Middle-Income Countries: Development Challenges and Growing Global Role

Some observers have questioned the rationale for continued engagement by international development institutions in middle-income countries. But these countries face development challenges calling for continued partnership with the international development community. Moreover, partnership with these countries is increasingly important in global collective action. There has been much debate recently about the role of international development institutions such as the World Bank in middle-income countries. Some observers have suggested that middle-income countries have reached a stage in their economic development that calls into question the rationale for development institutions' continued engagement in these countries. But Fallon, Hon, Qureshi, and Ratha find that middle-income countries continue to face significant development challenges. The nature of these challenges varies substantially, but all of these countries face an agenda calling for continued partnership with the international development community. Middle-income countries still have high levels of poverty. They are home to more than three-quarters of the world's poor (those living on less than US$2 a day). Poverty is pervasive in some middle-income countries, while in others the problem is one of major concentrations of poverty in backward areas. And recent crises have revealed the fragility of some of the gains against poverty in these countries. On the policy front, some countries have made great strides in reform, but many lag considerably behind, and even among the advanced reformers the unfinished policy agenda is substantial. The countries' institutional capacity to manage reform varies greatly. So does their integration with the global economy. Many middle-income countries still have little access to international capital markets, and even those with better access must contend with volatility in private capital flows. Beyond the need to assist middle-income countries in addressing these challenges, the case for continued engagement by international development institutions derives from the increasing importance of these countries for a range of global public goods. With their growing role and integration in the global economy, partnership with middle-income countries is a key element of global collective action in such areas as reducing global poverty, maintaining international financial stability, improving global economic governance, protecting the global environmental commons, and fighting systemic health threats. This paper - a product of the Economic Policy Unit, Poverty Reduction and Economic Management Network; and the Development Prospects Group - is part of a larger effort in the Bank to analyze the evolving development agenda in the middle-income countries and its implications for the Bank's role.

Poverty in its various dimensions is still high in the MICs. Indeed, more than three-quarters of the world's poor (as measured by those living on less than US$2 a day) live in these countries. Some MICs continue to face pervasive poverty, while in others the problem is one of major concentrations of poverty in backward areas. Also, recent crises have shown the fragility of some of the gains in poverty reduction in these countries. On the policy front, while some countries have made great strides in reform, many lag considerably behind, and even amongst the advanced reformers there is still a substantial unfinished policy agenda. The institutional capacity to manage reform varies greatly. There is also considerable variation in terms of the integration of these countries with the global economy. Many MICs still do not have much access to international capital markets, but even those with better access must contend with volatility in private capital flows.
In addition to assisting the MICs in addressing these challenges, the case for continued engagement of international development institutions derives from the increasing importance of these countries in relation to a range of global public goods. With their rising role and integration in the global economy, partnership with MICs is a key element of global collective action in areas such as global poverty reduction, maintenance of international financial stability, improvement of global economic governance, protection of the global environmental commons, and the fight against systemic health threats.  Tables   TABLE 1.  Electronic copy available at: https://ssrn.com/abstract=632725 I. INTRODUCTION

1.
This paper presents a summary profile of the middle income countries (MICs) and outlines the main problems and challenges they face. The MICs are a heterogeneous group. All of them face significant development challenges, but the nature of these challenges varies substantially within the group. The key themes running through the paper are: • There is much that remains to be done in the MICs in terms of poverty alleviation. Some of these countries continue to face pervasive poverty, while others have large concentrations of poverty. Moreover, the hard-won gains in poverty reduction in many countries remain vulnerable to economic volatility, as recent experience with crises has shown.
• Progress on policy reform varies considerably, with many MICs still at a relatively early stage of reform. Even in the more advanced MICs, a substantial reform agenda remains unfinished. There is considerable disparity in countries' institutional capacity to design and implement reform.
• Progress on integration with the global economy varies greatly. Many MICs still do not have much access to international capital markets. Volatility in private capital flows is a problem even for those with better access.
• MICs are increasingly important in relation to a range of global public goods.
Growth, stability and associated reforms in the MICs have important public goods aspects as they entail benefits for the rest of the international community.
2. The paper is structured as follows. Section II documents the rising role of the MICs in the global economy and their increasing significance for a number of global public goods. Sections III to V assess the status of the MICs with respect to three important dimensions of development: poverty reduction; policy and institutional reform; and integration with the global economy. Section VI proposes a classification of the MICs that reflects the diversity within this broad group in terms of the progress made and the remaining agenda relating to these dimensions of development. Section VII summarizes the conclusions of the paper.

