of the host country. That
important determinant of development and    is, to realize the growth potential of new
growth. Seminal work by Manklw, Romer, and  technology, the country must possess a high
Weil (1992) demonstrated a significant im-  enough stock of human capital to be able
provement in the explanatory power of the   to assimilate the technology. For example,
Solow growth model when it included mea-    Borensztein, De Gregorio, and Lee (1998)
sures of human capital. Similarly, many en-  show that the magnitude of the effect of FDI
dogenous growth models have benefited from  on growth depends on the available stock of
the inclusion of an education variable (see,  human capital in the host country. Within an
forexample, Romer 1990). Barro (1991) found  endogenous growth model, the researchers
Figure 3.16 Greater literacy is associated with higher growth
(1970 literacy rate, in percent)
120
100
80                °
60         _
*            C~~~~~~~~~`
20o ° °
e         *           Cu                          *
20
-100       -50         0         50        100        150        200        250
Percent GDP per capita growth rate, 1970-2000
Source World Bailk data
106



D O M E S T I C PO LI C I ES TO UN LO C K G LO B AL O P P O R T U N I T I E S
obtain a positive and highly significant coeffi-  human capital suffer the opposite effect
cient on the interaction variable between FDI  (Herrera and Garcia 2000). Thus, high levels
and human capital. The results suggest that  of human capital may help increase the overall
"the flow of advanced technology brought   level of investment through a crowd-in mech-
along by FDI can increase the growth rate of  anism. Countries with higher human capital
the host economy only by interacting with   also have lower fertility rates and higher ratios
that country's absorptive capacity."        of physical investment to GDP. Some evidence
Several other studies have looked at the  suggests that additional government expen-
relationship between FDI and human capital.  diture on education induces additional private
For example, Coughlin and Segev (1999),     expenditures on education. For instance,
Noorbakhsh, Paloni, and Youssef (2001),     Foster and Rosenzweig (1996) show that in
and Kolstad and T0ndel (2002) show     a    India higher returns to primary schooling
positive link between FDI inflows and the   actually induce increased private investment
stock of human capital in the host country.  in schooling.
Balasubramanyam  (2001) notes that human       Despite the overwhelming consensus that
resources are a key determinant of FDI.     human capital is one of the keys to sustained
Countries with the highest levels of both  economic growth, finding a robust empirical
schooling and FDI grew much faster than     relationship between education and growth
countries with the lowest levels in the pe-  has proven difficult (see Easterly 2001 for a
riod 1970-89 (figure 3.17). Human capital is  review). One striking example lies in compar-
also important as an interaction variable be-  ing East Asia to Sub-Saharan Africa: Between
tween FDI and domestic private investment.  1960 and 1985 East Asia's per capita GDP
Countries with high levels of human capital  grew more than 4 percentage points quicker
seem to experience crowding in of domestic  than incomes in Africa, yet Africa's educa-
investment by FDI, while countries with less  tional capital growth was actually higher than
Asia's (Pritchett 1999). However, part of the
answer to this puzzle emerges from the multi-
dimensionality of the investment climate:
Figure 3.17 Education raises the         education matters only if people are given
productivity of FDI, which leads to higher  opportunities to use their skills in productive
growth                                   industries in a supporting enabling environ-
(per capita GDP growth rate, in percent)  ment. Easterly (2001) contends that econo-
5                                        mies with low    black-market premiums14
4            /High schooling             on foreign exchange grow faster with higher
4          /                             schooling levels, while economies with high
3        *                               black-market premiums grow slowly regard-
2                   Medium schooling      less of the levels of education. That is,
2                         L so            "schooling pays off only when government
__  1_1        ~~~~~Low sch oling
actions create incentives for growth rather
mU w 1 wthan redistribution."
High      Medium      Low
FDI-to-GDP ratio            Policies to promote competition
Note The low, medium, and high categories for FDI-to-  R1'hile competition and entrepreneurship
GDP ratio are below 0.OI percent, 0 01 percent to 02 2AI
percent, and more than 0 2 percent, respectively For the  VV are essentially private sector activities,
schooling variable, the low, medium, and high categories  they require markets that function well. And
are below 0 4, 0 4 to 1, and more than 1, respectively
Souirce Borenszteun, De Gregorio, and Lee (I998).  it is up to governments to ensure an environ-
ment in which markets remain contestable
107



G LO BA L E CO NO MI C PROS PE CT S 2 0 0 3
and entrepreneurship is rewarded, which is    Competition and regulatory agencies can
not easy. Entrenched interests are powerful,  be instrumental in reducing abuses of market
and it is often hard to determine whether   power and in ensuring that markets remain
any particular program is largely in the pub-  contestable. Agencies can work toward this
lic interest or in the interest of a much smal-  general vision by focusing on two objectives:
ler, but more vocal, private constituency. In  protecting consumers while ensuring that the
general, though, following certain basic prin-  regulatory and market rules are credible to
ciples can help promote competition and     investors. These objectives, however, may be
growth.                                     difficult to balance when interests compete for
Governments and government agencies      regulatory favor. Moreover, there is the risk
should operate with transparent rules, should  that a new regulatory agency will become
minimize corruption, and should respect     another avenue for corruption, especially in
property rights. They should also make it eas-  countries with very poor investment climates.
ier to start and run businesses. The maze of  An agency will be better able to accomplish its
bureaucratic paperwork that is often required  objectives of correcting market failure while
to start businesses in developing countries  avoiding government failure if it meets sev-
seriously deters entry into many industries.  eral criteria. In particular, it must operate in
Moreover, such administrative hassles can be  a transparent manner, be accountable, be in-
especially pernicious: in some cases they may  dependent from  short-term  political pres-
punish small, local entrepreneurs who lack the  sures, have limits on its discretion, and have
resources to overcome such high hurdles.    adequate capacity to do its job.
Having more government agencies that can      The downside associated with failing to
block a firm's path will lead inexorably to  meet these criteria can be severe. For example,
more points at which a firm is required to pay  investment will be difficult to attract if regula-
bribes to move the process forward.         tory policies can be easily changed to benefit
The government's role extends beyond     any given politician's short-term  objectives.
setting up a generally investment-friendly en-  Likewise, an agency that is not transparent
vironment. Until the past decade or so, SOEs  and accountable runs the parallel risks either
have had monopolistic positions in many in-  of frightening away investors or of being cap-
dustries throughout the developing world.   tured by the industry it is supposed to regulate
The recent wave of privatizations not only has  at the expense of consumers. Without limits to
led to large efficiency improvements in these  its discretion, meanwhile, an agency may seek
firms and their provision of services, but also  to expand its influence into new areas and
has opened those industries to competition.  may become primarily another obstacle to
The greatest improvements in service have oc-  development and an avenue for rent seeking.
curred in industries in which the government  Finally, if the agency lacks the capacity to do
promoted competition along with priva-     its job, it will simply be ineffective.
tization and in which it avoided giving the   This range of criteria highlights the point
privatized firm any special monopoly rights.  that-especially in regulatory and competi-
Privatizations are often difficult and contro-  tion agencies-one size does not fit all. The
versial. However, governments should be     optimal type of regulatory and competition
aware that while they can usually increase the  agency (if any) depends not only on the con-
price that investors are willing to pay for a pri-  ditions of the market (for example, to what
vatized firm by giving the firm a monopoly,  extent an individual firm can exercise market
that same exclusivity usually lowers subse-  power to thwart entry), but also on the extent
quent investment. That is, investors will be  to which the country is likely to be able to
paying for the stream of monopoly profits,  credibly design and run an institution that
not for the right to invest more.           meets these criteria. Larger, more stable
108



D O M E S T I C  P O L I C I E S  r O  U N L O C K  G L 0 B A L   0 P P O R T U N I T I E S
countries with effective existing bureaucracies     that they imperfectly reflect a country's trade policy
are more likely to be able to meet all the crite-   regime. The high correlation of components of the
ria. Other countries may facc great difficulties.   Sachs and Warner index with policy and institutional
variables yields an upward bias in the estimation of
trade restriction effects. Meanwhile, tariff and nontar-
quately staff an agency can be quite sizeable,      iff barriers, the two variables that directly mcasure
potentially making it unrealistic for a small,      trade openness, have little cxplanatory power when
poor country. Some have suggested that when         considered separately in cross-country regressions
resources and skills are scarce, countries could       5. Patent citations refer to a requirement in some
work together to create regional agencies in        patent offices that inventors include in their patent
order to share the costs and responsibilities.      application the citations of the patented technology
with severe problems of corruption       that they used in developing their inventioni (see
Branstetter 2000) These citatiois are used as evidence
and with a lack of transparency, meanwhile,         of technological spillovers.
may have difficulty convincing consumers and           6 However, Keller (1998) finds that the role played
investors that a new agency would behave dif-       by import shares in determining productivity levels is
ferently from how the government behaved in         limited. Using the Coe and Helpmian (1995) model
the past. A governmcnt intent on overcom-           with randomly generated import shares, he also finds
ing this reputation and on encouraging com-         a positive relationship between foreign R&D  and
productivity
petition  may make some progress in two             poutvt
7. The entry cost ilmeasure used in figure 3 6 refers
ways: increasing the amount of publicly avail-      to the costs of obtaining the necessary permits and
able information on both firms and govern-          licenises and the other procedures required to set up a
ment agencies, and taking special steps to          new establishment See Djankov and others (2002) for
ensure the transparency of any new initiatives      further details
while emphasizing the discretionary limits of          8 See Hallward-Driemeier, larossi, and Sokoloff
those agencies.                                     (2002) for a longer discussion; Levinsohn (1993),
Haddad and Harrisoni (1993).
9. Of the various regulatory agencies that are seen
as obstacles, customs officials ranked second-only
Notes                                               behind labor regulators-as a major constraint to doing
1 Note that these investment categories are not  business in India.
strictly comparable because the FDI flows are taken    10 Delays are similar for clearing imports through
from balance of payments statistics and include for-  customs With such uncertainty, firms are likely to need
eign inflows intended for both new investimient and  to keep greater invenitories of materials on hand, thus
acquisition of existing assets Meanwhile, the other  incurring significant storage costs and tying up
investmenit figures are derived from national accounts  resources that could otherwise he put to more produc-
and refer only to new investmenit. The "domestic pri-  tive use.
vate" category is calculated as a residual and, there-  11. This will not always be the case, of course A
fore, may not match figures available from  other   country could easily enact a regulatory regime that
sources.                                            deters investors and increases the risk premium  Yet,
2. However, his corruption indicator is correlated  on average, regulator)' certainty seemed imiportanit to
with other explanatory variables so that the coefficient  investors.
on corruption is not significant once other explanatory  12. See Scherer and Ross (1990), chapters 7 and 8,
variables are included in the equation              for a description of this case and others about
3 The rise in FDI is moderated because improve-  collusion
ments in institutions are also associated with a reduc-  13. Some commentators have suggested that it is a
tion in FDI as a share of total capital inflows because  mistake to encourage developing and emerging market
other types of capital inflows are more sensitive to  economics to enact competition laws because the risks
instititional quality                               of misapplication are high as a result of weak institu-
4. Not everyone is persuaded by these cross-     tional capacity. Such laws may also become another
country regression results For example, Rodriguez and  form  of government intervention in markets and
Rodrik (1999) argue that some indicators of openness  may give rise to corruption However, suchi objections
are highly correlated with other indicators of economic  could also be applied to other policy areas suchi as rax
performance-including macroeconomic policy-or       collection, bank regulation, and so forth. The main
109



G LO BA L EC ONO MI C       PRO SP EC TS     2 0 0 3
implication instead is that it is important, first, to de-  Economic Growth?" Journal of International
sign a system of checks and balances, including mea-  Economics 45(l)1 115-35
sures for accountability and transparency, and, second,  Boubraki, Narless, and Jean-Claude Cosset. 1998a
to support institutionial building of capacity.       "The Financial and Operating Peformance of
14. The black-market premium here is seen as a     Newly Privatized Firms: Evidence from Devel-
proxy for available opportunities for legal and produc-  oping Countries." Journal of Finance 53(3):
tive employment                                       1081-110.
-_____ 1998b. "Privatization in Developing Coun-
tries: An Analysis of the Performance of Newly
References                                            Privatized Firms." Public Policy for the Private
Ahmed, Habib, and Stephen M. Miller. 2000.            Sector. Washington, D.C.: World Bank.
"Crowding-Out and Crowding-In Effects of the  Branstetter, Lee. 2000 "Is Foreign Direct Investment a
Components of Government Expenditure." Con-       Channel of Knowledge Spillovers) Evidence from
temporary Economic Policy 18(1). 124-33.          Japan's FDI in the United States." University of
Ambrose, William, Paul Hennemeyer, and Jean-Paul      California-Davis, Economics Department Work-
Chapon. 1990. "Privatizing Telecommunications    ing Paper Davis, Calif.
Systems: Business Opportunities in Developing  Braun, Bradley, and W. Warren McHone. 1992.
Countries." International Finance Corporation     "Science Parks as Economic Development Policy.
Discussion Paper. Washington, D C                 A Case Study Approach." Economic Develop-
Arthur, Brian. 1994. Increasing Returns and Path De-  ment Quarterly 6(2) 135-47.
pendence in the Economy. Ann Arbor: University  Bresnahan, Timothy. 1989. "Empirical Studies of
of Michigan Press.                               Industries with Market Power." In Richard
Aschauer, David Alan. 1989. "Does Public Capital      Schmalensee and Robert Willig, eds , Handbook
Crowd Out Private Capital?" Journal of Mone-      of Industrial Organization Amsterdam: Elsevier
tary Economics 24(2): 171-88.                     Science Publishers, pp. 1011-57.
-        2000 "Public Capital and Economic Growth  Brown, Cynthia J. 2001. "Gender in Mexico's Maquila
Issues of Quantity, Finance, and Efficiency " Eco-  Industry." World  Bank, Washington, D.C.
nomic Development and Cultural Change 48(2).      Processed.
391-406                                      Canning, David. 1999. "Infrastructure's Contribution
Balasubramanyam, V. N 2001. "Foreign Direct Invest-   to Aggregate Output" World Bank Policy Re-
ment in Developing Countries. Determinants       search Working Paper Washington, D.C.
and Impact " New Horizons for Foreign Direct  Canning, David, and Esra Bennathan. 2000 "The
Investment. Paris: Organisation for Economic      Social Rate of Return on Infrastructure Invest-
Co-operation and Development.                     ments " World Bank Policy Research Working
Barro, Robert J. 1991. "Economic Growth in a Cross-   Paper Washington, D.C.
Section of Countries." Quarterly Journal of Eco-  Carnaghan, Robert, and Barry Bracewell-Milnes.
nomics 106(2): 407-43.                            1993. "Testing the Market Competitive Tender-
Bartone, Carl R., Luiz Leite, Thelma Triche, and      ing for Government Services in Britain and
Roland Schertenleib 1991. "Private Sector Par-    Abroad." IEA Research Monograph: London.
ticipation in Municipal Solid Waste Service:  Caves, Richard. 1996. Multinational Enterprise and
Experiences in Latin America." Waste Manage-      Economic Analysis. New York: Cambridge Uni-
ment and Research 9' 498                          versity Press.
Bergman, Edward M., and Edward J. Feser. 2001.    Charkrabarti, Avik 2001. "The Determinants of
"Industrial and Regional Clusters: Concepts and   Foreign Direct Investment: Sensitivity Analyses
Comparative Applications." Regional Research      of Cross-Country Regressions " Kyklos 54(1):
Institute, West Virginia University, Morgantown.  89-114.
Blalock, Garrick. 2001. "Technology From Foreign  Choudhri, Ehsan, and Dalia S. Hakura. 1999. "Inter-
Direct Investment: Strategic Transfer Through     national Trade and Productivity Growth- Explor-
Supply Chains." Haas School of Business Work-    ing the Sectoral Effects for Developing Countries."
ing Paper. Berkeley, Calif.                      International Monetary Fund Working Paper
Bora, Bqiit. 2002. "Investment Distortions and the    Series. Washington, D.C.
International Policy Architecture." World Trade  Chunlai, Chen 1997. "The Location Determinants of
Organization Working Paper. Geneva.               Foreign Direct Investment in Developing Coun-
Borensztein, E., Jose De Gregorio, and J. W Lee. 1998.  tries." University of Adelaide Working Paper.
"How Does Foreign Direct Investment Affect        Adelaide, Australia.
110



D O M E S T I C  P O L I C I E S T O  U N L O C K  G L O B A L O P P O R I U N I T I E S
Claessens, Stijn, and Leora Klapper. 2002  "Bank-      Mayer, ecis , Privatizationz antd  Economic
rmptcy Around the World. Explanationi for its     Performance. Oxford, U.K.: Oxford University
Relative Use " World Bank Ilolicy Research Paper  Press.
\Vashington, D.C.                             Easterly, William 1999. "The Ghost of Financing Gap
Clarke, George R G , Claude Menard, and Ana Maria      Testing the Growth Model Used in the Interina-
Zuluaga 2000 "The Welfare Effects of Private      tonial Fnalicial Institutionis " Joirnral of Devel-
Sector Participation in Guinea's Urban Water       opinett Econoniiics 60(2) 423-38
Supply" Woild Bank Policy Reseaich Woiking            2001. Thbe Elusive Quiest fol Gioivth
Paper Washington, D C                              Cambridge, Mass M?IT Press.
Claugue, Christopher, Philip Keefer, Stepihen Knack,  Easterly, Williaam, and Ross Levine 1997 "Africa's
and MIancur Olson. 1999. "Contiact-Intensise       Growth Tragedy Policies and Ethnic Divisions
Money. Contract Enforcement, Property Rights,      Qutarterly  Journal of Economics   112(4):
and Economic Performanice " Jouirnial of Eco-      1203-50
nomnic Grovutb 4- 185-21 1                          . 2001. "It's Not Factor Accumulation. Stylized
Coe, David T., and Elhallan Helpillan 1995 "Interna-   Facts and Growth M\odels." World Bank Eco-
tional R&D  Spillovers " European Econoniic        nomic Reviewu 15(2). 177-219
Reviewv 39(5): 859-87.                        Easterly, William, and Sergio Rebelo  1993 "Fiscal
Coe, David T., Elhanan Helpman, and Alexander          Policy and Economic Growth   An Empirical
W   Hoffmaister  1997. 'North-South R&D            Investigation " Jouirnzal of Monietary Economics
Spillo%crs " Econouiicjoiurnal 107(440). 134-49.   32(3) 417-58
Connor, J 1977 Maiket Poever of Mtltiznationals.   Eaton, Jonathan, and Samuel Kortum. 1996. "Trade in
New York. Praeger.                                Ideas Patenting and Productivity in the OECD "
Cooper, R 2001. "Groxvth and Inequality The Role of    Jouirnial of lInterniational Economics 40(3-4)
Foreign Trade and Investmenit." Paper presented    251-78.
at ABCDE Washington, D C                      Ellisonl, Glenin, and Edward L Glaeser 1997. "Geo-
Coughllin, Clerus, and Eran Segev 1999  "Foreign       graphic Concenrtiation in U S Manufacturing
Direct Investment in Chinia: A Spatial Economet-  Indusnies A Dartboaid Approach." Journial of
rics Study." Federal Reserve Bank of St LouIs      Political Econom1y 105(5): 889-927
Workinig Paper St Louis                       Emery, James, MNIelvin T Spence, Louis T Wells, and
Cuadros, A., V. Orts, and NI T. Alguacil. 2001         Timoth) Buelhrer. 2000. "Administrative Barriers
"Openness and Growth: Re-examininig Foreign        to Foreign Investmenit: Reducing Red Tape in
Direct Investment, Trade, and OutpuLt Linkages in  Africa." Internatioial Finance Corpoiation Dis-
Latin  America " University  of Nottiighiam        CLussion1 Paper Washington, D C
CREDIT Research IPaper Nottingham, U.K.       Encarnation, Deniis, and Louis T. Wells 1986
Davenport, H J 1935 The Econiomics of Alfred Mar-      "Evaluating Foreign Investnmenit " In Theodore
shall. Ithaca, N Y' Cornell University Press.      Moran, ed., Investing in Development Neruv
de Ferranti, David, Guillermo E. Perry, Daniel         Roles for Private CaptialR New Brunswick, New
Lederman, and William   F. Nlaloney  2002         Jersey. Transactioni Books
Froni Natural Resources to the Kio0uvledge    Englisi, Phillip, and Luc de Wulf 2002  "Export
Economy. World Bank, NWashington, D C              Development Policies and InstititiolIs." In B.
Devereux, Michael P., and Rachel Griffith. 1998        Hoekman, A. Mattoo, and P. English, eds , De-
"Taxes and the Location of Production Evidence     velopIpient, Trade, antd the WTO A Handbook.
from a Panel of U.S. Multinationals." Jotirnal of  Washington, D C, World Bank
Plibitc Econonmics 68(3)' 335-67.             Everhart, Stephen, and Mariusz Sumlinski 2001
Djankov, Simeon, Oliver Hart, and Tariana Nenova.      "Tiends in Piivate Investment in Developing
2002 "Efficient Insolvency." Background Paper     Countries: Statistics foi 1970-2000 and the
for Doing Business in 2003 report Piivate Sector  Impact on Piivate Investment of Corruption
Advuisory' Services World Bank, Washington, D C   and the Quality of Public Investment." Interna-
Diankov, Simeon, Rafael La Porta, Florencio Lopez-    tional Finance Corporation DIscussion Paper
de-Silanes, and Andrei Shleifer 2002. "The        Washington, D.C.
Regulation of Entry." Quarterly Joufrnal of   Feeny, Simon, and Mark Rogers 2000 "The Role
Economics 117(1): 1-37.                           of Market Share and Concentiation in Firm
Domberger, Simon, and John Piggott. 1994. "Privati-    Profitability  Implicationis  for Competition
zation Policies anid Public Enterprise A Survey "  Polic)y " Economi0Zc Aialysis antd Policy 30(2)
In Matthew Bishop, Johli Kay, and Cohn P           115-32



G LO B AL EC ONO MI C      PRO SP EC TS     2 0 0 3
Felsensteini, Daniel 1994 "University-Related Science  Herrera, Santiago, and Conrado Garcia. 2000
Parks- 'Seedbeds' or Enclaves of Innovation?"    "Aggregate Investment and Foreign Direct Invest-
Technovation 14(2): 93-110.                      ment in Latin America and East Asia- Is There a
Finger, J. M. 1993. Antidumnping. Ann Arbor: Univer-  Crowding-Out Effect." World Bank Working
sity of Michigan Press.                          Paper. Washington, D.C
Fink, Carsten, Aaditya  Mattoo, and  Randeep     Hines, James R. 1996. "Altered States: Taxes and
Rathindran. 2002. "Liberalizing Basic Telecom-   the Location of Foreign Direct Investment in
munications: Evidence from Developing Coun-      America " American Econonmic Review 86(5):
tries " World Bank Working Paper Washington,     1076-94.
D.C                                          Hoekman, Bernard, and Kamal Saggi. 2000. "Assess-
Foster, Andrew D, and Mark R. Rosenzweig. 1996       ing the Case for Extending WTO Disciplines on
"Technical Change and Human-Capital Returns      Investment-Related Policies." World Bank Work-
and Investments: Evidence from  the Green        ing Paper. Washington, D.C.
Revolution." American Economic Revietv 86(4)  Hoekman, Bernard, Hiau Looi Kee, and Marcelo
931-53.                                          Olarreaga 2001. "Markups, Entry Regulation, and
Friedman, Eric, Simon Johnson, Daniel Kaufmann,      Trade- Does Country Size Matter)" World Bank
and Pablo Zoido-Lobaton 2000. "Dodging the       Policy Research Working Paper. Washington, D C.
Grabbing Hand The Determinants of Unofficial  Holmes, Thomas J. 1998 "The Effect of State Policies
Activity in 69 Countries " Journal of Public Eco-  on the Location of Manufacturing Evidence from
nomics 76(3): 4S9-93.                            State Borders." Journal of Political Economy
Galal, Ahmed, and Bharat Nauriyal. 1995. "Regulat-   106(4): 667-70S.
ing Telecommunications in Developing Coun-   Hulten, Charles R 1996 "Infrastructure Capital and
tries " World Bank Policy Research Working       Economic Growth How Well You Use It May Be
Paper. Washiisgton, D.C                          More Important Than How Much You Have."
Galal, Ahmed, Leroy Jones, Tandon Pankal, and Ingo   National Bureau of Economic Research (NBER)
Vogelsang. 1994. Welfare Consequences of Sell-   Working Paper. Cambridge, Mass.
ing Public Enterprises- An Empirical Analysis  Jaffe, Adam  1989  "Real Effects of Academic
New York Oxford University Press.                Research." American Economic Review 79(5):
Ghura, D., and Barry Goodwin 2000. "Determi-         957-80.
nants of Private Investment A Cross-Regional  Jaffe, Adam, Manuel Trachtenberg, and Rebecca
Empirical Investigation." Applied Economics      Henderson. 1993 "Geographic Localization of
32(14): 1819-29.                                 Knowledge Spillovers as Evidenced by Patent
Guisinger, S , and others 1985. Investment liicentives  Citations." Quarterly Jotirnal of Econonzics
and Performance Requirements New York:           108(3) 577-98.
Praeger.                                    Jayanthakumaran, K, and John Weiss. 1997. "Export
Haddad, Mona, and Harrison, Ann. 1993    "Are        Processing Zones in Sri Lanka: A Cost-Benefit
There Positive Spillovers from  Direct Foreign   Appraisal." Journal of International Develop-
Investment? Evidence from  Panel Data for        nient 9(5) 727-37.
Morocco " Journal of Developnient Econoniics  Johansson, Helena, and Lars Nilsson 1997. "Export
(Netherlands) 42: 51-74                          Processing Zones as Catalysts." World Develop-
Hakura, Dalia, and Florence Jaumotte. 1999 "The      nient 25(12). 2115-28
Role of Inter- and Intra-Industry Trade in   Kang, Sung Jin, and Yasuyuki Sawada. 2000. "Finan-
Technology Diffusion." International Monetary    cial Repression and External Openness in an
Fund Working Paper. Washington, D.C              Endogenous Growth Model." Journal of Interna-
Hallward-Driemeier, Mary, Guiseppe larossi, and      tional Trade and Economic Development 9(4):
Kenneth Sokoloff. 2002. "Exports and Manufac-    427-43
turing Productivity in East Asia: A Comparative  Keefer, Philip 1996 "Protection Against a Capricious
Analysis with Firm-Level Data," National Bureau  State: French Investment and Spanish Railroads,
of Economic Research (NBER) Working Paper        1845-1875." Journal of Economic History 56(1)-
no 8894. Cambridge, Mass.                        170-92.
Hausmann, R, and E. Fernandez-Arias 2000. "The   Keller, Wolfgang. 1998  "Are International R&D
New Wave of Capital Inflows Sea Change of Just   Spillovers Trade-Related? Analyzing Spillovers
Another Title." Inter-American Development       among Randomly Matched Trade Partners."
Bank Working Paper. Washington, D.C.             European Economic Review 42: 1469-81
112



D O M E S T I C  PO LI C I ES  IO  UN LO C K   G LO B AL O P P O R I U N I T I E S
2000 "Do Trade Patterns and Technology         of Economic Research (NBER) Working Paper
Flows Affect Productivity Growth"' World Bank     Cambridge, Mass.
Economic Revietv 14(1) 17-48.                Love,James H 1996 "EntryandExit.ACounty-Level
Khemani, R. Shyain. 2002 "The Role of Competition     Analysis." Applied Ecouoni,cs 28(4): 441-5 1
Law-Policy " Unpublished World Bank staff    Lumenga Neso, Olivier, Marcelo Olarreaga, and
report                                            Maurice Schiff. 2001 "On Indirect Trade Related
Kikeri, Sunita, and John Nellis. 2001 "Privatization  R&D Spillivers " World Bank Workinig Paper
in Comperitive Sectors- The Record So Far"        Washington, D.C
Private Sector Advisory Services, World Bank,  Maloney, W F 2001 "Missed OpportuLlities. Ininova-
Washington, D.C.                                  tion anid Resource-Based Growth in Latin Amer-
Knack, Stepheln, and Philip Keefer. 1995 "InstituLtios  ica " World Bank Working Paper. Washingtoni,
and Economic Performance. Cross-Country Tests     D C
Using  Alternative  Institutionial Measures"  Mankiw, N. Gregory, David Romer, and David N
Economnics and Politics 7(3) 207-27               Wel. 1992. "A Contribution to the Empirics of
Kolstad, Ivar, and Luie Tondel 2002  "Social De-      Economic Growthi " Quarterly Journal of Eco-
velopment and Foreign Direct Investments in       noniics 107(2) 407-37.
Developing Counrries." Chr Michelson Institute  McMillai, John, and Christopher Woodruff. 2002
Report 1 1. Bcrgen, Norway                        "The Cenitral Role of Entrepreneurs in Tranisition
Kreye, Otto, Folker Frobel, and Jurgen Heinrichs      Economies " Jomrnal of Economic Perspectives
1987   "Export Processing Zones in Devel-         16(3) 153-70
oping Countries. ResLitts of a New Survey "   Megginsoni, William L., and Jeffrey M  Netter. 2001.
Multinationial Enterprise Programme Working       "From State to Market. A Survey of Empirical
Paper 43. Internationai Labour Organisation,      Studies on Privatization " Jour,ial of Econoniic
Geneva.                                           Literature 39(2). 321-89
Krugmaln, Paul   1991   Geography and   Trade     Moran, Theodore 1998 Foreign Direct lnt'estuient
Cambridge, Mass: MIT Press                        and Development. Washington, D C.: Instittite
1998. "The Role of Geography in Develop-      for Interinational Economics
ment " Annial World Bank Cooferenice on De-          2002  Beyond Siveatshops Foreign Direct
velopment Economics. Washington, D C.             Investment and Globalizationz ii Developinig
Kwoka, John E. Jr. 1996. Power Structture Ownership,  Cointries Washington, D.C.: Brookings InstitIte
Integration, and Competition in the U.S Elec-     Press.
tricity Industry. Dordrecht, The Netherlands:  Morisset, Jacquies, and Olivier Lumenga Neso 2002
Kluwer Academic Publishers                        "Administrative Barriers to Foreign Investment
Lall, Sanijaya. 2001. "Competitiveness Indices and    in Developing Countries " World Bank Policy
Developing Cotintries An Economic Evaluation      Research Workinig Paper. Washington, D.C.
of tie Global Competitiveniess Report." World  Morisset, Jacques, and Neda Pirnia 2000. "How Tax
Developmiienit 29(9): 1501-25.                    Policy and Incentives Affect Foreigin Direct
Lall, Sanlaya, and Paul Streeten. 1977 Foreign 1nvest-  Investment: A Review." World Bank Policy
ment, Transnationals, and Developing Coointries   Research Working Paper. Washington, D C
Boulder, Colo : Westview Press.               Muendler, Marc-Andreas 2002 "Trade, Technology,
Lansbury, Melanie, and David Mayes. 1996 "Entry,      and lProductivity: A Stidy of Brazilian M1anufac-
Exit, Ownership and the Growth of Productivity."  turers, 1986-1998 " Dissertation. Berkeley, Calif.
In David Mayes, ed., Sources of Productivity  Navarettim  Giorgio Barba, and  Isidro Soloaga.
Growth. Cambridge, U K : Cambridge University     "Weightless Machinies and Costless Knowledge
Press.                                            An Empirical Analysis of Trade and Technology
Levinsohn, J. 1993. "Testing the Imports-as-Market-   Diffusion." World Bank Policy Working Paper
Discipline Hypothesis." Jouirial of International  Washington, D.C
Economzics 35(1/2). 1-22.                    Newfarmer, Richard 2001. "Foreign Direct Invest-
Li, Wei, and Lixin Cohn Xu 2001. "Liberalization and  ment: Policies and Institutions for Growth." In
Performance  in Telecommunicationis Sector        New Horizons for Foreign Direct Investment
around the World." World Bank Policy Research     Paris: Organisation for Economic Co-operation
Working Paper. Washington, D.C.                   and Development.
Lipsey, Robert E. 1999. "The Location and Character-  Newfarmer, Richard, and Lawrence Marsh. 1992
istics of U S Affiliates in Asia." National Bureau  "IIndustrial Structure, Market, and Profitability."
113



G LO B AL   EC ONO MI C     PRO SP EC TS     2 O 0 3
In C  Frischtak and Richard Newfarmer, eds,       Bureau of Economic Research (NBER) Working
Transnational Corporations Market Struicture      Paper. Cambridge, Mass
and Industrial Performance. London: Routledge  Sargent, John, and Linda Matthews 1999. "Exploita-
Noll, Roger. 2000 "Telecommunications Reform in        non or Choice? Exploring the Relative Attractive-
Developing Countries." Stanford Institute for      ness of Employment in Maquiladoras " Journal
Economic Policy Research (SIEPR) Policy Paper.    of Business Ethics 18 213-27
Stanford, Calif.                              Saxenian, AnnaLee, and Jinn-Yuh Hsu. 2000. "The Sil-
Noll, Roger, Mary Shirley, and Simon Cowan 2000       icon Valley-Hsinchu Connection: Technical Com-
"Reforming Urban Water Systems in Developing      munites and Industrial Upgrading." University of
Countries." In Anne Krueger, ed , Economnc Pol-   California-Berkeley Working Paper.
icy Reform. The Second Stage. Chicago: Univer-  Scherer, E M., and David Ross 1990. Induistrial
sity of Chicago Press                             Market Structuire and Economic Performance
Noorbakhsh, Farhad, Alberto Paloni, and Ali Youssef.   Boston: Houghton Mifflin.
2001. "Human Capital and FDI Inflows to       Schrank, Andrew  2001. "Export Processing Zones
Developing Countries New Empirical Evidence."      Free Market Islands or Bridges to Structural
World Developmient 29(9): 1593-61 0.              Transformation." Development Policy Review
OECD (Organisation for Economic Co-operation and       19(2). 223-42.
Development) 1996. Trade, Employment, and     Sheshinski, Eytan, and Luiz Felipe Lopez-Calva 2000
Labour Standards. A Stuidy of Core Workers'       "Privatization and Its Benefits Theory, Evidence,
Rights and Inteinational Trade Paris.             and Challenges." Consulting Assistance on Eco-
2001. "Summary of Cartel Cases Described by     nomic Reform (CAER) II Discussion Paper 35
Invitees " Paris OECD Global Forum on Compe-      Cambridge, Mass Harvard Institute for Interna-
tition. October 5                                 tional Development
Perrazzini, Ben. 1996. "Competition in Telecoms-  Shirley, Mary M , and Patrick Walsh 2000. "Public
Implications for Universal Service and Employ-    vs. Private Ownership: The State of the Debate."
ment." Puiblic Policy for the Private Sector:     World Bank Working Paper. Washington, D.C.
Washington, D.C  World Bank.                  Sjoholm, Fredrik  1996. "International Transfer of
Porter, Michael. 1990 The Comipetitive Advantage of    Knowledge: The Role of International Trade
Nations. New York Free Press.                     and Geographic Proximity " Weltwirtschaftliches
Pritchett, Lant. 1999. "The Tyranny of Concepts        Archiv 132(1): 97-115.
Cumulative  Depreciated  Investment Effort    Smarzynska, Beata. 2002 "Does Foreign Direct In-
(CUDIE) Is Not the Same as Capital Accumula-      vestment Increase the Productivity of Domestic
tion " World Bank Policy Research Working         Firiis' In Search of Spillovers through Backward
Paper. Washington, D C                            Linkages." World Bank Policy Research Working
Recanatini, Francesca, Scott Wallsten, and lixin Cohni  Paper. Washington, D C
Xu. 2000 "Surveying Surveys and Questioning   Smarzynska, Beata, and Shang-Jin Wei. 2000. "Cor-
Questions: Learning from World Bank Experi-       ruption and Composition of Foreign Direct
ence." World Bank Policy Research Working         Investment Firm-Level Evidence." World Bank
Paper Series. Washington, D.C.                    Working Paper. Washingtoni, D.C.
Rodriguez, F, and D. Rodrik. 1999 "Trade Policy and  Stein, Ernesto, and Christiani Daude. 2001. "Institu-
Economic Growth: A Skeptic's Guide to the         tions, Integration, and the Location of Foreign
Cross-National Evidence." Centre for Economic     Direct Investment " New Horizons for Foreign
Policy Research Working Paper. London, U.K         Direct Investment Paris' Organisation for Eco-
Rodrik, Dani. 1999. The New Global Economny and        nomic Co-operation and Development (OECD)
Developing Couintries: Making Openness WXork.  Stern, Nicholas. 2001 "Investment and Poverty: The
Washington,  I) C . Overseas  Development         Role of International Financial Institutions."
Council                                            Econonncs of Transition 9(2): 259-80
Romer, Paul. 1990 Endogenous Technological Change  Tybout, James R. 2000  "Manufacturing Firms in
Jouirnal of Political Economy 98(5) 57-102.        Developing Countries: How Well Do They Do,
Ros, AgustinJ. 1999 "Does Ownership or Competition     and Why?" Journail of Economic Literature
Matter? The Effects of Telecommunications         38(1): 11-44.
Reform on Network Expansion and Efficiency."         2001   "Plant and Firm Level Evidence on
Jourmal of Regulatory Economics 15(l): 65-92.      'New' Trade Theories." National Bureau of
Sachs, Jeffrey, and Andrew Warner 1995 "Economic       Economic Research (NBER) Working Paper
Convergence and Economic Policies." National      Cambridge, Mass
]14



D O M E S T I C  P O L I C I ES TO  UN LO C K  G LO B AL O     PPO R T UN IT I ES
UNCTAD (United Nations Conference on Trade and        Asian Economies." Journzal of Policy Modeling
Development) 1996. "World Investment Report:      24(5) 411-35.
Investmenit, Trade, and Internationial Policy  Wei, Shang-Jin 2000 "Howv Taxing Is Corruption on
Arrangemiienits." New York                        International Investors?" Reviewv of Econoniics
1997 "World Investment Report: Transna-        and Statistics 82(1). 1-11
tional Corporations, Market Structmre, and Com-  Weiss, Leonard. 1989  Concentration and Price
petition Policy." New York                        Cambridge, Mass N41T Press
1998 "World Investment Report Trends and   Wellenius, Bjorn 1992 "Telecommunications. World
Determinants " New York.                          Bank Experience and Strategy." World Bank
Van Ryckeghemil, Willy. 1995 "Domestic Policy Fac-    DIscussionI Paper Washington, D.C.
tors Influenicing FDI Inflows in Latin America " In  Wells, Louis T, Nancy J Allen, Jacques Morisset, and
W  Grabendorff, ed , Foreign Direct Investnent    Neda Pirnia 2001  "Using Tax Inicenitives to
it Developing Countries The Case of Latin         Compete for Foreign Investment Are They Worth
Amnerica Madrid: IRELA                            the Costs?" Foreign Investment Advisory Ser-
Villela, Luiz, and Alberto Barreix 2002 "Taxation and  vice Occasional Paper. World Bank, Washington,
Investment Promotion." Background Nore for        D C.
Global Economic Prospects 2003 Washington    World Bank. 1995 Butreautcrats in Business The Eco-
Inter-Americani Development Bank.                 noniics and Politics of Government Ownierships
Wallsten, Scott. 2000 "Telecommunicationis Privatiza-  Oxford, U K Oxford University Press.
tion in Developiig Countries. The Real Effects  -    2001  Global Development Fniance 2001
of Exclusivity Periods." SIEPR Policy Paper.      Coalition Btilding for Effective Developnicnt
Stanford, Calif.                                  Finance Washington, D C
2001a "An Econometric Analysis of Telecom  -   . 2002a. "Improving the Investment Climate in
Competition, Privatization, and Regulation in    India." Investment Climate Unit Working Paper
Africa and Latin America " Jotirnal of induistrial  Washington, D C.
Economics 49(1) 1-20                                 2002b  World Developnient Report 2002:
-   .   2001b. "An Empirical Test of Geographic       Building Institutions for Markets. Oxford, U K
Knowledge Spillovcrs Using Geographic Informa-    Oxford University Press
non Systems and Firm-Level Data " Regional   World Economic Forum   2002 Global Conipetitive-
Science & Urban Econouiics 31(5) 571-99           ness Report 2001-2002. Oxford, U K., and New
2001c. "Ringing in the 20th CentuIly- The      York. Oxford University Press.
Effects of State Monopolies, Private Ownership,  WTO (World Trade Organization) 1998 "Report of
and Operating Licenses on Telecommunicationis    the Working Group on the Relationship between
in Europe, 1892-1914 " World Bank Research        Trade and Investment to the General Council
Working Paper Washington, D C                     WT/WG/TI/2. December 8.
-   2001d "Thle Role of Government in Regional  Xu, Bin, and Jianmao Wang 2000 "Trade, FDI, and
Technology Development: The Effects of Public    Interniationial Technology Diffusion " Journial of
Venture Capital and Science Parks." SIEPR Work-   Economic Integration (Korea) 15(4) 585-601
ing Paper Stanford, Calif                    Zak, Paul 2001 "Institutions, Property Rights, and
2002 "Of Carts and Horses. Regulation and      Growth." Th7e Gritter Institute XVorking Papers
Pri atization in Telecommunicationis Reforms "    on Law, Econoniics, and Evolutionary Biology
Policy Research Working Paper. Washington, D C    2(1): Article 2.
Wang, Eric 2002   "Public Infrastructure and Eco-  ZhouL, Kate Xiao. 1996. Hoto the Fartners Cbaniged
nomic Growtlh: A New Appi oach Applied to East    Cbina. Boulder, Colo . Westview Press
115






International Agreements to
Improve Investment and
Competition for Development
F    ast-growing developing countries have  Ministers of the World Trade Organization
commonly been successful in setting up  (WTO) set an agenda for investment and com-
F    investment regimes that facilitate private  petition when they met in Doha, Qatar, in
investment and marshal competition to ensure  November 2001, and decided to launch nego-
growth in productivity. As with trade reform,  tiations on a multilateral framework that cov-
most of the benefit from new sound invest-  ers investment and competition. These negoti-
ment and competition policies comes from  ations are subject to a decision to be made
unilateral reforms of domestic policies. This  by explicit consensus on modalities at the
chapter explores the potential of international  Cancun Ministerial Conference, to be held in
collaboration-collaboration principally in  2003. The purpose of the new framework is
the form of international agreements-to help  "to secure transparent, stable, and predictable
developing countries consolidate sound invest-  conditions for long-termi cross-border invest-
ment climates.                            ment" that will expand trade and "enhance
International agreements that are associ-  the contributioii of competition policy to in-
ated with multilateral or regional arrange-  ternational trade and development."'
ments can potentially provide additional     The international community, and develop-
benefits when coupled with domestic reforms.  ing countries in particular, therefore faces two
Benefits can take several forms. For investment  questions: What types of new multilateral mi-
policies, international agreements usually have  tiatves on investment and competition policy
the objective of eliciting more investment by  can promote more-and more productive-
locking in reforms and providing additional in-  investmiient, and hence more rapid develop-
vestor protections. They can also reduce policy  ment? And, which issues are best tackled
externalities that have "beggar-thy-neighbor"  through voluntary initiatives and multilateral
consequences. Moreover, participating in in-  cooperation, and which are best handled
ternational negotiations can prompt partners  through binding commitments, such as those in
to undertake reciprocal reforms that would  the WTO and regional arrangements? The an-
not otherwise occur, as well as strengthen the  swers to these questions require a separate dis-
hand of domestic reformers. For competition  cussion of investment and competition policy.
policy, international agreements might lead to
removal of restraints that inhibit competition,  Can coordinated investment policies
thereby unleashing new price competition that  increase flows to developing countries
benefits all countries. A central purpose of this  and reduce beggar-thy-neighbor policies?
chapter is to identify collective actions that  An overall purpose of coordinating an invest-
have the greatest development effects.    ment policy is to expand the flow of investmenit
117



G LO BA L EC O NO MI C PRO SP EC TS 2 O 0 3
around the world, to minimize distortions that  reducing trade barriers in rich countries. The
hurt neighbors, and to help improve economic  second issue is the curbing of emerging com-
performance. Coordination might contribute    petition among countries in order to lure
to achieving these goals through three main   foreign investment through incentives. Unfor-
channels: (a) protecting investors' rights in  tunately, information on the extent of invest-
order to increase incentives to invest, (b) liber-  ment incentives is inadequate to assess their
alizing investment flows to permit enhanced   effects. Thus, a high priority for international
access and competition, and (c) curbing poii-  collaboration is to systematically compile this
cies that may distort investment flows and    information.
trade at the expense of neighbors.               Finally, participating in international in-
Analysis suggests several broad conclusions.  vestment agreements may have benefits over
As with trade reforms, unilateral reforms to  and above unilateral reforms if those agree-
liberalize foreign direct investment (FDI) are  ments are accompanied by reciprocal market
likely to have the greatest and most direct   access in areas of importance to developing
benefit for the reforming country. Beyond     countries. These benefits can become clear
this, new international agreements that focus  only in the course of negotiations.
on establishing protections to investors cannot
be predicted to expand markedly the flow of   Collective action can improve competition
investment to new signatory countries. This is  Greater competition is associated with more
because many protections are already covered  rapid development. Lowering policy barriers
through bilateral investment treaties (BITs),  to trade and foreign investment in develop-
and even these relatively strong protections do  ing countries, as shown in chapter 3 of this
not seem to have increased flows of investment  volume, is a powerful, procompetitive force.
to signatory developing countries. These facts  International agreements on competition pol-
suggest that expectations for new flows associ-  icy might bring benefits beyond unilateral
ated with protections emerging from any mul-  actions-provided that the agreements address
tilateral agreement should be kept low.       the major restrictions that adversely affect
International agreements that allow coun-   developing countries.
tries to negotiate reciprocal market liberaliza-  Restrictions on competition in the global
tion and to promote nondiscrimination can     marketplace that will most hurt development
reinforce sound domestic policies and con-    can take three forms. First, policy barriers in
tribute to better performance. Because most of  markets abroad limit competition from devel-
the remaining investment restrictions are on  oping countries in these markets. Particularly
services, the existing General Agreement on   harmful are the $311 billion in agricultural
Trade in Services (GATS) provides an oppor-   subsidies and textile quotas, as well as the
tunity to meet this objective. Similarly, curb-  high border protection, tariff distortions (such
ing beggar-thy-neighbor policy externalities  as tariff peaks and escalation), and protection-
can benefit developing countries, especially  ist use of antidumping. Those policy barriers
if agreements focus on two critical issues.   are common in all countries-rich and poor
The first issue is the reduction of trade barrn-  alike. All of these restrictions limit the ability
ers that-by depriving developing countries    of exporters in developing countries to com-
of market access and    discouraging  their   pete in international markets.
exports-will lessen the attractiveness of        Second, private restraints on competition
opportunities for both foreign and domestic   can adversely affect prices for consumers and
firms to invest in developing countries' export  producers in developing countries. For exam-
industries. In this regard, reducing trade barn-  ple, companies that are based in high-income
ers in developing countries is as important as  countries have cartelized  some markets;
118



IN T ER N Al IO NA L AG R Ee MEN r s IO   I MP RO V E INV EST MEN T AND CO MNP ET IT IO N
proven cartels have taxed consumcrs in devel-  International efforts to promote
oping countries by up to $7 billion in the  investment
1990s. Actions that facilitate prosecution of  A  ny pro-development effort to coordinate
cartels should be hligh on the priority list.  Ainvestmlecnt policies through agreemenit
Such actions can range from more systematic  has as its objectives increasing the flow of
arrangements to exchange informationi, to   investment, minimizing distortions among
granting developing countries the ability to  counltries, and helping countries participate
sue under foreign antitrust laws when their  ni the potential gains from  investment and
trade is adversely affected. Indeed, developing  investment-related trade. Chapter 3 of this
countries would benefit from  much greater  volume singled out domestic policies that in-
efforts to identify and to document restrictive  fluence the quantity and productivity of pri-
business practices that adversely affect prices  vate investment, both domestic and foreign.
of their trade.                             Governments that have provided     stable
Third, many governments in high-income   macroeconomic policies and effective prop-
countries officially sanction trade restraints by  erty rights for investors, and that have low-
exempting their companies from domestic an-  ered policy barriers to competition have, by
titrust laws. For example, many governments  and large, enjoyed greater success in creating
permit their comilpanies to cartelize exports.  the conditions for sustained growth. Interna-
Although these cartels are shrouded in the  tional efforts to support these policies can
secrecy of government registries, national ex-  take several forms: bilateral, regional, and
port cartels may well raise prices to develop-  multilateral. They can be binding, as in the
ing countries. Efforts should be made to make  case of the WTO and the North American
transparent any information on national ex-  Free Trade Agreement (NAFTA), or nonbind-
port cartels. If cartels were found to have  ing, as in the case of the Organisation for Eco-
adverse price effects, everyone would benefit  niomic  Co-operation  and  Development
from reducing these officially sanctioned pri-  (OECD). The current regimen is a mixture of
vate restraints on trade. Similarly, antitrust  binding and nonbonding efforts.
exemptions of ocean transport have given rise
to price-fixing arrangements that systemati-  Today's international investment
cally hurt consumers everywhere, includinig  framework is a patchwork quilt sewn
consumilers in developing countries.        together over many years
Competition policies in developing coiun-  The growing waves of FDI observed in recent
tries themselves can, in many cases, be irn-  decades have been accompanied by a steady
proved  through  increased  transparency,   rise in international agreements on invest-
nondiscrimination, and procedural fairness.  ment. Agreements are typically founded on
However, international cooperation in this  the presumptions that cross-border investmiienit
complex area of regulation has to recognize  provides benefits to both investing and reci-
that countries have different capacities and  pient couintries, that rules can minimize dis-
iiistitutional settings, which warrant caution  putes and provide for their resolution, and
in recommending-much less in mandating-     that agreed-on rules can enhance both the
across-the-board policies. This is an area where  quantity and quality of investment. The
voluntary programs that facilitate learning and  Havana Charter, designed to create the Inter-
adoption of best practice in developing coun-  national Trade Organization (ITO) at the end
tries can pay high dividends.               of the 1940s, proposed the inclusion of invest-
This chapter analyzes first the investment  ment provisions together with trade provi-
policy issues, and then the global competition  sions. The investmenit provisions were quite
issues.                                     limited in scope because many countries-
119



G LO BA L EC O NO MI C PROS PE CT S       2 0 0 3
particularly developing ones-feared foreign    of norms and principles that govern interna-
control over their natural resources and strate-  tional trade.
gic industries.2 Since then, a patchwork quilt
has emerged, made of differing bilateral       Bilateral agreements.   Recent years have
treaties, regional arrangements, and multilat-  witnessed a surge in BITs. The number of BITs
eral instruments relating to cross-border in-  quintupled during the 1990s, reaching 2,099
vestment. This regulatory quilt stands in sharp  by the end of 2001 (see box 4.1). During 2001
contrast to the more comprehensive system      alone, 97 countries concluded 158 BITs (see
Box 411           What is a BIT?
t   he number of BITs mushroomed in the 1990s     treatment, and treatment according to customary
(see box figures). These agreements typically    international law. In addition, BITs prescribe specific
contain broad definitions of foreign investment,  investment protections, which cover topics such as
inclusive of nonequity forms, various types of invest-  the transfer of funds, expropriation, and nationaliza-
ment assets (including portfolio investments), and  tion. They typically provide for the settlement of
intangible assets such as intellectual property. BITs  disputes between the treaty partners and between
generally avoid a direct regullation of the right to  investors and the host state. Provisions for so-called
establishment, referring this matter to nationial laws  investor-state arbitration normally refer to pre-
(and thus recognizing implicitly the right of host  existing arbitration rules, notably those under the
countries to regulate the entry of FDT). Most BITs  International Center for the Settlenment of Investment
BITs are increasing ...                           ... even among developing countries
2,000                                             500
400
1,500
1,500               _                            ~~~~~~~~~~~~~~~~~~~~~With transition Europe __
300
1,000                                                      South-South BlTs
200
North-South \\_
500                                              100
1960s   1970s   1980s   1990s                    1960s    1970s   1980s    1990s
Source UNCTAD (2000)                              Source UNCTAD (2000)
also do not explicitly address ownership and control  Disputes (or TCSID, which is affiliated with the
issues, though they often cover some operational  World Baiik); the United Nations Commission on
restrictions, such as the admission of key managerial  International Trade Law (UNCITRAL); or the Inter-
personnel. Only a few BITs discipline the use of per-  national Chamber of Commerce (ICC).
formance requirements.
Most BITs prescribe national treatment, most-
favored nation (MFN) treatment, fair and equitable  Source: World Bank staff.
120



IN TE RNAT IONA L AG RE EM ENT S TO I MP R O V E INV 1: SIM ENT AND CO MP ET IT IO N
UNCTAD 2002). For much of the post-World  agreements include technology transfers, envi-
War II period, BITs tended to be negotiated on  ronmental protection, taxation, conflicting
a North-South basis. More recently, however,  requirements, and standards for the conduct
there has been strong growth in the number  of multinationial enterprises.
of South-South BITs. In 2001, for example,
treaties between developing countries ac-  Multilateral accords. Significant multilateral
counted for 42 percent of new BITs (UNCTAD  rules for investment were put in place dur-
2002). BITs covered an average of 50 percent  ing the Uruguay Round, which concluded in
of all foreign investment flows to developing  1994. All of the following agreements either
countries in 1999-2001.                   directly or indirectly address key investnment
issues: the Agreement on Subsidies and Coun-
Regional arrangements. Investment discl-  tervailing Measures (ASCM), the Agreement
plines have figured prominently in regional  on  Trade-Related  Investment Measures
trade and integration agreements, particularly  (TRIMs), the GATS, the Agreement on Trade-
the most recent ones. Some of these agree-  Related Aspects of Intellectual Property Rights
ments embed foreign investmenlt into a    (TRIPs), and the plurilateral Government Pro-
broader framework of rules that are aimed at  curement Agreement.
promoting economic cooperation and deeper    Numerous multilateral agreements and
integrationl. This framework includes the  arrangements that have been concluded out-
European Union; NAFTA; the free trade     side the WTO also affect investment and can
agreement linking the G-3 countries (Mexico,  make a positive contribution to enhancing
the Republica Bolivariana de Venezuela, and  investnment climates in developing countries.
Colombia); the recently concluded Singapore-  Among others, these arrangements include
Japan agreement; and the European Free    efforts to curb bribery and corruption (OECD,
Trade Area. Other agreements-such as the  Organization of American States LOAS]); rules
OECD's Codes of Liberalizationi of Capital  governing the conduct of multinational enter-
Movements, the Colonia Protocol on the Pro-  prises (OECD Guidelines on Multinational
motion and Reciprocal Protection of Invest-  Enterprises, United Nations IU.N.1 Global
ments within the Southern Cone Common     Compact); guidelines on corporate social
Market (Mercosur), and the Asia Pacific   responsibility  and corporate governance
Economic Cooperation (APEC) Non-Binding   (OECD, World Bank); and cooperation on
Investment Principles-are less comprehensive  best practices in investment promotion activi-
with regard to their treatment of the trade-  ties (U.N. Conference on Trade and Develop-
investment interface.                     ment (UNCTAD), World Bank).
A distinguishing feature of regional agree-
ments with investmiienit disciplines is their ten-  New efforts exist for collective action on
dency to address both investment protection  investment
and liberalization (entry) issues, together with  The rising tide of FDI around the world has
disciplines on post-establishment operating  been, in part, a consequence of a progressive re-
conditions and means to settle investment  ceptivity of developing countries to FDI flows.
disputes (both state-to-state and investor-  Just as tariffs have fallen, so too have restric-
state disputes). The architecture of the most-  tions on incoming investments (particularly in
advanced regional free trade and integration  manufacturing) been lifted. Governmenits once
agreements reflects the complex interrelations  hostile to transnational corporations (TNCs)
among investment, trade, services, intellectual  now actively seek their participation-and
property rights, competition policy, and the  even compete for it. One indicator of this is the
movement of business people. Other impor-  change in investment regulations. Between
tant issues that are dealt with in some regional  1991 and 2001, a total of 1,393 regulatory
121



G LO BA L EC O NO MI C PROS PE CT S     2 0 0 3
changes were introduced in national FDI      facilitate access and entry, establishing in-
regimes, of which 1,315 (or 95 percent) were in  vestor protections as an incentive to invest,
the direction of creating a more favorable envi-  and curbing investment-distorting policies
ronment for FDI (figure 4.1). During 2001    that affect trade and investment location.
alone, a total of 208 regulatory changes were
made by 71 countries, only 14 of which (or   Liberalizing investment promotes
6 percent) were less favorable for foreign in-  market access-
vestors (UNCTAD 2002). This opens the ques-  The inclusion of investment in international
tion of whether this evident willingness to im-  negotiations may lead to greater openness of
prove the investment regime could be leveraged  investment regimes that can be accomplished
to achieve some additional benefits, through  unilaterally. If investment is negotiated as part
reciprocating in multilateral negotiations, an  of a broader set of trade negotiations, rather
issue that we take up below.                 than in isolation, then the traditional mecha-
The potential-and the challenge-of co-     nism of reciprocal access concessions can help
operation on investment policies become      create support for greater openness at home
clearer if it is broken down into the three core  and abroad. For example, exporters in devel-
subagendas that parallel the investment cli-  oping countries who obtain improved access
mate discussions in chapters 2 and 3. These  to foreign agricultural markets can be a coun-
policies relate to liberalizing investment to  tervailing force against those who resist the
Box 4.2          The Mu1tilateral Agreement on
Invemtn,enit (MAI)
B ecause investment regulations extend beyond tar-  of the GATS's MFN requiiement without making
J)iffs into domestic regulation, the political difficul-  any reciprocal concessions). Also at play was the
ties of enticing large groups of countries to harmo-  reluctance of OECD countries to open up sensitive
nize their domestic rules are not trivial. Those  sectors to foreign investment (for example, to mar-
difficulties were evident in the latest-and failed-  itime transport and audiovisual services). Labor and
attempt at crafting a multilateral accord. In 1995,  environmental groups objected to the fact that the
developed countries pushed to establish a Multilat-  MAI would give TNCs more power to ignore work-
eral Agreement on Investment (MAI) within the   ers' interests and environmental concerns while
OECD that had the objective of setting "state of the  providing rheni with extensive rights to challenge
art standards for investment regimes and inivestment  domestic regulatory conduct before international
protection with effective dispute settlement proce-  arbitration panels. Meanwhile, many developing
dures.`3 These efforts were unsuccessful, and the  countries, left out of the discussions because of the
MAI was not established.                        MAI's venue in the OECD, protested their unwilling-
One reason for the MAI's demise was the wan-  ness to accept rules that they had no voice in design-
ing support within the business community as it be-  ing (Gilpin 2000).4 By the fall of 1998, negotiations
came apparent that the level of investment protection  on the MAI were formally abandoned, thereby offer-
afforded to MAI signatories would almost certainly  ing sobering insights on the complexity and political
be lower than that offered in BITs. It was apparent  sensitivities involved in attempts at comprehensive
that prospects for significant investment liberaliza-  investment rulemaking.
tion would be held back, first, by concerns of free
riding by non-OECD WTO members (which stood                                                       l
to receive many of the benefits of the MAI by virtue  Souirce WVorld Bank staff
122



IN T ER NAT IONA L AG RE E MEN TS TO        I MP R o V E IN VE S TM ENT AND CO MP ET IT IO N
Figure 4.1 Countries are increasingly liberalizing their investment regimes
National regulatory changes In FDI regimes, 1991-2000
(number of regulatory changes)             (number of regulatory changes)
80                                         250
70                         _                         More favorable to FDI
60 r                                       200
50         /                               150 -           /   Less favorable to FDI
40   / 
30_100 _ _ _ t _ _ _
20                                          50
10                                             _
°l   I  l    I   I   I   I   I   I   I      01 I  O     I     I   I  I  I 1       I 
I, R   R  o 9r  0R                              5 R       a    -    5 O   0   0 4
Soiarcc UNCTAD (2002)
elimination of investment barriers in telecom-  invest primarily in other developing countries.
munications. At the same time, the need to    Estimates suggest that nearly one-third of
fight these battles about the domestic politi-  foreign investment flows to developing coun-
cal economy makes a country a credible ne-    tries originated in other developing countries,
gotiator for improved access. The process, if it  up from negligible amounts in the early 1990s
works, could produce a double benefit: liber-  (World Bank 2002a), so South-South FDI
alizing countries would benefit from the in-  flows have grown.5 (See figures 4.2 and 4.3.)
creased competition that is associated with
FDI, and their firms would have improved
access to foreign markets. A key issue-which     Figure 4.2 South-South FDI is rising
can be determined only during the negotiation    FDI flows to developing countries
process-is the extent to which an investment     (billions of dollars)
agreement leverages reciprocal commitments
among trading partners. Because reciprocal       100
gains are difficult to gauge, an important pre-      High OECD (North-South) countries
requisite for each country is to ensure that any  80
domestic policy commitment makes sense            70                       Sout h-South
when seen through the lens of promoting           60
national development.                             50
Even though most foreign investment orig-      40
inates in rich countries and is destined for      3                High non-OECD countries
other rich countries, there may well be some      20
scope for reciprocal agreements that benefit            '
developing countries, even within the narrow
domain of investment. Because developing           0                I ,
1 994  1 995  1 996  1 997 l1998  1 999 2000
countries are increasingly becoming active as
investors themselves, they have a mIutual in-    Souirce Aykur and Ratha (2002)
terest in clear rules of access. They tend to
123



G LO B AL EC ON OM IC PROS PE CT S 2 0 0 3
the host country's business environment either
Figure 4.3 Share of South-South FDI in     translates into higher risk premiums (in the case
total FDI is rising                         of pricing corporate assets) or imposes addi-
South-South FDI is rising and so is its share in total  tional information costs on enterprises. To be
FDI                                         sure, transparency, alone, can add little if the
(percent)                                   underlying laws and rules are inadequate or
100 -                                       unpredictable.
90 - no-E,E   I .                             Case studies suggest that companies may,
70 -                                       for example, be willing to invest in countries
60 -        North-South FDI/total FDI      with legal and regulatory frameworks that
50 -                                       would not otherwise be considered "investor
40 -                                       fnendly"-provided the compames are able to
30                                         fredy-rvddte-opne                 r  bet
20 -            South-South FDI/total FD   obtain a reasonable degree of clarity about the
10 -                                       environment in which they will be operating.
019 94 1995  1996  1997 1998 1999 2000    Conversely, there appear to be certain thresh-
old levels for transparency beneath which the
Source Aykut and Ratha (2002)               business conditions become so opaque that
virtually no investor is willing to enter, re-
gardless of the extent of the inducement.
The preceding argues that the potential       These policies do not lend themselves well
for benefits investment agreements to gener-  to including sanction-based dispute resolution
ate merits examination. Coordinated efforts to  procedures in legally binding agreements.
liberalize investments can subsume two issues:  Thus, international collaborative efforts should
first, transparency, and second, nondiscrimi-  perhaps take other forms such as increasing
nation in treatment of foreign investment in  developing countries' participation in nonbind-
market access.                                ing best-practice instruments or developing as-
sistance to strengthen institutions. To the ex-
Transparency. Transparency involves mak-      tent that transparency obligations are anchored
ing relevant laws and regulations available   in WTO agreements, monitoring by multilat-
to the public, notifying parties when laws    eral peer review and surveillance may provide
change, and ensuring uniform administration   the best means for promoting governance-
and application. In addition, transparency can  enhancing reforms in host countries.
be increased by offering affected parties the
opportunity to comment on laws and regula-    Nondiscrimination in treatment of foreign
tions, which implies communicating the policy  investment in market access. The practice of
objectives of proposed changes, allowing time  placing foreign and domestic sellers on an
for public review, and providing a means to   equal competitive footing is a hallmark of
communicate with relevant authorities.        trade agreements. This objective is no less
A nontransparent business environment in   important in investment agreements. Promot-
a host country raises information costs, diverts  ing liberalization in international investment
corporate energies toward rent-seeking activi-  essentially boils down to securing nondiscrim-
ties, and may give rise to corrupt practices. This  inatory terms of entry and operation. This
environment weighs down both domestic and     approach has elements of both MFN treat-
foreign businesses, though in many cases it may  ment (that is, nondiscrimination as between all
be particularly discouraging to foreigners who  foreign entities) and national treatment (that
are usually less privy to locally available infor-  is, nondiscrimination between "like" domestic
mation. This heightened risk of operating in  and foreign entities).
124



IN T ER NAT IO NA l AG RE E MEN TS TO I MP R O V E INV ES I MEN I AND CO MP ET IT IO N
Box 4.3 South-South flows: who invests and who
receives?
M      ost FDI flows within developing couintries are  has also expanded Its boundaries from East Asia to
M       between the Association of Southeast Asian  Latin America and to parts of Africa. Since the sec-
Nations (ASEAN) countries, and, recently, among  ond half of the 1990s, almost 30 percent of total
the Latin American countries, especially the Mercosur  FDI inflows into India are from other developing
members (UNCTAD 1999). There are signs that    countries-the principal sources being Mauritius,
FDI flows from East and Southeast Asia to Latin  Malaysia, and Korea (Aykut and Ratha 2002). Out-
America and Africa are picking up. According to the  flows from Latin America in 2001 were directed pri-
Chinese Ministry of Foreign Trade and Economic  manly at other countrics in the region (UNCTAD
Cooperation, China attracted $3 billion in invest-  2002). Chile continued to be the major player in
ment from 22 developing countries in 1998. Though  interregional investment, followed closely by Mexico
this figure made up only 7 percent of total FDI in-  and Argentina. Some South African TNCs have
flows to China, the flows originated in a wide spec-  recently moved to a strategy of international growth,
trum of countries (in terms of size and per capita  partly through cross-border mergers and acquisi-
income levels) and extended to varying sectors  tions. A noteworthy example of a global player is
(Aykut and Ratha 2002). In addition, Chinese TNCs  South African Breweries, which operates 108 brew-
are becoming prominent in world markets. China  eries in 24 countries including China, large parts of
has invested, not only in Asian countries, but also in  Africa, and Europe (UNCTAD 2002). FDI outflows
Bangladesh, Brazil, India, the Islamic Republic of  from the Central and Eastern European countries
Iran, and Poland, in addition to countries in Africa.  such as Croatia, Estonia, and Slovenia are also
The Republic of Korea, an OECD member,      headed primarily to neighboring countries. A ten-
invested nearly one-third of its direct investment in  dency to invest in neighboring countries that are at
developing countries (excluding those in Africa  similar or lower levels of development is another fea-
and the Middle East) in 1998. By 1999, Korea had  ture of South-South FDI (Aykut and Ratha 2002).
invested nearly 50 percent of its aggregate invest-
ment in other developing countries. Malaysian FDI  Source: World Bank staff.
Departures from nondiscrminiiatory treat-  establishment phase can thus send to foreign
ment essentially take one of two forms: before  investors powerful signals of the credibility of
entry in the "pre-establishment" phase of an  a host country's reform efforts.
investmenit, and aftcr entry in the "postestab-  By far the most contentious aspect of liber-
lishment" operating conditions of a business.  alization is the pre-establishment commitment
Governments everywhere have been reluctant  to openness, given the tendency to maintain
to extend full pre-establishment privileges to  restrictions on entry in a few sensitive sectors.
all potential entrants in every sector. Securing  Most countries now permit liberal access to
nondiscriminatory conditions of treatment is  foreign investors in manufacturing. The same
equally important in the postestablishment  holds true-if to a lesser extent-in mining
phase, because foreign investors will typically  and agriculture. Indeed, as a result of various
have significant start-up costs and will be  investment incentive schemes that are not
averse to sudden, unanticipated changes in  available to domestic firms, foreign investors
regulatory conditions that may tilt competi-  In manufacturing often enjoy treatment that
tive conditions in favor of local competitors.  is better than that available to domestic in-
Nondiscrirriination commitments in the post-  vestors. Most governmental measures that
125



G LO B AL EC O NO MI C PROS PE CT S       2 0 0 3
overtly discriminate against foreign investors  tially arise from the liberalization of invest-
and that restrict FDI inflows are maintained in  ment in services: the GATS. The GATS has
the service sector and concern key industries  several features that are attractive to countries,
such as telecommunications, broadcasting and  potentially making it a useful tool to widen
related audiovisual services, satellite services,  nondiscriminatory access in a reciprocal frame-
energy services, financial services (especially  wvork. By having a positive list approach-in
banking and insurance), civil aviation, and   which countries voluntarilv schedule sectoral
maritime transport.6 Sauve (2002) estimates   commitments to apply national treatment
that 80-85 percent of restrictions affecting in-  and to grant market access-governments
ternational investment are maintained in ser-  enjoy considerable flexibility to exempt sec-
vice sectors. Among the most dynamic sectors  tors that they deem of special national interest.
of the global economy, services are also where  Once commitments are undertaken, countries
some two-thirds of cross-border FDls have     accord all suppliers-foreign and national
been directed in recent years (see chapter 2,  alike-the same conditions of entry and oper-
this volume).                                 ation in a nondiscriminatory fashion. To date,
One telling proxy of the potential of ser-  however, the GATS has fallen short of its
vices for investment liberalization is provided  itberalizing potential. The coverage of com-
by the negative lists of measures drawn up by  mitments for a large number of countries is
prospective signatories of the ill-fated MAI.  limited. About two-thirds of the WTO mem-
The lists identify those sectors in which the  bership has scheduled fewer than 60 sectors
negotiators wished to restrict access by foreign  (of the 160 or so specified in the GATS list)
investors (see figure 4.4). A similar trend is ev-  (see Stern 2002). In many cases, commitments
ident under the NAFTA. Simply put, the mar-   do not reflect the actual degree of openness
ket access or agenda for investment is largely  (Mattoo 2000). In other cases, countries have
centered on services (Hoekman and Saggi       not moved actively to schedule sectors-even
2000; Sauv6 and Wilkie 2000).                 when domestic policies are open to foreign
A multilateral vehicle already exists for   investments. Finally, sometimes countries'
realizing the positive externalities that poten-  commitments serve to protect the privileged
Figure 4.4 Revealed preferences: governments shield services more often than
manufacturing from the winds of investment competition
All setr<   <=,Business  16%
All sector
21%/         \                                                    Communication 16%
Construction  2%
Services                          Financial   21%
63%                              Tourism      2%
Recreation   2%
Transport   37%
Goods                                                            Other        4%
16%
Total      100%
Note Listed are iioniconiforiiing measures reserved siitder che draft Multilateral Agreementit on tinvestrnent
Soutrce Sauvce (2002)
126



I N T E R N A I I O N A L A G R E E M c N T S I 0  I M P R O V E I N V E S I M r N T A N D C O NI P E I I T I O N
position of incumbenits, domestic or foreign,  caveat to this argumlienit: To the extent that
rather than to enhance the contestability of  the power imbalance is redressed in a multi-
niarkets                                     lateral agreemelit in favor of weaker states,
Countries could take greater advantage of  then the constituencies within the global busi-
the opportunity offered by the GATS to lend  ness communilty may well prefcr-as was the
credibility to reform programs by committing  case in the MAI negotiationis-the stronger
to maintain current levels of openness or by  level of investimicilt protection flowing from
precommitting to greater levels of future    BITs, and may lose interest in a multilateral
openness. To advanice the process of services  agreemcilt.
reforms beyond levels undertaken indepen-       Third, a multilateral set of disciplines on
dently and to lcad to more balanced outcomes  investimicit protection would arguably hclp
from the developing countries' points of vicw,  developing countries send a positive signal to
countries could better harness the power of  potential forcign investors regarding the per-
reciprocity by devising negotiating formulas  manence of policy changes, the expected
that widen the scope for tradeoffs across sec-  standard of treatmenit afforded to foreign in-
tors (both goods and services) and across    vestors, and recourse to a dispute-settlemeilt
modes of delivery, notably temporary move-   procedure.
menit of workers (Mattoo 2002).                 While these factors suggest that investmenit
flows might increase because of such an
-but protecting investment may not           arrangemilenit, care should be taken not to
increase flows                               overstate the response of investors Five facts
A foundation of any country's investment cli-  argue for caution. First, the absence of a body
mate is the protection of property rights for  of multilateral disciplines on investincilt
its investors. An agreement that encourages  protection has hardly deterred cross-border
countries to improve investor protections has  investimient activity. Indeed, FDI has far out-
the potential for improving investment flows  stripped trade and output growth over the
from abroad and for elicitinig more domestic  past decade and a half (see figure 4.5).
investiienit. The interinational communiity, in  Second, the absence of an agreemenit has
general, and developing countries, in particu-  not prevented substantial unilateral reform
lar, might find three benefits from multilateral  (see discussion above, and figure 4. 1).
disciplines on investment protection.           Third, a more precise iiidicator is the his-
First, an agreement on common standards    torical experience of the BITs in clicitinig new
would promote efficicncy by carr-ying potcn-  investmcnt. Does the signing of BITs increase
tially' significant economies of scale in makinig  the flow of FDI? Hallward-Driemileier (2002)
rules: one multilateral agreement could be-  finds few independent effects of BITs on sub-
come a "one-stop" substitute for the complex  sequent increases in investment (box 4.4).
and legally divergent web of existing BITs.     Fourth, it is not clear whether multilateral
Second, a multilateral regime for invest-  investment disciplines-whether in the U.N,
ment protection could help counterbalanice   WTO, or OECD-will embody investimielit
the bargaining asymmetries built into BITs   protections that are superior-and, therefore,
and into regional agreements conducted along  additive-to BITs. In the case of the WTO, the
North-South lines. In some cases, the negoti-  Doha Ministerial Declaration reflects a signif-
ating asymmetries that are common to bilat-  icantly more-limilited approach that cicarly
eral agreemenits have led to treaties in which  does not view a multilateral framework on in-
developing countries have taken on substan-  vestment as a substitutc for bilateral and re-
tive obligationis without any reciprocity other  gional arrangemcnts. Recent negotiating briefs
than the promise of increases in future private  in the WTO indicate that some malor coUIn-
investmenit. However, there is an important  tries have withdrawn support for investor-state
127



G LO B AL EC ONO MI C       PROS PE CT S     2 0 0 3
Figure 4.5 FDI is growing faster than exports and output
Developing countries: merchandise exports, output, and FDI inflows, 1980-2000
(index, 1980 = 100, average annual growth rates in parentheses)
3,000 -                                   L.-I ~~~~~~~~~~~~~~~~~FDI inflows
2,500 - -                                    ~~~~~~~~~~~~~(1981-90 5 3%)
…-                      - ~~~~~~~~~(1991-2000 20 8%)
1,500
1,000                                                        GDP
L1-I-+--1-I--l-l---1--I--b--4-~ 1IfiIf  If(1 9B1-90 3 2%)
500                                         ----            (1991-2000 4 4%)
I          ~~Merchandise exports
0 -                                        ~~~~~~~~~~~~~~~(1 981-90 3 3%)
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000   (1991-2000 9 6%)
Source UNCTAD (2001), Handbook of Statistics World Bank (2002), World Developmenit Iindicators, and WTO (2001),
Interinationial Trade Statistics
dispute settlement, which would tend to lessen    national investment. Three negative policy
the additive value of investor protection in a    externalities-when   one country's policies
multilateral accord.                              adversely  affect another-merit discussion.
Dispute settlement is another critical-and     The first and most powerful of these negative
as yet unresolved-issue that will influence       policy externalities are investment-distorting
the content of any multilateral agreement to     trade   barriers. Tariffs, tariff escalation,
strengthen  investor protections. Most BITs       and other forms of protection     discourage
contain dispute resolution mechanisms that        investment-both foreign and domestic-in
allow investors to challenge government rul-      export industries in developing countries.
ings before arbitration panels or international   Said differently, if developing countries con-
courts. In the context of the WTO, while there   front impediments to market access abroad,
is generally little support for the inclusion     the effect of the barriers is to lower the poten-
of investor-state arbitration provisions in a     tial stream of earnings in their export activities.
prospective multilateral investment agree-        This change reduces the incentive for foreign
ment, WTO     rules on investment protection      and domestic investors to invest in produc-
could entail complications even when admin-       tion for export in developing countries. Quota
istered through state-to-state dispute settle-    arrangements, antidumping actions, subsidies,
ment. For example, what would be the appro-       overly restrictive rules of origin, and other
priate remedy in an     instance of unlawful      trade restrictions distort not only trade, but
expropriation of a foreign investment? These      also investment, and these distortions are
difficult and contentious issues will take time   arguably the largest negative policy externality
to resolve in any international agreement.        affecting investment in developing countries.
Two other sets of policy externalities figure
Beggar-thy-neighbor investment                    prominently in investment decisions: perfor-
distortions must be minimized                     mance   requirements-to    compel multina-
Governments have adopted policies that may       tional companies to locate a greater part of the
affect the location and performance of trans-     value added chain in the domestic market-
128



IN T E RN AT IO NAL   AG RI: E MENTS TO      I MP R O VPE INV ES TM ENT AND CO MP ET IT IO N
B3ox 4.4          Do BITs increase investment flows?
Only a bit
B ITs are instrumenits used by countries to protect
their foreign investors, while host countries view    The share of FDI received by developing
BITS as an importanit means of attracting foreign     countries is relatively unaffected by the
investors. BITs can provide the basis for resolving   signing of a BIT
disputes; they can also impose potentially extensive  (share of annual FDI flow)
obligations on the part of the governments hosting    0 3
the investment. For example, almost all treaties stip-
ulate comiipensation for the expropriation of invest-
ments. ln some cascs, trcaties proscribe any govern-   0 2
ment action-even environmental actions or other
regulations-that would reduce the value of the
privare investment and they establish grounds for     0 1
compensation. Such compensation could cither
entail extensive liabilities for the host government
or compel it to refrain from making certain policy         -3   -2 T-    Year T-1 T2    +3
choices. Against this backdrop, the question of                         signed
whether BITs actually increase FDI is important.          Years before signing  Years after signing
Surprisingly little empirical work has been done
to test BlTs' role in attracting FDI. UNCTAD, in a     Source ia.lIlward-Orieine,cr (2002)
recent stUdy, found little evidence that BITs increased
FDI (UNCTAD 1998). That work looked at a single
year of investments and tested whether the number  host country was affected by the presence of a BIT.
of BITs signed by the host was correlated with the  The evidence here is that concluding a BIT is posi-
amount of FDI it received. Hallward-Driemeicr     tively associated with receiving a larger share of a
(2002) redid that test, but applied it to 20 years of  source country's FDI outflows, but that the result is
data, looking at the bilateral flows of OECD memn-  not statistically significant.
bers to 3 i developing countries. The Hallward-       Some countries have looked to BITs as a way of
Driemeier test covered the vast majority of FDI   signaling their respect for property riglts. Particularly
flows, as well as those relationships that were histor-  if their reputation for protecting suchi rights is weak,
ically the bulk of such treaties. Overall, the evidence  they have seen the signing of a BIT as a way of assuag-
is, at best, weak that BITs increase the amount of  ing the concerns of foreign investors. Conversely, the
FDI. By the end of the 1990s there were many more  crecibility of such a signal may not be that strong. It
BlTs, and FDI had increased dramatically. However,  may be that the domestic rule of law must be suffi-
controlling for a time trend, there was little indepen-  ciently strong before foreigners are willing to consider
dent role for BITs in accounting for the increase in  the terms of the BIT as being enforceable. To test he-
FDI. Countries that had concluded a BIT were no   tween these hypotheses, the study ran regressions that
more likely to receive additional FDI than were  included measurements of the rule of law, government
countries without such a pact.                    effectiveness, and regulatory quality. These measures
Another question is whether a BIT would draw  were then interacted with the presence of a BIT. The
attention to a particular location, thus leading to an  results indicate that in weak investment climates, the
increase in flows in the aftermath of negotiations.  BIT does not serve to attract additional FDI. However,
However, comparing flows in the three years after a  in cotintries with stronger investment climates, the
BIT was signed to those in the three years before, there  presence of a BIT does weakly increase the amount
was no significant increase in FDI (see box figure).  and relative share of FDI that the host receives.
A third question is whether the relative amount
of FDI that a source country allocated to a particular  Source- Hallward-Dricmeier (2002)
129



G LO B AL EC O NO MI C PRO SP EC TS 2 O 0 3
and investment incentives-usually through  Round's ASCM introduced limited disciplines
tax breaks or direct transfers from the state to  on the granting of investment incentives.
attract FDI. Even when these policies benefit  These disciplines are largely indirect because
the domestic economy, they both have the po-  they apply solely to export subsidies and other
tential for adversely affecting trade and invest-  goods-related transactions-that is, a govern-
ment flows with neighbors. Therefore, further  ment may invoke the agreement's provisions
internationial cooperation to curb their nega-  only when certain types of investment incen-
tive effects can create positive benefits for all.  tives used by certain types of members can be
Unlike restrictions on entry that primarily  shown to distort trade in goods.9
affect services, performance requirements and  Strengthening disciplines on investment-
investmenit incentives usually affect manufac-  distorting incentives could benefit developing
turing. In general, performance requirements  countries because those disciplines would re-
have been the instrument of choice for devel-  duce the scope for this zero-sum tax competi-
oping countries that are seeking to ensure that  tion. However, progress in crafting a set of
TNCs' activities generate the greatest possible  multilateral disciplines on investment incen-
spillovers for their economies. OECD coun-  tives has been negligible to date. One reason
tries have been the predominant users of in-  for this stalemate is that in large federal gov-
vestment incentives to attract investment,  ernments many investment incentive programs
though in recent years numerous developing  originate at the subnational level as instru-
countries have followed suit (see chapter 3,  ments to promote regional development.
this volume; see also UNCTAD 2002).        Another reason is that many emerging devel-
The trade-distorting effects of performance  oping countries have themselves become heavy
requirements-termed   TRIMs-havc     for   users of incentives in recent years. Conse-
some time been subject to negotiated disci-  quently, investment incentives have not figured
plines at both the regional and multilateral  promineitly among topics to be discussed in
levels. WTO disciplines on performance re-  international forums such as the WTO. The ill-
quirements were codified with the TRIMs    fated discussions in the MAI were also unsuc-
Agreement in 1995. Among performance re-   cessful in broaching investment incentives.
quirements, the most prevalent measures      Nonetheless, competition among govern-
relate to local content, joint ventures (or  ments for FDI through incentives is becoming
domestic equlty participation), exports, tech-  increasingly common in many parts of the
nology, and employment requirements. The   world. Developing countries often find them-
initial rationale for export requirements was  selves in competition with each other, but few
in part to relieve the pressure on the trade bal-  examples can be found of developing coun-
ance that inward investment-particularly  tries in direct competition with developed
import-substituting investment-was generat-  countries. Also, competing developing coun-
ing. Local content requirements were designed  tries are often middle-income countries. Four
to maximize vertical linkages and develop-  reasons seem to explain these patterns.
ment of local skills.7 Current discussions of  First, studies show  that the bulk of
chaniges to the TRIMs Agreement are associ-  incentive-bidding activity among governments
ated with the review process that is mandated  takes place within regions, rather than glob-
under Article 9 of that agreement.8 At present,  ally (Oman 2000; Chariton 2002). Only a
these debates are not on the Doha Agenda.  handful of developing countries situated close
In contrast with disciplines on performance  to developed nations experience direct compe-
requirements, disciplines on investment incen-  tition with the deep pockets of the treasuries
tives are-with the exception of the European  of rich countries. Mexico's automotive indus-
Union's comprehensive set of disciplines on  try under NAFTA is perhaps the most promi-
state aids-more   limited. The Uruguay     nent example of this situation.10
130



I N T E R N A T I O N A l  A G R E E M E N T S  r o I i P R O o   1. I N V 1 S T M E N I A N D C 0 Ni P I I I IO N
Second, locational competition tends to be  information that can be used to assess the
strongest between close neighbors with similar  trade- and investment-distorting consequenccs
economic conditions, factor endowments, and  of incentives-and, more broadly, to evaluate
policy regimes. Competition is also strongest  their net development benefits.
in high-skill, technologically intensive indus-  Taken together, the existing multilateral
tries, particularly for firms producing goods  agreements do provide Iimited discipline on
for export. Automakers, silicon chip produc-  certain types of beggar-thy-neighbor policies
ers, and pharmaceutical firms are among the  that are currently in use around the world
most sought-after investimnents. Only a Iimited  With respect to curbing incentives, even
number of higher-income developing countries  though potential benefits for countries exist
are likely to qualify for such a category of  from  a multilateral accord, the absence of
investment.                                  evident momeintumil at the multilateral level-
Third, competition is likely only when in-  when combined with a regional pattern of
vestors are somewhat indifferent about where  possible tax competition and trade effects-
to locate an investment among alternative lo-  suggests that regional arrangements may be
cations. This indifference implies that only the  more promising for international collective ac-
more relatively advanced economics (emerg-   tionis. Howvever, data are lacking. Multilateral
ing or transition economies) could have cause  efforts to improve informiationi on investment
to bid against developed nations.1           incentives, perhaps through a WTO mecha-
Fourth, overt bidding wars between coun-   nism, would help remedy that lacuna and
tries are relatively rare-even though bidding  allow better analysis of the extent of invest-
may be intense within particular countries-  ment distortions.
and are typically limited to a few sectors. They
generally occur when individual projects are ex-  Summary: Getting the biggest
ceptioially large and when the sectors in ques-  development benefit from international
tion (for example, automobiles or electronics)  collaboration on investment
are considered a high priority for national or  Developing countries can benefit from inter-
regional economic strategies (Charlton 2002).  national collaboration to liberalize market
To be sure, striving for a ban on all inceni-  access for investment, to address investor pro-
tives may bc counterproductive because, in   tections, and to minimize investment distor-
some cases, incentives can offset local disad-  tions. Five concltisionis emerge
vantages or can be used to capture spillovers   First, in each of these areas the primary
from  inward FDI (see Hoekman and Saggi      benefits of attracting high-quality investment
2000). In the case of Ireland and Portugal,  from sound investimienlt policies are likely to
for example, incentive programs have played  result from  unilateral enacting of domestic
a significant role in attracting investmenit to  reforms. Long a truism for trade liberalizing
less-developed regions. In the case of Brazil,  reforms, this conclusion-given the apparent
some evidence shows that incentives competi-  lack of investor responsiveness to interna-
tion i-nay have contributed to reducing re-  tional agreements-is increasingly germane to
gional disparities, because FDI in some sectors  investment. Many of the remaining restric-
(particularly automobile manufacturing) is in-  tions are on services. As we have seen in chap-
creasingly located outside the traditional in-  ter 3, progressive liberalization in services can
dustrial heartland around S5o Paulo (Cano    produce substantial economy-wide benefits
1998). While it is probable that, with re-   and should be a priority for consideration as
spect to incentives, stories of failures and  part of any development strategy. Better
excessive expenditures outnumber successes,  telecommunicationis, banking, auditing ser-
agreements must contain some elements of     vices, retail and wholesale trade, and the other
flexibility. A first step is generating adequate  service industries have multiple linkages to the
131