3.
For the purpose of this study, MICs are defined as those eligible for IBRD lending, including IBRD/IDA blend countries. There are 77 countries in this group, of which nine would fall in the low-income category as traditionally defined in the World Development Indicators (WDI). 1 The MICs are a heterogeneous group comprising a wide range of countries in terms of size, level of development and access to international capital markets. For example, MICs as defined here include Nigeria with a 1999 per capita GNP of $310 at one extreme and South Korea with a per capita GNP of $8,490 at the other. They also include Palau with a population of only 19,000 and China with over 1.2 billion people. In terms of access to international capital markets, the MICs include some investment grade countries such as Chile and Mexico, but the majority of countries in this group are rated below investment grade or simply not rated by major credit rating agencies.

4.
The MICs are becoming an increasingly important force in the world economy. Nearly 70 percent of the world's population lives in the MICs. The weight of the MICs in global economic activity has been rising. In the 1990s, these countries accounted for 19 percent, almost one-fifth, of world GDP ( Figure 1). According to Global Economic Prospects projections, their share will rise to more than one-fifth in the next decade. In purchasing power parity (PPP) terms, their share in world GDP is almost twice as high ( Figure 2). Their share in world exports has also been increasing, rising from 12 percent during the 1970s to 20 percent in the 1990s. These countries also receive a lion's share-96 percent in 1990-98-of net private capital flows to developing countries. This share in private flows remained steady during the debt crises of the 1980s and rose significantly dur ing the 1990s while their reliance on official flows declined ( Figure 3). Private capital flows to MICs rose sharply in the 1990s until the East Asian crisis (Figure 4).   1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

US$ billions
Source: DEC Development Group data.

5.
In 1970-1999, MICs consistently registered higher GDP growth rates than both high-and low-income countries ( Figure 5), and this trend is likely to continue. 2 Many of these countries, especially the fast-growing economies in East Asia, embraced trade liberalization and export promotion as a means to raise GDP growth rates. This is reflected in the rising ratio of exports to GDP in these countries ( Figure 6). The share of MICs in world population remained constant at 70 percent over the past two decades and is expected to stay at much the same value up to 2015. Population growth rates vary considerably, however, between regions.

6.
With 70 percent of the world's population and a 20 percent and rising share of world GDP, growth, stability and associated reforms in the MICs are public goods with considerable potential benefit for the rest of the international community. The MICs are of systemic importance in key areas with the characteristics of global public goods, including poverty reduction, financial stability, protection of regional and global environmental commons, public health, orderly flows of trade, movement of capital and labor across borders, and control of trans-border crime including money laundering and drug trafficking. Reform-minded MICs are producers of global public goods; reformaverse MICs may thus be said to be potential producers of global public bads.

7.
Poverty. A major reduction in poverty in these countries would contribute greatly to global peace and stability. Many among the better-off in developing and developed countries alike value an improvement in the living standards of those less fortunate than themselves. Since over 75 percent of the world's poor live in MICs, gains in poverty reduction in these countries through pro-poor growth would convey significant global externalities. In addition to contributing to global peace and stability, utilizing the untapped potential of the large number of the poor in these countries would expand global markets and demand.

8.
Economic and Financial Stability. Since MICs account for more than 90 percent of all capital flows to developing countries and around one-fifth of world GDP and trade, economic developments in these countries have potentially large implications for regional and international financial stability. The spillover effects of the East Asian crisis and the attendant international contagion provide a clear demonstration that the macroeconomic and financial policies of the MICs can have a major global public goods element. Spillovers from the Asian crisis not only affected MICs in other parts of the world (Russia, Brazil), but also impacted global markets. The crisis led, for example, to a downturn in growth, wages and employment in Latin America and elsewhere in 1997-98 through a downturn in external demand and capital inflows. It also had a negative effect on some low-income countries in that it led to real declines in key export prices, thereby compromising past development achievements.

9.
Many MICs have made great strides forward in terms of their economic and financial stability over the past decade or so. One can argue that the experience of economic difficulties in the 1980s has resulted in the global public good of convergence regarding the merits of macroeconomic stability and key elements of good practice in economic management. The 1990s have afforded further opportunities for learning from crises and have led to a degree of convergence with respect to good practice in financial and banking sector management. It is very important that MICs, with the support of the international community, continue to build on these achievements.

10.
Environment. The MICs have a large impact on the state of the world's forests, biodiversity, and to a lesser degree, greenhouse gas emission. For example, in 1996, China and India alone accounted for 20 percent of global carbon dioxide emissions. Measures to protect the environment raise investment costs considerably, and there is an expectation from both MICs and poor countries that they be compensated for conservation and abatement measures that are of global interest. Though progress has been made, there is a need for innovative financing and transfer solutions to foster environmental investment of global interest in MICs. The Global Environment Facility (GEF) offers a limited solution through funding of incremental costs on a grant basis and attracts cofinancing from different sources, including IFC and the private sector. The recent Forestry Policy Review by the Operations Evaluation Department (OED) of the World Bank makes a case for concessional funding for MIC investments in sustainable forest management. By developing a market for emissions certificates, the Clean Development Mechanism may eventually transfer significant resources to certain MICs. Still, there is a need for innovative thinking about how to use markets to help supply global environmental goods like carbon sequestration.