G LO BA L EC ONO MI C PROS PE CT S 2 O 0 3
B3ox 45          DliscipUnes on corporaitions can also improve
the investment climate
Currently, proposed investment rules in the    Affairs Center, and by international organizations
(CWTO focus exclusively on disciplines for gov-  and banks, such as the World Bank, the International
ernments, but they say little about responsibilities  Monetary Fund (I_MF), the Asian Development Bank,
of corporations (see Moran 2002). Improper corpo-  the U.N. Development Programme, and the U.S.
rate behavior-bribery or improper accounting-  Agency for International Development, in developing
can corrode the social fabric of developing and de-  approaches to counter corruption are noteworthy.
veloped countries alike. In the wake of the Enron,  Other programs have a more technical focus.
Arthur Andersen, and WorldCom accounting scan-  The World Bank's work on corporate governance
dals in the United States, efforts to improve corpo-  emphasizes disclosure, transparency, the rights and
rate transparency and good conduct assume a new  treatment of shareholders and stakeholders, and the
importance. Many such activities outside the WTO  duties of board members. Using the OECD's Princi-
are under way.                                 ples of Corporate Governance as a benchmark, the
To help combat bribery and corruption, the  Bank prepares corporate governance assessments for
OECD has recently established the Convention on  its client countries to assess their institutiolal frame-
Combating Bribery of Foreign Public Officials in  works for corporate governance. In addition, the
fnternational Business Transactions. The convention,  World Bank and the IMF together initiated the
put into force in 1999, currently includes all 29  Financial Sector Assessment Program and the
OECD members and five nonmenibers (Argentina,  Reports on the Observance of Standards and Codes.
Brazil, Bulgaria, Chile, and the Slovak Republic) as  More broadly, the U.N. adopted the Global
signatories. The convention makes bribing a foreign  Compact in July 2000 to allay concerns about the
public official a criminal offense. It also encompasses  social effects of globalization on the developing
noncriminal rules for prevention, overall trans-  world. About 100 major multinationals and 1,000
parency, and cooperation berween countries, and it  other companies across the world's regions are cur-
ends the practice of allowmg tax deductibility of  rently engaged in the Global Compact. Projects
foreign bribes. Many countries, however, have yet to  relate to making microcredit more accessible, reduc-
modify their national legislation to implement the  ing carbon dioxide emissions, fighting against human
convention fully. Regional forums of cooperation can  immunodeficiency virus/acquired immune deficiency
also help. For example, the Inter-American Conven-  syndrome (HIV/AIDS), and expanding of basic edu-
tion against Corruption was established in 1996 in  cation in local communities. In a similar vein, the
the OAS; in April 2001, the Summit of the Americas  OECD significantly revamped its Guidelines on
created an implementation mechanism for the Inter-  Multinational Enterprises in 2000 by adding
American Convention. Experience shows that, for  recommendations about eliminating child and forced
anticorruption initiatives to be effective, participa-  labor, improving internal environmental manage-
tion by civil society, private agencies, and the general  ment, addressing human rights, finding methods to
public is critical. In this context, cooperative efforts  combat corruption, and improving disclosure and
by nongovernmrental organizations (NGOs), suclh as  transparency.
Transparency International, the Global Coalition for
Africa, the Novartis Foundation, and the Public  Source World Bank staff.
rest of the economy, and can be sources of  ulations that are consistent with local capaci-
productivity growth for the whole economy.  ties and national objectives. The international
But the pace and form of investment liberal-  community can assist with these efforts
ization necessarily must vary across sectors  through multilateral and bilateral develop-
and across countries, because they require reg-  ment assistance, government-to-government
132



I N T E R N A I I O N A L A   G R E E M E N  Tr O T 0 o  M P R O V E I N V E S I M E N T A N D C O M P 1: I I 1 I 0 N
information exchanges, and private efforts to  quota arrangements, and other barriers-
inform and assist governments.              barriers that are common among developing
Second, international agreements that    countries as well as between rich and poor
focus on liberalizing conditions of entry by re-  countries-stifle developing countries' exports
moving barriers that discriminate against for-  and the investment needed to supply them.
eign competition may help consolidate domes-  Reducing these trade barriers would precipi-
tic reforms at the same time that they open  tate new investment in exports as these activi-
new avenues for reciprocity abroad. Because  tiCs expand, and some portion of this new
of the sensitivity of investment regimes, espe-  investment can he predicted to come from
cially in services, any agreement has to allow  abroad. The second set of externalities con-
for country diversity and must permit govern-  cerns disciplines for investment incentives that
ments the flexibility to design liberalization in  distort the allocation of investmenit. Coopera-
ways consistent with their development strate-  tive measures at the multilateral level have the
gies. Because the GATS provides this flexibil-  advantage of being conceptually clean and
ity and addresses most of the remaining out-  broad based. However, because investments
standing restrictions, multilateral efforts could  tend to affect countries in close regional prox-
concentrate on expanding the still-limited cov-  imity, countries may find it easier to work on
erage of the GATS by increasing the number  rules that curb disadvantageous competition
and quality of commitments that allow com-  on investment incentives through regional
mercial presence. Harnessing the full force of  arrangements. A prerequisite for collective
reciprocity-both across modes (especially by  action is information on the extent of invest-
putting on the table any temporary movement  ment incentives and their effects; thus, a mul-
of workers) and across sectors-may help     tilateral inventory of investment incentives is
motivate this expanded coverage.            a high priority. One option is to set up an an-
Third, an international agreement that    nual surveillance process, perhaps under the
seeks to substantially increase investment  auspices of the WTO or as part of the IMF's
flows by increasing investor protections seems  annual surveillance.
destined, on the basis of available evidence, to  Finally, if new investment arrangements
fall short of expectations. Some key issues are  leverage reciprocal commitments for reforms
already covered by relatively strong investor  abroad on other issues on the trade agenda,
protections in BITs. Moreover, it is not clear  particularly new market access, then agree-
that any investor protections emerging from  ments would certainly help developing coun-
multilateral negotiations would add markedly  tries. These matters can be decided only in the
to existing protections found in bilateral  course of negotiations.
agreements. Finally, merely creating new pro-
tections does not seem to be strongly associ-
ated with increased investment flows. For   International agreements to
these reasons, the overall additional stimulus  promote competition and
of multilateral rules that apply to new Invest-  competition policy
ment over and above unilateral reforms would  promoting development requires not only
probably be small-and virtually nonexistent    policies to encourage investment, but also
for low-income developing countries.        policies to ensure that investmenit is produc-
Fourth, international agreements can use-  tive; among these policies, competition is one
fully discipline two forms of beggar-thy-   of the most powerful. Most pohcies to pro-
neighbor policy externalities that are particu-  mote competition are domestic, and an impor-
larly adverse to developi-nent. The first and  tant conclusion of chapter 3, this volume, is
most important are investment-distorting    that the reduction of policy-related barriers to
trade measures. Tariff escalation, tariff peaks,  competition is essential to raising domestic
133



G LO BA L ECO NO MI C PROS PE CT S 2 0 0 3
productivity. Among the many domestic policy  arrangements that systematically hurt devel-
barriers to competition, the most prominent  oping countries. Competition policies in de-
often involve aspects of globalization, such as  veloping countries themselves can, in many
tariffs, restrictions on FDI (especially in ser-  cases, be improved through increased trans-
vices), state monopolies, and competition-  parency, nondiscriminationi, and procedural
limiting regulations in postprivatized sectors.  fairness. All of these policies are subjects of
Competition policy that disciplines private re-  international negotiation, but they have quite
straints in domestic markets is also important.  different potential effects on development.
However, competition laws have to be appro-
priate to local circumstances because they rely  The most important restraints on
heavily on the strength and independence of  competition are policy barriers to trade
the judiciary, the enforcement capacity of  Exporters  from   developing  countries-
legal authorities, and probity in public admin-  particularly exporters of agricultural prod-
istration. A well-intentioned law in an map-  ucts, textiles, and labor-intensive manufac-
propriate institutional environment can be-  tures  and  services-confront significant
come a source of bureaucratic harassment and  restraints on their ability to compete in global
corruption.                                 markets. Developing countries generally face
Governments working together in a multi-  higher barriers to exports than do industrial
lateral or regional framework may be able to  countries (World Bank-IMF 2002). Japan and
enact policies that widen the scope of compe-  the United States provide maximum protec-
tition and thereby confer benefits beyond   tion against imports from developing coun-
those obtained from unilateral reforms. Analy-  tries, while European Union protection is
sis has to begin with the restraints on compe-  skewed against imports from middle-income
tition in the global marketplace that most  countries. Developing countries, with average
adversely affect developing countries and that,  barriers higher than those in rich countries,
if removed, would provide the biggest stimu-  also raise barriers against competition from
lus to development.                         other developing countries. Taken together,
Three categories of restraints on competi-  protectionist measures such as high tariffs,
tion in the global marketplace are particularly  tariff peaks, restrictive tariff rate quotas on
adverse. First are those that involve policy  low-tariff imports, and domestic and export
barriers to trade that disadvantage exporters  subsidies are ubiquitous and raise barriers to
in developing countries by directly limiting  competition from all developing countries. Be-
their ability to compete in markets. The most  cause the world's poor people usually produce
important barriers affect agriculture, textiles,  agricultural and labor-intensive products, the
and other labor-intensive manufactures and  world trading system is tilted against the poor.
services. Second are private restraints on inter-  The average poor person selling into the
national competition that can raise prices to  global marketplace confronts tariffs that are
consumers or to producers in developing     twice as high as those faced by people who are
countries. These restraints include interna-  not poor (World Bank 2002c; see also Oxfam
tional cartels that are commonly illegal in  2002).
OECD countries when they affect OECD mar-     Subsidies and trade barriers in agriculture
kets. Third are officially sanctioned restraints  are particularly pernicious. In developed
that may adversely affect developing coun-  countries tariff rates in agriculture are twice
tries' import or export prices. We discuss  those of manufactures. Sheltering of agricul-
below the effects of exemptions from antitrust  ture by hefty subsidies aggravates the effects
laws that governments grant to their firms  of these   tariffs  (OECD  2001; World
national export cartels, and the price-raising  Bank-IMF 2002). The costs of such price sup-
effects of ocean transport and aviation     ports are borne by low-income consumers in
134



IN T ER NAT IO NA L AG Re E MEN T S T O   IM P R OV E I N V E S T MI E N T A N D C O M P E T I T I O N
protected markets-those consumers who       oping countries' terms of trade through its
spend a large proportion of their income on  effects on long-term world prices. Protection
food, while the supports benefit only a hand-  of agriculture is also common in developing
ful of large farmers. The U.S. subsidies to  countries-comparable in weighted ad val-
cotton producers, for example, cost taxpayers  orem equivalent terms-but is much lower
nearly $4 billion a year-three times the U.S.  when subsidies are taken into account (see
aid budget for Africa-while adversely affect-  World Bank-IMF 2002).
ing low-inconme West African economies that    Policy barriers restrain competition in
produce cotton. High protection and support  clothing and textiles with similarly adverse
of the sugar industry in the European Union  effects on developing countries. Developing
and the United States is another example of  countries account for about 50 percent of
these harmful policies. Total OECD support  world textile exports and 70 percent of world
for agriculture amounted to 1.3 percent of the  clothing exports (World Bank-IMF 2002).
gross domestic product of those countries in  Under the Uruguay Round Agreement on
2001, with the producer support estimates12  Textiles and Clothing, quota restrictions are to
the highest in the European Union in absolute  be abolished gradually during 1995-2005.
terms (see figure 4.6). Prices received by  The slow pace of removing restrictions on
OECD farmers were on average 31 percent     competition in textiles and clothing has re-
above world prices (measured at the border)  sulted in sizable losses in export earnings and
(World Bank-IMF 2002). Though efforts       productive employment in many developing
have been made to lower protection for agri-  countries. The combined negative income ef-
culture in OECD countries, the recently en-  fect for developing countries caused by quotas
acted 2002 U.S. Farm Bill increases support  and tariffs o1 industrial-country imports
spending to a projected $45 billion, or 21 per-  amounts to $24 billion annually, and the
cent of producer income during fiscals      export revenue loss is $40 billion (World
2002-07 (see appendix 2). This increase may  Bank-IMF 2002).
well aggravate secular deterioration in devel-  Impediments to competition take other
forms as well. Between 6 and 14 percent of the
tariff lines of Canada, the European Union,
Japan, and the United States are subject to
Figure 4. 2001  ountriesptu$23          tariff peaks, in some cases at rates well over
producers                                 100 percent (Hoekman, Ng, and Olarreaga
2001). Developing-countries' exporters may
Producer support estimate by the OECD countries  be displaced by high tariff peaks in Canada
totaled $230 billion In 2001              and the United States (in textltes and clothing)
and in the European Union and Japan (in agri-
culture, footwear, and food products). Even
though France exports 12 times more to the
L4~7,242 mll1on                    United States than Bangladesh, U.S. tariff
European union        revenues on imports from Bangladesh were
$93,083 million       roughly the same tariff revenues on imports
from  France  (Gresser 2002). Escalating
tariffs-in which protection is lower for pri-
United States                  Inary products but increases as the local value
$49,001 million                added increases-discourage development of
forward processing. Chilean firms, for exam-
Soturce World Bank-1\4F (2002)            ple, can export fresh tomatoes to the United
States, paying a tariff of 2.2 percent; however,
135



G LO BA L E CO NO MI C      PRO SP EC TS 2 O 0 3
if they dry and package the tomatoes, the U.S.    need to convince the importing government to
tariff is 8.7 percent; and if they make salsa     impose or tolerate entry restrictions, or need
out of the tomatoes for export, the duty is       to  be able to raise private entry barriers
11.6 percent (Schiff 2001). By reducing the       (Hoekman and Kostecki 2001).
demand for higher-processed imports from de-         In practice, post-World War II cases of suc-
veloping countries, tariff escalation prevents    cessful predatory dumping are the exception,
developing countries from diversifying exports     not the rule. More than 90 percent of all an-
into areas of their competitive advantage.        tidumping investigations would never have
These tariff structures are common in poor as     been launched if a competition standard-
well as in rich countries (see World Bank         potential threat of injury to competition-had
2002c: 45).                                       been used as a criterion (Messerln 2000).'3 As
Another restraint on competition is fre-        it has evolved, antidumping has become a fa-
quent recourse to    antidumping and     other    vored vehicle for restricting competition from
types of contingent protection. Antidumping       imports, and it is applied with increasing fre-
laws were originally created to counteract        quency by developing countries against each
predatory practices of foreign sellers into a     other. Since 1995, countries have initiated
home market. ThLs was the original rationale      more than 1,800 antidumping investigations
for U.S. antidumping legislation of 1916. The     (table 4.1). Although industrial countries have
fear was that a foreign firm (or cartel) could    traditionally been the main users of such mea-
deliberately price products low    enough to      sures, developing countries have been more
drive existing domestic firms out of business     active in recent years, led by India, Argentina,
and to establish a monopoly. Once estab-          Brazil, and South Africa. In the seven years to
lished, the monopolist could more than re-        2001, developing countries initiated almost
coup its losses by exploiting its market power.   two-thirds of all investigations, well in excess
For predation to work, the monopolist or car-     of their share in world trade. However, devel-
tel would not only have to eliminate domestic     oping countries have also been the target of
competition, but would also have to be able to    nearly 60 percent of investigations, mostly mi-
block entry by new    competitors. It would,      tiated by other developing countries. The re-
therefore, need to have a global monopoly,        cent steep rise in antidumping investigations
Table 4.1 Many antidumping investigations were initiated during the 1995-2001 period
Affected countries
indusmat     United    Eutropean    Dicloping   Transition          l
Initiating country         countries   States      Union       counttnes     counai es   Total
Number of investigations    511         102         313         1,086         248        1,845
Industrialcountries        128         17          67           363          114         605
Of which
United States           79           0         46           146           30         255
European Unton          15           6          0           165           66         246
Developing countries        379          85         242          718          131        1,228
Transition countries         4          0           4             5           3           12
Percentage of investigations  28          6          17           59           13         100
Industrial countries        21          3          11            60          19          100
Of which
United States           31           0          18           57           12          100
European Union           6           2          0            67           27          100
Developing countries         31           7          20           58           l           100
Transition countries         33           0          33           42           25          100
Souirce WTO Secretariat, as reported In World Bank-IMF 2002
136



IN T ER NAT IO NA L AG RE E MEN TS T O    I M P R O V E I N V E S T M E N T A N D C O M P E T I T I O N
puts the predictability and nondiscriminatory  countries. These cartels are typically illegal
application of trade policies at risk.      when they adversely affect a country's own
Removing these restraints on competition  commerce. However, OECD     governments
from developing countries would have a big  have no authority to prosecute cases when
development payoff. These issues and detailed  cartel activities function outside their national
policy recommendations have been well ana-  jurisdictions and cannot be shown to affect
lyzed elsewhere (see, for example, World Bank  prices of imports or domestic goods.
2002c). Suffice it to say that dismantling both  The 1990s saw the uncovering of several
worldwide trade barriers and agricultural sub-  interinational cartels. Prosecutions of interna-
sidics could increase long-term growth in de-  tional cartels picked up after 1993 when the
veloping countries by as much as 0.5 percent  United States revised its anticartel enforce-
annually, which, when taken together with   ment practices to grant amnesty to the first
termis-of-trade improvements, could reduce  cartel member that cooperated with authori-
the number of people living in poverty by as  ties. Before 1993, approximately one firm a
much as 13 percent by 2015. One-third to    year applied for leniency under anrticartel
one-half of the welfare gains would accrue to  laws, and big cases were rare; now, one firm a
the developing world (World Bank 2002c).    month applies for leniency. U.S. fines against
Because of the growing importance of South-  domestic and international cartels during the
South trade and the remaining high barriers  1990s totaled $1.7 billion. The publicity asso-
among developing countries, removing the    ciated with these prosecutions (many of which
barriers to competition among themselves    affected international markets as well as the
would produce substantial gains (see World  United States) encouraged prosecutions by
Bank 2002a; and World Bank-IMF 2002).       other enforcement agencies, including those in
These facts underscore the importance of the  several middle-income countries (for example,
Doha Development Agenda of the WTO and      Brazil and Korea). Antitrust authorities in the
the various regional efforts arounid the world  United States and European Union alone
that could lower trade barriers to developing  prosecuted 40 international cartels during the
countries' exports. Because not all countries  1990s.
will benefit from some reforms (such as re-    Cartels that have been uncovered through
moving the textile quotas), a broader reform  law enforcement have had a substantial role in
that covers all trade issues and is linked to  increasing the prices to developing countries.
development assistance is vital.            Although estimates vary, the average interna-
tional price increases caused by international
Private restraints on international         cartels have been estimated to be on the order
competition can raise prices                of 20-40 percent. The estimated price in-
to developing countries                     creases resulting from cartels, as shown in six
Besides policy barriers to competition, large  high-profile international cartel prosecutions
international companies with market power   (table 4.2), vary widely-from 10 percent for
can form cartels that fix prices, allocate mar-  stainless steel tubes to 45 percent for graphite
kets, and restrain competition. Although trade  electrodes. Cumulative overcharges to devel-
reform and the expansion of potential com-  oping countries over the life of the cartels in
petitors in markets around the world have   the six cases ranged from $3 billion to $7 bil-
undoubtedly reduced the scope for private   lion, depending on whether SITC or HS codes
cartels, the numerous international cartels un-  are used. Developing countries imported 12
covered in the 1990s suggest that market    products that had a value of sales of $11 bil-
forces alone do not offer complete protection  lion in 2000 and that were sold by interna-
against price-fixing and market-allocation  tional cartels prosecuted during the 1990s
arrangements that raise prices to developing  (figure 4.7); if price collusion were to raise
137



G LO B AL     EC ONO MI C        PROS PE CT S        2 0 0 3
Table 4.2 International cartels can be expensive: estimates of sales and overcharge
Possible overcharge to
Cartel sales                         devcloping countries
Ycars of   Number                                   Pocc
Product                cartel    of firms      srrc           HS         increase     SITC            HS                Fines
Vitamins             1990-99'              $26.4 billion  $10 8 billion  35%       $3 05 billion  $1 71 billion  Almost $2 billion
Citric acid          1991-95       111     $9 9 billion   $447 million   20%/,     $402 million  $67 million     Over $250 million
Bronsine             1995-98        2      $598 million   $409 million   15%      $46 million    $8 milliols     $7 million
Seamless steel tubes  1990-95        8     $26 6 billion  $21 7 billion  10%      $1 63 billion  $1 19 billion   99 million euros
Graphite electrodes  1992-97       23      $9 billion     $7 billion     45%      $1 35 billion  $975 million    Over $560 million
Lysine               1992-95        5      $4 8 billion   $913 million   10%       $294 million  $43 million     About $200 million
SITC = Standard International Trade Code, HS = Harmonized System Classification
Notes Figures for each cartel span the entire period of the conspiracy Sales are approximated using export statistics from countries of origin of indicted
firms and thus exclude domestic sales If participating firms are multinationals and the locations of their subsidiaries are known, sales are calculated by
taking into account the exports of countries of subsidiaries When that information is unavailable and production is understood to be global, sales are
calculated by using exports of all countries producing the cartel product Overcharge refers to Imports to developing countries / (1 + price increase) x
price increase Sales calculations provided are based on the SITC Revision 11 and the HS 1988
a Because the cartel ended in February 1999, sales and overcharge estimates are aggregated from 1990 to 1998
Soturce Connor (2001), Levenstein and Suslow (2001), OECD (2000), and World Integrated Trade Solution database
prices by an average 20 percent, the total over-           questions about the effectiveness of prosecu-
charges would have reached almost $2 billion               tion as a deterrent for cartel behavior. More-
in 2000.                                                   over, 24 of the 40 cartels prosecuted by the
Despite   the   rise  in  prosecutions, reining         United States and the European Union lasted
in international cartels remains difficult. The            for at least four years, indicating that market
fines imposed     by authorities often      fall well      forces are    not always     adequate     to  rapidly
short of the     estimated    overcharges, raising         eliminate    cartels.   The    history    of  cartels
Figure 4.7 Imports affected by cartels rose from 1981 to 2000 for both rich
and poor countries
Total imports ot twelve products where proven cartels existed
(in billions of dollars)
35
30
25
25                ~~~~~~Imports to industrial countries_               ___-           =    _    __
20
-                               .:; -                        ---- 
10
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Year
Souirce World Banik staff
138



I N T E R N A -r I o N A L A G R E E M E N T S  T O  I M P R O V E  I N V E S T M E N r A N' D C O M P E1  I T I O N
Box 4.6           The lysine cartel, 1995-2001
Lysine is a food additive used in hog and poultry    The cartel had a significant effect on both po-
feeds. The global lysine cartel lasted from 1992  tential producers and users of lysine. Lysine produc-
to 1 995. During that period the five participants  tion in 1994 was at least 20 percent less than under
controlled more than 97 percent of global capacity.  competitive conditions, resulting in lower production
Cartel members engaged in price-fixing, allocating  among the feed and meat industries that depend
sales quotas, and monitoring volume agreements. In  on lysine. Moreover, the cartel limited potential
1994, at the peak of the cartel's effectiveness, the  developing-country competitors by Lising price dis-
price of lysine reached about $1 .20 per pound, ap-  crinminationl across regions, and it froze the relative
proximately $0.50 above the competitive price level.  positions of the leading firms in the market, when
Estimates of the overcharges to U.S. customers  compared with the very fluid situation before the
during this period vary and are as high as $141 mil-  conspiracy. Although a few relatively small produc-
lion. Although no formal analysis of non-U.S. over-  ers entered the market during the 1990s (mainly in
charges is available, the observed lower prices in  Hungary, the Slovak Republic, and South Africa),
Asia suggest overcharges in the rest of thc world  most new entrants began production only after the
were lower than those in the United States. Accord-  lysine cartel had been broken up In 1995. China, in
ing to Connor (2001) a teasonable projection of the  particular; has been a source of increasing lysine pro-
global overcharge by the lysine cartel would be in  duction. Nevertheless, the five original participants
the $200 million to $250 million range. A more con-  in the cartel continued to control 95 percent of
servative estimate assumes a 10 percent overcharge  global capacity at the end of the decade.
on $1.4 billion in global sales during the life of the
cartel, for a total of $140 million (OECD 2000: 1 6)  Souirce Connor (2001)
indicates that some operate intermlilttently over  ability of prosecution to be very low (Eveniett,
decades. 14                                   Lehmann, and Steil 2000).
One option for curbing illegal international
New initiatives to discipline illegal         cartels ts to extend further the extraterritorial
international cartels                         reach of industrial nations' anticartel laws
Firms will be deterred from price fixing and  (Hoekman and Mavroidis 2002). When a
forming cartels if the fines for doing so, multi-  competition authority in an industrial econ-
plied by the probabdiity of being caught (that  omy uncovers a cartel that affects markets
is, the expected value of the cost), exceed the  both inside Its owin borders and in other coun-
extra profits that result from this anticompet-  tries, then that authority could take enforce-
itive behavior (that is, if the potential punish-  ment action on behalf of all affected nations.
ments for creating cartels exceed the benefits).  A stronger version would have the competi-
Reforms that raise the sanctions on cartels and  tion authority take action even if the cartel
that increase the probability of successfully  affected a foreign market without affecting the
prosecuting cartels will tend to dissuade more  home rnarket. In both cases, the authority
firms froi-n forming cartels, whether domestic  could request help in collecting evidence from
or international. The secret nature of most   enforcement bodies in other nations. Fines
cartel agreements poses a special problem     and sanctions against the cartel would be de-
because it implies that governments must ac-  termined on the basis of its detrimental effects
tively search for evidence or must encourage  on all affected economies.
cartel members to come forward with evi-        Yet another option is to granc govcrnmcnts
dence; otherwise, firms wvill perceive the prob-  of developing countries-or their citizens-
139



G LO B AL EC ONO MI C PROS PE CT S 2 0 0 3
standing in the major OECD countries so     provisions have, unfortunately, been applied
those affected could initiate private injury  to illegal conduct uncovered during cartel in-
suits against companies headquartered under  vestigations. These restrictions on information
the jurisdiction of a particular antitrust au-  exchange are especially worrisome at a time
thority. Because most antitrust actions are dri-  when so much evidence about international
ven by private complaints and through private  cartels is being collected through national
suits, such legal changes would markedly   leniency programs, thereby suggesting that the
strengthen the hand of consumers and busi-  potential for information exchange could be
nesses in developing countries to curb private  considerable.
restraint practices. The principal attraction of  Another approach is a multilateral agree-
such a proposal is that it would allow devel-  ment. Proponents of including competition on
oping countries to benefit from the sophisti-  the multilateral agenda have gravitated to-
cated investigative powers and regulatory   ward a relatively narrow focus. They are seek-
expertise in the OECD competition authori-  ing disciplines on (a) the so-called core issues
ties. The enforcement record in the 1990s sug-  of nondiscrimination, national treatment, and
gests that the overwhelming malority of cartel  transparency; and (b) private "hard core" in-
members have their headquarters in industrial  ternational cartels. These disciplines would
economies. A drawback to the proposal is that  apply to all WTO members, both industrial
extraterritorial application is a perennial  and developing, with technical assistance and
source of tensions among countries, and the  capacity building envisaged. Most recent dis-
incentives are low for OECD governments to  cussions have emphasized the need for volun-
take actions against their own firms for effects  tary international cooperation (Anderson and
in foreign markets.                         Jenny 2001).)S
A more modest option for reform could       In summary, policies that help developing
focus on notification and information ex-   countries discipline international cartels more
changes by national enforcement authorities.  effectively would have a potentially large ben-
This exchange would build on the growing    efit, for consumers in rich and poor countries
number of bilateral cooperation agreements  alike.
on competition matters, thus expanding their
scope to include many more economies. The   Officially sanctioned private restraints can
objective here is to raise the probability of suc-  hurt trade to developing countries ...
cessfully prosecuting cartels by encouraging  Officially sanctioned restraints on trade make
the sharing of conspiracy-related information  up the third major category of competition
between enforcement authorities. The modali-  restrictions that adversely affect developing
ties for this type of international cooperation  countries. These restraints take the form of
have received considerable attention in recent  exemptions from domestic antitrust laws and
years, not the least of which is the OECD's  pertain to certain types of international activ-
nonbinding Recommendation on Hard Core     ity. Many governments legally permit their
Cartels. However, this approach essentially of-  own private firms to cartelize export mar-
fers gains only to those economies that have  kets-as long as markets affected are outside
effective competition laws, and many devel-  the country, and export cartels do not provide
oping economies do not. Furthermore, the    an opportunity for producers to fix prices
amount of information that can be exchanged  at home. Indeed, numerous economies have
on cartel cases today is highly constrained  explicitly  exempted  export cartels from
because most countries have laws against   their domestic competition laws-essentially
sharing confidential information. The original  providing some legal cartel privileges for
intent of those laws was to protect legal busi-  their national firms, but not foreign firms
ness secrets and pLans, and the confidentiality  (table 4.3). U.S. soda ash producers have
140



IN T ER NAT IO NA l       AG RE E ME NT S       TO   I MP R O V E   INV EST ME NI        AND     CO MP ET IT IO N
Table 4.3 National exemptions to competition law for exporters
Country         Type of exemptioon                           Reporting requirement
Australia       Contracts for the export of goods or         Submission of full particulars to the national
supply of services outside Atistralia       authority within 14 days
Brazil          Joint ventures for exports, as long as there  Approval by the national authority
are 11o effects on the Brazilian market  I
Canada          Export activities that dc) not affect domestic  None
cornpetitioti
Croatia         Agreements that contain restrictions that aim  Notification of the agreement to national autlority
to improve the competitive power of          . h.-,  days after coniclusions of the agreenenit
undertakings on the international market
Estonia         Activities that do not affect the dtinestic niarket  None
Hungary         Activities that do not affect the domestic market  Nonie
Japan           Agretmients regarding exports or amonig      NotticaElon of and approval by the industry
domestic exporters                           administrator
Latvia          Activities that do not affect the domestic market i None
Lithuania       Activities that do inor affect the domesti mLiarket  None
Mexico          Associations and cooperatives that export    None
New Zealand     Arraigements that relate exclusively to exports  Authorization of the national authority
and that do not affect the domestic niarket
Portugal        Activities that do not affect the domestic niarker  |..
Swedeii         Activities that do not affect the doniestic market  Nonie
United States   Webb-Pomerene Act activities that do not      i. I    i's Act, filing of agreements with the
affect domestic competition                  i'  1.1 . ,Iade Commission
Export Trading Companies Act strengthened    F.    1 . , 0 .rJ  Companies Act- Certificates of
immunities granted by Webb-Pomerene Act      Review provided by U.S. Department of Commerce
Source Eveniett and Ferrarini (2002), drawn from OECD (1996), OECI) (2000), and <http Hwww gettingthedealthrough com>
(accessed May 2002)
taken advantage of these provisions in U.S.            legitimate in the past, most small- and medium-
law   to  form   an  export cartel, which      has     sized enterprises in industrial economies today
subsequently    been   the  target of European         export without a need for cartels, so the ratio-
and   Indian  enforcement actions. Generally,          nale is moot.
these cartels may attempt to raise prices in               Another exemption from       OECD     antitrust
their export markets       to  the  detriment of       laws is maritime transport, which         inadver-
overseas consumers. Their success depends on           tently put developing countries at the mercy
the number of other foreign competitors in             of price fixing. The exemption        in U.S. law
these markets. Because competitioll is more            extended    to  maritime   transport has facili-
likely to  be limited   in the smaller markets         tated, through shipping conferences, collusive
of developing countries, it is probable that de-       arrangements in ocean-liner shipping. Agree-
veloping   countries    are  adversely    affected     ments    among    private   shlipping  companies
disproportionately.                                    have a long history, beginning with trade be-
Because cartel registers are secret in Europe       tween the United Kingdom        and India in the
and Japan, and virtually secret in the United          1870s. Such arrangements have taken differ-
States, information on their extent, products,         ent forms, includinig the conclusion of agree-
and geographic coverage is nil. The legal ex-          ments on     uniform    freight tariff rates and
emptions are known, and the latest available           conditions of service, the establishment of ex-
information-from      the OECD     in 1974-has         clusive or preferential working relationships
indicated  a broad    proliferation. The    initial    between shipping Iines, or the integration of
rationale for export cartel exemptions was that        shipping networks through strategic alliances.
small exporters     could    join  to  share   the        The    power of such       arrangements     has
allegedly substantial costs of marketing their         eroded in recent years because outside ship-
products abroad. Even if such arguments were           ping lines have gained a significant share of
141



G LO BA L EC O NO MI C PRO SP EC TS 2 O 0 3
Table 4.4  Breaking up floating cartels       beyond their borders. Government support
could help developing countries               for beggar-thy-neighbor export cartels is
(Econiomic effects of eiding private restrictioiis  anachronistic in an era of global trade rules.
on ocean-liner conipetitton)   ______________  Reciprocal international agreements offer the
Effect                          Amount        promise of reducing foreign distortions to
Reduction in price of ocean transport  20%    domestic markets in return for commitments
Projected total savings for U S imports  $2 I billiorI  to desist from such practices. Agreements on
Proiected savinlgs for developing-country     international cartels ivolve givig up some
imports                        $2 3 billion
_   rents from exporting in return for the benefits
Souirce Fink, Matroo, anid Neagu (2001), WVorld Bank  of more competitive markets at home.
(2002c)                                          A multilateral accord to curb export cartels
would probably benefit developing countries.
the market and regulators have moved to en-   An alternative and less-ambitious approach is
courage greater price competition. Nonethe-   to narrow the coverage to sectors in which it
less, Fink, Mattoo, and Neagu (2001) con-     can be demonstrated that small- and medium-
clude that a breakup of cooperative working   sized enterprises cannot compete internation-
agreements and price-fixing arrangements      ally without a mechanism to share burdens
among the major private carriers could reduce  such as marketing costs, and so on. Because
transport prices by 20 percent on U.S. routes,  the extent of injury to foreign consumers is
for a savings of $2 billion or more (see      not known, a minimalist policy toward export
table 4.4; see also Francois and Wooton 2001).  cartels involves disclosure. If export cartels are
If developing countries could save the same  allowed to retain their legality, governments
percentage of their import costs, then their  should agree to require that firms seeking to
total import bill would fall by $2.3 billion.  establish an export cartel publicly register as
This figure is probably an underestimate of   such-and that those registries be updated an-
the effect of breaking up private constraints  nually and made accessible to the public over
on ocean trade services for developing coun-  the Internet. Furthermore, if these cartel ex-
tries. Their freight charges are more likely to  emptions were specifically to aid small firms,
be subject to price-fixing than are freight   then there is no argument for permitting large
charges on industrial-country routes because  firms to participate.
low traffic volumes limit the number of com-     Similarly, countries could agree to end anti-
mercially viable competitors. For example, the  trust exemptions for maritime transport and,
European Commission found that the Associ-    at the same time, give standing so exporters
ated Central West African Lines abused its    in developing countries that are harmed by
dominant position by providing rebates to     subsequent cartel activities can sue under an-
shippers that complied with its policies, as  titrust statutes. This change would have sig-
well as carrying out other anticompetitive    iificant effects by unleashing competition in
practices. 16                                 this sector and by altering an arrangement
that today drives up the cost of exporting
and international agreement could          from many developing countries.
rein in their adverse effects
Multilateral efforts to curb national export  International collaboration can strengthen
cartels, as well as to rein in private restraints  domestic competition policies
in regulated industries that have been rooted  Domestic policies in developing countries
in exemption from antitrust laws, are particu-  have a significant effect on competitive con-
larly well suited to the WTO. Most govern-    ditions. Chapter 3 underscored the particu-
ments today either encourage or acquiesce to  lar importance of policy barriers to competi-
national cartels that adversely affect markets  tion, particularly in trade, in restrictions on
142



IN i ER NAT IO NA L AG RE E MEN TS TO I MP R O V E INV EST MEN t AND C O Nl P E r I I o N
incomliing FDI, and in restrictions on new   tion law is technical and requires the usc of
entry (foreign or domestic) in regulated indus-  skills that are in short supply In many devel-
tries. Chapter 3 also concluded that the po-  oping countries. Building capacity to apply
tential role of a domestic competition agency  competition legislation effectively will take
was shaped largely by the domestic institu-  time. Given that competitioni law is applied
tional environment. In some countries with   on a case-by-case basis, dealing with systemic
strong legal and judicial systems, a compcti-  trade and investment barriers and with gov-
tion agency could help augment competition;  ernment regulations that rcstrict competition
in other countries with weak legal and judicial  may generate a higher rate of return (see
systems, establishinig a competition agency  chapter 3). Kee anid Hockman (2002) have
could be counterproductive if they become a  investigated the effect of the existence of a
source of rent-seeking and corruption.       competition  law   on  estimated  industry
International discussions on trade policy  markups over cost. They used cross-country,
have, since their inceptioni, seen domestic  cross-industry time series panel regressions
competitioni policy as an issue associated with  that include data on the nunmber of firms bv
market access. Competition policy is intrinsi-  industry (turnover), sales (market size), and
cally related to the principles of national treat-  import competitioni. They concluded that
ment and MFN treatment insofar as competi-   antitrust legislation on its own has no effect
tion law allows recourse to address certain  on markups, but that imports and entry have
kinds of discrminiiatory policies and arrange-  a major and statistically significant effect in
ments that deny foreigners access to markets. 1 7  reducing markups (see chapter 3). Competi-
The launchiniig in 1997 of the WTO Work-   tion law is found to have an indirect effect,
ing Group on Trade and Compctition Policy    however, by reducing the first order marginal
signaled the beginning of the most recent in-  effect of imiports and by reinforcinig the mar-
ternational discussions about the interface  ginal cffcct of domestic competition. That
betweeni trade and competition, as well as the  effect is stronger In the more-developed and
possibility of multilateral cooperation on   larger economies.
competition law. Not all domestic competition   The effect of government policies that re-
matters give rise to internationial trade prob-  strict competition for nontradables may be
lems, and vice versa. There are sitiations   more important from a developmenit perspec-
when the lack of, or inappropriate applicatioii  tive thani is antitrust enforcement, because
of, competition law can impede trade and     those policies affect the price and quality of
market access, however. After five years of  key intermediate inputs that determine the
discussions, governments have progressively  competitiveness of industrics on world mar-
retreated from ambitious applications (such as  kcts (for example, Fink, Mattoo, and Neagu
harmonization) to proposals that focus on    2002; Francois and Wooton 2001). Depend-
core principles, transparency, nondiscrimnia-  ing on the capacity of governnmenit, a role may
tion, and procedural fairness. Governments   exist for a competitionl agency that reviews
may perhaps also focus on provisions address-  new policy and rcgulatory barriers to compe-
ing illegal interinationial cartels (see discussion  tition (see chapter 3, thlis volume, as well as
above). Aside from these general principles,  Anderson and Holmes 2002).
the exact content of national competition       As Winters (2002) notes, administration of
laws could vary considerably in the range of  competition law is complex, and its misappli-
conduct and structural disciplines that they  cation can have a costly and chilling effect on
include.                                     investment. Issues relating to the institutional
From a national point of view, for compe-  design, the indepenidence of investigatinig au-
tition law to be a priority it must yield a  thorities, the effective judicial review  and
higher payoff than other choices. Competi-   appeal mechanisms, and the availability of
143



G LO BA L EC O NO MI C PRO SP EC TS 2 0 0 3
expertise-both legal and analytical-are all  the national administration will operate out-
critical issues for the effective application of  side its direct control. Until a few decades ago
antitrust law. Therefore, the development of  most European Union member states had no
competition law in many countries has oc-    experience in the field of antitrust. Before a
curred gradually over a long period, and con-  government determines national priorities,
tinues to evolve. The necessary administrative  both the costs and benefits of competition en-
apparatus cannot be put into place within a  forcement ought to be considered, including
short time frame. The institutional guarantees  the possibility of perverse outcomes through
necessary for a competition authority to be  capture or corruption.
independent from eventual political influence   This discussion suggests that the reciprocal
(so that it can concentrate on its mandate) re-  bargaining and enforcement framework of
quire government acceptance that branches of  the WTO is less well suited to collective action
Box 4.7          HInternatioinial cooperation aai±d
competition policy
S everal entities outside the WTO have activities that  ICN, created on October 25, 2001, to deal with in-
iare germane to competition policy. For example,  ternational antitrust enforcement through regular
the OECD launched a Global Forum on Competition  consultations between government officials, private
in October 2001 to stimulate a comprehensive policy  firms, and NGOs from around the world. According
dialogue about competition, and that goes beyond its  to its mandate, the ICN will "formulate proposals
previous activity of providing technical assistance.  for procedural and substantive convergence through
The Forum, backed by the OECD's Committee on     a results-oriented agenda and structure." Its special
Competition Law and Policy, engages in high-level  status stems from the fact that it is maintained by the
discussions with key officials from member and non-  enforcement authorities themselves, has voluntary
member countries, including countries that do not  membership, and is not bound by rules, but rather
have well-developed competition enforcement author-  by a community of interests. The first annual confer-
ities. The objective of the Forum is, first, to encourage  ence was held in Italy during 2002 and sparked dis-
common understanding and sharing of experiences  cussions on reforms to the merger review process;
among a larger number of competition officials and,  the advocacy role and activities of competition agen-
second, to generate benefits through cooperation,  cies (especially in developing and emerging
conflict prevention, and voluntary convergence. Its  economies); and recommendations on best practices.
first meeting successfully highlighted the role of  Individual enforcement authorities will have the flex-
competition policy and of its authorities In economic  ibility to make decisions on the most suitable means
reform; it also fomented greater international cooper-  of implementing the recommendations. The ICN will
atnon on such matters. The latest semiannual meeting  address complex issues, and newly established com-
in February 2002 discussed the merits of competition  petition authorities will no doubt benefit from the
policy for developing economies, international coop-  collective experience of other member agencies.
eration in merger and cartel cases, capacity building,  Though it is too early to gauge the success of
and technical assistance. In addition, the forum bene-  the Global Forum on Competition and the ICN in
fits from contributions of regional organizations such  terms of fostering global cooperation, they play a
as the Common Market for Eastern and Southern    useful role in disseminating information on best
Africa and international organizations such as the  practices for implementing a competition law policy.
World Bank, UNCTAD, and the WTO.
Another example of an entity outside the WTO                                                  I
with activities germane to competition policy is the  Source: World Bank staff
144



IN T ER NAT IO NA L AG RE E MEN TS 'IO I MP R O V E INV EST MEN T AND CO MP ET I 'II ON
on competition law than internationial col-  cific objectives that will be important to de-
laboration through development assistance   veloping and reinforcing positive domestic
and other venues. To be sure, international  policies rather than distorting them.
negotiations can help reinforce progressive   For investnient policy, international agree-
domestic reforms in competition law  (see   ments may help increase flows of foreign
Birdsall and Lawrence 1999).18 However, in  investment, but evidence suggests that these
this complex area of domestic regulation, one  benefits are likely to be limited unless they
size does not fit all, and, as maniy WTO mem-  focus on creating nondiscriminatory terms
bers have noted, cooperation on competition-  of liberalization and on eliminating adverse
law policy requires establishing a domestic  policy externalities. Agreements that curb
enforcement capacity that at present is beyond  beggar-thy-neighbor investment policies that
the reach of many developing countries. Other  distort investment location are particularly
channels can help disseminate best practices to  important in two areas. One critical area is
countries wishing to strengthen their competi-  investment-distortinig trade barriers-that is,
tive conditions. Several agencies and forums  border protections, agricultural subsidies, tar-
have work programs on interiationlal compe-  iff escalation, and other practices that bias in-
tition policy. These agencies include the   vestment flows away from developing coun-
OECD, UNCTAD, and the International        tries' export activities because such barriers
Competition Network (ICN) (see box 4.7).    discourage imports from those countries A
The OECD and UNCTAD have developed          second critical area is disciplining competition
their own guidelines or recommendations for  among governments to lure foreign invest-
tackling international cartels, but they have no  ment through wasteful investment incentives.
powers of enforcement or investigation. The  An important initial step is developing an in-
nascent ICN has focused more on interna-    ventory of the extent, costs, and distorting
tional mergers and acquisitions, and it is in-  consequences of those incentives. Agreements
tended to facilitate informationi exchange and  should be carefully designed to limit their
disseminiation of best practices.           scope to areas where international externali-
ties exist. In the case of the WTO, the design
should focus on reducing discrimination and
Conclusions                                 increasing market access. International coop-
For both investment and competition policy,  eration on the design of domestic regulation is
Fdomestic reforms that are implemented      more effectively provided through develop-
unilaterally in the national interest of promot-  ment assistance-whether bilateral, regional,
ing a sound investment climate and a more   or multilateral.
competitive economily are likely to yield the  For competition policy, an agreement
most direct and positive effect on growth and  would potentially have large benefits if it
poverty reduction. The interinational commu-  addressed those restrictions on competition in
nity can assist the reform  process through  the global marketplace that most adversely
multilateral and bilateral development assis-  affected developing countries: policy barriers
tance, government-to-government informa-   to competition that hurt exporters, private re-
tion exchanges, and private efforts to inforimi  straints in the form of international cartels,
and assist reform-minded governments. Coun-  and officially sanctioned private restraints em-
tries may be able to use regional and multi-  anating from  antitrust exemptions. Much
lateral agreements to motivate progressive re-  more information is needed in this area on the
forms at home at the same time that they use  prevalence and effects of policies that restrict
reforms to leverage reforms abroad to pro-  competition. The international community
mote development. Yet to be effective, these  can collaborate with developing countries by
agreements must be designed to achieve spe-  providing technical and financial assistance
145



G LO BA L EC ONO MI C       PRO SP EC TS     2 O 0 3
to foster mutual learning, information ex-        the "financial contribution" that could cover the range
changes, and cooperation      on competition      of fiscal and financial incentives that are used by de-
policy.                                           veloped and developing countries. These disciplines
have yet to be tested. In the case of services, the GATS
provides a mandate for developing "necessary multi-
lateral disciplines to avoid such trade-distortive
Notes                                             effects." The work has progressed slowly
1. WTO 2001a, Paragraphs 20 and 23 in the Doha    10. There is evidence of significant investment diver-
WTO Ministerial Declaration. The need for enhanced  sion away from the Caribbean Basin countries and to-
technical assistance and capacity building in these  ward Mexico, but Mexico's adherence to NAFTA has
areas was also recognized.                        almost certainly been a more important motivating fac-
2. See Ostry 1997; see also Hart 1996.         tor than the use of fiscal or financial incentives, which
3. United States 1998, as cited in Gilpin 2000: 184.  it can generally ill afford.
4. See Smythe 1998.                                11. This is not to deny the potential risk of "invest-
5 South-South FDI is calculated by comparing    ment poaching," including within developing coun-
developing countries' FDI inflows with recorded out-  tries. Studies have indeed documented the negative
flows from other regions. This figure may be more ac-  welfare implications that derive from incentive pack-
curate than others because developing countries often  ages that merely transfer investment from one location
underreport FDI outflows In addition, round tripping  to another without creating new jobs or improving
of a country's own capital can overestimate the FDI  productivity. In the case of Brazil, for instance, the con-
figure (World Bank 2002a).                        sensus among researchers is that heavily indebted
6 For a cogent description of the predominance of  states have granted very large tax breaks to automotive
services in the NAFTA reservation lists, see Rugman  companies to build factories that the companies had
and Gestrin 1994 See also Gestrin and Rugman 1993.  intended to build in Brazil anyway (Rodriguez-Pose
7 By 2003, all members must have completely    and Arbix 2001).
phased out performance requirements that were in     12. Producer support estimates are the annual
place at the time of the agreement and that were grand-  monetary value of gross transfers from consumers
fathered through a notification process. All 27 notifi-  and taxpayers to support agricultural producers. These
catnons of policies nor consistent with the agreement  numbers are taken from World Bank-IMF 2002.
were from developing countries. Almost half of noti-  13. The fact that predation has very little to do with
fied measures related specifically to the automotive sec-  antidumping as it is practiced is perhaps best illustrated
tor. Many of these performance requirements have  by the United States, which has two antidumping
already been phased out during the transition period.  statutes. One, the Antidumping Act of 1916, maintains
Ten countries that requested an extension of the tran-  a predation standard for antidumping, the other, the
sition were granted an additional four years, to 2003.  Tariff Act of 1930, as amended, has a price and cost-
8. WTO members are faced with two options.     discrimination standard. Invariably cases invoke the
First, they can agree to re-open the agreement, which  second act and not the first.
seems unlikely. Second, they can seek to reduce or elab-  14 Epstein and Newfarmer 1980, for example,
orate on the length of the Annex Illustrative List The  found that a cartel for heavy industry operated off and
issue is that, even though both notifications and dis-  on from December 1939 through the mid-1970s, with
putes have, to date, centered primarily on the "illus-  overcharges of more than 20 percent on sales of steam
trated" list (notably on local content and, less so, on  turbines and other products.
trade-balancing requirements), the agreement arguably  15. WTO members with established competition
prohibits a greater range of as-yet unspecified perfor-  enforcement seem to insist that a precondition for co-
mance requirements. Introducing greater specificity in  operation is that developing countries adopt legislation
the language could enlarge the effective coverage of the  and establish enforcement capaciry- "[C]ooperarion
agreement or confine it to the illustrated list.  with respect to competition matters [is] only possible
9. Within the framework of the ASCM the scope  when a competition regime [is] already in operation,
for discipline lies in the challenge of an investment in-  that is, when there [is] a domestic competition law of
centive that can be shown to be specific, to be within  some sort and a domestic competition authority ex-
the meaning of the agreement, and to be contingent on  isted with sufficient powers to effectively enforce that
export or on having an "adverse effect" on the trade of  law . While cooperation could be provided within a
another member. The difficulty of such a challenge de-  voluntary framework of mutual interest, it would not
pends on the specific types of policies that are in ques-  be possible for a developing country to eradicate anti-
tion. One of the key factors in determining a subsidy is  competitive practices which had an impact on their
146



IN I ER NAT IO NA L     AG RE E MI ENTS     IO s   MP R O V E INV ES I MIN T AND          C O M P E I I r I o N
markers unless it also developed a national competitioni  Evenett, Simon J , and Benno Ferrarini 2002 "Devel-
law" (WTO 2001c 27, para. 79).                          oping Country Interests in Internationial Carrel
16. See World Bank 2002c for a fuller discussion of  Enforcements in the 1990s " Background paper
conferenices on ocean liners,                           for Global Economnic Prospects 2002. Iunvesting
17 See WTO 2002.                                     to Unlock Global Opportuntties World Bank,
18 Birdsall and Lawrenice (1999) write "iWhen de-   WXashington, D C
veloping countries enter into modern trade agreements,  Evenett, Sinon J , A Lehmann, and B. Steil 2000
they often make certain commitments to particular       Antitriest Goes Global: What Futture for Trausnza-
domilestic policies-for example, to antitrust or other  tional Corporatio;P Lonidoni Royal Institute of
competition policy. Agreeing to such policies can be in  Inteinational Affairs, and Washington, D C
the interests of developing countries (be) ond the trade  Brookings Institution Press
benefits directly obtained) because the commitment  Finger, J. Michael, and A. Winters 2002 "Reciprocity
can reinforce the internal reform process Indeed, partic-  in the WITO." In Bernard M  Hoekman, Philip
ipation in an international agreemenit can make feasi-  English, and Aaditya Mattoo, eds., Developymenit.
ble internal reforms that are beneficial for the country  Trade, anjd the WTO A Hanidbook Washington,
as a whole [andl that might otherwise be successfully   D C  World Bank
resisted by interest groups.                       Fink, C, A Mattoo, and I C. Neagu 2001. "Trade in
International Nlai itire Services How  Much
Does Policy Marter`" IWorld Batik Econtomiic
Review 16 81-108
References                                         Francois, Joseph F, and I Wooron 2001  "Trade in
Anderson, Robert D., and F Jenniy 2001. "Curient        International Transport Services- The Role of
Developments in Competition Policy in the World    Competition i" Review of International Econoni-
Trade Organization." Antitrnst 16(1l) 40-44 Fall   ics 9(2). 249-61 (May).
Anderson, Robert D., and Peter Holmes 2002 'Com-   Gestrin, Michael, and Alan M. Rugman 1993 "The
petition Policy and the Future of the Multilateral  NAFTA's Impact on the North American Invest-
Trading System " Jottrnal of Iinternational Eco-   ment Regime " C. D. Howe Commentary no 42,
nomic Lavw 5(2) 531-63                             C D. I-fowe InstitLute, Toronto March.
Aykur, D , and D. Ratha. 2002 "South-South Flows   Gilpin, Robert 2000    The Cballenge of Global
in the 1 990s " Background Paper for Global        Capttalisn- The World Economy rn the 2 Ist Cen-
Developmtenit Finance  2002  NWorld  Bank,         tury. Princeton, N J Princeton University Press
Washington, D C. April.                        Gresser, Edward. 2002  "America's Hidden Tax on
Birdsall, N , and Robert Z. Lawrence 1999. "Deep        the Poor The Case for Reforming U S Tariff
Integration and Trade Agreemenits. Good for        Policy." Progressive Policy Institute Policy Report
Developing Countries?" In Kaul, GCunberg, and      Progressive Policy Institute, Washington, D C.
Stern, eds , Global Puiblic Goods. Internationjal  March.
Cooperation in the 21st Century. 128-151.      Hallward-Driemeter, M. 2002. "Bilateral Investment
[[City']]. Oxford University Press.                Treaties Do They Iicrease FDI Flows'" Back-
Cano, Wilson. [998. Deseqtilibrios Regionais e Con-     ground Paper for Global Economic Prospects
centrado Industral no Brasil, 1930-1995 2d ed.    2003. lnitestuog to Unlock Global Opportzitittes
Campinas, Sao Pauilo- lnstituto de Economia da     World Bank, Washington D.C.
UNICAMP                                        Hart, N4 1996, "A Multilateral Agrecmenit on Foreign
Chariton, Andrew. 2002  "Incentive Bidding Wars         Direct Iivestmenit Why Now?" In P' Sauv& and
for Mobile Investment' Economic Consequences       D. Schwanen. eds , Investment Rules for the
and Potential Responses." OECD Developmenit        Global Econonmy Enbancing Access to Markets,
Center, Paris,                                     Policy Study 28: 36-99. Toronto. C D  Howe
Connor, John M  2001 Global Price-Fixing. Oir Ctis-    Institute.
tomners Are the Eneniy Boston. Kluwer Academic  Hoekman, Bernard M., and Michiel Kosteck-i. 2001
Publislsing.                                       The Polrtical Economy of the World Tradintg
Epstein, B., and Richard S. Newfarmer. 1980 "Inter-     Systeon: The WTO and Beyond Oxford and
national Electrical Association. A Continuing      New York: Oxford University Press
Cartel." Report to the Committec on Interstate  Hoekmnan, Bernard AI., and Petros C. Mlavrordis
and Foreign Commerce, U.S. House of Repre-         2002   "Economic Development, Competition
sentanves U S   Government Printing Office,        Policy, and the WTO " P'aper presented at the
Washington, D C June                               Roundtablc "Inforiilng the Doha Process New
147



G LO BA L EC ONO MI C       PRO SP EC TS     2 0 0 3
Trade Research for Developing Countries."     Oman, Charles. 2000. Policy Competition for Foreign
Cairo, May 20-21                                  Direct Investment: A Stidy of Competition
Hoekman, Bernard M., and K Saggi. 1999. "Multi-        among Governments to Attract FDI. Paris:
lateral  Disciplines  for  Investment-Related     Organisation for Economic Co-operation and
Policies " Policy Research Working Paper no.      Dcvelopment
2138  Development Research Group, World       Ostry, Sylvia. 1997. "A New Regime for Foreign
Bank, Washington, D.C                             Direct Investment." Occasional Papers no 53.
. 2000. "Assessing the Case for Extending WTO    Group of Thirty, Washington, D.C.
Disciplines in Investment-Related Policies." Jour-  Oxfam (Oxfam International). 2002. "Rigged Rules
nal of Economic Integration 15(4) 629-53          and Double Standards: Trade, Globalization and
(December)                                        the Fight Against Global Poverty." London.
Hoekman, Bernard M., Francis Ng, and M  Olarreaga.  Rodriguez-Pose, Andres, and Glauco Arbix. 2001.
2001. "Eliminating Excessive Tariffs on Ex-       "Strategies of Waste: Bidding Wars in the
ports of Least-Developed Countries." Working      Brazilian Automobile Actor." Internatzonal Jour-
Paper no. 2604. World Bank, Washington,           nal of Urban and Regional Research 2(1): 134-54
D C. May                                          (March).
Kee, Hiau L., and Bernard M. Hoekman. 2002. "Com-  Rugman, Alan M., and Michael Gestrin. 1994.
petition Law as Market Discipline." Development   "NAFTA's Treatment of Foreign Investmiient." In
Research Group, World Bank, Washington, D.C.      Alan M. Rugman, ed., Foreign Investment and
Processed                                         NAFTA 47-79. Columbia, S. C.: University of
Levenstein, Margaret, and V. Suslow. 2001 "Private     South Carolina Press.
International Cartels and Their Effects on Devel-  Sauve, Pierre. 2002. "Collective Action Issues in In-
oping Countries." Background paper for the        vestment Rule-Making." Background Paper for
World Bank's World Developnient Report 2001.      Global Economic Prospects 2003: Investing to
World Bank, Washington, D C                        Unlock Global Opportunities World Bank,
Mattoo, Aaditya 2000. "Developing Countries in         Washington, D C.
the New Round of GATS Negotiations: Towards   Sauve, Pierre, and Christopher Wilkie. 2000 "Invest-
a Pro-Active Role." World Economy 23(4):          ment Liberalisation in GATS." In Pierre Sauve
471-89.                                           and Robert M. Stern, eds., GATS 2000: New Di-
. 2002 "Shaping Future Rules for Trade in        rections in Services Trade Liberalization 331-63.
Services: Lessons from the GATS." In Takatoshi    Washington, D.C.: Centre for Business and Gov-
Ito and Anne Krueger, eds., Trade in Services.    ernment, Harvard University, and Brookings
Cambridge, Mass: NBER.                            Institution Press.
Messerlin, Patrick A. 2000 Measuring the Costs of  Schiff, Maurice. 2001  "Chile's Trade Policy: An
Protection in Europe Washington, D.C Institute    Assessment." World Bank, Washington, D C.
for International Economics.                      November. Processed
Moran, Theodore. 2002. "The Relationship between  Smythe, Elizabeth. 1998. "Agenda Formation and the
Trade, Foreign Direct Investment, and Develop-    Negotiation of Investment Rules at the WTO:
ment: New Evidence, Strategy, and Tactics under   The History of a Campaign." Paper presented at
the Doha Development Agenda Negotiations "        the Annual Meeting of the International Studies
Paper prepared for Asian Development Bank's        Association, Washington, D.C. February 18.
Study on Regional Integration and Trade: Emerg-  Stern, Robert M. 2002. "Quantifying Barriers to Trade
ing Policy Issues for Selected Developing Member  in Services." In Bernard M. Hoekman, Philip Eng-
Countries, September.                             lish, and Aaditya Mattoo, eds., Development,
OECD (Organisation for Economic Co-operation and       Trade, and the WTO: A Handbook Washington,
Development). 1974. Export Cartels. Report of     D.C.- World Bank.
the Committee of Experts on Restrictive Business  UNCTAD (United Nations Conference on Trade and
Practices. Paris: OECD.                           Development). 1998. World Investment Report.
. 1996. "Antitrust and Market Access. The        New York: United Nations.
Scope and Coverage of Competition Laws and           1999. World Investment Report. New York:
Implications for Trade." Paris: OECD.             United Nations.
2000. Hard Care Cartels. Paris: OECD.            . 2000 Bilateral Investment Treaties 1959-
2001. "Agricultural Policies in Emerging and    1999 United Nations: New York and Geneva.
Transition Economies: Special Focus on Non-   -        2001a. World Investment Report. New York:
Tariff Measures." Paris: OECD                     United Nations.
148



IN T ER NAT IO NA L AG RE EM EN TS TO           I MP RO V E INV ES TM ENT        AND    CO MP ET IT IO N
2001b Handbook of Statistics. New York     World   Bank-IMF    (World   Bank-linternational
United Nations.                                    Monetary Fund) 2002    "Market Access for
2002   World Investment Report United          Developing Country Exports-Selected Issues
Nations                                           September.
United States 1998 Economitc Report of the Presi-  WTO   (World Trade Organization) 2001a  "Doha
dent. Washington, D C  U S Government Print-      WTO Ministerial Declaration." (WT/MIN(01)/
ing Office February                               DEC/1) November
Winters, Alan. 2002. "Doha and the World Poverty         2001b. "International Trade Statistics
Targets " Paper presented at the ABCDE Confer-       2001c. "Report of the Working Group on the
ence World Bank, Washington, D C. April.           Interactioni between Trade and Competition
World Bank 2002a. Global Development Finance.          Policy to the General Council." (WT/WGTCP/5).
Washingtoni, D C  World Bank.                      October 8
2002b   World Development Indicators            . 2002. "Core Principles, Including Trans-
Washington, D.C.: World Bank                       parency, Non-Discrimination, and Procedural
2002c. Global Economic Prospects' Making        Awareness." Background note by Secretariat.
Trade Work for the World's Poor. Washington,      (WT/WGTCP/W/209). Geneva, September 19
D C.: World Bank
149






Appendix 1
Regional Economic Prospects
East Asia and Pacific                        The recovery so far has been underpinned
by both external and domestic demand. A
Recent developments                        gradual pickup in world conditions from
T       HE EAST ASIA ECONOMIC RECOVERY     2001's severe global slowdown supported a
that began in late 2001 continued to  revival of exports, including exports in the imn-
strengthen in the first half of 2002, but  portant high-tech sector. U.S. dollar prices for
momentum slowed after the summer and uni-  many important East Asian primary commod-
certainties have increased. Output growth in  ity exports such as rice, rubber, palm oil, co-
developing East Asia is estimated to have risen  conut products, and lumber also rose sharply
to 6.3 percent in 2002 from about 5.5 percent  after late 2001, although they rose from the
in 2001, led by China growing at more than  very low levels reached following several years
7 percent. Output growth has rebounded     in which steep declines in the terms of trade
most rapidly in those economies, such as the  inflicted large income losses in many East
Republic of Korea, Malaysia, Taiwan (China),  Asian countries. Intraregional exports have
Singapore, and Thailand, that had been hard-  been strong, most notably those to China,
est hit by 2001's steep fall in world trade and  which jumped an amazing 50 percent in the
high-tech demand. Annualized growth in the  first half of 2002. Supportive monetary and
first half of 2002 over the last half of 2001  (in several cases) fiscal policy conditions have
jumped to a 5- to 10-percent range in most of  fueled domestic demand. Household spending
these economies, although industrial produc-  in the region has been especially robust, boost-
tion and trade data indicate that output  ing spending on cars and other consumer
growth softened in the second half of the year.  durables, but private investment still lags.
Elsewhere in the region growth had in any  Debt-equity ratios, though they have fallen
case slowed less among transition economies  in some instances, remain high by internia-
such as China and Vietnam, because of con-  tional standards in most cases, and together
tinued strength in domestic demand and a   with low corporate profitability, continue to
less marked slowdown in exports. Growth in  depress the investment climate and undermine
Indonesia and the Philippines had fallen less  the prospects for accelerated medium-term
sharply to a 3- to 3.5-percent pace in 2001  growth.
from a 4- to 5-percent pace in 2000, and     World semiconductor sales and orders for
the cyclical rebound in these countries (while  high-tech equipment in the United States re-
more pronounced in the Philippines), has also  bounded strongly in late 2001, but growth
been less clear-cut than elsewhere in the  rates had already peaked before the summer
region.                                    of 2002, and the levels remained far below the
151



G LO B AL EC ONO MI C PROS PE CT S 2 0 0 3
crisis. Local East Asian financial markets have
Figure A1.1 Emerging high-tech Asia:     been remarkably resihent in the face of the
semi-conductor dollar sales and          global equity and emerging market turmoil.
industrial production*                      On balance, countries in the region appear
(percent change, 3m/3m saar)             comparatively well placed to weather the pre-
125-       Semiconductor      - S0       sent high levels of external risk, especially
100 - g  /sales (left axis)  ^  - 40     if they can build on recent successes and con-
75 - [     Indusrial roduction  30      tinue to advance reforms that further reduce
50 -       ( // Industritaxi  p r   o  duclion - 20  their vulnerability to external shocks and fos-
25 -v \/ \ R   0 ter sources of domestic productivity growth.
-25- - -10                Improvements in macroeconomic conditions
-S0 __     __              -20      have been key to cushioning the impact of
Jan  Jul  Jan  Jul  Jan  Jul          external shocks and volatility in international
2000 2000 2001 2001 2002 2002         capital markets on East Asian countries over
Note Through August 2002 'Rep of Korea, Malaysia,  the last three years. Most countries have
Singapore, and Taiwan (China)
Source SIA and national sources through Datastream,  strengthened national balance sheet positions
World Bank staff estimates               by running continuous current account sur-
pluses since the 1997 crisis, reducing foreign
debts (including short-term debt), and build-
ing up foreign reserves. The adoption of more
record levels of 2000 (figure A1.1). A sharp fall  flexible exchange rate regimes has also allowed
in North American semiconductor equipment   countries to adapt to external shocks more
bookings in August and September suggests   smoothly (while the recent fall in the U.S. dol-
that this year's incipient revival in the depressed  lar has alleviated concerns about a loss of
global high-tech industry may have faded. Sev-  competitiveness in countries with currencies
eral East Asian countries report some slippage  on a dollar peg, such as China and Malaysia).
in August-September export and industrial     A potential source of macroeconomic vul-
production growth from the high rates reached  nerability in East Asia, however, is the higher
around midyear.                            level of public sector debt (including contin-
Steep falls in global equity markets in June  gent liabilities) accumulated after the 1997
through September, evidence of renewed      crisis in several countries. Public debt levels
weakness in industrial country growth, con-  are quite high in Indonesia and the Philippines,
tinued turmoil in emerging markets, rising  and are not insignificant elsewhere. Public
world oil prices, and terrorist attacks in the  debt-to-GDP levels are now starting to fall
Philippines and Indonesia are among the fac-  because of rising currencies and lower interest
tors that have substantially increased uncer-  rates, and well-judged fiscal policy measures
tainty about prospects for the region.      can build on this trend so as to secure medium-
term fiscal sustainability.
Near-term outlook                              Private investment in the crisis-affected
How well are East Asian countries able to ride  countries still remains weak, compared with
out new external and internal shocks that may  pre-1997 crisis levels. However, these peak
arise? Here the experience of the last two  levels were based on overoptimistic expecta-
years is cautiously encouraging. Neither the  tions and abundant foreign capital. Continued
serious export decline of 2001 nor the hefty  economic growth, macroeconomic stability,
terms-of-trade losses of recent years led to a  low interest rates, rising currencies, and policy
new wave of corporate and financial failures  efforts to improve the investment climate
in the countries previously hit by the 1997  should lay the groundwork for an investment
152



R EG I O N A L  E C O N O M I C  P R O S P E C L S
revival in due course. Equity markets in East     ness confidence could      reduce Indonesia's
Asia also fell sharply in this period, but over   growth by 1 percentage point in 2003. Gov-
the last one or two years they have generally      ernments in Southeast Asia must now grapple
outpaced industrial country equity markets.       with a major security challenge that also has
In countries with flexible regimes, currencies    potentially divisive domestic political implica-
generally rose against the dollar during 2002      tions-all at a time when many countries are
while remaining steady or depreciating mod-       moving into the next turn of their electoral
estly against the yen. The stronger trend in      cycles. A loss of focus on development and
currencies occurred despite the fact that do-     reform priorities could result.
mestic interest rates in most countries were         In this environment, acceleration of growth
significantly lower than average 2001 levels,      is unlikely, although there are also no signs of
and is striking evidence of improving investor    sharp deterioration. Regional GDP is expected
perceptions of the region.                        to grow 6.1 percent in 2003 and 6.4 percent in
The October terrorist attack     in Bali is     2004. Growth in the region, excluding China, is
also expected to reduce near-term   growth in      expected to reach 3.8 and 4.3 percent in 2003
Indonesia. Receipts from tourism represent4to     and 2004 respectivcly. Median inflation is cx-
5 percent of gross domestic product (GDP) in       pected to remain low, albeit probably above the
Indonesia, and in Southeast Asia generally.       historically low 2 5 percent in 2002 With cx-
Indonesian   tourist arrivals could    fall by     port volumes growing at a rate of around 9 per-
around 20 percent, based on the experience of     cent, the current accounts remain on average at
the 1997 terrorist attack on tourists in Luxor,    a comfortablc surplus of 2.5 percent of CDP.
the Arab Republic of Egypt, and the impact of
previous political unrest tn Indonesia itself.    Long-term outlook
Lower levels of tourist arrivals and the adverse  East Asian countries face growing demands
impact of the attack on consumer and busi-         for better-quality public goods and services,
Table A1.1 East Asia and Pacific forecast summary
(percent per year)
Baseline forecast
Growth rates/ratios      8T                                     2002    2003     2004   2005-i5
Real GDP growth                       SOS                I       6 3     6 1      6 4      6 2
Private consuimptioti per capita      £         iJ       v       5 7      5.2     s 6      5 6
GDPper capitr a                                          4S      5 4      5 2     5 6      5 4
P'opulation                          l         DJ!D    I       0 9      0 9     0 8      07
Gross domestic InvestmentlGDP°       St          1      31L11   32 8     33 9    34 9     30 4
Inflatiornb                           £5        IJD       j      2 5      3 6     3 6
Central gvt bridget balance/GDP      _f        =B      =3 -U 336        -3 4    -3.3
Export market growthc                 G)               =IN LM   ) 3 6    9 2      8 7
Export volumed                   -                       It     15 9     15 7    113
Terms of trade/GDPe                                     A4      -0 5    -0 I     -0 3
Current account/GDP                    }        ,        5i    -  2 7    2 7      2 5
Memorandumrt items
GDP growth East Asia excluding Chinla                    I       3 6     3 8      4 3      5 0
Not available
a Fixed investmenit, miieasuired in real terms
b Local currency GDP deflator, median
c Weighted average growth of import demand in export markets
d Goods and nionfactor services
e Change in terms of trade, measured as a proportion to GDP (percentage)
Source World Blank baseline forecast, November 2002
153



G LO BA L EC O NO MI C PRO SP EC TS 2 0 0 3
not least for better law and order. Protecting  Efforts to strengthen regional cooperation
the lives and property of the public and up-  in East Asia have gained ground in recent
holding rule of law are essential complements  years, alongside the region's long-standing
to the ongoing emergence or consolidation  commitment to multilateral trade liberaliza-
of representative or democratic government in  tion and integration. The Association of
the region, as well as a basic underpinning for  Southeast Asian Nations (ASEAN) Free Trade
economic development. The Bali terrorist at-  Area (AFTA) came into force at the begin-
tack underlines the importance of strengthen-  ning of 2002 among the six original members.
ing security and law and order, through both  Association members now seek to build on
national and cooperative regional efforts.  their success in reducing tariff barriers and
Continued public sector and governance re-  fostering deeper integration by tackling issues
forms are important if greater fiscal discipline in  such as non-tariff barriers, product standards,
East Asia is to be combined with meeting pub-  customs procedures, trade in services, trans-
lic demand for more and better public goods  port and logistics, and investment flows. At
and services, such as law and order, health, ed-  the same time the Initiative for ASEAN Inte-
ucation, basic research, technology dissemina-  gration (IAI) has been launched to help new
tion, and infrastructure generally. Many of these  and less-developed members (Cambodia, the
things are valuable to consumers in their own  Lao People's Democratic Republic, Myanmar,
right, are key elements of a productive invest-  and Vietnam) build capacity and integrate
ment climate, and provide a crucial underpin-  more fully into the AFTA over time. Regional
ning for political legitimacy and stability. There  discussions and cooperation in the ASEAN+3
is now open discussion and acknowledgment of  (ASEAN, China, Japan, and Korea) and other
these issues in the region, though the pace of re-  new forums also continue to develop.
form is generally slow, except in the Republic of  Against this background, per capita GDP
Korea, Malaysia, and Singapore.           in the region could grow solidly at a rate of
Countries have generally continued to make  around 5.5 percent during the coming decade,
some progress (at varying rates) on corporate  although this implies a slowing by 1 percent-
and financial restructuring, reforms to improve  age point compared with the previous decade
financial supervision and regulation, and  of rapid catching up. As population growth
strengthening corporate governance. However,  also slows during the coming 15 years-being
most countries need to deepen corporate and  0.5 percentage point lower than during the
financial sector reform efforts, as well as fur-  1990s-the long-run trend in GDP growth is
ther the broader institutional agenda of foster-  estimated to be 6.2 percent, 1.5 percentage
ing private sector development and strengthen-  point below the trend during the 1990s.
ing the investment climate. It is notable that
the sectors which are more dynamic are those
where the corporate restructuring agenda is  South Asia
less urgent, for example, small- and medium-
scale enterprises or large exporters who have  Recent developments
moved more quickly to resolve their debts.  j      rowth in the South Asia region is proj-
This is especially important because China's    ected to average 4.6 percent in 2002.
World Trade Organization (WTO) accession is .      This downward revision from  the
leading to a restructuring of regional produc-  Global Economic Prospects 2002 forecast of
tion networks, a surge in foreign direct invest-  5.3 percent GDP growth for the year largely
ment (FDI) to China, and an increase in com-  reflects adverse weather conditions, continued
petitive pressures in other countries in the  internal and external security concerns, and
region, especially those competing directly  the more protracted than anticipated recent
with China in low wage sectors.            global economic slowdown. As a consequence,
154



REG IO NA L EC ONO MI C PRO SP EC TS
given slower demand growth, inflation has
remained low throughout most of the region,   Figure A1.2 Industrial production in
notwithstanding high oil prices. Further, gov-  South Asia
ernment budget balances are expected to post   (3-month movlng average, y/y percentage change)
rising deficits in most countries. On an ag-   25     Bangladesh           Pakistan
gregate level, current account balances-        20           India
which strengthened during 2001, reaching a      15                        __ __I-it
surplus in the case of India-are projected     10
to show a slight improvenient during 2002.      5
This is partly due to lower levels of imports,  0
which are generally projected to more than     -s
offset the negative impact of slow export      -10                v
volume growth dampened by weak external        -15                v
de i and.                                       s,    ,.       qO. IO i4 iv i  ii
Because agriculture is a key sector in the re-     - #   -S -S W     -S � ,  @
gion's economies, important to both employ-
Source International lMonetary Fund, international
ment and growth, the recent drought has re-    FiTaa1cial Statistics
strained output, although the late rains (in
some cases flooding) have somewhat amelio-
rated the shortfall for the season. As a conse-
quence, 2002 harvests are expected to be gen-
erally smaller than in 2001, an outcome that is  2002, although it shows some firming relative
cxpected to put some additional pressure on  to the contraction witnessed in 2001 for most
government coffers through increased trans-  of the region's economies. During the second
fers for drought relief.                    half of 2002, preliminary data suggest some
Security issues are an ongoing concern, al-  further acceleration of export volume growth,
though there has been some easing of tensions  although not as strong as projected in Global
in parts of the region, paving the way for  Economic Prospects 2002, given the more
improved growth conditionis. The diminished  protracted recovcry in world trade volumes.
fighting in Afghanistan from late 2001, and  Import volume growth has been dampened by
the more recent introduction of a new govern-  sluggish demand for raw materials and inter-
ment in June 2002, has led to a rebound in ac-  mediate goods by exporters and by weak do-
tivity to begin rebuilding the economy from  mestic demand conditiois. The overall impact
extremely low levels and contraction in 2001.  of these factors has led to a net improvement
Throughout the region, the weakening of  in the region's trade balance. This improve-
economic conditions has complicated cfforts  ment, combined with a rise in the region's ag-
at fiscal consolidation. Although some coun-  gregate worker remittances-reflecting port-
tries were able to achieve budget targets by  folio shifts following the September 11 attacks
cutting expenditures-including spending on  and, in the case of Bangladesh, a sharp in-
development programs-to offset revenue      crease in private inflows due to improvements
losses, India notably did not cut expenditures.  in the speed with which remittances are trans-
Citing concerns about the magnitude of      ferred through official channels-has con-
India's domestic debt-estimated at 70 per-  tributed to an increase in the region's foreign
cent of GDP-Standard and Poor down-         reserve holdings. Given the rcccnt weakness of
graded Indian government paper to junk bond  the U.S. dollar, a number of currencies in the
status in October 2002.                     region have appreciated during 2002. How-
External demand in South Asia's major    ever, a general monetary stance of maintaining
markets has been subdued in the first half of  competitiveness suggests that these short-term
155



G LO B AL EC O NO MI C       PRO SP EC TS      2 0 0 3
cross-currency dynamics will reverse and that       provides the largest share of employment.
regional exchange rates will resume a path of       Higher rural incomes will feed through into
depreciation.                                       stronger private consumption      growth, but
prospects for investment-particularly in the
Near-term   outlook                                 private sector-are more mixed, as continued
Our near-term outlook is premised upon con-         concerns over stability will at least partially
tinued   improvement      in   both   domestic      offset the impetus to invest generated       by
political stability and regional security issues-   stronger domestic and external demand. In-
although   tensions will remain     evident. A      dustrial production will marginally underpace
marked improvement in Sri Lanka's political         real GDP growth, as the service sector contin-
and policy environment, for example, should         ues to provide the main near-term       contri-
allow that country to enjoy a sharp accelera-       bution to growth over much of the region.
tion of GDP over the near term-and, indeed,            Our global projections of a recovery in
into the medium   term. In contrast, continued      world trade prospects translate into strength-
domestic political pressures would cause Nepal      ening external demand      conditions for the
to underperform    relative to regional growth      economies in the South Asia region. Partly be-
averages in the near-term forecast horizon.         cause of the above assessment of the domestic
We are also assuming a return to more            economy, the region is forecast to post a firm-
normal weather patterns over the next two           ing of growth to an average of 5.4 percent in
years, in contrast to the recent drought condi-     2003 and 5.8 percent in 2004. Trade balances
tions discussed above. Agriculture accounts         should benefit, although import growth will
for approximately one-fourth of regional out-       also strengthen along with external demand,
put, but the impact of improved harvests will       given  the high   import content of exports
be even more pronounced       for the poorest       and high import intensity of consumption.
households, for whom       the rural economy        Coupled with a modest deterioration in the
Table A1.2 South Asia forecast summary
(percent per year)
Baseline forecast
Growth rates/ratios                 1991-2000   2000     2001     2002     2003    2004    2005-15
Real GDP growth                         5 2       4 8      4 4      4 6     5.4     5 8      5 4
Private consumption per capita          2.0       1.4     4.1.      2 6     3.6     4 0      3 2
GDP per capita                          3.3       2 9      2.6      2 9     3.8     4 2      4 l
Population                            1.9       19       1.7      1 7     1 6      16       1.3
Gross domestic mvestment/GDPa          21 9      24.2     24.9     25 8    25 8    25 5     24 8
Inflationb                              7.9       3.9      6.t      5 0     5.1     6 8
Central gvt budget balance/GDP        -10.3     -9 7    -10.3    -10 3     -9 8    -9.2
Export market growth'                  128       13.1      04       28      73      79
Export volumed                         10 9       7.4      5 3      8 3     8 8     8 5
Terms of trade/GDP'                    -0.1     -0.8       0.4      0.0     0 2     0 2
Current account/GDP                    -1.5     -0 8     -0 3      -0 1    -0 2    -0 2
Memorandum items
GDP growth South Asia excluding India   4 3       4 2      3 8      3 9     4 8     5 2      5 3
Not available
a Fixed investment, measured in real terms
b Local currency GDP deflator, median
c Weighted average growth of import demand in export markets
d Goods and nonfactor services
e Change in terms of trade, measured as a proportion to GDP (percentage)
Source World Bank baseline forecast, November 2002
156



REG IO NA L EC ONO MI C P R O S P E C I S
terms of trade, these growth patterns are ex-  Long-term prospects
pected to translate into a slight increase in the  Long-term growth in South Asia is forecast
regional average current account deficit to  to average about 5.5 percent, in line with
GDP ratios.                                 the growth forecast in Global Economic
Macroeconomic policies in the region have  Prospects 2002. This forecast is somewhat
historically been biased toward accommodat-  higher than the 5.2 percent average real
ing growth and have generated fiscal imbal-  growth posted during the 1990s. The higher
ances. Generally, revenue mobilization remains  projected growth over the coming decade,
a challenge in the region. The unsustainable  through 2015, reflects a number of underly-
nature of fiscal deficits has been widely rec-  ing assumptions, such as the assumption of a
ognized by governments across the region-   larger contribution of growth by the private
although this does not imply that a consensus  sector. This in turn reflects the expectation of
has been reached. Therefore, the introduction  progress with fiscal consolidation and contin-
of prudent fiscal policies over the next two  ued structural reforms, including reforms in
years could be tempered by political consider-  trade, banking, privatization, and infrastruc-
ations, and by the need to address the lagged  ture. These factors, combined with the im-
effects of the current downturn in agricultural  provement in human capital indicators in
output. Indeed, budget deficits in many coun-  recent years-such as rising literacy rates and
tries could deteriorate in the next fiscal year.  school enrollments, and declining infant mor-
One of the key problems is the scale of public  tality rates-will lead to an increase in pro-
debt-large servicing requirements limit the  ductivity. Despite a projection of declining
scope for cutting expenditures, especially in  infant mortality rates, overall the South Asian
conditions of extreme poverty. The main issue  population growth rate is projected to deceler-
for the region is extending the tax base, but  ate because birthrates are expected to decline
progress to date has not been as rapid as an-  faster. Lower population growth in the com-
ticipated. The introduction of tax reforms in  ing decade, along with the forecast growth
India, for example, has been delayed.       rates, implies that per capita GDP growth will
Most regional economies have a general    be close to 4 percent per year.
monetary policy posture involving deprecia-
tion aimed at shoring up foreign reserves and  Risks
promoting exports through increased compet-  Downside risks to the forecast include a weaker
itiveness, which is forecast to also have a pos-  than anticipated recovery in global demand,
itive impact on current account balances. This  translating into slower export growth and
conduct of monetary policies had been facili-  lower regional GDP. Similarly, a continuation
tated by relatively weak inflationary pressures  of adverse weather conditions would result in
across the region. In general, consumer price  lower than anticipated growth caused by
inflation in the region has been relatively low,  lower agricultural output. Domestic political
despite the droughts and firm oil prices of re-  uncertainty could slow down implementation
cent years. Several countries have subsidies for  of fiscal and  other structural reforms.
fuel and some foods, a policy that cushions  Regional political tensions also pose impor-
the impact of imported inflation on consumer  tant threats to the growth forecasts, poten-
prices. However, these are being lifted in some  tially aggravatilg economic disruptions and
cases (for example, Pakistan). While the recent  increasing poverty rates, as well as generating
slowdown in demand will restrain price rises  less severe effects, such as declines in tourism
initially, inflation is forecast to rise moderately  revenues and the reduction of foreign assis-
later in 2003 and in 2004 as growth momen-  tance. Generally, the region faces a number of
tum builds, and as the average prices of non-  vulnerabilities, as evidenced by the covariant
oil imports rise.                           shocks over recent years-such as droughts in
157



G LO B AL EC ONO MI C PRO SP EC TS 2 0 0 3
Pakistan, floods in Bangladesh, natural disas-  of 1.2 percent. Negative spillover effects from
ters in parts of India, and civil war in Sri  the meltdown in Argentina began to affect
Lanka. Vulnerability to such shocks has con-  neighbors in the region in the second half of
tributed to the high incidence of transient  the year.
poverty in a country such as Pakistan, and   The external environment for most of Latin
probably explains the large year-to-year fluc-  America was more adverse than expected at
tuations in income and poverty that are ob-  this stage of the global economic recovery.
served. Such vulnerability can also have a  Despite low interest rates in industrial coun-
profound impact on the growth prospects of  tries, capital flows to developing countries fell,
a country, since uninsured risk can affect the  and the decline was especially pronounced
ex ante behavior of economic agents.      in the Latin America and the Caribbean
(LAC) region, partly because of the crisis in
Argentina and its small neighbors. After
Latin America and the Caribbean            falling for most developing countries between
October 2001 and April 2002, spreads on
Recent developments                        external debt rose sharply for many LAC
U       nihke most other developig regions  countries-Chile, Mexico, and small Central
where economic growth strengthened  American and Caribbean countries had only a
Lin 2002, GDP contracted 1.1 percent     small rise (figure Al.4). Few countries were able
In Latin Amenica, about 1.6 percentage ponts  to attract the necessary capital flows to sus-
lower than antclpated in the spring. This  tam a strong growth recovery in the absence
growth deceleration from 0.4 percent in 2001,  of rising domestic savings. U.S. growth, after a
however, was the result of enormous eco-   strong start in the first quarter of 2002, weak-
nomic contraction in a handful of countries,  ened significantly thereafter, while European
fueled in large part by the external environ-  growth was anemic, resulting in lower growth
ment and aggravated by domestic factors.   expectations for LAC principal markets in the
Growth performance in the region, excluding  short term. The region's export market growth
Argentina, is expected to be 0.7 percent in  rate was a disappointing 1.2 percent in 2002
2002, somewhat lower than last year's growth  and was not helped by the price fall in key
commodities exported by the region (for ex-
ample, sugar fell by 26.5 percent, arabica
Figure A1.3 GDP growth in LAC,          coffee by 8.8 percent, bananas by 7.4 per-
1997-2002                               cent, aluminum by 6.5 percent, and copper
(percent y/y)                           by 0.2 percent). Moreover, tourism revenues
12                                     were weak because of reduced air travel from
8          Argenhina  OiherLAC         North America (affecting the Caribbean coun-
4                                      tries), and the collapse of income in Argentina
o,     ,                               significantly affected tourism  in Paraguay
-4          \y            \            and Uruguay as well as workers' remittances
-8                          \           to Bolivia and Paraguay.
\8           Domestic factors are important in explain-
-12                                     ing the weak economic performance in a small
-16
-20                             V       set of countries, and these countries contri-
Q1    Q1    01   01    01    01       buted most to the region's dismal growth
1997  1998  1999  2000  2001  2002    performance in 2002. In Argentina, the lack of
Note *Other LAGC= Bra7dI, Chile, Mexico  political consensus on a sustainable macro-
Souirce Datastream and World Bank staff estimates  econ1omic framework has delayed an Interna-
tional Monetary Fund (IMF) program, shrunk
158



REG IO NA L EC ONO MI C PRO S rEC IS
Figure A1.4 Spreads for selected LAC countries
(basis points above US Treasures)
2,500 -                                                                     - 7500
2,000 -                                                                      -6 000
Argentina (right axis)
1,500 -                                   A- 4500
Countrnes outside Bank sf esti tes
Latin Amen                             dca         e cu       
d,000- pt                          Ia           o      th a3,000
--                         /      ~~~~~~~~~~~~~~Mexico
500   p        T   a       o   -                                             1.5M0
0                                                                         0
Jan         Apr    Jul       Oct      Jan       Apr       Jul          Nov
2001      2001     2001      2001      2002     2002      2002         2002
ucSoerrce Dasasorea t and World Ba thk s eaff estimaoes
net capital flows gcbn furtlor, and led to a  reserves aond CatSIlg  the currency to float,
deep, protracted economic contraceiof in g x-  and forcing thB authoritnes to ethebark obe   fis-
cess of 10 percent. This affected other Merco-  cal consoidation and tmonetary tighted ig.
surafflates deeply, especially Ui-uguay and, to  GDP growth is estIloated to havc fallon by
a lessdr ixtent, Paragtay and Bonvia. In Brazil,  about 10 percant. Brazil was forced to tightea
uncertainty associated with the elections as  fiscal pobicy but was able to lower short-tari
well as with th. global outlook weighed on   pohcy rates owly teoporarily and experienced
mvestor confidence f1 Spite of sorsnd macro-  sigcalficant cicreases an long rates. The xpctial
cconomic poelcies, siowing th pac ogeou       Urgroth Brazul ina conveStr ent at the beginnng
recovert . tn   the Repcbrca Boglvariana de  of the yea  d gave way to a  sore lmued second-
Venezi. ela, asn acute poitical crisis that culmd-  half growth. en Colotibca, the fiscal accounts
nated in a short-ived coup coIntributd to a  renai vauthcrable, with the governdecit mnak-
resurgence in large-scalc capital outfhows, lttle  tg sfiall progress an reducing the fiscal deficit
lnvestmeat outside of the oll sector, and a stiep  to brag debt dynamics oiinto a sustaable
decline in growth.                           path in the wake of presidential elections,
The conflueTnce of these factors weakened  escalatin1g civil war, and lower-tliaii-expected
the acceleration in growth that one would ex-  growthi. Peru, in contrast, was able to make
pect in the context of recovering global activ-  progress in addressinig the fiscal deficit andc
ity. Policy responses, constrained by hijgh debt  in keeplnlg the public debt situation under
levels, high external borrowviig requirements,  conitrol; growtlh accelerated in 2002. Ecuador
and contracting external fni1ancing, were Uil-  benefited from relatively hiigh oil prices, but
avoidably contractionary. In Uruguay, a fin-  thc authorities had diffiCLIlty in reducing
aiicial crisis elnsued as Argentnnies withdrew  the fiscal deficit, a move rnecessary to rnana-
a large amount of dollar deposits, depletn1lg  tain tnacrocconioiiiic discipliile in a dollar
159