11.
Health. Systemic threats also emanate from MICs in the health field. For example, the spread of AIDS and multi-drug-resistant tuberculosis from such countries as Russia and China, and AIDS from Thailand, India, South Africa, and others, is significant. India and China account for 47 percent of world tuberculosis cases.
Fortunately, it appears that the MICs are more willing to take out official loans to address "spillover" diseases than environmental issues of global interest, as national benefit is perceived to be larger in the case of many health problems than it is in the environmental field. Russia and China, for example, have borrowed substantially from the World Bank to address tuberculosis and AIDS.

12.
Knowledge. Knowledge is a key public good. Timely creation and dissemination of information is crucial for development. The success of the MICs relative to other developing countries is explained in part by their greater effectiveness in acquiring and utilizing knowledge and best practice from other, more advanced parts of the world. Tapping the global knowledge base will remain crucial for the MICs as they strive for further economic progress. But the MICs themselves are now becoming an increasingly important source of knowledge for others. How they have coped with challenges of development and integration into the global economy offers useful lessons for the lowerincome developing countries, both in terms of best practice to emulate and pitfalls to avoid. There are thus important knowledge externalities associated with the work of the World Bank and other institutions in the MICs. The involvement of these institutions in the MICs not only provides value to the recipient country but adds to the global knowledge base, thus also benefiting other client countries. For example, for the World Bank as a global knowledge institution, experience and knowledge acquired through work in the MICs is important for it to be able to provide lessons of global experience to its lower-income clients.

13.
The MICs are also an increasingly important source of research and development whose benefits extend beyond their boundaries. Examples are Brazil and India in agricultural research, and India in information and communications technology. But on the whole, the R&D capacity of the MICs has much catching up to do, a challenge magnified by the widespread exodus of talent from many of these countries.

14.
Global Governance. Adoption of international standards, codes or best practices are an important element in the reform of the international financial architecture and global governance. Given the systemic importance of MICs for the global financial system, their participation in this effort is crucial to its success.

15.
Recognition of the MICs' rising systemic importance has been accompanied by calls for better representation of these countries in global forums. The creation of the G20, for financial matters, is a response to this call and represents recognition of the growing importance of some MICs. The question of burden-sharing in the provision of global public goods is linked to the issue of representation. One contribution by MICs is their agreement that the multilateral development banks (MDBs) may go forward with grant programs (which, at the margin, increases loan charges).

16.
In sum, the MICs are becoming a major supplier of a range of global public goods. They are an increasingly important element of global collective action to address market failures in the production of these goods. Accordingly, effective engagement of these countries on the part of the international community will be essential to the success of the global public goods agenda.

17.
Poverty in various dimensions is still high in the MICs, although its incidence varies widely across countries. Even in countries with low national poverty rates, large pockets of poverty persist in sub-national regions within countries. Moreover, as recent experience of countries impacted by crises suggests, gains in poverty reduction may be fragile, and susceptible to reversal in the event of economic volatility.

18.
Currently, the MICs are estimated to account for more than three-quarters of the world's poor. Even if the poor in China and India are excluded, some 40 percent of the remainder of the world's poor live in the MICs. More than half of the total population of the MICs is estimated to live on less than $2 per day (Table 1). 3 Even in the better-off MICs with per-capita incomes exceeding $2996 (IBRD Categories IV and V), over onefifth of the population is poor. Indeed, poverty and related indicators in these countries are little if any better than those with per-capita incomes in the $1,446-$2,995 range (IBRD Category III). Poverty is much higher among the lower-income MICs, with the poor accounting for over 80 percent of the population in the IBRD/IDA Blend countries. The severity of poverty among the poor as measured by the poverty-gap ratio tends to increase with poverty incidence and is most marked among the Blend countries where the poor receive on average incomes 40 percent below $2 dollars a day. The Regional groupings show that both poverty incidence and severity among MICs are lower in ECA and LAC than the average for other Regions.

19.
Social indicators also vary considerably across the MICs. While primary school enrollment s cover virtually all of the school age population in almost all but the poorest MICs, such as India and Pakistan, illiteracy remains a problem, particularly among females. To a substantial degree this reflects the legacy of incomplete enrollments in Standard deviation .