G LO BA L EC O NO MI C PRO SP EC TS 2 0 0 3
economy. In Central America, low export    $10 billion lower than at the end of last year,
earnings from falling coffee prices and weak  due primarily to sharp declines in Argentina
demand in the United States, and relatively  and Uruguay.
high fiscal and external deficits, limited fiscal
expansion. And some Caribbean countries    Near-term outlook
were faced with little scope for fiscal expan-  The region's growth prospects are expected to
sion or monetary easing in light of declining  improve in 2003, supported by strengthening
tourism  revenues; weak prices of bananas,  of the global economy, particularly in trade
sugar, and aluminum; and, in the case of   volumes, commodity prices, and capital flows.
Jamaica, a large debt overhang that was ex-  The region's GDP is now expected to grow by
acerbated by the financial sector crisis in the  1.8 percent in 2003-still almost 2 percentage
mid-1990s. In Mexico, the government stuck  points lower than the spring forecast and in
to fiscal discipline-helped by higher-than-  line with the downgrading of world growth,
expected oil prices-and is making slow     provided there is a turnaround in the current
progress with key reforms, while being only  uncertain political and financial market out-
partially successful in increasing non-oil gov-  look. Greater fiscal adjustment in a number of
ernment revenues.                          countries with high debt and relatively large
Overall, policies in Latin America were in-  financing requirements is a necessity for reduc-
strumental in keeping inflation from accelerat-  ing economic vulnerabilities. This, along with
ing in most countries. However, countries that  reforms currently on the agendas of many
had a sharp adjustment in their exchange rates  countries, is needed to restore investor confi-
due to a sharp fall in net capital flows did face  dence (which will lower interest costs), attract
inflationary pressures. Inflation (as measured  more equity external financing, and reinvigo-
by consumer price index-CPI) rose to fairly  rate growth.
high levels in Argentina (about 45 percent), in  In the baseline forecast, it is assumed that
Venezuela (in excess of 20 percent), and in  Argentina will put in place an externally sup-
Uruguay (about 20 percent). Given high debt  ported macroeconomic program that will be
levels and worsening public debt dynamics-  reaffirmed by the new government after the
due to high interest rates and depreciating  election in March 2003. Depending on its
currencies-most authorities were unable to  actual timing, this would lead to a revival of
pursue countercyclical policies, and unemploy-  growth, possibly by midyear-but may not
ment remained high throughout the region   show up in strong annual growth until 2004.
(21.5 percent [Mayl in Argentina, 19 percent  If this scenario were to materialize, it would
[September] in Uruguay, 16.2 percent [unel  improve prospects for the smaller Mercosur
in Venezuela, 17.2 percent [September] in  countries. The base case also depicts a new
Colombia, and 9.7 percent [Septemberl in   Brazilian government that maintains prudent
Chile).                                    macroeconomic policies and succeeds in
Slow world growth and external financing  restoring market confidence. Regional GDP
constraints compelled an adjustment in the re-  growth is expected to rise to 3.7 percent in
gion's external accounts. Import volumes fell  2004.
for a second consecutive year (mostly due to  There are a number of positive factors for
sharp declines in Argentina, Uruguay, and  the region that give credence to the baseline
Venezuela), and the current account deficit  forecast. Mexico has weathered the recent
narrowed from over $50 billion (or 2.9 per-  global downturn very well. Sound policies
cent of GDP) in 2001 to below $25 billion (or  have kept investors confident and inflows
1.5 percent of GDP) in 2002, with most of  large, and the upturn in U.S. growth should
the adjustment coming after April. The level  bolster a rapid expansion in 2003. Chile also
of foreign reserves in August was around   has fared well in light of negative effects from
160



REG IO NA L     EC oNO MI C      PRO SP EC IS
Table A1.3 Latin America and the Caribbean forecast summary
(percent per year)
14   _ .             Baseline forecast
Growth rates/ratios               jI.'1-2i iii   .wi      2001    2002     2003     2004    2005-1S
Real GDP growth                         3 3       3 7      0 4    -I 1      1.8      3 7      3 8
Private consumption per capita          2.3       2 4    -1.1     -2 8     -0 3      1 9      2 5
GDP per capita                          1.6       2.1      1 2    -2 6      0 3      2 3      2 6
Population                            1 7       1.6      1 6      1 S      14      14        1 2
Gross domestic investment/GDPa         19.8      19 7     19 3     18 0    17 3     18 1     22 0
Inflationb                             12 S       6.4      6.9      5 0     4 2      4 4
Central gvt budget balance/GDP        -3.1       -1 8    -1 8     -2 6     -3 1     -3 0
Export market growth,                  114       11.9    - I 5      1 2     82       8 8
Export volumed                          8.6      10.5      14      4 7     11 1     10.9
Terms of trade/GDIp                     0 2       0 6     -0.5    -0 1     -0 3     -0S
Curreit account/GDP                   -28        -24     -2.9  |- 5        -1 3     -I 8
Memnorandum itenms
GDP growth LAC excluding Argentinia     3 1       4 5      1 2      0 7     19       3 6      3 8
Cenitral America                      4 4       2 9      1.6      21      3.1      3 6      4 0
Caribbean                             3.5       6 2      2.7      3 6     4 3      4 3      4 2
Not available
a Fixed investment, measuired in real terms
b Local currency GDP deflator, median
c Weighted avcrage growth of import demanid in export issarkets
d Goods and nionifactor services
e Change in terms of trade, measured as a proportion to GDI' (percentage)
Souirce World Banik baseline forecast, November 2002
Argentina and adverse terms of trade, and            monetary overhang and weak public finances.
expected firming of copper prices should un-         While it is assumed that the required political
derpin growth. The policy stancc of the new          consensus will be attained before the March
Colombian administration has been positive           2003 elections, there are risks that the con-
for investors, and it has strong support from        sensus may not materialize that quickly. The
the United States for assistance in resolving        incoming Brazilian admitlistration will have
the guerrilla war. Expected strengthening of         to take confidence-building measures to re-
metals prices next year should help Peru to          verse the current market uncertainty. Should
reduce the rise in external financing as the         Brazil fail to maintain a sustainable macro-
economy recovers from      the 2001   recession.     economic policy framework, its correspon-
And the small island states in the Caribbean         dent weak economic performance will have
have shown resilience in the face of strongly        a major impact on regional economic pros-
negative   external and    domestic   factors-       pects. Venezuela    will be facing    lower oil
Jamaica has stayed the course with tight fiscal      prices, which will make necessary adjustment
policy (a primary surplus of 9 to 11 percent of      even more painful. And Caribbean countries
GDP in the past three years); Trinidad and           face the possibility of a delayed recovery in
Tobago's economy continued to grow despite          tourism, which      would    reduce   the  fiscal
political uncertainty.                               space to tackle the ever-present risk of natural
However, risks are on the downside in the         disasters.
short term. Argentina     faces enormous ob-
stacles in its financial system, and while the       Long-term   prospects
situation  has stabilized somewhat in 2002,          Per-capita GDP growth over the long term
the authorities will face a difficult situation      (2005-15) is projected to average 2.6 percent
in containing   inflation  in 2003-with      the     a year, 1 percentage point higher than the



G L0 B AL EC ONO MI C PRO SP EC TS 2 O 0 3
growth achieved by the region in the 1990s.  prices. Finally, the region still lags in financial
Improvement in macroeconomic management    deepening (which could help raise national
in a number of countries throughout the    saving rates), infrastructure, and quality of
1990s, albeit emanating from crises, should  institutions-areas that need to be improved
provide the basis for a good investor climate.  before the region can attain high sustainable
This has already encouraged FDI into the re-  growth rates.
gion at the fastest pace of all regions, even
though this has been due partly to privatiza-
tion of public enterprises and mergers and ac-  Europe and Central Asia
quisitions. The next wave of FDI is likely to be
in greenfield activities, as witnessed in Costa  Recent developments
Rica's attracting high-tech firms. Regulation    he slack external environment, espe-
and supervision of financial sectors have been   cially in Western Europe, is contribut-
strengthened since the early 1990s, but there  ,  ing to a general slowdown of growth
remains room for further improvement. Poten-  in most of the countries of the Europe and
tial gains from global trade have increased  Central Asia (ECA) region in 2002 relative to
with trade liberalization during the 1990s,  2001. However, the region has weathered the
leading to high and rising ratios of trade to  recent global economic downturn relatively
GDP (Chile, Mexico, the Caribbean Commu-   well, largely because of fairly strong domestic
nity and Common Market, and other small    demand throughout most ECA countries, and
economies are examples). Moreovcr, the re-  sustained high oil prices to the benefit of the
gion has been proactive in deepening trade in-  oil-exporting Commonwealth of Independent
tegration, especially with North America,  States (CIS) countries. Whereas almost all ECA
through negotiations to establish a Free Trade  countries are facing a moderation of growth in
Association of the Americas (FTAA) and bilat-  2002, Turkey is expected to post a recovery-
eral agreements (Chile and Central America  a massive swing from the over 7 percent col-
with the United States).                  lapse in GDP that it posted in 2001-which is
At the same time, the region remains more  raising the region's average growth for the
vulnerable than many other developing re-  year. As a consequence, aggregate growth for
gions. First, a high debt overhang from the  the ECA region is projected to expand from an
1980s remains a problem to finance in many  estimated 2.3 percent in 2001 to 3.6 percent in
countries. In the 1990s, some countries con-  2002. Growth in the ECA region (excluding
tinued to rely on significant debt financing,  Turkey) is forecast to decelerate to 2.3 percent
particularly in the public sector. Public debt-  in 2002 from 2.9 percent in 2001.
to-GDP ratios rose in some countries and the  The Central and Eastern European (CEE)
maturity of that debt shortened in duration,  countries, in particular, have been affected by
increasing their vulnerability to shlifts in in-  weakening demand from Western Europe-
vestor sentiment as they question debt sustain-  their main export market-as well as by ex-
ability. LAC countries may have to learn to  change rate appreciation in many of the sub-
live with less debt in the future, adjusting pub-  region's economies. As a result, export volume
lic expenditures as required. Countries need to  growth has slowed significantly, although it
create fiscal space during good times (boom  still remains at impressive levels in a number
years) to be able to conduct countercyclicaL  of countries (such as the Czech Republic,
policies in future downswings in economic  Hungary, and Poland), as they have increased
activity. Second, many countries, especially the  market penetration (see box 1.2 in chapter 1).
low-income coffee producers, also need to  Thus, the consequent drag on GDP is not as
further diversify their export base to reduce  high as it might have been. Furthermore, eas-
vulnerability to large swings in commodity  ing iniport volumes-which largely reflect
162



R E G I O N A L E C O N OM N I C P R 0 S P E C I S
Figure A1.5 German imports and CEECs exports in dollars, 1999-2002
(3-month moving average, yly percentage change)
25
20
15
10               /-/.\\                                         CEEC-exports
5
-10
-15                                                     i   prs
German rnports'
-20
Jun Aug Oct Dec Feb Apr Jun Aug Oct Dec Feb Apr Jun Aug Oct Dec Feb Apr Jun Aug Ocl
1999 1999 1999 1999 2000 2000 2000 2000 2000 2000 2001 2001 2001 2001 2001 2001 2002 2002 2002 2002 2002
Sozurce inrernational ANionerar) Fund, Interiearional F inanoccal StItistics
high import intensity of exports in all of the  Turkey remains a key question mark. Indica-
CEEs-are helping to reduce the impact of the  tors for early 2002 polit to a recovery, facili-
slowdown in export volumes. Strong domestic   tated in part by the new reform   program,
demand, witnessed in most of the CEE coun-    lower interest rates, and improved confidence
tries, is partially offsetting the loss in impetus  relative to 2001. However, uncertainty linkedl
to growth from the slowdown in the external   to the continued implementation of the cur-
sector. Domestic demand has been generally    rent economic program   of the new govern-
supported by lower inflation, easing iilterest  nient, as well as heightened political instability
rates, and expanding private constimption. In  in the Middle East, could contain the building
some countries, fiscal policy has become more  growth momentuim, as interest rates have
expansionary, notably in the Czech Republic,   begun to rise and market sCentimilent is becorn-
Hungary, and the Slovak Republic. Poland is   ing more cautious. Lowering interest rates re-
the main exception to this picture of relatively  mains important to achieving a sustainabie
strong domestic markets, with tight monetary   public debt.
conditions, weak wage growth and record          Relatively firm  oil prices-fueling fiscal
unemployment rates translatinig into anemic   linkages in oil producers and bolsterlitg intra-
domestic detmanid. While inflationiary pres-  CIS trade-have sustained the recent recovery
sores have generally been subsiding in the CEE  in the CIS subregion, partially buffering it
subregion, they reainai significant (in double  from the deterioration in external conditionis.
digits) in Romania, Turkey, and the Federal   Nevertheless, given the moderation in external
Rlepublic of Yugoslavia. Many of the CEE      demand, exchange rare appreciation in a
counltries conltinue to rin relatively high cutr-  number of counti-es (such as the Russian Fed-
rent account deficits of above 4 percent, al-  eration), and some easing in oil prices in the
though sufficient external financing has been  first half of 2002, growth is moderating
sustained. Continued significant FDI inflows,  throughout the CIS subregion. Hydrocarbon
in particular, are helplig to finance the current  exporters-Azerbaijan, Kazakhstan, Russia,
account deficits, in addition to generally suip-  and Turkmenistan-in particular have experi-
porting domestic demand and regional growtth.  enced robust groNvthi, although growth has
163



G LO B AL EC O NO MI C PRO SP E CT S 2 0 0 3
slowed significantly from 2001. Increased in-  Russia's current account surplus has nar-
vestment in production capacity in the energy  rowed, although it remains very large.
sector has also supported economic activity in
these countries. Among the energy-importing  Near-term outlook
CIS countries, other factors have contributed  Growth in the CEE countries is expected to
to the continued strong growth, including rel-  begin increasing again in 2003, assuming a
atively good harvests in the Kyrgyz Republic  firming of European Union (EU) import de-
(combined with increased gold production),  mand in 2003, which is projected to strengthen
Tajikistan (combined with increased alu-   significantly in 2004. For Turkey, assuming rel-
minum production), and Ukraine. Not only   ative political stability and the continued pur-
because Russia represents the largest weight in  suit of the current reform program by the new
the subregion, but also because it represents a  government, the recovery is expected to be sus-
major export market for many of the CIS   tained in 2003. Growth in the CEE subregion
countries (in addition to some of the CEEs),  is forecast to accelerate from 2.3 percent in
continued strong growth in Russia has also  2002 to 3.11 percent and 4.3 percent in 2003
sustained growth in the subregion. As in many  and 2004, respectively. As the EU accession
CEE countries, domestic demand has been un-  process moves forward, it is expected that the
derpinned in the CIS countries by falling infla-  first round of new members in particular will
tion rates and lower interest rates. While in-  continue to receive significant FDI inflows (in
flationary pressures have eased in all the CIS  addition to EU transfers), which will remain an
countries, they remain a source of concern in  important source of external finance and un-
some cases, particularly in Belarus and Uzbek-  derpin long-term growth.' While a number of
istan. Inflation rates in Russia and Talikistan  important hurdles remain-linked in particu-
are also expected to post in the double digits  lar to the negotiations on the existing common
in 2002, but at more moderate levels. While  agricultural policy (CAP), on the EU's budget,
energy-related receipts boosted revenues, fis-  and on transfers to new member countries-
cal positions remained relatively prudent in  the process is still on schedule.2 Growth is ex-
the energy-exporting CIS countries, although  pected to slow in the CIS subregion (through
they became somewhat more expansionary     fiscal and trade linkages) through 2004, as-
in Azerbaijan and Kazakhstan. To provide   suming significant declines in oil prices in both
for budgetary smoothing, Azerbaijan and    2003 and 2004 (from a projected $25 per bar-
Kazakhstan have created national funds for  rel in 2002 to a forecast of $23 per barrel and
windfall oil rents. With the decline in energy  $20 per barrel in 2003 and 2004, respectively).
prices posted in 2001 (in both real and nomi-  CIS GDP is forecast to decline from a pro-
nal terms) from the spike posted in 2000, the  jected 4.4 percent in 2002 to 3.5 percent and
fiscal surplus in Russia narrowed, although re-  to 3 percent in 2003 and 2004, respectively.
cent firming in oil prices should contribute to  While Russia's current account is expected to
some stabilization of budget receipts for 2002  remain in surplus over the near term, it is
over 2001. In other CIS countries, fiscal policy  forecast to narrow markedly as energy prices
has been more expansionary, in general, gen-  soften. In 2003 and 2004, growth is expected
erating widening deficits (for example, in  to average 3.4 percent and 3.6 percent, respec-
Ukraine). Throughout the CIS subregion, cur-  tively, for the ECA region as a whole.
rent account surpluses (for instance, in Russia
and Ukraine) are narrowing and deficits (in  Long-term prospects
most of the remaining CIS economies) are   In the CEE, during the second decade of tran-
widening somewhat, because of exchange rate  sition, a number of factors are expected to
appreciation, rising imports, and, until more  contribute to stronger growth than posted
recently, moderating oil export earnings.  during the previous decade. These include
164



R E G I O N A L E C O N OM l I C P R O S P E C r s
Table A1.4 Europe and Central Asia forecast summary
(percent per year)
.                                   Bfasehine forecast
Growth-, . i.                   i-**i-ii.     2'""'    2i.,i      2       22
Real Gi ,.-. .rh                  -I -         .        :         1 ;-
Private oi.iuiiptL-) pcs iapin.    H   -      2t 1     1 >      4 o       4 u      . o       J 37
GDP per (apita                     -I          6 4      2 2      3 5       3 3      3 5       3 5
Population                          -        0 1      0.1       01       01       01        01
Gross domestic investment/GDPa '    *.        22 1     21.9     20 6      20 8     212       28 6
lInflationb                      .     -       -4       7 0      3 2       5 8      5 7
Central gvt budget halance/GD'     -4 4      _5 4      -6 5     -6 2     -6 0      -5 5
Export market growthc              ii' -      12.9      6.7      2 1       6 4      8 1
Export voluimed                     ' 1       10.9      8 8      6 4       8 2      7 5
lerms of trade/GDI"                ,,,       -1.6       3.0     -2 5     -1 3       0 1
Currenr account/GDP                -         -4 9      -18      -2 4     -2 4      -2 3
Menmoraidim itemns
GDP growth transition cotintriesf  -2 .        6 4      4.6      3 5       3.3      3 5
Ceiitral and Eastern Europef        r        3.8      2 9       2 3      3 1      4 3
Cis                              -44         84       59        44       35       30
Not available
a Fixed investment, measured iii real terms
b Local currenicy GDP deflator, median
c Weighted average growth of inmport demanid in e'spoir markets
d Goods and inonfactor services
e Change in terms of trade, measured as a proportion io GDP (percentage)
f Excluding Turkey
Souirce World Bank baseline forecast, November 2002
higher investment rates and ongoing restruc-         achieved higher growth rates until diversifica-
turing of the capital base. Continued improve-       tion from energy becomes much more broadly
ments in the policy environment, including           based. Given the degree of energy dependence
greater macroeconomic stability, are expected        in many of the CIS economies, particularly
to underpin the projected higher growth rates.       Russia, the projected softening of oil prices-
The EU accession process and coming mem-             to an average nominal price of about $18
bership will continue to act as an anchor for        to $19 per barrel for the 2005-1O     period, in
structural reforms and will help attract signif-     the underlying forecast-implies a ratcheting
icant inflows of FDI. While structural reforms       down of the subregion's growth from       recent
are being pursued in many CIS countries, in          high rates.
general, implementation is not as advanced
or as widespread as in the CEE subregion's           Risks
economies, and in some cases there is signif-        Weaker than anticipated recovery in the EU
icant resistance to structural reforms. This         area would reduce external demand for re-
implies lower long-run growth in comparison.         gional economies, particularly for the CEE
The recent boom     in hydrocarbon rents has         countries, and    thus   translate  into  lower
provided an impetus to growth, facilitated the       growth. In many CEE countries high current
introduction of a number of reforms in many          account deficits could become an important
of the oil-exporting countries, and contributed      downside risk for international credit, if FDI
to an increase in investment outlays (particu-       inflows suddenly dry up (which a pronounced
larly in the energy sector). However, given          delay  in the   EU   accession  process could
the volatility of energy market prices, these       trigger). For CIS     hydrocarbon     exporters,
ecoilomies will not be able to sustain recently      which are highly dependent on energy prices,
165



G LO B AL EC ONO MI C PROS PE CT S 2 0 0 3
a sharper decline in oil prices than forecast  strong. Meanwhile, countries where severe
poses a significant downside risk. In Russia, as  drought put a damper on agricultural pro-
a consequence, a more rapid decline in growth  duction fared worse and, as usual, the worst
than projected would result in lower external  outturns were to be found in countries experi-
demand for a number of the other economies  encing civil and political strife. But slowdown
of the CIS and of some CEEs (that is, the  was widespread throughout the region.
Baltic states). Greater political uncertainty or  The subdued economic performance in
a reversion from  the reform  program  in  2002 is attributable to the region's export de-
Turkey could undermine its fledgling recovery  pendence on European markets, where growth
and result in much lower than anticipated  was weak. As a result, exports remained basi-
growth-and affect some of its trade partners  cally stagnant in real terms. On the plus side,
(including Bulgaria). Some of the poorest CIS  there was a turnaround in non-oil commod-
countries (Georgia, the Kyrgyz Republic, and  ity prices, albeit from historically depressed
Talikistan) have relatively high external debt  levels. After reaching a low In October 2001,
(up to 200 percent of exports), placing them at  non-energy commodity prices rebounded
risk of default if growth does not materialize  sharply, gaining 26 percent in export weighted
as anticipated and external assistance Is not  terms by August 2002. The main contributor
forthcoming. Greater political instability and  was cocoa, which was up 80 percent, but other
heightened tensions in the Middle East could  commodities gained as well-robusta coffee
prove destabilizing to economic (and political)  up by 20 percent, cotton up by 33 percent,
positions in the neighboring ECA countries,  copper by 7 percent, and gold by 10 percent.
particularly in Turkey and in some of the  Because the upward momentum began late in
southern-tier CIS countries. Aside from poten-  the year, on an annual basis many commodity
tially undermining tourism  revenues, there  prices exhibited declines in 2002, and most re-
could be trade disruptions and a general rise in  main well below peak levels of the mid-1 990s.
the perception of risks in the region leading to  Nevertheless, modest real price gains over the
higher interest rates.                     medium term will deliver a measure of relief
to external balances. Oil prices, too, remained
stronger than expected, an outcome that,
Sub-Saharan Africa                         though unwelcome to non-oil exporters, was
a positive benefit to the region as a whole.
Recent developments                          The disappointing results for tourism,
S ub-Saharan Africa was not immune to      which accounts for I 1 percent of regional ex-
the chilling effects of the global recession.  port receipts, reflected not only the weakness
The slowdown may have been less pro-   of Europe's economy, but the aftermath of
nounced there than elsewhere, but the collapse  September 11. Some important destinations-
in world trade and steep declines in commod-  such as Mauritius, which benefits from its rep-
ity prices had a depressing effect on economic  utation as a safe destination-saw an increase
performance. Services exports, prtimarily  in the flow of arrivals in 2002. Most other
tourism, were further affected by the Septem-  countries, however, only partially recovered
ber 11 terrorist attacks. In real terms, GDP  from the impact of September 11. According
growth slowed from 3.2 percent in 2000 to  to estimates by the World Travel and Tourism
2.9 percent in 2001 and 2.5 percent in 2002.  Council (WTTC),3 Sub-Saharan travel and
With the exception of Nigeria, oil exporters  tourism exports slowed to just 1.5 percent
generally outpacecl the region, especially those  growth in 2001 before recovering to 4.3 per-
where production was increasing-Angola,    cent in 2002. Both years represent growth sig-
Equatorial Guinea, Sudan-as oil prices,    nificanitly below potential. The WTTC esti-
though down from 2001, remained relatively  mates September 11 to have cost 3.2 percent
166



R E G I O N A L. E C O N O M I C P R O S P E C T S
of exports and 1.3 million jobs over the period
2001-02.                                                  Figure A1.6 Real GDP growth of
In the domestic sphere, agricultural pro-             Sub-Saharan Africa oil and non-oil
duction was disrupted by drought and, in the             exporters
case of Zimbabwe, political disturbances. As              (percent)
a result, nearly 30 million persons were left             5
in need of emergency food aid. Approximately                                       Ol4
half of them   were in southern Africa, which                              Non oil
experienced a second successive year of poor              3
harvests. Malawi, Zambia, and         Zimbabwe
were most affected, and because agriculture               2
constitutes a large share of their economies,             1 *             0         2         2
incomes and consumption spending were de-
pressed. But Ethiopia suffered the worst, with            o           ,
drought putting at risk an estimated 15 mil-                   2001      2002       2003      2004
lion persons. However, the effects were quite            Souirce World Bank staffestimates
localized. South Africa experienced a bumper
maize harvest, which gave a strong boost to
domestic spending.                                    expenditure side, domestic demand grew at a
South Africa sustained a recovery through          2.3 percent annual pace in the first half, re-
the  first half of 2002, with       broad-based        flecting buoyant fixed investment, strong pri-
strength in agriculture, as well as in export-        vate consumption, and a moderately expan-
oriented  mining    and  manufacturing, which         sionary fiscal stance, though weak inventory
benefited from   the rand's weakness. On the          accumulation slowed the pace. In the external
Table A1.5 Sub-Saharan Africa forecast summary
(percenzt per year)
Baseltne forecast
Gros.th rates/ratios                   1991-2(00    2000     2001    2002     2003     2004    2005-15
Real GDIP growth                           2.2       3 2      2 9      2.5     3 2      3 8      3 7
Private consumption per capita           -0 6       -1 4      0.7      0 3     0 8      1 3      1 2
GDP per capita                           -0 4        0.7      05       01      0 9      15       1.5
Population2 2                                      25       2.4      2 4      2 3     2 3      2 2
Gross domestic investment/GD)P'           16 9      17 9     18.7     18 9    18 6     18 2     21 1
Inflatiollb                                9 8       6.3      5.4      4 3     39       4 2
Cenitral gvt budger balance/GDP            3 0      -0 6     -0 3    -0 4     -0 5     -0 3
Export market growth'                     14 1       10.8     0.1      2.4     71       7 8
Export voluimed                       -- -           4 3      28       1 1     53       5 8
Termiis of trade/GDPe                      00        2 1     -1 1    -1 5      0 2     -0 8
Curretnt account/GDP                     -2 1       -2 3    -2 2     -3 0     -1 .8    - 1 2
Memoranzdumrz atelt1s
GDP growth SSA excluding South Africa        -       3 3      3 6     2 7      3 7      4 2
Oil exporters                                      453      4 4      2.0     3 6      3 8
CFA counrries                            2.6       2.3      3.2      2 9     3.3      3 8
Not available SSA is Sub-Saharan Africa CPA is Comiimuniautie Finaniciere Africaine
a Fixed investment, ineassired in real ternis
b Local currency GDP deflator, imedian
c WXeighted average growth of import demand in export markets
d Goods and noiifactor services
e Chalige in terms of trade, measured as a proportion to GDP (percentage)
Souirce World Bank baseline forecast, Novemiber 2002
167



G LO BA L EC O NO MI C PROS PE CT S 2 0 0 3
accounts, tourism has been slow to recover  nal performance. Overall, the forecast antici-
from the effects of September 11, but the  pates real growth rising to 3.2 percent in 2003
weaker rand boosted net visibles trade, and  and 3.8 percent in 2004, around 1.1 percent in
the current account balance moved back into a  per capita terms. Faster growth in Europe will
small surplus. The main concern for the South  boost export volumes, and price trends will
African economy was an uptick in inflation-  remain broadly favorable as supply-demand
not surprising after a nearly 40 percent depre-  balances for most commodities tighten with
ciation of the rand in 2001, but nevertheless  the recovery in the world economy. Cocoa, al-
putting upward pressure on interest rates.  ready at a 15-year high, is the main exception.
Robust wage gains were only partially offset  Oil prices are expected to ease, though they
by higher productivity, resulting in a substan-  remain significantly above late-1990s lows.
tial rise in unit labor costs. Nevertheless, mdi-  While large, real price gains for commodities
cations are that monetary restraint will prevail,  important to Sub-Saharan Africa are unlikely
and the outlook is for inflation to ease. The  to be sustained in the medium term, the wide-
rand stabilized at an average of 10.9 against  spread introduction of structural reforms and
the dollar in the first three quarters of the year,  market liberalization are expected to raise
after weakening to above 12 in late 2001.  export competitiveness, and the region should
In Nigeria, adherence to reduced Organi-  remain a significant commodity supplier for
zation of Petroleum  Exporting Countries  the foreseeable future. On balance, with im-
(OPEC) quotas more than offset the impact of  port price inflation remaining low, non-oil
stronger-than-expected oil prices, while bud-  exporters' terms of trade are forecasted to
get gridlock constrained government spend-  strengthen marginally in 2003-04.
ing, resulting in a real GDP decline estimated  Along with a modest improvement in ex-
at 0.6 percent from 2001. Unfortunately, the  ternal sectors, the forecast assumes a stronger
strength of oil revenues may have also relieved  domestic performance. The expectation is for
pressure to implement urgently needed struc-  a return to more normal weather conditions,
tural reforms, as evidenced by the discontinu-  which will underpin a recovery of agriculture
ation of the informal monitoring arrangement  and consumption spending, while faster
with the IME Politically, the situation remains  growth stimulates a cyclical upturn in invest-
tense. Local elections have been postponed  ment. Overall, domestic demand will remain
twice, and national elections are now sched-  the primary source of GDP growth, as rising
uled for April 2003. President Obasanjo has  import demand-constrained mainly by the
come under strong pressure by the National  availability of financing-holds net exports in
Assembly on budgetary issues. If oil prices de-  check. Indeed, current account balances are
crease, as expected in the near term, Nigeria's  expected to narrow marginally as foreign di-
problems will mount, though there are indica-  rect investment eases back from recent levels
tions that these problems may be offset by  with a slowdown in privatization offerings
a significant medium-term expansion of the  such as Air Madagascar, though at the same
energy sector, consistent with expected OPEC  time more favorable economic conditions
policies. The United States is particularly  overseas should have a positive impact on
interested in expanding oil purchases from  remittances.
Nigeria in an effort to diversify sources of  For the region's oil producers, weaker ex-
supply to decrease dependence on the Middle  port prices will be offset by higher production.
East.                                      Major new offshore developments are sched-
uled to come onstream in the Gulf of Guinea
Near-term outlook                          (Angola, Equatorial Guinea, and Nigeria)
The forecast calls for growth to accelerate in  over the next few years, and despite the fall
2003-04 on the strength of a pickup in exter-  in prices oil sectors should remain highly
168



REG IO NA L EC ONO MI C P R O S P E C T S
profitable. As in the past, however, more  ment, will promote diversification, encourage
general spillovers to non-oil sectors will be  higher savings and investment, and stimulate
relatively muted. As a result, growth of oil  productivity growth. Despite a dismal overall
producers will improve from  2 percent in  record, the region boasts a number of exem-
2002 to 3.6 percent in 2003 and 3.8 percent in  plary performers (Botswana and Mauritius)
2004. Falling terms of trade will widen cur-  and cases in which policy reform has pro-
rent account deficits, though oil prices are  duced  dramatic turnarounds  (Ethiopia,
expected to remain comparatively strong,   Mozambique, Tanzania, and Uganda). At the
above $20 per barrel, containing the pressure  same time, a panoply of deep-rooted problems
on external performance and fiscal accounts.  will continue to plague economic perfor-
mance. There are no quick solutions to low
Long-term prospects                        levels of human and physical capital, poor
The forecast anticipates per capita growth av-  infrastructure, HIV/AIDS, civil strife and neg-
eraging 1.5 percent over the 2005-15 period,  ative perceptions of international investors,
near the levels achieved in the mid-1990s. The  and excessive export specialization will con-
outlook is optimistic-such a result would sig-  tinue to expose external sectors to high price
nify a reversal of the region's long-term histor-  volatility. Nevertheless, if the new realism that
ical decline-but even so growth will fall short  appears to be taking hold on the continent
of what is needed to reduce poverty and    proves to be more than just rhetoric, there is
achieve the Millennium Development Goals.  adequate reason to believe the moderate im-
Sub-Saharan Africa will continue to lag behind  provement in overall performance that the
other developing regions, and by 2015 the  forecast anticipates can be achieved.
number of persons living on less that $1 per
day is forecasted to rise by nearly 30 percent,
from 315 million to 404 million.           Middle East and North Africa
The outlook is predicated on a continuation
of broad trends toward better standards of  Recent developments
governance and economic policies. The great       espite a continuation of high oil
majority of countries in the region have begun    prices, growth in the Middle East and
preparing Poverty Reduction Strategy Papers,  D   North Africa (MENA) region slowed
and nine are in different stages of implemen-  in 2002 to 2.5 percent, down from 3.2 percent
tation of the strategies. Early results seem  in 2001 as the events of September 11, 2001,
promising. At the same time, NEPAD (the New  continue to reverberate throughout the region.
Partnership for Africa's Development) and    For the oil-exporting countries, growth re-
other regional initiatives are enhancing the  mained above 2 percent. The larger increases
credibility of governments and strengthering  in growth that might have been expected
intraregional cooperation. There has been  from both high oil prices and increased public
encouraging progress as well toward resolving  expenditures were offset by a slowdown in
intractable conflicts in central Africa, even  production and exports, a result of tightened
though events in C6te d'lvoire, Madagascar,  OPEC quotas put into place in 2001 to sup-
and Zimbabwe underscore the fragility of   port higher oil prices. To counter these effects,
the region's political processes. Improved pol-  some oil exporters have implemented expendi-
icy environments should stimulate faster pro-  ture programs financed with increased oil rev-
ductivity growth and diversification from  enues. In the absence of a strong private in-
agriculture, and reduce export dependence on  vestment response to the reform program it
primary commodities.                       had implemented, Algeria, for example, put a
These internal developments, coupled with  four-year economic recovery program in place
modest improvements in the external environ-  worth about 10 percent of 2001 GDP, aimed at
169



G LO B AL EC ONO MI C PRO SP EC TS 2 0 0 3
positive agricultural growth of 2001 in
Figure A1.7 Oil prices and current       Morocco slowed down somewhat in 2002
accounts                                 because of the less favorable weather condi-
(dollars/barrel)       (billions of dollars)  tions. To prevent a potentially steep increase
29 -                           - 40     in the current account deficit, fiscal and mon-
27- -   X              ,O______  - 3_5   etary policies were tightened in Tunisia; this
25 -                    Oi -prices       action contained domestic demand pressures
23                             -                                         pessre
21               -       ~_  \  - 20     and reduced pressures on external balance
19*- _ _- 15                             but exacerbated the impact of the slowdown
17 -     Current account         1 0     stemming from the external shocks. Though
15-                 I            o       tourism  has begun to recover gradually in
2000  2001  2002  2003  2004        Egypt, the recovery of the private sector has
Source. World Bank staff estimates       been hampered by unresolved problems with
the exchange rate regime, which have their
roots in the late 1990s.
supporting growth in domestic demand while  Short-term prospects
enhancing social and economic infrastructure.  Growth prospects for MENA are clearly con-
Diversified exporters faced worsening con-  Gent pospether Military ction-
ditions in 2002, with GDP growth falling to  tingent upon  whether military actions are
2.2 percent, a decline of 2 percent from 2001  taken in the region. Assuming that there is no
External factors leading to this decline include  conflict over the next year (and thus that there
the deterioration in export market growth for  is a gradual resumption i confidence in the
Egypt, Morocco, and Tunisia, as well as sharp  region), the region's growth is forecast at
declines In the tourism sector in North Africa  3.7 percent for 2003-04. A recovery would be
and the several countries in the Levant, fol-  expected for both oil-exporting countries and
lowing the events of September 1l, 200n In  diversified exporters. The public expenditure
Egypt, the number of tourists fell 41 percent  programs being implemented by oil-exporting
Egyp, te nmbe oftourstsfel 41perent countries over the next several years to im-
year-on-year in the last quarter of 2001, and  cove  overe next            al   support
even by the end of the first half of 2002,  prove minfrastructure, agriculture, and support
tourism had not yet returned to precrisis lev-  reforms in social sectors would help to icrease
els. In Morocco, tourist arrivals were down by  the public sectors contribution to growth.
over 20 percent in the first half of 2002. For  Growth is expected to average 3.6 percent
some countries in the region, however, the
declines in tourism from abroad have been
partially offset by MENA nationals diverting  Figure A1.8 Export volume growth
their tourism to within the region. Tourist  (percent)                    (percent)
receipts may likely show an even larger decline
than tourist numbers, as many firms in the    8 -  \   Export market growth (right axis) - 14
7 .  _ _ _ _  _ _  _  _ _   _ _  -12
tourist sectors have discounted heavily. Ex-  6                              - 10
ports to the EU have also suffered, in particu-  5Exportgrowth5 - B
lar Moroccan textiles and electronics exports  3-                            -
to the EU. Jordan has been somewhat insu-     2 -                            -2
lated from this situation, as its export demand  1 -    \/                   -0
from India and the United States continued to  0 - _   _     _    _          - -2
grow quite strongly in 2002.                     2000  2001  2002  2003  2004
Internal factors have also contributed to  Souirce World Bank staff estimates
a decline in growth in several countries. The
170



R E G I O    N A l  E C O N O M I C  P R O S P E C F S
Table A1.6 Middle East and North Africa forecast summary
(percenit per year)
Blaselinie f-rccast
Gr,n,ah rates/ratios             1991-2000     2000.    2001       2002      2003      2004     2005-15
RealCGDPgrowth                      32          42        32        25        35        37         32
Private consumption per capita      0 3         1 4       1 1       0 6       0 8       0 8        1 1
GlP per capita                      1.0         2 2       1 3       0 6       1 5       17         14
l'opulation                       2 2         2 0       1 9       1 9       1 9       19         1 8
Gross domestic investnient/GDPI    51-2        21 6      22 0  i   22 5      22 7      22 9       24 3
inflantuin                         -6 0         4 8       4 2       4 1       4 0       4 0
Central gvt buidget balance/GDP    -1.1       -1 4       - 1       -2 1      -2 4      -2 4
Export market growvth,             10 1        13.3     -I 0        3 0       7.8       8 2
Expoit voluned                      5 0         7 6       2.4       2 7       5 7       58
Terms of trade/GDP'                 0  -        8 4      -0 9      -0 9      -II       -14
Current accoUt1/GD)P               -2. t        6.8       5.2       3 3       2 3       0 9
Memtorandumi itenis
GDP growvth Oil exporters           24          3 6       2 4       2 4       3 7       3 6
Diversified exporters             4 0         3.7       4 3       2 2       2.7       3 6
Not available
a Fised iivestnent, measuired in real terimis
h Local currenicy GDI' detlator, median
c Weighted averagc giosth of import demiland in export miarkets
d Goods anid iioni-actor services
e Chanige in terms of trade, measured as a propoitioni to GDP (p rcentage)
Soiiice World Bank baseline forecast. Novemilber 2002
for the oil-exporting countries, supportecd by        would    be expected    to  rise considerably in
likely increases in OPEC oil quotas in 2003 to        2003-04, anid     the tourism    industry  would
meet growing demand in a period of tight              also be expected to recover from    its slidc since
crude inventories. Additionally, Algeria's crude       September 11, 2001, as the external environ-
production capacity is set to increase, and its        nient gradually improves through 2004. Some
gas exports are expected to experience growth          of the diversified exporters, sLich as Morocco
unfettered by quotas, iclicating higher growth         and Tunisia, wvill not be able to rely on gov-
in the short term. However, much           of the     ernnent stimulus for growthi because of the
growth increase in Algeria in the medium term          need to pursue fiscal consolidation and keep
would come from the large public sector stirn-         large inflexible expenditures, especially the
ulus currently under way. This would revive           government wage bill, under control. Over-
domestic demand for the duration of the ex-           coming    the vulnerabilities unveiled     by  the
penditure program. In Saudi Arabia, the stock         external shocks would call for a faster pace of
market is performing well, and credit to the          structural reforms to improve prospects for
private sector is rising, auguring favorably for      private sector investment. Morocco must rely
2003-04. The Islamic Republic of Iran will see        on   privatization   receipts to   lower budget
the revival of the agricultural sector after sev-      deficits, but this finanicing option will dimiii-
eral years of drought, with the expansion in          ish over the medium     term  as the better candi-
gas exports and the continuing relaxation of           dates for privatization   are sold off and the
import restrictions buttressing growth in the         revenues from the former parastatals decrease,
non-oil sector.                                       implying higher deficits in the future unless
Growth in diversified exports would be ex-         expenditure reforms occur. This will require,
pected to increase to 2.7 percent in 2003 and         among other thmigs, public sector reforms,
3.6 percent in 2004. Export market growth             in particular reforms related to the wage bill,
171



G LOB AL EC ONO MI C PRO SP EC TS 2 O 0 3
that underlie high expenditures. Weaknesses  Jordan is reaping the benefits of a more open
will remain in the private sectors of several  trade regime, and many of the North African
countries. Jordan is weathering the downturn  countries are pressing ahead with more liberal
more successfully than other countries be-  trade relations. However, budget deficits in
cause of the high demand for its exports in  these countries could lead to problems, partic-
India and the United States, but weak con-  ularly in those countries that rely heavily on
sumer demand and subdued rates of lending  privatization revenues to finance increased ex-
to the private sector indicate that domestic  penditures. The temporary nature of receipts
weakness will continue. Egyptian private sec-  from privatization makes reforms in public
tor activity is weak, reflecting credit condi-  expenditures (and the public sector policies
tions and tight monetary policy (although  that underlie them) and taxation of para-
the recent 100 basis point fall in the discount  mount importance in the future. Furthermore,
rate to 10 percent will help somewhat); falling  many of the Mediterranean countries are cur-
bank earnings in 2002 indicate that private  rently facing sluggish growth or stagnation in
sector investment will remain sluggish in the  the private sector, with low levels of growth in
short term. In the current policy context, the  private investment. This indicates the need for
government will continue expansionary fiscal  strengthening the investment climate and re-
policy despite already high deficits and will  moving bottlenecks in access to finance and
put off the implementation of reforms that  backbone productive inputs that hamper pri-
might have the potential for high social costs.  vate sector investment. In Egypt, long-term
Recent political events obviously cast a  prospects appear weak. The policy reform
shadow over prospects in the region. The con-  agenda has slowed, mirroring the slowdown
tinuing uncertainty stemming from the poten-  in the domestic economy. The current policy
tial actions against Iraq would significantly  mix includes an expansionary fiscal policy to
affect the gradual recovery in the short-term  counter the tight monetary policy necessary
forecast. Oil exporters would benefit from  to support the exchange rate.
increased quotas and higher oil prices initially,  In the oil-exporting countries, Iran has uni-
but these impacts would probably be short  fied its exchange rate, allowed the formation
lived. Diversified exporters would suffer  of private banks, reaffirmed its commitment
more, particularly from  declines in the  to privatization, and is pushing ahead with
tourism industry. Even the expectations sur-  fiscal reforms. Algeria is considering the
rounding military action within the region  deregulation of the power industry and open-
could significantly affect growth in the short  ing the sector to private investment, and has
term. Moreover, the effects of military action  announced the privatization of public compa-
on confidence in already fragile global capital  nies. In Saudi Arabia, customs tariffs have
markets may lead to increased spreads and a  been reduced to pave the way for a customs
flight to quality, particularly from countries in  union with the Gulf Cooperation Council in
the region close to the field of war.      2003, negotiations are under way with the
WTO, and new legislation is being prepared
Long-term prospects                       to increase competition in domestic capital,
Even if relatively stronger growth perfor-  labor, and insurance markets. However, many
mance is managed in the short term, growth in  of these changes are occurring very gradually,
the long term is expected to average just over  and some are being greatly delayed. Algeria
3.2 percent.                               has not pushed ahead with its announced pri-
The policy environment affecting long-term  vatizations, and reform in the power sector is
growth is gradually improving in many coun-  very slow. Many of the oil-exporting countries
tries in the region, albeit at a gradual pace.  in the region still have large and inefficient
172



R E G I O N A L E C o N O M I c P R O S P E C I S
public sectors, and low rates of private invest-  2005-06, when Caspian oil production is cx-
ment in the non-hydrocarbon      sectors. Re-     pected to comc onstream.
forms in these sectors leading to efficiency
gains and higher potential growth rates will be   Notes
required.                                             I At both the June 2002 summilt in Seville and
In the long tcrm, the region has to continue    October 2002 sumiiilt in Brussels, the Europeani
to address several obstacles. The region relies   Community (EC) confirmed that eight of the ECA EU
very heavily on a narrow    range of external     accession candidate countries (the Czech Republic,
revenue sources, particularly oil, remittances,    Estonia, Huligary, Latvia, Lithuanila, Poland, the
and tourism, and this reliaice introduces the     Slovak Republic, and Slovenia) are on track ro con-
potential for vulnerability * i export earings.   clude accession negotiations at the end of 2002 and
Although   several countries have adjusted        to sigin formal accession treaties in 2003. The official
n i ele , f d  target is for enlargement to happen in time for the
nominal exchange rates in recent years, fixed     mid-2004 European Parliament electionis Bulgaiia and
exchange rate regimes in several countries        Romania are at an earlier stage in the accession pro-
may adversely affect export competitiveness       cess and are expected to accede somewliat later Turke)
and offset gains made from trade and customs       is also an accession candidate, although it has yet to
reforms   Receipts from   tourism  and remit-      begin formal negotiatiolns.
tances are vulnerable to the sluggish income         2. Following the forthcoming EC Copenhagen sum-
mit in mid-December 2002, the existing 15 EU member
growth il    source countries and    recuarring    countries' national parliaments will vote to approve or
political conflicts and potential military action  reject enlargement.
in the region. Oil revenue windfalls are usu-        3 WTTC. 2002 The Impact of Travel and Tourism
ally temporary, and real long-term   oil prices    on Jobs and   the Economy-2002     http.//www
are expected   to  decline, particularly  after    wttc org
173






Appendix 2
Global Commodity Price Prospects
rude oil prices increased about 3 percent  agricultural prices will be more strongly influ-
G     min 2002 as a result of tight supplies and  enced by supply increases and by recent
Middle East tensions. Non-oil prices in-  weather disturbances such as El Nifio and
creased about 5 percent, led by a 9 percent in-  droughts. The index of nominal non-oil
crcase in agricultural commodities, which more  commodity prices is projected to increase by
than offset a 4 percent decline in metals and  5.8 percent in 2003 and by nearly 8 percent by
minerals (figure A2.1). Uncertainty about the  2005 in real terms. (Specific forecasts for com-
strength of the global economic recovery con-  modity price and price indexes for 2002, 2003,
tributed to the decline in metals and mineral  2005, 2010, and 2015 in current and constant
prices, but the effect of uncertainty on agri-  dollars are given in tables A2.13-A2.15 later
cultural prices was offset by lower supplies of  in this appendix.)
selected commodities, such as grains and      Agricultural commodity prices appear to
oilseeds, because of drought. The weakness in  have reached a cyclical low, after declining
the U.S. dollar supported commodity prices.  since mid-1997, and by 2005 nominal prices
Crude oil prices are expected to remain  are expected to increase about 13 percent over
firm in early 2003 because of the potential for  2002 levels. The increases will leave nominal
military action againist Iraq and tight supply  prices of most agricultural commodities well
conditions resulting from production restraint  below 1997 highs. Prices of specific agricul-
on the part of the Organization of Petroleum  tural commodities have declined much more
Exporting Countries (OPEC). Once Middle    than the average decline because of large sup-
East tensions ease, oil prices are expected to  ply increases, weak demand, or both. Some of
decline because non-OPEC oil supplies will in-  those prices are not expected to recover to
crease and Iraqi oil will return to the market.  1997 levels for the foreseeable future. Because
The average price of crude oil is projected  of large supply increases from Vietnam and
to decline from $25 per barrel in 2002 to  Brazil and because of slow growth in demand
$23 per barrel in 2003. By 2005, crude oil prices  despite low prices, robusta coffee prices, for
are prolected to decline to $19 per barrel.  example, have fallen to nominal lows not seen
Non-oil prices are in the early stages of  since the 1 960s. In 2002, cotton prices fell to
price recovery. That recovery is expected to  nominal levels, which were last seen in 1986
last about three years before nominal prices will  and the mid-1970s. Palm oil prices declined
begin to weaken. The strength of the global  by more than half from 1998 to 2002 and
economic recovery will strongly influence the  reached nominal levels last seen in 1986.
timing and strength of the recovery in metals  In real terms,] robusta coffee prices fell
and mineral prices. However, the recovery of  85 percent from 1980 to 2001, and cotton
175



G LO B AL EC ONO MI C PRO SP EC TS 2 0 0 3
metal to sustain price increases that can be at-
Figure A2.1 Commodity price trends       tributed to low stocks and expectations of tight
(index, January 1997 =100)               supplies. Gold prices also rose strongly in
160                                      2002, mainly because of the buyback of pro-
1Crude oil      ducers' hedge positions. However, the decline
Agriculture   T                 in equity markets, weakening of the U.S. dollar,
120                IIIIAA, ?  land nervousness about military activity in the
100                                      Middle East also contributed to the price rise.
80                   A                     \J Crude oil prices began 2002 below $20 per
80                   v_ \ +; 5L barrel because of weak demand, increasing
60                 mti r  \              supplies from non-OPEC producers, and over-
40 _________ _______________          quota production in several OPEC members.
Nevertheless, OPEC production restraint has
20                                       been sufficient to bring prices back to the top of
o,     ,                                OPEC's targeted range of $22 to $28 per barrel.
Jan  Jan.  Jan  Jan  Jan  Jan          Significant OPEC cutbacks, which commenced
1997  1998  1999  2000  2001  2002     in early 2001, started to draw down crude oil
Source World Bank                        stocks during the second half of 2002 and gen-
erally supported higher prices. In addition,
increasing uncertainty about a supply disrup-
prices fell 61 percent from 1980 to 2002. Real  tion from a possible U.S. attack on Iraq helped
palm oil prices declined 60 percent from 1980  push prices higher-to near $30 per barrel.
to 2001. The extreme price declines in agri-  Real commodity prices declined signifi-
cultural commodities resulted from a number  cantly from 1980 to 2001, with the World
of factors, including large increases in pro-  Bank's index of agricultural prices down
ductivity, slow growth in demand caused by  S3 percent, crude oil prices down 46 percent,
falling population growth rates and income  and metals and mineral prices down 35 per-
elasticities, and policies that support prod-  cent (figure A2.2). Such declines in commodity
uction in high-income countries. Several large
commodity exporters experienced deprecia-
tion of exchange rates. That depreciation,     Figure A2.2 Real commodity prices
which was linked to Asia's economic crisis,   (index, I 980        o Iprc)
further contributed to price declines.
Metals and mineral prices fell about 4 per-  140
cent in 2002 as a result of weak demand, high  A12           Agiculiure
stocks, and continued production increases. A                      Metals & minerals
recovery in prices following the October 2001  100
lows stalled in 2002 as the economic recovery  80
slowed and as industrial demand failed to re-
bound as expected. Most metal markets were     60
in surplus, and stocks remained high. A num-
ber of metal producers closed their production  40                 V   - 1
facilities in an attempt to prevent further stock-  20   Crude oil
building and price declines. Despite such ef-
forts, production increased in a number of      °          I
countries. That increase, coupled with an ab-   1970      1980      1990      2000
sence of strong growth in demand, pressured   Souirce World Bank
prices lower. Nickel has been the one major
176



G LO BA L CO MM OD I TY PR ICE P R O S P E C T S
prices, relative to manufactures prices, pose  modities, such as coffee, cotton, and sugar.
real challenges for developing countries that  Real prices will rise an estimated 11 percent
depend on primary commodities for a sub-     from 2002 to 2015. However, the rise in real
stantial share of their export revenues. The de-  prices is a reflection of current low prices
cines are expected to continue in the longer  rather than a change in the long-term trend of
term as productivity increases in commodities  declining prices relative to manufactures.
continue to outpace those in manufactures.      There has been considerable disparity
among commodities: prices of some com-
modities (cocoa) reached multiyear highs in
Agriculture                                  2002, while others (coffee and cotton) have
A  gricultural commodity prices are expected  recently reached new lows or continue to
to increase about 9 percent in 2002 after    decline. The disparity is related partly to the
falling 9 percent in 2001. The increase follows  different levels of carryover stocks, and partly
sharp declines from 1997 to 2001 that reduced  to the effects of weather conditions on supply.
the World Bank's index of annual agricultural  Droughts in Australia, Canada, and the
prices by 38 percent. Prices are expected to in-  United States reduced yields and contributed
crease 13 percent from 2002 to 2005 in nomi-  to increases in grain and oilseed prices.
nal terms. That increase will recover a little  The United States enacted a new farm bill,
more than one-third of the 1997-2001 decline.  which will be in effect from 2002 to 2007. The
The recovery of prices is expected to be mod-  bill raised price supports for many commodi-
est because of weak growth in demand, con-   ties and included some commodities that had
tinued rapid increases in production and     not previously been included under govern-
productivity, and high stocks in some com-   ment programs (see box A2.1). The European
Box A2.1            U.S. Farm Bill
O   n May 13, 2002, the United States enacted a new  new subsidies for dairy farmers as well as for pro-
farm bill, the Farm Security and Rural Invest-  ducers of lentils, chickpeas, peanuts, honey, wool,
ment Act of 2002. The new bill covers a six-year  and mohair. It expands the Conservation Reserve
period, from 2002 to 2007. Low commodity prices  Program, which pays farmers to let environmentally
had led to a series of annual bailouts to supplement  sensitive land stand idle, and it establishes a new
regular subsidy programs under the previous law. The  Conservation Security Program to pay crop farmers
new farm bill essentially extends those temporary  for improved environmental practices.
bailouts through the six-year life of the bill.     Under the 1994 Uruguay Round Agreement on
The key features of the new farm bill are higher  Agriculture, the United States agreed to limit spend-
price supports for major crops, the revival of target  ing on domestic agricultural support programs, which
prices to give more subsidies to producers when  are considered trade distorting, to $19.1 billion per
world prices fall, and a large increase in conservation  year. Since payments are not fixed, but are deter-
programs. The bill continues fixed annual payments  mined by the levels of market prices as well as the lev-
to grain and cotton farmers. It creates a new target  els of support, it is not possible to know whether pay-
price system similar to the one abolished in 1996,  ments under the new farm bill will exceed the agreed
to provide supplemental payments when prices fall  limit. If it appears that this limit will be met or
below certain levels-except that acreage set-asides  exceeded, the U.S. Congress has instructed the U.S.
are no longer necessary for farmers to qualify for  Department of Agriculture (USDA) to take steps to
payments under the new bill. It allows farmers to  reduce payments so as not to exceed this limit.
update planting records that are used in calculating
certain program payments. The bill also establishes  Source: Bank staff
177



G LO B AL EC ONO MI C PRO SP E CT S     2 0 0 3
BL)ox A2.2          E.U. Common Agricultural Policy
Dresident Chirac of France and Chancellor Schroeder  E.U. Agriculture Commissioner Franz Fischier
JIof Germany reached a budget agreement on the  had proposed radical CAP reforms in July 2002 in
Common Agriculture Policy (CAP) in Brussels     the Midterm Review of Agenda 2000. The reforms
October 26-27,2002. The agreement limits CAP    would shift income support away from production
budgets to increases of 1 percent annually from 2006  of surpluses and toward meeting of tough environ-
to 2013 from an estimated budget of 45.6 billion  mental, animal welfare, and food safety standards.
euros in 2006. Total direct and indirect support to  According to the proposal, E.U. farmers would get
E.U. agriculmtre was estimated at 117.9 billion euros  a single decoupled payment based on historical
in 2001 by the OECD; more than half of that support  references-regardless of whether they continue
comes from higher food prices paid by consumers.  production on the same scale. Direct spending on
Participants in the Brussels summit proposed  farmers would be cut by 3 percent per year over
that agricultural support to new E.U. accession  seven years, and the savings would be spent on rural
countries increase from 25 percent of current   development. Aid to large farms would be capped.
member-support levels when those countries join in  This proposal has proved controversial, and several
2004 to 40 percent in 2007 and parity by 2013. The  European states have indicated their opposition to
agreement puts a Iliit on CAP spending increases  changing the current system.
even after the 10 accession countries join in 2004,
a limit that could necessitate CAP reforms as the  Sources: Agra Europe Ltd., London and Europeani Commission
accession countries' support levels increase or that  Information about the Common Agricultural Policy can be
could require shifting of funds from farmers in  found on the European Union Web site.
current member countries.                       hrtp://europa.eu.int/polUagr/index-en htm
Union reached an agreement that limits future  In real terms, coffee prices are currently less
budget increases for the Common Agricultural  than one-third of their 1960 level. The decline
Policy through 2013 (see box A2.2).          reflects mostly the surge in supplies, but the
equally important longer-term problem is
Beverages                                    weak demand. According to the U.S. Depart-
The World Bank's index of beverage prices    ment of Agriculture (USDA), per capita annual
(comprising coffee, cocoa, and tea prices)   coffee consumption in the major importing
increased about 17 percent in 2002, largely  countries has been stagnant, at about 4.5 kilo-
because of a 70 percent increase in cocoa    grams of green coffee equivalent, during the
prices. In contrast, coffee and tea prices   past decade.
remained weak. The sharp increase in cocoa      Global coffee production in the 2002-03
prices reflects production problems and the  season is expected to increase 10.7 percent
recent coup attempt in C6te d'lvoire, a major  from  last season's  110.7  million  bags
producer of cocoa. The weakness in coffee    (table A2.1). Brazil, the dominant producer
prices can be attributed to large stocks, weak  with one-third of global output, is expected to
demand, and large production increases by    produce a record 46.9 million bags, while
major exporters. Tea prices declined as a re-  Colombia and Vietnam, the second and third
sult of abundant supplies and weak growth in  largest producers, will each reach about 10 mil-
demand.                                      lion bags.
A number of unsuccessful attempts have
Coffee. Coffee prices fell to record lows and  been made to arrest the price decline. The
became the most visible symbol of the declines  Association of Coffee Producing Countries,
in agricultural commodity prices during 2002.  which has urged coffee producers to join its
178



G LO B AL CO M MO DI TY PR ICE PRO SP EC TS
Table A2.1 Coffee production in selected countries
(1m1illio/i bags)
1997-98      1998-99     1999-2000      2000-01     2001-02      2002-03
Brazil            22.8         35.6         30.8          34.1        33.7         46.9
Colombia          12.2         10.9          9.5          10.5         11.0        10.9
Core d'lvoire      3.7          2.2          5.7           4.3         3.3          3.3
Indonesia         7.8          6.9           6.7           6.5         6.0          5.8
Mexico             5.1          5.0          6.2           4.8         4.7          5.2
Vietnam            6.9          7.5         11.0          15.3         12.3        10.5
World             96.4        108.4        113.3         117.0        110.7       122.6
Souirce: U.S. Departmenit of AgriCLlture.
export retention scheme for the past three sea-  Production in two major producers, C6te
sons, ceased operating on February 1, 2002.    d'lvoire and Ghana, is estimated to be down
A plan backed by tihe Internatiolial Coffee    4 percent anld 2 percent, respectively, in the
Organization, which called for removal of low-  just-ending 2001-02 marketing season. The
quality coffee beans from the market, was not  extreme price increases in response to such
well supported by same coffee-produciig couni-  relatively small changes in output were partly
tries because it did not compenisate producers  caused by speculative buying by commodity
of low-quality beans. A number of countries    funds. In addition, uncertainty about the relia-
have also undertaken their own price-support   bility of supplies prompted strong demand
schemes or stock-holding nmechanisms. Brazil,  from processors.
for example, has subsidized put options to        Cocoa prices are expected to remain at
effectively guarantee a minimum price to pro-  their 2002 level next year. They will decline
ducers. While such scIleiles may be partially  12 percent in 2004 as production continues
successful in the short run, they could exacer-  to increase. This forecast is based on the as-
bate the oversupply problem in the long run.   sumptions that (a) the strong prices enljoyed
We project a recovery in both robusta and   this season have already given inlcentives to
arabica prices in 2003 and a further recovery  growers to maintaini their trees anid to increase
in arabica in 2004. Nevertheless, we recognize  production; (b) part of the recent surge in
the risk that it may take longer for the recov-  prices may have been caused by speculative
ery to materialize if the recent supply surge  activities of a short-term nature that are un1-
persists. Over the long term, real coffee prices  likely to be carried over into the next year;
are expected to recover, but they will remain  and (c) the recent coup attempt in C6te
well below the historical highs of the 1970s   d'lvoire has been repelled.
and more recelt highs of the 1990s. By 2015,      In response to high prices, growth in de-
real arabica and robusta prices are projected  mand for cocoa in the current and next mar-
to increase about 75 percent from the 2002     keting season is expected to slow from  the
levels. Prices would still be about only half of  1990-2000 average of 2.4 percent. But it
their 1990s peaks.                             should then return to historical growth rates
(table A2.2). By 20-15, real prices are projected
Cocoa.   Cocoa prices led the recovery of agri-  to decline 25 percent from 2002 levels.
cultural comnmodity prices, after falling to a
three-decade low in February 2000. Since then,  Tea. The three-auction average tea price fell
cocoa prices have more than doubled to a 16-   6 percent in 2002 as supplies continued to in-
year high amid supply disruption in major      crease relative to demand and stocks remained
producers from political instability and from  high (table A2.2). Production in major ex-
producers' responses to extremely low prices.  porters (India, Kenya, and Sri Lanka) was up
179



G LO B AL    EC ONO MI C      PROS PE CT S       2 0 0 3
Table A2.2 Global balance for beverages
Annual growth ratc (%)
1970     1980    1990     1999     2000    2001       1970-80  1980-90  1990-00
Coffee (thousand bags)
Productioni            64,161  86,174   88.849  113,345  117,049  110,773     2 11     1 36      1 20
Consumptioni           71,536  79,100   96,300  104,670  106,580  108,450      1 01    1 97      0 22
Exports                54,186  60,996   76,163   92,256  89,968   88,788      078      241       1 68
Cocoa (thouisand tons)
Production              1,554   1,695    2,506    3,073   2,812    2,750      0 46     4 62      1 16
Grindings               1,418    1,556   2,335    2,967   3,014    2,823      0 16     4.48      2.58
Stocks                   497      675    1,791    1,341    1,111   1,101       238    1389      -466
Tea (thousand tons)
Production              1,286    1,848   2,516    2,900   2,960    3,030      4 09     2 87       1 49
Exports                  752      859    1,132    1,259    1,330   1,389      2 35     2 39      1 62
Notes Tinme referceice for coffee (production and exports) aisd cocoa are based on crop year showin under the year that
production begins October to September for cocoa and April to March for coffee Coffee coiisumption and tea data are based
oni the calendar year
Souirces International Coffee Organization (ICO), International Cocoa Organization (ICCO), Food and Agriculture Organization
(FAO) of the United Nations, International Tea Committee (ITC), U S Department of Agriculture, and World Bank
4 percent in 2001-the last year for which
data are available. Other exporters, such as              Figure A2.3 Food prices
China and Vietnam, have also been increasing              (index, 1990 = 100)
exports rapidly, and such increases could fur-            130
ther weaken prices.
Prices are projected to increase modestly              120
from the 2002 lows (up 3 percent in 2003), but
they will remain depressed relative to the highs          110     AJ
in 1997 and 1998. If emerging exporters, such                          \ /
as Vietnam, continue to increase exports, there           100
is a significant risk that prices could continue                           \
to  fall. However, higher petroleum        export
prices in the Russian Federation and in major              80                w       \v    \v
consuming countries in the Middle East have
historically supported demand, and we expect               70 1     1      1     1      1
tea prices to begin a gradual recovery. By 2005,            Jan    Jan   J Jan  Jan    Jan   Jan
we project nominal tea prices to rise 10 percent            1997   1998  1999   2000  2001   2002
from 2002 levels, which would leave nominal               Source World Bank
prices down 20 percent from 1997 levels.
Food                                                  exporting countries. By 2015, real prices should
The index of food prices has not changed for          decline about 2 percent from 2002 levels.
several years after declining sharply during
1997-99 (figure A2.3). The index rose about           Fats and oils.    Prices of fats and oils recov-
4 percent in 2002 and is expected to rise 7 per-      ered 13 percent in 2002 after falling 40 per-
cent in 2003 and 2 percent in 2004 because of         cent from    1997 to 2001. The increase was
higher grain and oilseed prices following this        greatest in vegetable oils such as palm    oil (up
year's drought in major grain- and soybean-           35 percent) and coconut oil (up 30 percent)
180



G L O B A L C 0 MM 0 D I T Y P R I C E P R O S P E C T S
because of lower production. Meal prices re-     Table A2.4   Palm oil production
mained weak, with soymeal down 3 percent         (million tonls)
because of weak demand for livestock and         Year         Indonesia    Malaysia    WZorld
poultry feeds. Prices of most fats and oils are
expected to increase during 2003-05, and the     1990-91        2.41         6.10      11.03
index of prices is expected to increase 13 per-  201-06                                I 753  11.94  23 54
cent in nominial terms from 2002 to 2005.        2001-02        8.20        11.65      23.98
Global production of the major fats and       2002-03        8.50       '11.82      24.53
oils is expected to increase about 2 percent     Soturce; Oil World.
in 2002-03, while consumption is expected
to increase by 3.2 percent, causing stocks to
decline and prices to continue increasing. Palm  Table A2.5   Global grain stocks-to-use
and soybean oil production is the largest among  (percentages excluding Chia)
the vegetable oils. Together they represent      Year        Maize   Rice   Wheat   Total grains
40 percent of total vegetable oil production.
World soybean production is expected to re-      1998-99      1.'     9.6    18.6      14.0
main constant in 2002 because of drought in      1999-2000    11.4    t.s    17.7      13.7
the United States, after growing by 5.3 percent  2000-01     11.S    12.9    19.0      14.3
per year since 1990. This stoppage in growth     2002-03     610     10.9    21.1      1257
has led to higher soybean prices and reduced     1990s low    6.0     7.8    13.9      9.7
stocks in 2002 and is expected to support        Source: U.S. Department of Agriculture.
lmigher prices in 2003. Other major producers
(table A2.3) are expected to increase soybean
production despite econonmic problems and un-    increases in response to price increases. There
certainties.                                     is a risk that grain prices could continue to rise
Palm oil production has more than doubled     even more sharply than projected       if the
since 1990 (table A2.4), with the largest in-    drought conltinues in the major exporting
creases coming from Indonesia and Malaysia.      countries, or if other major grain producers
However, production is expected to increase a    have lower-than-expected production. Wheat
more niodest 2 percent in 2002-03.               prices are projected to rise an additional
19 percent in nominal terms by 2003 after
Grains.   World grain stocks, relative to use,   increasing nearly 20 percent in 2002. Prices are
are expected to fall significantly during the    then expected to decline 6 and 12 percent in
current crop year (table A2.5), and the declines  2004 and 2005, respectively, as production
are expected to keep grain prices rising through  responds to the higher prices. Maize prices rose
2003. Prices should then decline as production   12 percent in 2002 and are expected to rise an
additional 25 percent by 2003 before declining
Table A2.3   Soybean production                  in 2004 and 2005. Rice prices rose 11 percent
(million tonls)                                  in 2002 and are expected to rise an additional
22 percent by 2005.
Year    Argentina  Brazil  United Satres  World     Stocks in the major grain exporting coun-
1990     '1.5     15.8      52.4      '104.1     tries-the United States, the European Union,
1995     12.4     24.2      59.2      124.9      Canada, Australia, and Argentina-are ex-
2000     27.8     39.0      75.1      175.1      pected to fall to the lowest level in 2003-03,
2001     29.5     43.5      78.7      183. 
2002     30.0     48.0       71.5     183.3      relative to total use, since 1997-98. The de-
dline is mostly attributable to the droughts
Note: Argentina, Brazil, anld the Uniited States accounit for  ine  i ted  atesiCnaa, andrAustr
about 80 percenit of global produictioni.        In the United States, Canada, and Australia,
Souirce: U.S. Departmenit of Agriculture.        which are expected to reduce grain yields by 9,
181



G LO B AL EC O NO MI C PRO SP EC TS 2 0 0 3
7, and, 6 percent, respectively, in 2002-03  keting year that just ended in August, and an
compared with yields in the previous year. An  increase in carryover stocks to nearly 50 per-
El Nimio weather pattern has contributed to the  cent of annual consumption. Brazil, the largest
unfavorable weather pattern in Australia and  exporter, is expected to have a sugar cane crop
could further reduce production next year.2  that could exceed the previous year's crop by
The lower yields in the United States, Canada,  8 or 9 percent. Imports are expected to be
and Australia have been partially offset by  weak because of large production in import-
record grain yields and production in the   ing countries. Hence, prices are unlikely to
European Union. Economic problems in        recover significantly in 2003.
Argentina have contributed to lower produc-    Brazil has nearly 30 percent of the export
tion and exports from that country, but the  share in recent years and has been the primary
largest effect of the economic turmoil is ex-  source of increased global exports, with pro-
pected to be in the next crop year, because most  duction and exports growing rapidly in the
of the planting and input-use decisions had  past decade (figure A2.4). The other major
already been made before the economic crisis  exporters, Australia and Thailand, increased
fully emerged.                              production by 50 and 70 percent, respectively,
Grain production in developing countries  from 1990-91 to 1997-98, when sugar prices
is projected to be down 1.8 percent in      were attractive. However, they have cut pro-
2002-03, with production generally strong in  duction as prices have declined.
Asia, Latin America, and the Middle East, but  Sugar prices are expected to begin to re-
lower in Eastern Europe and Russia. Produc-  cover in 2004 as low prices reduce global
tion in China is expected to be up 2.3 percent,  supplies. However, prices are expected to re-
while production in India is expected to be  main relatively weak for the next several years,
down 4.9 percent because of a poor monsoon  with fluctuations depending on the year-to-
season.                                     year balance of production and consumption.
There is considerable variation in the stock  By 2005, nominal sugar prices are expected
situation in individual grains, with the global  to increase 17 percent over 2002 levels. In the
stocks-to-use percentage for maize at the low-  long term, nominal prices are expected to
est levels of the 1990s, while rice and wheat  return to the center of the trading range,
percentages are above previous lows. How-   and real prices are expected to average about
ever, grain prices are highly correlated, and  18 cents per kilogram (8.2 cents per pound).
price increases in one grain would normally
be reflected in the prices of others. Higher
grain prices woulcl benefit developing-country
net exporters such as Argentina (which is      Figure A2.4 Sugar exports
expected to export more than 22 million tons  (million tons, raw equivalent)
of grain in 2002-03) while harming net im-    40     Rest of world
porters such as Mexico and the Arab Republic  35        1      
of Egypt, which are expected to import 13 mil-  30
lion and 10 million tons of grain, respectively,  25
in 2002-03.                                   20
15
Brazil
Sugar   Sugar prices fell to 15 cents per kilo-  10
gram in 2002 (down 21 percent from 2001) to    5
return to the lower end of the trading range of  0 l  l    l    l
A   . .   , s     >^             ~~~~~1990 1992 1994 1996 1998 2000 2002
10-30 cents per kilogram of the past 20 years.
The decline follows an estimated 5 percent in-  Source US Department of Agriculture
crease in world sugar production in the mar-
182



G LO B AL    CO MM OD IT Y       PR ICE     PRO SP EC TS
Table A2.6 Global balance for foods
Annual growth rates (%)
1970    1980     1990      1999     2000     2001    1970-80  1980-90  1990-2000
Grains (million tons)
Productioni               1,079   1,430    1,769     1,871     1,839   1,860    2.88      1.55     1.04
Consumption               1,114   1,451    1,717     1,869    1,868     1,890   2.58      1.78     1.02
Exports                    119     212       206      245       233      231    6.35     0.13      0.94
Stocks                      193    309       490      529       500      470    7.24     3.83     -0.56
Soybean (tbousand tons)
Production               42,133  62,173  104,093   159,904  175,098  183,724    6.84      1.87     5.08
Consumption              45,968  68,052  104,307   160,541  172,166  184,228    6.53     2.04      4.99
Exports                  12,342  20,822   25,388    46,683   55,074   57,127    5.24     0.80      2.88
Stocks                    3,394  10,266   20,569    27,908   30,803   30,218    13.83   -0.66      0.20
Sugar (thousand tonrs
(rawv equsivalent])
Production               70,919  84,742  109,403   138,094  143,220  136,111    2.80      1.59     3.26
Consumption          '   65,395  91,062  106,807   130,281  133,104  134,712    3.30      1.40     3.00
Exports                  21,931  27,571   34,078    38,710   42,015   38,495    3.26     0.83      3.12
Stocks                   19,614  19,494   19,299    31,702   35,939   35,474    3.96    -0.77      4.52
Fats and oils (nilliont tots)
Production                39.78   58.09    80.84    113.42   117.09   119.42    3.68     3.54      3.70
Consumption               39.82   56.80    80.87    111.98   116.94   120.74    3.55     3.69      3.64
Exports                    8.83   17.76    26.89     35.55    38.10    39.57    7.05     4.19      3.39
Stocks                     5.18    9.25    12.15     14.26    14.47    13.19    7.09     2.44      0.69
Note: Time referenices for grains, soybeans, and sugar are based on marketing years, shown under the year in which production
begins, and they vary by counitry. For fats and oils, crop years begini in September.
Source: U.S. Department of Agriculture and Oil World.
The global balances for major foods are             2002 levels, while real prices are projected to
given in table A2.6. The balances show that the        rise 18 percent by 2015 over 2002 levels.
rate of growth of production and consumption
of grains has slowed during the 1990s com-             Cotton.    Cotton prices declined an additional
pared with    previous decades, while growth           5 percent in 2002 after declining 19 percent in
rates have increased for soybeans and sugar.           2001 because of large production increases in
The growth rates for fats and oils were rela-          the United States and China, the two largest
tively constant during the 1980s and 1990s.            producers (table A2.7). Prices in 2002 were
less than half of their 1995 highs, and they
reached 30-year nominal lows. The extreme
Agricultural raw     materials                         price weakness was caused by a number of
The index of agricultural raw materials prices         factors, such as slow growth in demand, large
(comprising    prices of tropical hardwoods,           production, and competition       from   synthetic
cotton, and natural rubber) declined sharply           fibers. Subsidies to cotton producers in the
during Asia's economic crisis and then sta-            United States and China have contributed to
bilized  before   declining   again  as supplies       the production surplus. During the past three
of commodities continued        to  increase  (fig-    seasons, U.S. support to its cotton producers
ure A2.5). Prices reached a low      in 2001 and       averaged almost $3 billion, and China's sup-
have since recovered because of higher cotton          port averaged $2 billion.
and natural rubber prices. Nominal prices are              Cotton production in the coming season is
projected to increase 16 percent by 2005 from          expected to be 19.2 million tons-10 percent
183



G LO B AL EC ONO MI C PRO SP EC TS 2 0 0 3
Table A2.7 Cotton production in selected countries
(thouisand tons)
Country           1998-99        1999-2000       2000-01        2001-02        2002-03
China              4,501           3,830          4,350          S,320          4,420
Franc zone           897            928             700          1,034           921
India              2,710           2,650          2,350          2,686          2,500
Pakistan           1,480           1,800          1,750          1,853          1,731
United Stares      3,030           3,835           3.818         4,420          3,826
Uzbekistan          1,000          1,150            960           1,055         1,015
World           j 18,551          18,887          19,126         21,422        19,157
Source International Cotton Advisory Committee
prices gained momentum at the beginning of
Figure A2.5 Raw materials                 2002, with average 2002 prices rising about
(index, 1990 =100)                         32 percent from 2001. The recovery is mainly
130                                       a response to adverse weather conditions in
Thailand and a slowdown in Malaysia's out-
120                                       put growth as natural rubber plantations
are being converted to more profitable palm
110                                       oil plantations. Demand, however, remains
weak as car tire manufacturing (the largest
100                                       demand for natural rubber) in Organisation
90  A       1tfor Economic Co-operation and Development
(OECD) countries is estimated to be down
80                      \       \         2 percent in 2002.
The strength in natural rubber prices is
70 1_1_____             I_     _hlikely to persist because supply controls by the
Jan   Jan  Jan  Jan   Jan  Jan          Tripartite Rubber Corporation-a trilateral
1997  1998  1999  2000  2001  2002       organization formed last year by Indonesia,
Soturce World Bank                        Malaysia, and Thailand following the col-
lapse of the International Natural Rubber
Organization-may restrict exports. We ex-
lower than in the previous season, with the  pected natural rubber prices to remain firm,
United States and China accounting for most  but not increase significantly, in 2003 from
of the decline. In the United States, the drought  2002 levels because of weak demand that ac-
has reduced production from    the record    companies the apparent slowing of growth in
2001-02 season. Global consumption is ex-    the global economy. By 2005, nominal prices
pected to increase about 2.6 percent, according  are expected to increase 6 percent from 2002
to the latest forecasts by the International  levels. Over the longer term, real prices are
Cotton Advisory Committee. Given lower pro-  projected to decline-down 3 percent from
duction in combination with higher consump-  2002 to 2015.
tion, we forecast the A Index cotton price to
increase 10 percent in 2003 and 16 percent   Tropical timber  The decline in Asian tropi-
in 2004. By 2015, real prices are projected to  cal timber prices since the mid-1990s appears
increase 30 percent relative to 2002 prices.  to have ended, and prices have begun to re-
cover from the lows reached at the end of
Natural rubber. After prolonged  weakness    2001. Nominal timber prices increased about
following the Asian crisis, natural rubber   9 percent in 2002 compared with 2001 prices
184



G LO BA L    CO MM OD IT Y       PR ICE    P R O S P E C r S
as a result of the improved       demand    from      slowed dramatically during the 1990s com-
Japan, the    weakening     of the   U.S. dollar       pared with the 1980s. Exports of cotton grew
relative to the yen, and the continued strong         only 0.2 percent during the 1990s, which con-
import demand      from  China. Prices are ex-        tributed to the sharp price decline. Growth
pected to continiue to recover in 2003        and     rates of natural rubber production, consump-
2004, with annual average increases of 8 per-         tion, and exports remained       nearly constant
cent per year, resulting from     improved eco-       during the 1990s compared with the 1980s.
nomic growth in Asia. African sapelli log prices      Tropical timber log production slowed while
have declined less than Asian log prices, as          production of sawnwood increased as timber-
demand has remained firm      in Europe. Sapelli      producing countries shifted to increased do-
nominal log prices are expected to increase           mestic   processing. Sawnwood        imports in-
about 5 percent from 2002 to 2005.                    creased while plywood imports slowed during
Real tropical timber prices are expected to        the 1 990s compared with the 1 980s.
recover from lows, but they are not expected
to reach new highs during the forecast period
to 2015. By 2015, real meranti log prices are         Fertilizers
projected to rise 47 percent, while sapelli log       ]ertilizer prices remained nearly constant
prices are projectedto rise by only 18 percent.       F   in 2002 after several years of large adjust-
The difference is due to the smaller decrease          ments (figure A2.6). Import demand remained
and, therefore, smaller rebound       of African      weak because of low      commodity prices and
sapelli logs prices compared with Asian mer-          increased local production. However, fertilizer
anti log prices.                                      production in major exporters contracted in
The   global balances for raw        materials     response to low      fertilizer prices. Hence, a
are given in table A2.8. The data show       that     market balance was achieved with little pres-
cotton production, consumption, and exports           sure on prices. Acreage used for global grain
Table A2.8 Global balance for raw materials
Annual growth ratc s%)
1970    1980   1990   1999    2000   2001   1970-80  1980-90  1990-2000
Cotton (tbousa,nd tons)
Production                       11,740 13,832  18,970  19,126  19,408  21,422  1.2   3.1      0.8
Consumption                      12,173 14,215  18,576  19,796  19.762  20,070  1.1    3.1     0.2
Exports                          3,875   4,414  5,081   6,142  5,750   6,430   0.9    2.8      0.5
Stocks                           4,605   4,895  6,645   9,559  9,274  10,630   1.7    2.8      1.4
Natural rubber (tboussand tons)
Production                       3,140   3,820  5,080   6,810  6,740   7,170   1.8    3.2      3.1
Consuimptioni                    3,090   3,770  5,190   6,660  7,330   7,030   1.6    3.2      3.3
Net exports                      2,820   3,280  3,950   4,670  4,940   5,160   1.3    2.1      1.8
Stocks                            1,480  1,480  1,500   2,540  1,950   2,090   0.6    0.2      3.7
Tropical timber (thousanzd cubic nseters)
Logs, productioni                  210    262     300    286     287    276    1.5     1.7     0.5
Logs, imnports                    36.1    42.2   25.1    18.3   21.1    21.0   0.2    5.1      5.4
Sawnwood, produiction              98.5  115.8   131.8  103.9  101.5    99.3   1.2     1.7     2.0
Sawnwood, imilports                 7.1   13.2   16.1    21.2   24.3    23.5   5.0     2.6     3.3
Plywood, produiction              33.4    39.4   48.2    52.6   55.4    54.9   1.2     2.0     0.5
Plywood, iniiports                  4.9    6.0   14.9    18.9   19.8    20.3   0.7    9.1      3.6
Notes: Time referenice for cottoni is based oni the crop year beginning in AuIguIst. For niatural rubber and tropical timber, time
refers to the calendar year.
Sources: Initernationial Cottoni Advisory Committee, linterisationial Study Rubber Grotip, FAO, and World Banik.
1.85



G LO B AL EC ONO MI C PRO SP EC TS 2 0 0 3
cent above 2002 levels, as the industry contin-
Figure A2.6 Fertilizer prices            ues to rationalize and reduce surplus capacity.
(dollars per ton)                           Prices for potassium chloride (also known
200                                      as muriate of potash, or MOIP) declined 5 per-
180                                      cent in 2002 from weak demand and large
160 \                                    surplus capacity. Price declines could have
140  Xi \    A         M   X  \          been much larger without aggressive supply
120             /                    s s I controls by major exporters. Increased domes-
> 0         A        /tic production in China is expected to weaken
100   \ /9      V                   future import demand and, along with a large
80                                       surplus in global production capacity, to keep
60  m              U                     price increases small, despite the increased use
MOP            Urea                  for grain production, which accompanies the
20                                      recovery in grain prices. By 2005, nominal
0,              ,        ,              MOP prices are projected to increase 10 per-
1980   1985   1990   1995   2000       cent from 2002 levels, and real prices are pro-
jected to fall 6 percent by 2015 compared with
Note TSP = trpipe super phosphate, MOP = muriate of  2002 prices.
potash                                   20     rcs
Sonirce Fertilizer Week                     Triple super phosphate (TSP) prices in-
creased 5 percent in 2002 after falling 27 per-
cent from 1998 to 2001. Production fell in
production, which accounts for more than half  2001 in response to low prices, and imports
of total fertilizer use, declined for the sixth  declined slightly because of increased local
consecutive year in 2002, but it is expected to  production in China and India. Demand
increase in 2003 and 2004 in response to re-  should increase along with increased grain
cent and expected grain price increases. Pro-  prices and area planted. Surplus capacity is
duction capacity remains substantially larger  smaller than for other major fertilizers and is
than demand for all major fertilizers, but it is  expected to decline over the next several years.
most extreme in potash, where surplus capac-  This decline will cause nominal TSP prices to
ity may be as high as 30 percent of demand,  increase by an estimated 13 percent by 2005.
according to industry estimates.            Real prices are projected to decline by 5 per-
Nitrogen fertilizer prices (as represented by  cent by 2015 from 2002 levels.
urea prices) were down about 2 percent in     The large surplus of global production ca-
2002, as exports from  major producers In   pacity in the fertilizer industry is largely a result
Eastern Europe fell because of rising natural  of the sharp declines in consumption in former
gas prices, currency changes that made ex-  Soviet bloc and Eastern European countries
ports less profitable, and increased local fertil-  following the collapse of the former Soviet
izer demand. This fall was partially offset by  Union and the transition of those countries
reduced demand in major importing countries  to market economies. Many countries (such
as a result of low commodity prices and in-  as Russia and Ukraine) were left with large
creased local fertilizer production. Urea prices  production capacities and reduced domestic
are expected to continue to increase because  demand-which led to export growth of nearly
of higher grain prices and reduced exports  4 percent per year since 1993 from the former
from Eastern Europe. By 2005, nominal urea  Soviet Union. Those increased exports dis-
prices are projected to increase 36 percent  placed traditional exports and depressed prices
from 2002, but then increases are expected to  of nitrogen and phosphate fertilizers. Global
slow, and real prices should decline. By 2015,  fertilizer consumption fell about 17 percent
real urea prices are expected to remain 19 per-  from  1988 to 1993 and has only recently
186



G LO B AL    CO MM OD IT Y       PR ICE    PRO SP EC T S
Table A2.9 Global balance for fertilizers
(million tolls)
Aninual growth rates (%)
1970     1980    1990     1998     1999     2000     1970-80  1980-90  1990-2000
Nitrogen
ProdLctioll              33.30   62.78    82.28    88.30    87.75    84.62     6.53     3.12     0.28
Consumption              31.76   60.78    77.18    82.77    84.95    S.62     6.86     2.60     0.56
Exports                   6.77   13.15    19.59    23.00    23.94   24.70      7.23     5.10     2.34
Phosphate
Plroductioni             22.04   34.51    39.18    33.09    32.51   31.70      3.72     1.70    -2.10
Constimption             21.12   31.70    35.90    33.35    33.46    32.65     3.85     1.39    -0.90
Exports                   2.92     7.51   10.50    12.59    12.70    12.11     8.37     5,01     1.44
Potash
production               17.59   27.46    26.82    25.01    25.01    25.54     3.97   -0.03     -0.49
Consumptioin             16.43   24.24    24.68    22.04    22.12    22.16     3.93     0.05   -1.07
Exports                   9.45   16.72    19.82    22.23    22.65   23.41      4.89     0.73     1.68
Note: All data are in mizrkerinig years.
Souirce: IFood and(l Agricultyre Organiization.
recovered to near thie 1988 peak. Table A2.9
gives the global balances for fertilizers.                Figure A2.7 Aluminum and copper prices
(dollars per ton)
2,700
Metals and minerals
P   rices for metals and minerals rallied from                       Copper
the October 2001 lows because of expec-               2,300
tations of a robust economic recovery that
would lead to a strong demand for metals.
However, the price rally stalled in the second           1,900                         A
quarter of 2002, as it appeared the recovery
would be more muted than anticipated. With               1500
weak   demanid    and  large inventories, most            1,500
metal prices have receded to at or below end-                                       Aluminum
2001 levels (see figure A2.7 for aluminum     and         1,100                           I I l
copper). Even with the rally, the index of Inet-             Jan.   Jan.   Jan.  Jan.   Jan.   Jan.
als and minerals prices during the first nine                1997   1998  1999   2000   2001  2002
months of 2002 averaged 5.6 percent lower                Source: I'latts Metals Veek.
than for the same period a year earlier.
Growth il demand has been very sluggish
in 2002, with little indication of strong growth
in the near term. Meanwhile, production con-          num, have helped support prices, but more
tinues to rise, despite efforts to shut capacity.     closures may be necessary to prevent further
As a result, the London Metal Exchange (LME)          stock building and even lower prices.
invenitories of most metals have continued to            The price recovery will likely be delayed
risc to relatively high levels (see figure A2.8 for   until 2003, and the strength of the recovery
aluminumil and copper). A number of produc-           will largely be determined by the timing and
tion cutbacks, notably in copper and alumi-           strength of the rebound in the global economy.
187



G LOBA L EC ON OM IC PRO SP EC TS        2 0 0 3
Figure A2.8 Aluminum and copper stocks
(thousand tons)
1,400
1,200
1,000
80   Aluminum
600
400
200                  ~~~~~~~~Copper
200
Jan          Jan           Jan           Jan          Jan          Jan
1997         1998          1999         2000          2001         2002
Souirce London Metal Exchange
There is a possibility that during the upturn  surplus over the next two years, which should
of the next economic cycle metal prices could  prevent any substantial increase in prices.
rise significantly, augmented by strong buy-  Chinese exports are expected to continue rising
ing from investment funds. However, this rise  over this period, contributing to the surplus.
would induce the development of new capac-   The market is not expected to move into deficit
ity and the restart of idle facilities, and prices  until 2005, but there are many risks in the near
would eventually decline. Real metals and    term, such as the strength of the economic
minerals prices are expected to decline in the  recovery, the reactivation of idle capacity, and
long term, as production costs continue to fall  the amount of Chinese net exports.
with the implementation of new technologies     Real prices for primary aluminum are ex-
and of improved managerial practices.        pected to decline in the long term, as new low-
cost capacity is developed to meet expected
Aluminum                                     growth in demand. However, investment in
Aluminum prices have fallen back near the    new aluminum plants will continue to require
lows of October 2001 because of relatively   low-cost power supplies. There is not expected
weak demand, rising production, and soaring  to be any significant constraint on alumina sup-
stocks. Prices have been partly supported by  ply in the medium term, because several new
reductions in capacity caused by high electric-  alumina capacity expansions are under way.
ity prices and rationing in the Pacific North-
west and Brazil, but reactivations in Brazil and  Copper
to a lesser extent in North America have con-  Copper prices led the rally in base metals during
tributed to the surplus. Production in China  the past year following a series of production
has grown significantly, and despite demand  cuts, with prices rising 20 percent from October
growth of more than 10 percent per year,     2001 to June 2002. Prices have since receded
the country became a net exporter this year,  because of prospects of weak demand in the
adding to the downward pressure on prices.   near term. However, the market is expected to
Growth in demand is expected to accelerate  be in reasonable balance this year as world mine
in 2003, but the market is expected to remain in  production declines about 2 percent because of
188



G LO BA L CO MM OD I TY PR ICE P R O S P E C T S
industry curtailments. LME inventories re-   60,000 tons of stock as collateral against a
main high, although they started to decline in  three-year loan from Western banks, which
May largely as a result of the strong growth of  may keep the material off the market for the
Chinese imports.                             duration of the loan. Demand for nickel has
Demand is expected to strengthen next     been relatively strong in the stainless steel sec-
year, and supply is expected to almost keep  tor, largely because of the shortage of scrap
pace, largely because of the recent commis-  supply.
sioning of Chile's Escondida Phase IV project  The nickel market is expected to move into
and the restart of idle capacity. The firm mar-  deficit in 2003 and over the next few years be-
ket balance should help support prices, but  cause production increases are expected to fall
high stocks may prevent sharply higher gains  short of a strong growth in demand. No major
next year. The market is expected to remain in  new projects are being commissioned until
modest deficit over the next few years, which  2005. Poor technical and financial perfor-
should support rising prices during the forth-  mance with pressure-acid-leach technology in
coming economic cycle. In the longer term, in-  Australia has been a major reason for the cur-
creases in new low-cost capacity are expected  rent lack of investment, which could result in
to result in the continued decline of real prices.  fairly strong prices over the next couple of
A major uncertainty over the forecast period  years. Over the long term, large new develop-
will be the volume of Chinese imports.       ments are expected to come onstream, such
as Inco's Goro project in New Caledonia (in
Nickel                                       2005) and Voisey Bay in Labrador, Canada (in
Nickel has been one base metal to sustain price  2006). Supply will originate from other new
increases this year, with a 38 percent gain  projects, expansions, and Norilsk's stockpiled
between October 2001 and September 2002.     material. New technologies will lead to lower
Relatively low stocks and Russian Norilsk's  costs, and real prices are expected to decline.
efforts to keep surplus supplies off the export  Table A2.10 shows the production, con-
market have supported prices that are signifi-  sumption, and LME ending stocks for alu-
cantly higher than would be expected at the  minum, copper, and nickel from 1970 through
bottom of the business cycle. Norilsk is using  2001.
Table A2.10 Global balance for metals and minerals
(thousanid tonts)
Annual growth rates (%)
19_0 1980 1_990   _ 9    _ 2000  201 _    1970-80  1980-90 1990-2001
Aluminum
Production               &lI,I                 NA M    4(*)      3 2     1.9    2.2
Consumption          N6=     1l  BONO3   3I    KM                3.2    1.8     1.8
LME ending stocks  ff              132    wi X    E      P   _p -  n a.  -0.3    9.2
Copper
Production        a!l !ULU-   8         f4Ie   86J03f  i         1.9    1.1     3.4
Consumption   1      !3   4,M   ffl" 10,(W     lUs    4          2.5    1.0     2.8
LME endimg stocks     f,    X          r   D 'P         -,       7.4   -56      14.6
Nickel
Production                          t.j i.i' i!'      inZJ       ri       6
Consumpts .i. i.                                                 r. a
LMEendiim .1.j.                      _ k                         111 -4i        15
Sources: World Bureau of Metal Statistics, Loridon Metal Excliange, and World Batik.
189



G LO BA L E CO NO MI C     PROS PE CT S 2 0 0 3
Gold                                             tions come onstream. An important deter-
Gold prices have averaged more than $300 per     minant of medium-term     prices will be the
troy ounce (toz) since April 2002, which is the  decision by central banks on whether official
first time since 1997 that prices have been      gold sales should be stemmed further when
above $300 for more than a month. Much of        the Washington Agreement expires in 2004.3
the strength has been from buybacks of hedged       Table A2.11 shows the demand for end sup-
positions by gold producers. In addition, in-    ply of gold from 1991 through 2001.
creased investment demand-partly in reac-
tion to declining U.S. equity markets and the
declining dollar-has helped support prices.      Petroleum
However, the recent rally in gold prices isi il prices slumped after September 11,
not expected to endure as producer buybacks      0J2001, because the economic recession,
end and central bank selling continues. At       mild weather, and reduced air travel weak-
present, hedging by producers is unattractive    ened demand. Also, OPEC made no attempts
because of low   interest rates, but at some     to prop up falling prices (figure A2.9). How-
point producer hedging could again become        ever, as OPEC    prices fell well below   the
attractive, which would push prices lower. Al-   organization's target range of $22 to $28 per
though the United Kingdom's central bank         barrel (OPEC   basket $17.53 per barrel in
sales program  ended in March 2002, other        December 2001), 10 OPEC countries, exclud-
central banks (such as Switzerland's) are pro-   ing Iraq, agreed to reduce production quotas
ceeding with their programs.                     6.5 percent at the start of 2002. This reduction
If prices remain above $300/toz, they will    was the fourth cut in quotas in less than a year,
weaken the price-sensitive jewelry demand        totaling S million barrels per day or 19 percent
market and will stimulate investment in new      (figure A2.10).
supply. Even when prices fall below $300 per        Prices started to rebound at the end of 2001
toz, mine production is expected to continue     on expectations that markets would tighten be-
to increase moderately as new low-cost opera-    cause of a recovery in world oil demand, OPEC
Table A2.11 Global balance for gold
(tons)
Percent per year
1991    1995    1996    1997     1998   1999     2000    2001      1991-2001
Demand
Jewelry          2,358   2,618   2,791   2,851   3,349   3,149    3,188   2,995        24
Other fabrication  518    457     503     484     560      595     564     487       -0 6
Bar hoarding      252     231     306     182     325      240     214     220       -1.3
Other           2,358     n a       6     n,a     n.a.     170     n a.    n.a.       n.a
Total demand    3,128    3,305   3,606   3,518   4,234   4,154    3,982   3,804        2.0
Supply
Mine production  2,159   2,279   2,274   2,361   2,479   2,568   2,580    2,595        1.9
Net official sales  111    81     173     279      626     464     471     468        15.5
Old gold scrap    482     617     625     640      628     616     608     695         3 7
Net hedging        66     163     535     142      504     506     n.a.    n a        n a.
Other             310     173     n a.     95     297      n.a.    322      46        n.a.
Total supply    3,128    3,305   3,606   3,946   4,154   4,154   3,982    3,804        2 0
n a = Not available
Sources. Gold Field Minerals Service and World Bank
190



G LO B AL CO MM OD IT Y PR ICE PRO SP EC TS
Figure A2.9 Oil price and OECD stocks
(dollars per barrel)                                                        (million barrels)
35                                        -                                        2850
30 L                           Al-" Stcks                   WB price             - 2800
v \                                    - ~~~~~~~~~~~~~2750
25 -          A\                                                                   } \/>\ /\ / \ / - 2700
V'V                                               ~~~~~~~~~~~~~~~~~~2650
20             A[-;       ;    >      A       &      2    \/                       2600
2550
15                                          vvv2500
2450
10                                                                                 2400
Jan           Jan          Jan           Jan           Jan           Jan
1997          1998         1999          2000          2001         2002
Sou(rce; World Bank and Interna-ional Energy Agency
Figure A2.10 OPEC-10 production and quotas
(million barrels per day)
30
Plus Iraq production
28
27
26
25    OPEe: - quotas'                 Perroduction
24                                                     .      . .
23.                                                     .  . .  . .
22
21
Jan            Jan           Jan.           Jan            Jan           Jan
1997           1998          1999           2000          2001           2002
Source: International Energy Agency and Organization of Petroleum Exporting Countries
output restraint, and declining stocks. In addi-  in March and approached $30 per barrel in
tion, perceived threats of a supply disruption  September as U.S. President George W. Bush
from a United States-led invasion of Iraq also  took his case for war against Iraq to the United
helped push prices higher, and those anxieties  Nations (U.N.). Market fundamentals also
deepened as the year progressed. The World      started to tighten heading into the peak-
Bank's average price rose above $20 per barrel  demand winter season.



G LOB AL EC ON OM IC        PROS PE CT S     2 0 0 3
Table A2.12 Global balance for petroleum
(million barrels per day)
Annual growth rates
1970   1980   1990   2001    2002   2003   197G80    1980-90  1990-2001
Consumption
OECD                    34 0   41.5    41.5   47 7   47.6   48.0     2 0       0.0       1.3
Former Soviet Union   . 5.0     8.9    8.4     3 7    3.8    3.9     6 0      -0,6      -7 2
Other non-OECD countries  6.8  12 3    16.1  25.1    25 3   25.7     6 1       2 7       4.1
Total                   45.7   62.6    66.0   76 5   76.6   77.5     3.2       0.5       1.3
Production
OPEC                    235    27.2   24.5    30.2   28.5   28.7     1.5      -10         1 9
Former Soviet Union   . 7.1    12.1    11.5    8.6    9.3    9.9     54       -05       -2.6
Other non-OPEC countries  17.4  24.6   30 9   38.2   38 6   39.1     3.5       2.3       1.9
Total                   48.0   63.9    66 9   76 9   76.4   77.7     2.9       0.5       1.3
Stock change, miscellaneous  2.3  1.3  0.9     0.4   -0.2    0.3
Memo item: Iraq          1.6    2.7    2.0    2.4     1.9    2.0     5.5      -2.7        1.5
Sources BP, International Energy Agency, and World Bank
Fundamentally, the market was in reason-       difficult to quantify such a premium, and no
able balance for much of 2002, and invento-       precise definition exists. Energy expert Philip
ries were at fairly typical levels, although      K. Verleger Jr. defines the premium    as the
stocks could fall to relatively low levels during  incremental amount a buyer is willing to pay
the winter without higher OPEC production.        for ensured prompt supply over deferred oil
World oil demand is likely to rise only mar-      given the level of inventories. He argues ac-
ginally this year (table A2.12), similar to the   cording to that definition that no war pre-
gain in 2001. Meanwhile non-OPEC supplies         mium existed at the end of September 2002.4
continue to increase strongly, rising by an esti-   The near-term   outlook for the oil market
mated 1.2 million barrels per day, with more      depends heavily on developments in Iraq and
than half of the gain expected to come from       on OPEC's production decisions. While there
Russia.                                           is agreement between the United States and
It is only through significant production      U.N. to allow weapons inspectors back into
restraint that OPEC has kept prices within its    Iraq, there is likely to be less agreement on
target range-notwithstanding some overpro-        how to proceed if Iraq refuses U.N. demands.
duction by members of the group. In addition,     Should an attack occur in the coming months,
Iraq's exports have been less than half of the    prices could spike sharply higher, depending
country's potential for much of the year, be-     on the prevailing level of inventories, the re-
cause of disputes with the U.N. about Iraq's      sponse from OPEC producers, and the draw-
surcharges, which the U.N. sought to eliminate    down of strategic reserves. During the 1990
with a retroactive pricing scheme. However,       war in the Persian Gulf, more than 4 mbfd of
buyers are exposed to large risks with this       oil from Kuwait and Iraq were removed from
mechanism, and crude oil purchases from Iraq     international markets, and prices exceeded
were curtailed.                                   $40/bbl. There was substantial surplus pro-
Expectations of an attack on Iraq have         duction within OPEC, and the organization
led to a wide range of estimates of a "war        raised output-but not immediately. Impor-
premium" on prices this year. Estimates range    tantly, prices did not fall until the war com-
from very little (prices reflect the market bal-  menced (and its success was quickly assured)
ance) to several dollars per barrel. It is very   and the strategic stocks were released.
192



G L O B AL CO MM OD IT Y  PR ICE P R O S P E C r s
Since Iraq is exporting only around I mb/d,  below $30/bbl, but the organization may have
much less oil is at risk, although it is conceiv-  to reduce output at winter's end to keep prices
able that Iraq could launch scud missiles into  within its price target. The demand for OPEC
Kuwait and Saudi Arabia and could tem-      oil is expected to rise only modestly in 2003.
porarily disrupt supplies. There is more sur-  An increase in non-OPEC supply of 1 mb/d
plus capacity within OPEC than in 1990, and  is expected to capture the bulk of the growth
sufficient spare capacity within Saudi Arabia  in world oil demand. Rising capacity withini
alone could easily replace lost oil from Iraq.  OPEC, requests for higher quotas (from
However, OPEC desires prices of at least    Algeria and Nigeria), and a recovery of Iraq's
$251bbl, and it is not clear how quickly its  exports could strain OPEC's efforts to support
members will raise production to prevent a  higher prices. But as long as the risk of a sup-
surge in prices. In such an environment, crude  ply disruption hangs over the market, prices
prices could be bid up sharply because of   are likely to remain well within OPEC's target
higher demand, speculation, and hoarding.   range.
Buyers might have to pay a substantial pre-    Oil prices are expected to decline from
mium for prompt supplies, and prices could  $25 per barrel in 2002 to $23 per barrel in
rise to 1990 levels.                        2003 as a result of rising supply competition
Once war ends, prices could fall precipi-  and below-trend growth in demand. By mid-
tously as a result of a higher OPEC produc-  decade, prices are expected to fall below
tion, a draw from strategic stocks, and the  $20 per barrel (figure A2.11). A risk to the
return of Iraqi exports. Disputes within OPEC  forecast is that OPEC could maintain strong
over market share could take prices well below  production discipline over the next few years
$20/bbl.                                    to keep prices at or above $25 per barrel. If
In the absence of an attack, OPEC's pro-  such efforts prove successful, they would
duction decisions will heavily influence prices.  add to the growing pressures on prices-by
The group will likely attempt to keep prices  negatively affecting demand and by stimulat-
at $25/bbl. Higher OPEC production will be  ing competing supplies-and prices would still
required during the wtnter to keep prices   be expected to fall below $20 per barrel by
Figure A2.11 Crude oil prices
(dollars per barrel)
50
45
40
35
30
25                                     1990 dollars                 Nominal
20
Real
15 
10
5            ,   2
O   ;I  I  I  I  I  I  i I  I  I  I  I   I   I   I  I  I  I  i I   I   I I I I I I I I I I I I I I I I I
1960   1965  1970   1975   1980   1985   1990   1995  2000   2005   2110   2015
Souirce W'orld Bank
193



G LO BA L EC O NO MI C PRO SP E CT S 2 0 0 3
mid-decade. By 2005-06, significant new sup-  deep water offshore and the Caspian Sea), and
plies from West Africa, the Caspian Sea, and  development costs continue to fall from new
elsewhere are expected to become available.  technologies (shifting supply curves outward).
Coupled with rising capacity within OPEC,  In addition, OPEC members are increasing
those supplies will exert severe downward  capacity and will add to the supply competi-
pressure on prices.                       tion in the coming years. Consequently, real
In the long term, growth in demand will  oil prices are expected to continue their long-
be only moderate, as it has been for the past  term decline.
20 years (table A2.12), but new technologies,  As mentioned at the beginning of this ap-
environmental pressures, and government   pendix, we will now present tables showing
policies could further reduce this growth.  actual commodity prices for 1970 through
Prices somewhat below $20 per barrel are suf-  2001, plus price projections for 2002 through
ficiently high to generate ample development  2015. Table A2.13 gives the commodity prices
of conventional and unconventional oil sup-  and forecasts in current dollars, table A2.14
plies, and there are no apparent resource con-  uses constant 1990 dollars, and table A2.15
straints far into the future. In addition, new  displays weighted indices of commodity prices
areas continue to be developed (for example,  and inflation.
194



G LO B AL    CO MM OD IT Y      PR ICE    PRO SP EC TS
Table A2.13 Commodity prices and price projections in current dollars
Actual                i               Projections
Conmmodiy                        Unita       1970    1980    1990    2000   2001     2002    2003    2005   2010    2015
Energy
Coal, Australia               $/mt            n a    n a    39 67  26 25   32 31    26 50   26 00   27.00  29 50   32 00
Crude oil, average            $/bbl          121    36 87   22 88  28 23   24 35    25 00   23 00   19 00  19 00   21.00
Natiural gas, Europe          $/mmbtu         n a    3 40    2 55   3 86    4 06     3 00    2.80   2 60    2 75    3 00
Natural gas, U S              $Imnibtu    f  017     1 55    1 70   4 31    3 96     3 25    3 20   3 00    3 00    3 25
Nonenergy commodities
Agriculture
Ifeveragcs
Cocoa                      c/kg       1  67 5   260 4   126.7    90 6   106 9 |82 0     182 0   160 0   157.0   168 0
Coffee, other milds        c/kg       1 114 7   346 6   197 2   192 0   137.3   133 0   141 1   187.4   242.5  280 0
Coffee, robusta           c/kg           91 4   3243    1182     913     607     639     706     838    1100   142.6
Tea, aUction1s (3) average  c/kg          83 5  165 9   205 8   187 6   159 8    150 0  155 0   165.0   175.0   180.0
Food
Fats and oils
Coconut oil              $/mt          397 2  673 8   336 5   450.3   318 1   415 0   450 0   600 0   645 0   670.0
Copra                    $/mt          224 8  452.7   230 7   304 8   202 1   268.0   375 0   450 0   480.0   500 0
Groundnut oil            $/rit         378 6  858 8   963 7   713 7   680 3    680.0  750.0   820.0   850.0   875 0
Palm oil                 $/mt          2601   583.7   289 8   310 3   285 7    385 0  390 0   400.0   450 0   475.0
Soybean meal             $/mt          102.6  2624    2002    1892    181 0    175 0  2000    205.0   2150    2200
Soybean oil              $/mt          286 3  597 6   447 3   338 1   354 0    440 0  450 0   430.0   460 0   505 0
Soybeans                $S/mt          116 9  296 2   246 8   2118    195 8 i 210 0   230 0   235 0   240.0   250 0
Grains
Maize                    $/mt           58 4   125.3  109 3    88 5    89 6    100.0  125 0   115 0   120 0   130.0
Rice, Thailand, 5%       i/nt          126 3  410 7   270 9   202 4   172.8    192 0  210.0   235 0   260.0   265 0
Sorghum                   i/mt          51 8   128 9  103 9    88 0    95 2    102.0  125 0   116.6   119 5   128.0
Wheat, US, HRW           $/m         f  549    1727   1355    1141    1268     1515   1800    1500    160.0   165.0
Othcr food
Banianas, U S            $/mt          1661   377 3   540 9   424 0   583 3    530 0  518 1   529.1   568 0   590 0
Beef, U S                c/kg          130.4  276 0   256 3   193 2   212 9    215 0  230 0   228 0   222.0   230 0
Oranges                  $/mt          1680    4002   531 1   363.2   5955     5880   5500    5000    5250    550.0
Shrimp, Mexico           c/kg           is a   1,152  1,069   1,513   1,517 1 1,040    1,150  1,650   1,700   1,720
SLigar, world            c/kg            8 2   63 16  27 67   18 04   19 04    15.00   15 00  17.60   21 00   22 00
Agricultural raw materials
Timber
Logs, Camerooni          S/cumn         43 0   251 7  343 5   275 4   266 1    265 0   275 0  300.0   338.0   385 0
Logs, Malaysia           $/cumn         43 1   195 5   177.2  190 0   159.1    163 0   170.0  215.0   260 0   295 0
Sawniwood, Malaysia      $/cum         175 0   396 0  533 0   594 7   481 4    528.0   560.0  625 0   720 0   820 0
Other raw materials
Cotton                   c/kg           676    2062    1819   1302    1058     1000   I1102   127.9   149.9   1600
Rubber, RSS1, Malaysia   /kg            40 7   142 5   86 5    69 1    60.0     79 4   81.6    83 8    87.7    95 1
Tobacco                  $/mt          1,076   2,276  3,392   2,976   3,005    2,770   3,000  3,250   3,275   3,300
Fertilizers
DAP                        $/mt          54 0   222.2   171 4   154.2   147 7    158 0  168.0   170 0   175.0   180.0
Phosphate rock             $/mt          1100   4671    4050    4375    41.77   4080    41 00   43.00   4500   4600
PotassiUm chloride         S/mt          320    1157     98 1   1225    118 1    1130   120.0   124.0   1270    1300
TSIP                       $/mt          430     1803   1318    1377    1269     1330   140.0   150.0   1500    1550
Urea, East Europe, bagged  $/mt           n a     n a   119 3   101 1    95 3     93.0  108.6   126.7   131.3   135.8
Metals and minerals
Aluminum                   $/mt           556    1,456  1,639   1,549   1,444    1,340  1,400   1,500   1,600   1,700
Copper                     $/nmt      1 1,416   2,182   2,661   1,813   1,578    1,545  1,650   1,900   2,000   2,050
Gold                       $/toz         35 9   607 9   383 5   279 0   271 0   310.0   300 0   275.0   300 0   3000
Iron ore                   c/dmtu     '   9 84  28 09   32 50   28 79   30.03   29 50   30 00   31 00   32.00   32 50
Lad                       c/kg           303     906     81.1    454     476     45.0    480     550     600     62.5
Nickel                     S/mt         2,846   6,519   8,864   8,638   5,945   6,700   7,500   7,500   6,700   6,800
Silver                     c/toz      i 177 0   2,064   482 0   499 9   438 6   460 0   480.0   500 0   525 0   550 0
Tin                        c/kg          367 3   1,677  608 5   543 6   448 4   405.0   4500    525.0   540 0   550 0
Zinc                       c/kg           29 6   76 1   151 4   112.8    88 6     77.0   84.0   100.0   105 0   110 0
ti a = Not available
a $ = U S dollar, c = U S cent, bbl = barrel, cumo  cubic meter, dmtu = dry nietric ton init, kg = kilograin, mmhtu = million British thermal unit,
mit = metric ton, anid toz = troy ouince
Note Projections as of Novemiiber 12, 2002
Soturce World Bank, Developmenit Piospects Group
195



G LOB AL     EC ONO MI C     PROS PE CT S 2 0 0 3
Table A2.14    Commodity prices and price projections in constant 1990 dollars
Actual                                Projections
Commodity                       Unita       1970    1980   1990   2000    2001    2002    2003   2005    2010   2015
Energy
Coal, Australia              $/mt           n.a.    n.a   39.67  26 97   33 68   2748    26 18  26 06   26.53  27.00
Crude oil, average           S/bbl          4 31  46 80   22 88  29 01   25.38   25.92   23 16  18.34   17.09  17 72
Natural gas, Europe           $/mmbtu       n a    4 32    2.55   3 96    4 23    3.11    2 82   2.51    2.47   2 53
Natural gas, U.S.             $/mmbtu       0 61   1.97    1.70   4 43    4.12    3.37    3.22   2.90    2.70   2.74
Nonenergy Commodities
Agriculture
Beverages
Cocoa                     c/kg         240 6   330 5  126.7    93.1  111 4    188.7   183.2  154.5  1412    141 8
Coffee, other milds       c/kg         408.8   440.0  197 2   197 3  143.1    137.9   142.1  180.9  218 1   236 3
Coffee, robusta           c/kg         325 7   411 7  118 2    93 8   63 3     66 3   71 0    80.9   98.9   120.3
Tea, auctions (3) average  c/kg        297.7   210 6  205 8   192 8  166.6    155.5   156 1  159.3   157.4  151 9
Food
Fats and oils
Coconut oil             $/mt        1,416.0  855 3   336.5  462 7   331.5   430 3   453 0  579.2   580.1  565.4
Copra                   $/mt          801.6  574 7   230.7  313 1   210 6   277 9   377.5  434.4   431 7  421 9
Groundnut oil           $/mt        1,349.5 1,090 1  963 7  733 3   709 0   705.0   755.1  791 6   764.5  738.3
Palm oil                $/mt          927 1  740 9   289.8  318 8   297 7   399.2   392 6  386.1   404 8  400 8 1
Soybean meal            $/mt          365 7  333 1   200.2  194.4   188 6   1814    201.4  197.9   193 4  185 6
Soybean oil             $/mt        1,020 8  758 6   447 3  347.4   368 9   456.2   453.0  415.1   413 7  426.1
Soybeans                $/mt          416.8  376.0   246 8  217.7   204 1   217.7   231.6  226.9   215 9  211 0 1
Grains
Maize                   $/mt          208.2  159.0   109.3   91.0    93.4   103 7   125.8  111.0   107.9  109.7
Rice, Thailand, 5%      $/mt         450 3   521.4   270.9  208 0   180 1   199 1   211.4  226.9   233.9  2236
Sorghum                 $/mt          184.7  163.6   103.9   90.4    99 3   105.8   125.8  112 6   107.5  108.0
Wheat,US,HRW            $/mt          1957   2193    1355   1172    132.2   157.1   1812   1448    1439   1392
Other food
Bananas, U.S.           $/mt          592 1  478 9   540.9  435.7   607.9   549.5   521.6  510 8   510.9  497 9
Beef, U.S               c/kg          465 0  350.3   256.3  198 5   221 9   222.9   231 6  220.1   199.7   194.1
Oranges                 $/mt          S99.1  508.0   531.1  373.2   620.6   609.6   553 7  482.7   472.2  464.1
Shrimp, Mexico          c/kg           n.a.  1,462   1,069  1,554   1,581   1,078   1,158  1,593   1,529   1,451
Sugar, world            c/kg          29 32  80.17   27 67   18.5    19.8    15.6    15.1   17 0    18.9    18.6
Agricultural raw materials
limber
Logs, Cameroon          $/cum         153.3  319.5   343 5  283.0   277.3   274 8   276 9  289 6   304.0  324.9
Logs, Malaysia          $/cum         153.8  248.2   177.2  195.2   165.8   169.0   171.2  207.6   233.9  248 9
Sawnwood, Malaysia      $/cum         623 9  502 7   533 0  611.1   501.7   547.4   563.8  603 3   647 6  691 9
Other raw materials
Cotton                  c/kg          241 1  261.7   181.9  133.8   1103    103.7   111.0  1234    1348    1350
Rubber, RSSI, Malaysia  clkg          145.2  180.8    86.5   71 0    62.6    82.3    82 1   80.9    78.9   80.2
Tobacco                 $/mt          3,836  2,889   3,392  3,058   3,131   2,872   3,020  3,137   2,946  2,785
Fertilizers
DAP                       $/mt         192.5   282.1  171 4   158.5   154 0   163.8   169 1  164.1   157.4  151.9
Phosphate rock             /int         39.2    59.3   40 5    45 0   43.5     42.3   41.3    41 5   40 5    38.8
Potassium chloride        $/mt         114.1   1469    98 1   125.9   123.1   117.2   1208   119.7   114.2  1097
TSP                       $/mt          153.3  228.8   131.8  141.5   132 2   137.9   140.9  144.8   134.9  130.8
Urea, East Europe, bulk   $/mt           n a    n.a.  119.3   1039    99.3     96.4   1094   1223    118.1  114.6
Metals and minerals
Aluminum                  $1mt         1,982   1,848   1,639  1,592   1,505   1,389   1,409  1,448   1,439  1,434
Copper                    $/mt         5,047   2,770  2,661   1,863   1,645   1,602   1,661  1,834   1,799  1,730
Gold                      $/toz        128.1   771.6  383.5   286.7  282 4    321.4   302.0  265.5   269.8  253.1
Iron ore                  c/dmtu        35.1    35 7   32 S    29.6   31.3     30.6    30.2   29.9    28 8   27.4
Lead                      c/kg          108.0  115.0   81 1    46.6   49.6     46.7   48.3    53.1    54.0   52 7
Nickel                    $/mt        10,147   8,275  8,864   8,876   6,196   6,947   7,551  7,240   6,026  5,738
Silver                    c/toz        631.0 2,619 4  482.0   513.7  457.1    476.9   483 2  482.7   472.2  464 1
Tin                       c/kg        1,309.6 2,129.3  608.5  558.5   467.4   419.9   453.0  506.8   485 7  464.1
Zinc                      c/kg         105.5    96.6   151.4  115.9   92 3     79.8    84.6   96.5    94.4   92 8
n.a. = Not available.
a. $ = U.S. dollar, c = U S cent, bbl = barrel, cum = cubic meter, dmtu  dry metric ton unit, kg = kilogram, mmbtu = million British thermal unit,
mt = metric ton, and toz = troy ounce.
Note: Projections as of November 12, 2002.
Souirce: World Bank, Development Prospects Group
196



G L O B A L C OIM IM O D I T Y P R I C E P R O S P E C T S
Table A2.15     Weighted indices of commodity prices and inflation (1990 = 100)
Actual                                          Prolectionsa
Index                            1970      1980     1990      200(0      2001  1    2002     2003      2005     2010      2015
Current dollars
1'etroleumli                      5 3     161 2    100 0     123.4      106 4      109 3    100 5      83 0     83 0      918
Noncinergy commoditiesh          43 8     125 5    100 0      86.9       79 0       82 9     87 7      94 2    102 7     109 9
Agriculture                    45 8     138 1    100.0      87.7       79 7       86 5     91 7      98 0    108 9     118 0
Beverages                    56.9     181.4    100 0      88 4       72.1       84 4      87 5     97 8     1151     130 6
Food                         46 7     139.3    100 0      84 5       86 0       89 8     96 3      97 8     104 3    108 2
Fats and oils              64 4     148 7    100 0      96 2       89.0      100 2     108 2    1131      120 9    126 1
Grains                     46.7     134 3     100 0     79 5       78 2       89 0     104 5     99 4     106.6    111 1
Other food                 32 2     13-1 3   100 0      77 7       87.9       819       819      84 4      89 4     92 0
Rav materials                36.4     104.6    1000       91.4       774        836       888      984      1102     1212
Timber                     318       79 0    1000      111.0       902        98 1     1039     1178      1366     1555
Other raw nmaterials       39 6     122 0     100 0     78.0       68.6       73 7      78 5     85 2      92 2     97 8
Fertilizers                    30 4     128 9    100 0     105 8       98 8  I   102 0    104 4     111 0    112 8     116 1
,Metals and minierals          40 4      94 2    100 0      83 0       75 1       72 4     76 5      83 2     86 4      89 5
Conistant 1990 dollars'
Iletroleunm                      189     204.6     1000      1268       1109       1133    1012       802      747       775
Noinenergy comniodities         156.3     159 3    100 0      89 3       82 3       86 0     88 3      90 9     92 4      92 7
Agricilturc                   163 3     173 3    100 0      90.1       83 1       89 6      92 3     94 6      97 9     99 5
Beverages                   2028      2303      100.0     908        751        875       881      944      1035     1102
Food                        166 5     176.8    100 0      86 8       89 6       93 1      96 9     94 5     93 8      91 3
Fars and oils             229 5     188 7     100.0     98 9       92.8      103 8     109 0    109 2     108 8    106 4
Graiis                    1666      1705     1000       817        815        923      1052      960       959      938
Other food                 114 9    170 5     100 0     79 9       916        84 9      82 4     81 5      80 4     77 6
Rawv materials              129 8     132 7    100 0      93 9       80 6       86 7      89 4     95 0      99 1    102 3
Timber                    113.3     100.3     100.0    1141        94 0      101 7     104 6    113 7     122 9    1312
Other raw materials       141 1     154.9    1000       80.1       71 5       765       790      822       829      825
Fertili,ers                   108.3     163 6    100 0     108 7      102.9      10S 7     105 1    107 2     101 5     98 0
Metals anid min[erals         143 9     119 6    100.0      85 3       78.3       75 1      77 0     80 3      77 7     75 5
Inflatioi indicesd
MUVindexe                       2805      7878    100.00     9732       9595       9645     9933     10359     111 18   11851
Percentage chanige per annum              10 88     2 41    -0 27       -1 40       0 53     2 98      2 12      1 42     1 29
US GDI' deflator                33.59     65.93   100 00    123.73     126 42  I 127 69     129.73   136 03    153 01   172 27
Percentage change per annLum               6 98     4.25      215        218        1 00      1 60     2 40     2 38      2 40
a Commodity price projections as of November 12, 2002
b 'IThe Wrorld Bank primiiary commodity price indices are compuited front 1987-89 export values in U S dollars for low- and middle-incomiie econoinies.
rehased to 1990 Weights for the subgroup indices expressed as ratios to the Tioneniergy index are as follows agriculture-69 1 percelnt, fertilizers-
2 7 percent, anid nmetals anld mincrals-28.2 percenit, bevcrages-16 9 percent, food-29 4 percent, and rawv materials-22 8 percent, fats and oils-
10 I percent, graiiis-6 9 percent, anid other food-12 4 percent, tinber-9 3 percent and other raw inaterials-13 6 percent
c Computed from unirounded data and deflated bv the imialnLufactuires unlit value (MUV) index
d Inflation indices for 2002-15 are proecticons as of November 8, 2002 MUV for 2001 is an estimate Growth rates for 1980, 1990, 2000, 2005,
2010, and 2015 refer to compound annuial rate of change between adjacent endpoint years, all others are annual growth rates from the previous year
e Unit vaLie index in U S dollar terimis of manufactures exported from the G-5 countries (France, Germany, Japan, the Ulnited Kingdonm, and the Uniited
States) weighted proportionially to the countries' exports to des eloping cointries
Souirce World Bank, Developmenit Prcispects GroLIp U S Departiment of Commerce for historical U S GDP deflator
197



G LO BAL EC ONO MI C        PROS P ECTS      2 0 0 3
Notes                                             Centigrade warmer in 1997 Thus the effects of this
1. As measured relative to the manufactures unit  year's El Nifio are expected to be smaller than in 1997,
value (MUV) index, which is the unit value index in U.S.  when drought in Southeast Asia led to wildfires and
dollar terms (1990 = 100) of manufactures exported  poor crop harvests.
from the G-S countries (France, Germany, Japan, the  3. The European Central Bank and 14 European
United Kingdom, and the United States) weighted by the  central banks agreed in September 1999 to limit sales
country's exports to developing countries.       to only 400 tons of gold per year, and not more than
2. An El Niflo occurs when the Pacific Ocean   2,000 tons in total, over the subsequent five years.
warms, as occurred this year. But this year's El Niiio is  4. Verleger, Philip K. Jr The Petroleum Economics
significantly weaker than the last one, which occurred  Monthly. August 2002, p. 11, and September 2002,
in 1997. The Pacific is about 1 degree Centigrade  P. 1.
warmer than usual this year compared with 3 degrees
198



Appendix 3
Global Economic Indicators



G LO BA L     EC ONO MI C       PRO SP E CT S      2 0 0 3
Table A3.1 Growth of real GDP, 1971-2015
GDP in 199S prices and exchange rates, average annual growth (percentt)
GDP in 2001          Growth percent
(current billions                                    Esnmate   Forecast
of dollars)  1971-80  1981-90  1991-2000   2001     2002     2003-15
World                                  30,790       3.7      3.0        2.6       1.1       1.7      3 1
High-income economies                  24,852       3.5      3.1        2.5       0.7       1.5       2.6
Industrial countries                 24,088       3 4      3 1        2 4       0.8       1 4       2.5
G-7 countries                      20,632       3.4       3 1       2 3       0 5       1.3       2 5
United States                     10,082      3.3       3 2       3 2       0.3       2 3       3 1
Japan                              4,166      4 5       4 1       1 3      -0.3       0 0       1 6
G-4 Europe                         5,678      2 9       2 4       18        L.4       0 7       2 2
Germanva                         1,856      2 7       2 2       1 7       0.7       0.4       1.9
Euro area                            6,090      3 2       2 3       2.0       1 5       0.8       2.3
Non-G-7 industrial                   3,456      3.2       2 9       3 0       2.0       2 2       3 0
Other high income                       764       7 7      5 1        5 1     -07         2 3       4 3
AsianNIEs                             531       95        74        61       -1.4       26        47
Low- and middle-income economies        5,938       4.8      2.6        3.2       2.9       2.8       4.6
Excluding CE Eur / CIS                5,101       5 5       3 0       4 7       3 0       2 7       4 7
Asia                                 2,205      5 2       6 8       7 0       5.2       5.8       6 0
East Asia and Pacific              1,573      6 6       7 3       7 7       5 5       6 3       6 2
China                            1,150      6 2       9.3      10 1       7.3       7 8
Indonesia                         145       7 9       6 4       4 2       3 3       3 2
South Asia                          632       3 1       5 7       5 2       4 4       4.6       5 4
India                             495       3 0       5 8       5 6       4.5       4.8
Latin America and the Caribbean      1,882      5 9       1 1       3 3       0 4     -1.1        3 6
Brazil                              503       8 5       1 5       2 7       1 5       0 7
Mexico                              618       6 7       1 8       3.5      -03        1.3
Argentina                           269       3 0     -1 5        4 5      -4.4     -11 9
Europe and Central Asia               977       3 5       1 7     -1.7        2 3       3.6       3 6
Russian Federationb                 310       3.7       1 5     -4 0        5 0       4 3
Turkey                              148       4 1       5 2       3 5      -7 4       4.1
Poland                              176       5 1     -1 7        3 7       1 0       1 0
Middle East and North Africa          568       6.5       2 5       3.2       3 2       2 5       3 3
Saudi Arabia                        188      10 3       0.4       2 2       1 2       1 1
Iran, Islamic Rep of                 114      1.8       2 7       4 1       4.8       4 5
Egypt, Arab Rep of                   96       6.6       5 5       4 4       2 9       1 0
Sub-Saharan Africa                    306       3 3       1 7       2 2       2 9       2 5       3 7
South Africa                        113       3 5       1 3       1 7       2 2       2.2
Nigeria                              41       4 7       1 1       2 6       4.0      -0 6
a Data prior to 1991 covers West Germany
b Data prior to 1992 covers former Soviet Union
Note Growth rates over intervals are computed using compound average methods
Souirce World Bank data and staff estimates
Figure A3.1 Real GDP growth
(percent)
100
80
1991-2000       2003-15
60
40 
-20
-40
High-income  East Asia   South    Latin America  Eastern  Middle East  Sub-
countries  and Pacific   Asia      and the   Europe and  and North   Saharan
Canbbean   Central Asia  Africa     Africa
Source World Baiik data and staff estimates
200



G LO BAL     EC ONO MI C     I ND ICA rOR S
Table A3.2 Growth of real per capita GDP, 1971 -2015
GDP in 1995 prices anid exchange rates, average atnnial growth (percent)
capita   ~Growthi percent
dolr) 1971-80  1981-90  1991-2000           i      0321
World                ~i~l                      1.8       1.3      1.2                14        i
High-income economies'                 3       2.6       2.5      1.8       14       ho1      oI
Industrial countries                         2 6       2 5      1 8    *0A         1.1 
G-7 countries                               2 7      2 5       1.7      111      14    
United States                             2 2      2 2       2 2     -1         l~~
Japan                                     3.3       3 5      1 1     -04      '1L        13
G-4 Europe          ~I$1                  2 6      2 1       15        FI      ®         .4
Germanyb                                2 6       2.0      1 4      l1ll     113   ~    4
Euro area           ~i11.                   2 7      2.1       1 7      I1       lI13       l
Non-G-7 industrial               ll~11      2 2      2 2      2 4       14       11)~      4
Other highi income                 1lI~      5 1       3 3      3.6               11.1   ~~
AsaianNIEs                       11i        7 2      5 9      4 7       i113               l4
Low- and middle-income economies     1j1       2.6       0.7      1.6      14        13   ~    4
Excluding CE Eur /CIS              ll9l1l.   3 1       0 9      2.9       13        14      o¶
Asia                                W       3 0      4 8      5 4    ~   4        13   ~   I
East Asia and Pacific            1        4 6      5 6       6 4      414     .941       34
China                          ¶1~      4 3       7.7      9 0    W65                  -
Indonesia                      13.      5 4       4 4      2 5       Xl       14
South Asia                       43       0 7      3 4       3 3       3       14)       .11
India                           iI)     0 7       3 6      3 7       3)      £4        o
Latin America and the Carihbean    6'       3 3     -0.9       1 6     111'12      ~      14 
Brazil                          l1?       5 9     -0 4       1 2      11u       1
Mexico                          6,=31     3.6     -0.3       1 7    ___
Argentiiia                      ?1l 3      3      -2 9       3 2    ___       1
Europe and Central Asia          131f1      2 5      0.7     -1 9                .14      14
Russian Federation'              3J~      3 1      0 8     -3 9        3        14       -
Turkey                         13¶®       1 7      2 8       1 9    08!7        3     
Poland                          43        4 2     -2 4       3.5      14        !4
Middle East and North Africa      1   ~     3 6     -0 6       1 0                        1
Saudi Arabia                              5 1     -4 8     -1 2               -113    
Iran, Islanmic Rep of                 --1 4       -0 7       2 4               &3 £1
Egypt, Arab Rep of                    -4.4         2 9       2.4      1143
Sub-Saharan Africa                 414      0 5     -1 2     -0 4       1        go3(i~1  1
South Africa                    1321      12      -1 2     -0.3      (LIP?     1111
Nigeria                          M311     1 7     -1 9     -0.2       113       3J
a Regional aggregates compuited as surn(GDP,)/surn(POP,), where "i" indicates country, in the region, and are unweighted by
population or other measures
b Data prior to 1991 covers West Germany
c Data prior to 1992 covers former Soviet Uniioni
Note Growth rates oiver intervals are computed using compound annual methods
Source World Bank data and staff estimTates
Figure A3.2 Real per capita GDP growth
(percent)
7 0                                               2003-15
6 0
4 0 _____                          1991-2000
2 0
10=F
0 0
-1 0
-2 0
-3 0
High-income  East Asia  South  Latin Amenca  Eastem  Middle East  Sub-
countries  and Pacific  Asia    and the  Europe and  and North  Saharan
Canbbean  Central Asia  Africa   Africa
Source Wosrld Batik data and staff estimates
201



G LOB AL      EC ONO MI C      PROS PE CT S       2 0 0 3
Table A3.3 Inflation: GDP deflators, 1971-2015
Deflators in local currency units, 1995=100, percentage change'
Growth percent
Growth percent                           Estimate   Forecast
1971-80    1981-90      1991-00        2001        2002     2003-15
World                                    9.0        5.8         3.7           2.3         1.7        1.9
High-income economies                   8.8         5.3         2.0           1.5         1.0        1.3
Industrial countries                   8.7        4.6         2 0           1.5         1.1        1 3
G-7 countries                        8 3        4.2         1 7           1 2         0 9        1 1
United States                      7.0        4 3         2.1           2.4         1.1        1.4
Japan                              7.8        2.0         0.1          -1.2        -0 9        0 0
G-4 Europe                         9.9        5.7         2.6           1.8          1.9       1.7
Germanyb                         5 3        2.6          2.6          1.4          1 4       1.2
Euro area                            9.6        6.1         2.8           2.3         2.1        1 6
Non-G-7 industrial                  11.1        7.1         3 3           2.9         2.1        1.9
Other high income                     19.3       33.1         3 8           0 4        -0 4        1.9
Asian NIEs                           9.5        4 7         2 4          -0.4        -1 0        1.5
Low- and middle-income economies        9.6         8.3        11.7           5.8         4.4        4.2
Excluding CE Eur / CIS                11.4       10.0         9 2           5 7         4 3        4 1
Asia                                10.5        7.2         6.9           6.3          3 4       4 5
East Asia and Pacific              9.6        5 5         5.6           6.6          2 5       3 4
China                            0.9        5.4         6.3           0.0          0.4        ...
Indonesia                       20.6        8 8         15 0         11.4         11 4
South Asia                        11 9        9 0         7.9           6.1          5 0       5.5
India                            8 9        8.5         8.1           4.0          2 8        ...
Latin America and the Caribbean     14.6       19.3        12.5           6.9         5 0        4 1
Brazil                            39.7      330.8       206 1           7.3          7.9
Mexico                            18.1       63.7        18.1           5 4         3.3
Argentina                        117 7      439 5        10.2          -1.1        36.5
Europe and Central Asia              0.3        2 4        50.3           5.9         4 7        4 0
Russian Federation'                0 3        2 3       104 5          18.0         5.8
Turkey                            32.8       46 6        71.7          47.2         27 9
Poland                             4 4       72 1        24.1           1.7          3.8
Middle East and North Africa        11 7        8 7         6.0           4.2         4 1        4 0
Saudi Arabia                      23 8       -3.1         2.9           7.0          4.8
Iran, Islamic Rep. of             20 2       15 6        25.6           8 6          5.1
Egypt, Arab Rep. of               11 0       13 1         8.6           4 5          4.1
Sub-Saharan Africa                  10 4        9.4         9 8           5.4         4.3        4.0
South Africa                      13.3       15 1         9.8           7 1         12 0
Nigeria                           13.4       16.6        28 6           5 9         4.5         ...
a High-income group inflation rates are GDP-weighted averages of local currency inflation, LMIC groups are medians, world is
GDP-weighted average of the two groups
b. Data prior to 1991 covers West Germany.
c Data prior to 1992 covers former Soviet Union.
Note. Growth rates over intervals are computed using compound annual methods.
Source. World Bank data and staff estimates
Figure A3.3 GDP inflation
(percent)
50 3
140
120
10   0                                         -        _ _ _  _ _ _  _ _ _  _ _
80 011991-2000                                -
4 0                   __7
20-                      -                        m
High-income  East Asia   South    Labn Amenca  Eastem    Middle East   Sub-
countries  and Pacific  Asia      and the   Europe and  and North   Saharan
Canbbean   Central Asia  Afnca     Africa
Source World Bank data and staff estimates
202



G LO BA L     E CON OM IC         IND I CATO RS
Table A3.4 Current account balances, 1971-2015
Expressed as shares of GDP (percent)
Cufrrent Acct         Sbares percent     * 
2001 (billions                                         Estimate   Forecast
of dollars)  1971-80  1981-90   1991-2000    2001      2002     2003-15
Worlda                                     -187      -0.1      -0.5       -0.4      -0.6       -0.7      -0.6
High-income cconomies                      -190      -0.1      -0.2        0.0      -0.8       -0.9      -0.7
Industrial countries                     -247      -0.3      -0.5       -0.1      - 1.0      - 1.1     -0.9
G-7 countries                         -289       -0.1      -0.4       -0.3      -1.4       -1.6      -1.3
United States                        -393       0.0      -1.9       -1.8      -3.9       -4.8      -3.3
Japan                                  89       0.3        2.3       2.5        2.1        2.8       2.3
G-4 Europe                             -4       0.1        0.3     -0.1       -0.1        0.6        0.4
Germanyb                              2       0.5        2.4      -0.7        0.1        1.7       0.4
Euro area                                10      -0.1        0.4       0.3        0.2       0.8        1.0
Non-G-7 industrial                       42      -1.5      -0.6        0.8        1.2        1.3       1.3
Other high income                          57      12.3       10.3       4.0        7.7        6.6       3.0
Asian NlEs                               50       1.8        6.9       5.4        9.5       10.3       4.2
Low- and middle-income economies              4       0.0      -1.7       -1.6        0.1        0.4     -0.3
Excluding CE.Eur / FSU                     11       0.2      -1.9       -1.6        0.2        0.6     -0.1
Asia                                       42      -0.7      -1.6       -0.1        1.9        1.9       1.6
East Asia and Pacific                    42      -0.8      -1.4        0.5        2.7       2.7        1.8
China                                  17       0.1        0.2        1.6       1.5        1.6       ...
Indonesia                               7      -2.3      -3.1       -0.4        4.8       2.6
South Asia        0                       0      -0.5      -2.0       -1.S        0.0      -0.1      -1.2
India                                    1      0.2      -1.7       -1.2        0.2        0.2       ...
Latin America and the Caribbean           -54      -2.8      -1.8       -2.8      -2.9       -1.5      -2.3
Brazil                                  -23      -4.4      -1.6      -2.2       -4.6       -2.8        ...
Mexico                                  -18     -3.9      -0.8       -3.7      -2.9       -2.7
Argentina                                -4      -0.4      -2.2       -3.2      -1.7         9.4       ...
Europe and Central Asia                    -7      -0.8      -0.5      -2.3       -2.0       -2.4      -1.6
Russian Federationc                      36       2.1        3.5       4.7       11.7        8.0       ...
Turkey                                     3     -2.1      -1.3       -1.1        2.5      -0.6        ...
Poland                                   -7      -0.9      -1.4       -3.6      -4.4       -3.7        ...
Middle East and North Africa               29       9.5      -1.7       -2.1        5.2        3.3     -1.0
Saudi Arabia                              13     22.0      -7.2       -6.6        6.9        5.5       ...
Iran, Islamic Rep. of                     7      11.8      -0.4        1.9        7.1       4.3        ...
Egypt, Arab Rep. of                       0      -3.4      -3.4        1.5      -0.4       -0.2        ...
Sub-Saharan Africa                         -7      -1.8      -2.7       -2.1      -2.2       -3.0      -2.0
South Africa                              0      -1.3        0.4      -0.2      -0.2        0.3        ...
Nigeria                                   0        1.S     -0.7       -0.4        0.4      -4.8        ...
a. CtLrrent account as defined in Balance of Payments (BOP) version 5.0, world represents statistical discrepancy; shares over inter-
vals are period averages.
b. Data prior to 1991 covers West Germnany.
c. Data prior to 1992 covers former Soviet Union.
Source: World Bank data and staff estimates.
Figure A3.4 Current account balance-to-GDP ratio
(percent of GDP)
3.0
2.0                                     1991-2000                    2003-15
1.0
0.0 r      L 
-2.0
-3.0
-4.0
High-income  East Asia     South    Latin America  Eastem   Middle East    Sub-
countries  and Pacific    Asia       and the   Europe and   and North   Saharan
Caribbean  Central Asia   Africa      Africa
Source: World Bank data atid staff estimates.
203



G LO B AL    EC ONO MI C      PROS PE CT S      2 0 0 3
Table A3.5 Exports of goods, 2001
Merchandise exports (FOB), millions of dollars, average annual growth rate 1992-2001 (percent); effective market growth (EMG) 1992-2001 (percent)
Exports  Growth   EMGb                I Exports   Growth  EMGb                | Exports   Growth   EMGb
World          6,056,976   7.0    10.6  Europe and                               Sub-Saharan
Central Asia (cgntinued)                Afnca Iconinued)
All developing                          Belarus           7,421    9.6    10 4   Sudan        I    1 797   14 4    7.3
countries     1,509,357   8.2     9.4  Bulgaria          4,872    2.2     9.6   Zambia            '789   -0 5    13.1
Czech Republic   33,690    8 7     9.0  Zimbabwe          2,063    7.8     9 2
Asia            574,958   11.3     8.8  Estonia           3,347   35.1     8 6
East Asia       509,540   11.7     8.5  Georgia             505   15.0     8.0   High-income
China           266,322   16 0     8 3  Hungary          28,244    8.7     9.1   countries     4,547,619   6.7    10.9
China  266,322  16 0     8 3  Kazakstan        10,569   44 8     9.8
Fij i              483     4.3     6 1  Kyrgyz Republic     548   10.4    19.5  Industrial
Indonesia        56,321    8 2     8 5  Latvia            2,091   13 6     9.6   Idsra
Malaysia         87,980    8 7     8.6  Lithuania         4,706   22 1     9.5    countries    4,010,686    6.7    11.1
Myanmar              .                  Poland           42,674    10 8    9.4
Papua New                                Romania         11,385   10 3     9 8   G-7 countries  2,793,252   5.7    11.3
Guinea           1,813    3 7     6.5   Russian                                 Canada          267,706    7 9     8.3
Philippines      31,242   12 3     7.9   Federation      104,501   4 3    10 9   France          323 886    6 5    16.1
Thailand         65,379    8 1     88    Slovak Republic  12,534   104    104    Germany         543:416    63     136
Vietnam          15,470   19 2     8 1  Tajikistan          858   48 6    16.2   Italy        | 264,390     4 9     9 9
South Asia       65,418    8.5    11.3  Turkmenistan       1,619  44 4     8.3   Japan           384,482    27      9.1
Bangladesh        5,790   13 1    11.2  Turkey           28,121    9.3     9 1   United Kingdom  276,076    5.8    12.5
India            43,268    9.6    11.3  Ukraine           17,319   10 1    9.7   United States   733,297    6.6     9.8
Nepal               698    8 1     89   Uzbekistan        3,428   48 0    15.0
Pakistan          9,665    3 0    10 1                                           Other industrial 1,217,435  6.7    7.8
Sri Lanka         5,998    8 5    13 4  Middle East and                          Australia        63,759    6 7     7.8
N. Africa      164,753    2.2     9.0   Austria          63,459    7 7    8 5
Latin America   351,608    8.2     9.7  Algeria          19,567    0.8    10 5   Belgium'        178,698   40 4     6.1
Argentina        26,670    7.1     9 0  Egypt, Arab                              Denmark          50,912    1.9     8.2
Bolivia           1,196    7.1    12.5   Rep of           6,830    3 6    10.2   Finland          43,006    71     10 0
Brazil           58,223    5.3    11.9  Iran, Islamic  I                         Greece           10,615    43      81
Chile            18,505    6 7    10.2   Rep of          24,517    2 1    10.1  Iceland            2,035    09      7.8
Colombia         13,281    23     110   Jordan            2,192    86      6.9  Ireland           83,242   123     10.9
Costa Rica        5,709   103     14.7  Morocco           7,139    1 7     9.5   Korea, Rep of   150,494   136      8.3
Dominican                                Oman             10,563   6 0     8 1   Netherlands     193,239    7 0    203
Republic         5,594   19 9    11.8   Saudi Arabia     78,342    1 5    8 1   New Zealand      13,918    3.5     7.9
Ecuador           4,923    4 6    10 3   Syrian Arab Rep   5,151   2.3     5 7   Norway           59,701    4 2     9 4
El Salvador       3,367   13 7    13.6  Tunisia           6,684    2 9    12.7   Portugal        114,427    8 8     9.6
Guatemala         2,975    5 6    10.0  Yemen, Rep of     3,769    9 9     8.5   Spain           114,427    8 8     9 6
Jamaica           1,520     1 0    6 8                                           Sweden           77,635    7 4    10 S
Mexico          158,449   13 2     8 3  Sub-Saharan                              Switzerland      86,497    3 8     9 0
Panama            5,919    2 8     8.9   Africa          92,057    3.3    12.5
Paraguay          2,251   -3.2    103   Angola            7,944    6 7    12 3   Other high-
Peru              7,518    3 3    10.4  Botswana                            ..,  income          536,932    7.0     9.1
Trinidad and                             C6te d'lvoire    3,741    6 4     9 0   Bahrain           6,260    5 0     7 4
Tobago           3,153    6 9     7.4   Cameroon          2,262    1.9    9 0   Brunei       I    3,156   -1.1     6 5
Uruguay           2,081    0 3    10.1  Ethiopia            396   13 9     6 9   Hong Kong,
Venezuela,                              Gabon             2,540   -1 1     7.9    China          189,842    7 2     9.9
R B de          25,928    5 4     8 3   Ghana            2,021    9 5    113    Israel           31,275    9 0    13 3
Kenya             1,712    4 9     7.7  Kuwait       1   17,968   29 0     7.5
Europe and                              Madagascar          796    9 6    11.5   Singapore       121.747    7.4     8 4
Central Asia   325,981    8.3    10.1   Nigeria          16,443    0 7    7 3   Taiwan, China   122,495    5.7     8.3
Armenia             327   -8 2    30.1  Senegal            1,024   3 6     6 5   United Arab
Azerbaijan        2,110    0 5     9.7  South Africa     30,198    2.8     6 1    Emirates        37,638    2 7     6 4
FOB is free on board
a Includes Luxembourg.
b Effective market growth (EMG) is a weighted average of import volume growth in the country's export markets
Source See Technical Notes
204



G LO BA L      E CON OM IC         IND I CATO RS
Figure A3.5a Merchandise exports as share of GDP, 2001
(penrcent)
40
30
20
01 
Industrial     Sub-       East Asia     South        Latin       Eastern       Middle
countries    Saharan     and Pacitic     Asia     America and  Europe and     East and
Atrica                             the Caribbean Central Asia  North Atrica
Source World Bank data
Figure A3.5b Annual growth rate of export volumes, 1992-2001
(percent)
15
10
World                             -            _            _
Industnal      Sub-       East Asia     South        Latin       Eastern       Middle
countnes     Saharan     and Pacific     Asia      America and  Europe and    East and
Afrca                              the Caribbean Central Asia  North Atrica
Source World Bank data
205



G LO B AL    EC ONO MI C      PRO SP EC TS       2 0 0 3
Table A3.6 Imports of goods, 2001
Merchandise imports (CIF), millions of dollars, average annual growth rate 1992-2001 (percent), merchandise imports share of GDP (percent)
| Imports   Growth   Share                 Imports   Growth   Share                 Imports   Growth   Share
World          6,079,466   6.2    18.1   Europe and                               Sub-Saharan
Central Asia (chntinued)                 Africa (continued)
All developing                           Czech Republic   36,746   14 1    67.0   South Africa     27,873    5.5    25 3
countries     1,375,828   6.2    19.7   Estonia           4,255   22.3    83.0   Sudan             1,439    2 4    11 5
Georgia           1,047   30.3    37.0   Zambia            1,018    2.8    27.9
Asia            502,434    9.2    22.7   Hungary          30,759   11 9    63 7   Zimbabwe            920   -5 9    10 2
East Asia       424,657    9.7    26.8   Kazakstan         9,322   25.2    46.9
EaChmAsia       4242,65   95.7    268    Kyrgyz Republic     552   -1 2    37.5   High-income
China           232,322    15.8   20 1   Latvia            3,351   14.7    46.9    countries     4,703,638   6.9    18.9
Fiji                503   -1 8    29.9   Lithuania         5,740   25 2    49 7
Indonesia        30,962    4 9    21.4   Moldova                                  idsra
Myanmaysa6,r5               68    7.     Poland           53,874   13 1    32.8    countries     4,188,176   6.9    17.4
Papua New                                Romania          14,124    10 1   35.6
Papua New                     ~~~~~~~Russian
Guinea             932   -3 2    31.5    Federation      65,387    4 1    21 1   G-7 countries  3,037,774    6.7   14.7
Philippines      28,496     8.7   39.9   Slovak Republic  13,978    10 6   65 8   Canada          226,372    6 7    32 1
Thailand         61,847     3 0   53.9   TaFikistan        1,290     1 0   22.0   France          307,429     5 5   23.5
Vietnam          15,059    17.0   45.8   TFYR                                     Germany         457,102     5 1   24.6
Macedonia                    ~~~~~Italy       243,055     3.9   22 3
South Asia       77,776    6.7    12.3       cedn          28                     Japan 3a24                  3     223
Bangladesh        8,601    9 9    17.1   Turkmenistan      2,073    35 4   348    Jaa21,9                      2     7
India            51,624     8 0    10.4  Turkey           41,460     6.4   29.7   United Kingdom  324,362     6 9   22 8
Nepal             1,047     7.1   20.8   Ukraine           15,959    5 0   42.2   United States  1,166,561    9 3    11.6
Pakistan          10,484    0.6   16.7   Uzbekistan        3,331    25 2   33 8
Si Lanka          6,020 74   2    33 2                                            Other industrial 1,150,401  7 3   33.3
Sri Lanka  ,020  72 332  Middle East                           Australia        61,856    7 0    17.2
Lattn America   3S2,347    10.2   18.8    and N. Africa  106,531     1.3    19.2  Austria          64,325     6 1    34 2
Argentina        19,100     9 7    7.1   Algeria          11,775     5 2   20.4   Belgium'         168,485   38 4   73 3
Bolivia           1,600    6 2     19.9  Egypt, Arab                              Denmark          43,956     3 3   27 2
Brazil           55,573    9 6    11 1    Rep of          15,016     3.4   16.1   Finland          30,341     3.8   24 9
Chile            16,412    7 6    24.7   Iran, Islamic                            Greece           29,684     6 0   25.5
Colombia         11,826    9 8    13.9    Rep. of         16,665   -5 0    16.1   Iceland           2,066     3 4   27 4
Costa Rica        S,995    12.9   37.3   Iraq                 ..     ...     .    Ireland          51,300     9.7   49.7
Dominican                                Jordan            4,061     5.5   46.9   Korea, Rep of    141,096    71     33.4
Republic         8,963    17 4   43 9   Morocco          10,273     3.9   28.3   Netherlands      174,344    6.9   45 8
Ecuador           4,674     6 7   34.3   Oman              4,735     3.8   24.6   New Zealand      12,447     5.3   25 8
El Salvador       5,055    13.8   36.2   Saudi Arabia     28,427     0 8   l5 1   Norway           33,682     2 4    20.0
Guatemala         4,672    10 0   23 9   Syrian Arab                              Portugal
Jamaica            2,906    5.9    37 3   Rcp              3,757     4.8   21.2   Spain            146,449    7 5    25.1
Mexico          168,440   J12.6   27.3   Tunisia           9,349     6.0   44.4   Sweden           62,331     S 4    29 7
Panama            6,890     3 4   67.6   Yemen, Rep of     2,473     1.9   26 6   Switzerland      89,251     3 2   36 3
Paraguay          2,793     3 4   36 4
Peru              7,408     6 9    14.5  Sub-Saharan                              Other high-
Trinidad and                              Africa          76,253     4.8   22.3    income         515,462     7.4    67.6
Tobago            3,216    9 5    38.2  Angola            2,477     9.2    27.1   Bahrain           4,929    1.8    83.3
Uruguay           2,971    6 0     15.0  Botswana          2,258    3.6    34 1   Brunei             1,427    1 9   48 4
Venezuela,                               C6oe d'lvoire     2,741   -0.2    28.3   Hong Kong,
R B de           16,677    4.4   14.1   Cameroon           1,359  -3.6    13.5    China           198,588    7.8   21.1
Ethiopia          1,019   -0.8    15 7   Israel           35,111    7 1    30.3
Europe and                               Gabon             1,034     3.5   18 8   Kuwait            6,764     3.0   19.8
Central Asia   265,272     2.5   30.6   Ghana             2,326     9.2   30.6   Qatar               .-
Armenia             841   -6 7    40 3   Kenya             3,279    0 3    27.9   Singapore       109,852     7 0   28.3
Azerbaijan         1,976    2 2   319    Madagascar          954     7.6   20 9   Taiwan, China    107,298    6 8   38.1
Belarus           8,149    6 0    26.0   Nigeria          10,598    2 7    25 7   United Arab
Bulgaria          5,771    3.3    46.1   Senegal            1,351    1.5   29.2    Emirates        39,769    11 1   82.6
CIF is cost insurance and freight.
a Includes Luxembourg
Source- See Technical Notes
206



G LOB AL       E CON OM IC        IND I CATO RS
Figure A3.6a Merchandise imports as share of GDP, 2001
(percent)
30
20                            A         -World
10i--U                                                                      U
Industrial    Sub-      East Asia     South        Latin      Eastern      Middle
countries   Saharan     and Pacific    Asia    America and  Europe and    East and
Africa                           the Caribbean Central Asia  North Africa
Sotirce World Bank data
Figure A3.6b Annual growth rate of import volumes, 1992-2001
(percent)
15
10                                                               World
5-
Industrial    Sub-      East Asia     South        Latin      Eastem       Middle
countries   Saharan     and Pacific    Asia    America and  Europe and    East and
Africa                           the Caribbean Central Asia  North Africa
Souirce World Bank data
207



G LO B AL     EC ONO MI C       PROS PE CT S        2 0 0 3
Table A3.7 Direction of merchandise trade, 2001
(percentage of tworld trade)
Low- and middle-income importers
Latin
High-income importers                                         Middle Amefica  All
Sub    East       Europe   East  and   low-
Other   All  Other   All      Saha-   Asia        and     and   the    and
United               midus- indus- high-  high-     ran   and   South Central North  Carib- middle-
Source of exports     States  EU-15  Japan  trial  rrial income income  Africa Pacific  Asia  Asia  Africa  bean mcome     World
High-income
economies            12.8   27.8    3.3    6.6  52.1   5.7    57.8      0.9    6.1    0.8   3.4    1.5    4.5  17.1        74.9
Industrial          11.0   26 4    2.4   6.3   47.3   4 4   51.8       0 8    3.7   0.5    3 3    1 4    4.3  13.9       65.7
United States       . .   2 6    1.0   3 4    7.4   1.1     8.5      0 1    0 8    0 1   0 2    0.2    2 8   4.2        128
EU-15              3.7   20.4    07    2.2   272    1.4    286       05     0.9    02    29     08     09    62        348
Japan              2.2    t.1    ...    03    4.1    12     53       0.1    12     01    01     01     03     1 8        71
Other industrial   4 5    2 0    0 4   0.3    73    03      7.6      0 0    0.3    0 1   0.1    0 1    0 1   0.8         8.4
Other high-income    1.9    1 4    0.9   0 3    4.8   1.2     6 0      0 1    2 4    0.3   0 1    0.1    0.2   3 2         9.2
Low- and muddle-
income economies      6.7    6.1    2.1    0.7  16.4    2.8   19.2       0.5   1.4    0.4   1.8    0.5    1.3    5.9       25.1
Sub-Saharan Africa   0.3    0.5    0.0   0.0    0 9   0 1     1.0      0 2    0 1    0 0   0 0    0.0    0 1   0 4         1 4
East Asia and Pacific  1.9  1 3    1.5   0 3    5.5   2.0     7.5      0.1    0 7    0.2   0 2    0.1    0 2   1 6         9.1
SouthAsia            0.3    03     0.0   0.0    07    01      0.8      00     01     01    00     00     00    03          1 1
Europe and
Central Asia        0.3    0 8    0 3   0.0    1.7   0.3     2 0      0 1    0 3    0 1   0 1    0 1    0 0   0 6         2.6
Middle East and
North Africa        0.3    0 8    0.3   0 0    1.7   0.3     2.0      0 1    0 3    0 1   0 1    0 1    0 0   0 6         2 6
Latin America
andtheCaribbean     3.6    07     0.1    02    46    01      47       0.0    0.1    00    0.1    01     09     12         5.9
World                 19.5   33.9    5.5    7.3  68.5    8 5   77.0       1.4   7.5    1.2   5.2    2.0    5.8   23.0      100.0
EU is European Union
a. Expressed as a share (percent) of total world exports. World merchandise exports in 2001 amounted to some $6,000 billion
b Other high-income group includes the Asian newly industrializing economies, several oil exporters in the Gulf region, and Israel.
Souirce International Monetary Fund, Direction of Trade Statistics
208



G L C B A L E C O N O bt I C I N D I C A T O R S
Table A3.8 Growth of current dollar merchandise trade, by direction 1992-2001
(average annual percentage growuth)
F                   F                       ~~~~~~~~~~~~~~~~Lowv- and n1dl-lnc.eome,Iportr5r   } 
*   fEDBEL      |                                       Ml~~~~~~~~~~~~~~~fddle AtoerLe   AlliZ
Sub   East        Europe  East  and   loss -
~~fIi ~~  A~~iD  Salia-  As,a       and    and    the   aild
@ilnited           ais^hb f xfi                   ran   and   South Central North  Canb- middle-
Souirce of exports   Stae        ftis    D       OM                   Afnca l'actfic  Asia  Asta  Africa  beatt tnconic  WI'd
High-income
cconomies            DS7           g a   4I     9                     1.7    9.1   5.2   8.2    0.4    8 7   7.0       4.
Industrial               D  A     LT    43          4L     L3BL      1 6    8 6   2 4   8 1    0 3   8 7    66 
United States            .J    1DA4   *0     43    ga     43       3 8    9 4   4 0   3 0    0 4    9 9   8 2
EU-IS            bril    1104    6                 "               1 6    7 6   1 1   9 3    0 1   7 4    59 
Japan                 lD                      k -  ,-W   D   1X8  -2 6    7 2  -1 0  -19    -2 5    3 2   43        2
Other industrial  _      &      hjU     6    S0    &A      2       2 7    6 0   5 8   1 4    3 4   6 1    46         7
Otherhigih-income  4       U      .10  ZLT 7A        '.T             26    100   121   115     16    80     94 4
Low- and middle-
incomc cconomiCs    W12'                              R      h3Ih    12.8   15 6  11.0   9.5    5 5   11 2  11.0
Sub-SaharallAfrica  @h2                i@ @d       AU       ?9B  -  124    221   179    90     89    168   146
East Asia anid P'acific           U t   WJ aL0      'PA D   AfT     16 4   16 4  14 6  12 2    8 2  22 0   15 1
South Asia                                 >   ' #.6 93 @dl         117    15S6  11 6  -2 3    4 8  24 3    8.4       L1--
Europe arid
CentralAsia       WP3                         LE Z3J  4    3       160    185    78   -27     34    0.0    75 5
Middle East and                                                                                                         I
NorthAfrica        ILj3   ^4           8ih f. 3.%   Ml       v     16 0   18 5   7 8  -27     34    00     75
Latin America
atidtheCaribbean EIFA6    a9 &lA 61 @                               77    131   100    83     57    103   100 
World                 f       l     il d<                     e .      4.3   10.1   7.0   8.6    1 4    9.2   7.9
LU is Euiropean Union
Note Growth rates are compound averages
Souirce Intcrnational Monetary Funid, Directoiz of trade' Statistics
209



G LO BA L    E CON OM IC      PRO SP EC TS       2 O 0 3
Table A3.9 Structure of long-term debt, 2000
Share of long-term debt (percent) concessional debt, nonconcessional debt at variable interest rates, nonconicessional debt at
fixed interest rates
Nonconcessional                                     Nonconcessional
Concessinnal  Variable  Fixed                       Concessionial  Vanable  Fixed
All developing                                       Europe and Central
countres               18.7       40.3     41.0      Asia (continued)  i
Bulgaria                  4 3       75 9     19.7
Asia                     28.5       38.3     33.2    Czech Republic            1.8       274      70 9
East Asia                20.1       44.9     35.0    Estonia                   1.6       23 6     74 8
China                    20 9       26 4     52 7    Georgia                  610         8 8     30 1
Indonesia                279        61.8     10.4    Hungary                   1.2       17.7     81 0
Korea, Rep. of            1.8       53.8     44.4    Kazakhstan                3.6       16.8     79 6
Malaysia                  6.1       60.7     33 2    Kyrgyz Republic          58.3       15 6     26.1
Myanmar                  79 6       10.4     10.0    Larvia                    5.5       60.4     34 1
Papua New                                            Lithuania                 3.2       21.0     75 8
Guinea                 36.1       12.3     51.6    Moldova                  19.3       45 0     35 6
Philippines              29.6       33 0     37 4    Poland                   12.1       56 0     31.9
Thailand                 14.9       52 0     33.1    Romania                   3 4       41 9     54 7
Vietnam                  68.2       16 6     15.3    Russian Federation  I     0 3       19.7     80 0
Slovak Republic           3.9       22.6     73.6
South Asia               S5.5       17.4     27.0    Talikistan               76.6        7.7     15.8
Bangladesh               97.9        00       2 1    Turkmenistan
India                    39.4       20.2     40.4    Turkey                    5.9       45.3    48.7
Nepal                    99.8        0.0      0 2    Ukraine                  23.3       35 0     41.6
Pakistan                 60.7       27.3     12.1    Uzbekistan               22 3       58 2     19.5
Sri Lanka                85 7        5 8      8.5
Middle East and
Latin America             4.7       56.1     39.2      N. Africa              36.9       28.8     34.3
Argentina                 1.5       45 5     53.0    Algeria                  13.3       49.8     36 9
Bolivia                  60.3       27.0     12 7    Egypt, Arab Rep of       84 5        6.2     9 3
Brazil                    1.1       76 1     22.8   Jordan                    56 1       30.2     13.7
Chile                     1 0       55 2     43.7    Morocco                  315        32 4     36 1
Colombia                  3.1       60.8     36 1    Oman                     16.7       30 6     52.7
Costa Rica               16.7       24.1     59 2    Syrian Arab Rep. of      93.0        0.0      7 0
Dominican Republic       42.7       33 7     23.6    Tunisia                  26.8       20.8     52 4
Ecuador                  16.7       29.9     53.4    Yemen, Rep of            95.9        1 8      2.3
El Salvador              38.9       34.4     26.7
Guatemala                40.3       30.1     29.6    Sub-Saharan Africa  j    47.5       11.2     41.4
Jamaica                  25 9       24.4     49.7    Angola                   22 4        9.9     676
Mexico                    0 8       44.5     54 7    Botswana                 64.2       10 2    25 6
Nicaragua                53.6       22.2     24 2    C6te d'lvoire            39.1       46 6     14.3
Panama                    5 6       45.1     49.3    Cameroon                 54.6       11.0     34 4
Paraguay                 34.4       44 0     21.7    Ethiopia                 90 3        0 2      9.5
Peru                     16.3       58 4     25.3    Gabon                    39.6        9 5     50 9
Trinidad and Tobago       0.6       39.2     60 3    Ghana                    82.2        4 6     13 2
Uruguay                   3.5       51 0     45.5    Kenya                    76.4        5 9     17 7
Venezuela, R. B de        0.2       60 3     39 5    Madagascar               66 9        5 0    28 1
Nigeria                   4.4        6 0     89 6
Europe and Central                                   Senegal                  86.8        9 S     3 7
Asia                    5.9       34.8     59.3    South Africa              0.0       20 2     79 8
Armenia                  71.7       18 0     10 3    Sudan                    50.0       17.9    32 2
Azerbaijan               50.4       23 4     26.3    Zambia                   79 3        6 1     14 7
Belarus                  12.0       59.6     28 3    Zimbabwe                 46.5       21 2    32 4
Note- Nonconcessional debt data are available only for countries which report to the World Bank's Debtor Reporting System For
aggregate figures, missing values are assumed to have the same average value as the available data.
Source. World Bank data, see Technical Notes
210



G LO BA L      EC ONO MI C         IND I CATO RS
Figure A3.9a Structure of long-term debt, by group, 2000
(percent)
80  =          1 ~Cncs;ional             =
80-
60
40
20
Severely         Moderately         Severely         Moderately          Other
indebted          indebted         indebted          indebted          countries
low-income        low-income      middle-income     middle-income
countries         countries         countries         countries
Source: World Banik data.
Figure A3.9b Structure of long-term debt, by region, 2000
(percent)
100
80         o       co+l ssion ilH
60 ~~~~~~~~~~~~~-                                 Variable me t
80
60
Sub-Saharan      East Asia     South Asia       Latin         Europe         Middle
Africa       and Pacific                  America and    and Central     East and
the Caribbean      Asia        North Africa
Source: World Banik data.
Figure A3.9c Top ten ratios of nonconcessional debt to GDP, 2000
(percent)
140
120 -
100-
80-
40 -
Source: World Bank data.
211



G LO BAL     EC ONO MI C       PRO SP ECTS        2 0 03
Table A3.10 Long-term net resource flows to developing countries, 2000
(millions of dollars)
Pnvate
Official
Total   Percent           D)ebt flo ss
'"ilhson $  COP     Trial     (net)    FDI    Portfolio  Total    ODA      Other
All developing countries  261,133   4.3   225,846    8,288   166,691  50,867    35,287   38,088  -2,801
Asia                     87,822     4.0    74,947 -14,686     55,223  34,411    12,874   10,918    1,956
East Asia                74.556     4.7    65,693 -18,721     52,130  32,285     8,863    6,649    2,214
China                    60,525     5 6    58,295   -2.30)2   38.399  22,198     2,230      992    1,238
Indonesia               -9,156    -6 0    -11.210   -7.039   -4.5S0     379      2,053    1,173     881
Korea, Rep of            13,875     3 0    13,215   -3,852     9,283   7,784      660      -73      733
Malaysia                  3.411     3 8     3,229    1,027     1,660    542        182       52     131
Myanmar                    244       .        188     -66       255       0         55       55       0
Papua New Guinea           335      9 6       128     -50       130      48        207      118      90
Philippines               2,401     3.2     2,459      140     2,029    290       -57       528    -585
Thailand                  -525    -0 4     -1,383   -5.793     3.366   1,044       858     1,129   -271
Vietnam                   1,790     5 7       581    -717      1,298      0      1,209    1,195       14
South Asia               13,265     2.2     9,254    4,035     3,093   2,126     4,011    4,269    -258
Bangladesh                1,207     2 6       269     - 14      280       3        938      931       7
India                     9,928     2 2     8.771    4,340     2.315   2,117     1,157     1,463   -306
Nepal                      237      4 3       -4       -8         4       0        240      240       0
Pakistan                   526      0 9       -53    -361       308       0        578      580        1 
Sri Lanka                  530      3 3       262       83      173       6        268      239      28
Latin America            99,315     4.9    97,304   12.839    75,088   9,378     2,010    3,245  -1,235
Argnina                  16,719     5 9    16,620    4,504    11,665    450        100    -222      321
Boia                      1,230    14.8       923      190      733       0        307      341     -34
Brazil                   43,934     7 4    45,672    7,877    32,779   5.016    -1,738      340  -2,078
Chile                     4,733     6 7     4,834    1, 141    3,675     1 8     -101         0    -101
Colomhia                  3,312     4 0     3,130      728     2.376     26        182      119      63
Costa Rica                 S73      3 6       610      201      409       0       -36      -31       -6
Dominican Repuhlic        1,103     5 6     1,142      115      953      74       -40      -22      -17
Ecuador                   1,172     8 6       904      194      710       0        268       98     170
El Salvador                467      3 5       338      153      185       0        129       56      73
Guatemala                  415      2 2       178     -52       230       0        238      173      65
Jamaira                     972    12 6       898      442      456       0         74       -8       81
Mexico                   11,035     1 9    11,536   -5.267    13,286   3,517     -502      -80     -422
Nicaragua                  797     38 5       395      141      254       0        401      429     -28
Panama                     946      9 4       947      344      603       0        -1      -15       14
Paraguay                     99     1 3      -16      -98        82       0        115       12     103
Peru                      2,291     4 3      1,553     668      680     205        738      613     125
Trinidad and Tohago        633      8 2       673       23      650       0       -40         0     -39
Uruguay                    719      3 6       574      276      298       0        145       -8     152
Venezuela, R B de         5,708     4 7     5,454      919     4.464     71        254       57     197
Europe and Central Asia  54,000     5 8    45,446   11,560    28,49S   5,391     8,553    8,138     416
Armenia                    270     14 1       159       19      140       0        III      119      -8
Azerhaijan                 305      5 8       175       45      130       0        130      168     -38
Belarus                     125     1 2       122       32       90       0          3       26     -23
Bulgaria                  1,363    11 4     1.114      107     1,002      5        249      337     -87
Czech Repuhlic            3,441     6 8     3,299   - 1,901    4,583    617        142      154     -12
Estonia                    514     10 3       485      126      387     -29         29       48     -19
Georgia                    207      6 8       155       24      131       0         52       64     -12
Hun ary                   1,643     3 6     1,721       29     1,692      0       -78        11     -89
Kazai hstan               1,979    10 8     1,900      650     1,250      0         80      111     -31
Kyrgyz Republic             112     8 2      -65      -62        -2       0        177      178      -1
Latvia                     669      9 4       581      176      407       0         86       50      36
Lithuania                  910      8 0       799     269       379     151        til       63      48
Moldova                    269    20 9        209       81      128       0         60       62      -2
Poland                   13,413     8.5    13,19.5   2,982     9,342    871       218       470    -251
Romania                   2,606     7 1     1,900      875     1.025      0       706       177     529
Russiani Federation       2,508     1 0     2,200   -1.589     2,714  1,075       308       661    -354
Slovak Republic           2.234    11 7     2,185      133     2,052      0         49       51      -3
Tajikistan                  134    135S        64       40       24       0         70       70       0
Turkmenlistan                .
Turkey                   1 2,21i7   6.1    11,416    7,733      982    2,701       801       37     763
Ukraine                    169      0 5       927      332      595       0      -759      -785      26
Uzbekistan                 303      2.2        18     -82       100       0       284       261      24
Middle East and N. Africa  1,470    0.2      1,074   -931      1,209    795       396     3,223  -2,827
Algeria                 -1,678    -3 1     -1,212   -1,226       10       4      -465      -81     -384
Egypt, Arab Rep of        2,312     2 3     1,967      114     1,235    619       345       541    -196
Iran ,Islamic Rep of    -2,253    -2 2      -610     -649        39       0     -1,643     -11   -1,632
Jordan                      807     9 6       455    -115       558      12        352      369     -17
Morocco                   -460    -1 4      -293     -449        10     147      -167        96    -263
Oman                        69      0 5        57       23       23      11         12       21      -9
Syrian Arab Rep             68      2.1       107      -4       ill       0       -40       -14     -26
Tu.nisa                   1,009     5 2       966      214      752       0         44      190    -146
Yemen, Rep of              -12    -0 1      -201         0     -201       0        189      210     -21
212



G LOB AL       EC ONO MI C        IND I CATO RS
Table A3.10 Long-term net resource flows to developing countries, 2000 (continued)
(millions of dollars)
Private
Official
Total   IPcrcent            Debt flowvs
iMillions S  GDP       Total     (net)      FDI    Portfolio    Total     ODA      Other
Sub-Saharan Africa          18,527     5.7       7,074     -494      6,676      893      11,453    12,563   -1,110
Angola                       1,407    15.9       1,206     -492       1,698       0        201        230     -29
Borswana                        11     0.2          27       -3         30        0        -16         10     -26
Cte d'lvoire                   56      0.5        -47      -159        106        6         103       237    -134
Camiieroon                    185      2.1        -21       -52         31        0        205        292     -87
Ethiopia                      587      9.2          42       -8         50        0        545        565     -20
Gaboni                        -24     -0.5         142       -8        150        0       -166          9    -175
Ghania                        483      9.7          71      -57        110       17        412        442     -29
Kenya                         374      3.6          53      -61        111        4        32-1       403     -82
Madagascar                    298      7.7          83         0        83        0        215        219       -3
Nigeria                       706       1.7        907     -177       1,082       2       -201        113    -314
Senegal                       349      8.0         106       -2        107        0        243        276     -33
Souta Africa                 2,957     2.3       2,736      911        961      864        221        219        3
Sudan                         563      5.0         392         0       392        0         171       173       -2
Zambia                        778     24.0         191       -9        200        0        587        620     -33
Zimbabwe                       108      1.5         29      -50         79        1          79       145     -66
FDI is foreign direct investment; ODA is official developmenit assistance.
Souirce: World Bank data; see Techniical Notes.
Figure A3.10a Distribution of long-term net resource flows, 2000
(percent)
100                                                       '
60-
40-
20 
Sub-Saharan     East Asia     South Asia       Latin        Europe       Middle East
Africa      and Pacific                 America and    and Central    and North
the Caribbean      Asia         Africa
Souirce: World Bank data.
Figure A3.10b Change in share of private long-term flows, 1990-2000
(percent)
80
60
40__
O2-0~                                      -             P              P             r-
Sub-Saharan     East Asia     South Asia       Latin         Europe      Middle East
Africa      and Pacitic                  America and    and Central    and North
the Caribbean      Asia          Africa
Sotirce: World Bank data.
21.3



Technical Notes
The principal sources for the data in this ap-  millions of current U.S. dollars, while growth
pendix are the World Bank's central databases  rates are based on constant price data, which
and several International Monetary Fund   are derived from current values deflated by
databases, combined with data sourced from  relevant price indices or unit value measures.
the OECD and from Oxford Economics Inc.   Effective market growth (EMG) in table A.3.5
(OEF), covering the industrial and other high-  is the export-weighted growth of each coun-
income economies. The cut-off date for data  try's trading partner imports.
updates'was November 15, 2002. Data revi-     Tables A3.7 and A3.8. The IMF's Direc-
sions and new releases since that time have not  tion of Trade database serves as the underly-
been incorporated in the tables. Regional ag-  ing source for the bilateral trade share- and
gregates are based on the classification of  growth information highlighted in these tables.
economies by income group and by region, fol-  Growth rates are compound annual averages,
lowing the Bank's standard definitions (see  and are computed from current U.S. dollar
country classification tables that follow).  measures of trade flows.
Debt and finance data (appendix ta-       Table A3.9. Long-term debt covers public
bles A3.9 and A3.10) cover the 137 countries  and publicly guaranteed debt but excludes
that report to the World Bank's Debtor Re-  IMF credits. Concessional debt is that with an
porting System (DRS), supplemented by data  original grant element of 25 percent or more.
for non-DRS countries, for which commercial  Nonconcessional variable interest-rate debt
market information has been utilized. Small  includes all public and publicly guaranteed
countries have generally been omitted from the  long-term debt with an original grant element
tables, but are included in the regional totals.  of less-than 25 percent, whose terms depend
Current price data are reported in U.S. dollars.  on movements in a key market interest rate.
This item conveys information about the bor-
Notes on tables                           rower's exposure to changes in international
Tables A3.1 through A3.4. Historic data  interest rates.
sourced from the databases noted above, while  Table A3.10. Long-term  net resource
projections are consistent with those high-  flows are the sum of net resource flows on
lighted in chapter 1 and appendix 1.      long-term debt (excluding IMF) plus non-debt-
Tables A3.5 and A3.6. Merchandise trade  creating flows. Foreign direct investment refers
data is sourced from combined IMF, World  to net inflows of investment from abroad.
Bank, OECD, and'OEF sources. Merchandise  Portfolio equity flows are the sum of country
exports and imports exclude trade in services.  funds, depository receipts and direct purchases
Imports are reported on a cost-insurance-and-  of shares by foreign investors.
freight basis. Trade values are expressed in
214