MICs High-income LICs
Source: World Bank staff estimates. earlier decades-in Brazil, for example, the net primary enrollment rate was 80 percent in 1980 as compared with 97 percent currently-and disadvantaged educational access of females. But it also reflects the low quality of primary schooling provided in many countries and smaller secondary enrollments. Life expectancy at birth in the MICs rose in most cases over the past decades to an average of 67 years and compares favorably with that of low-income countries (57 years), but still lies substantially below that of high-income countries (78 years). While some health indicators such as child immunization approach high-income country levels in most MICs, others are widely dispersed. Access to sanitation in countries such as Bulgaria, Korea, Romania and Poland lies in the 95-100 percent range, but reaches lows in the 10-20 percent range in Egypt, India and Uzbekistan. Access to drinking water is also substantially dispersed. The incidence of communicable diseases such as tuberculosis and the prevalence of HIV/AIDS are at levels somewhat closer to those of low-than high-income countries.

20.
Economic growth in MICs has in general been more volatile than in high-income countries, though less volatile than in low-income countries (Figure 7) Moreover, growth volatility increased somewhat in the MICs in the 1990s compared to earlier decades.

21.
High economic volatility hurts the poor. The 1990s have witnessed several financial crises, including the Mexican and the East Asian crises. During the East Asian crisis, poverty rose markedly (Table 2) as real wages fell and unemployment rose. In addition, there was considerable churning in employment across sectors. To some extent, families can smooth their incomes by increasing labor force participation and through the use of private transfers, though wealthier families are best able to smooth consumption. In crises, households in rural areas where poverty is usually the greatest can sometimes suffer deeper losses than do urban households. School enrollment declines, especially among poorer families, as does the use of health facilities. Crises have typically proved short-lived, though the ability of households plunged into poverty during crises to recover in line with the economy as a whole remains an open question. In view of the risk of economic volatility and its implications for the poor, strengthening social safety nets is an important agenda item in many MICs.

22.
The national rate of poverty inc idence for a given country often masks the fact that, particularly in large countries, the level of poverty is much more serious in some geographical regions than others. In China, for example, across the neighboring provinces spanning the regions between Hong Kong and Myanmar, the rural poverty rate in the mountainous province of Guizhou was 7-10 times higher than in the coastal province of Guangdong just a few hundred kilometers away. In Brazil, although the Electronic copy available at: https://ssrn.com/abstract=632725 national poverty rate at $2 per day is only 17 percent, poverty incidence is much higher in the backward regions in the North East.

23.
Sub-national poverty dispersion is closely related to the spread in per-capita income across geographical areas. Average per-capita incomes still vary considerably across sub-national regions despite a widespread absence of internal barriers to trade or factor mobility (Figure 8). Regional dispersion is much higher in large MICs and the lowest value in the coefficient of variation in income per capita obtained for any such country is higher than that observed anywhere in the developed world. The highest dispersions are usually in the larger developing countries where transport costs are an important factor-Argentina, Brazil, Indonesia and South Africa-but this is not always the case as dispersion is relatively low in India and high in Thailand where large income differentials persist between Bangkok and the rest of the country.

24.
The MICs continue to face considerable challenges in eradicating poverty. While many MICs recently enjoyed high growth, others are stagnating or even declining. The poverty impact of growth has also varied considerably. Chile (1987-94), Egypt (1975-95), and Pakistan (1991 are examples of countries that enjoyed episodes of strong pro-poor growth; but in India (1991-97), Thailand (1981, and Peru (1986-94) the impact of growth on poverty was much weaker. While there is a reasonable degree of consensus about the policy frame work needed to encourage economic growth, less is known about how to make growth more pro-poor.

25.
Addressing the problems posed by sub-national pockets of poverty is likely to remain a particularly thorny issue. Experience suggests that the dispersion of per-capita incomes across the sub-national regions of developing countries is likely to remain substantial for some time to come. Relying on market forces alone to remove subnational inequality is not enough, yet long experience with regional policy in most developed countries suggests that there is no easy way of ironing out sub-national differences

26.
The quality of the policy framework varies widely across the MICs. While many MICs have achieved or are close to achieving a good standard in most policy areas, a number of countries lag well behind. There is considerable unevenness in the 'scorecard' achieved by the MICs in different policy areas and there remains much room for progress in improving the enabling environment for investment and growth. While some countries have completed "first-generation" reforms associated with market liberalization, they have still to make adequate progress on a broad range of "second-generation" institutional reforms. 4 Thus, even among the relatively advanced reformers, a sizable policy agenda remains unfinished. Other countries in the group are still at the first stage of reforms. Moreover, the institutional capacity to manage policy reforms varies greatly across the MICs.

27.
In the 1950s and 1960s, many developing countries adopted inward-looking stateled development paths. Such countries pursued protectionism, government control of investment and state mo nopolies in key sectors, with varying degrees of success. In India, this resulted in persistently slow growth. In Latin America, the strategy delivered strong growth during the 1960s, but growth faltered later as a result of the oil shocks in the 1970s and the debt crisis in the 1980s. In contrast, the experience of East Asian countries demonstrated that it was possible to have a more interventionist government role and still achieve rapid economic growth over an extended period provided government direction was exercised within a broadly outward-oriented and marketfriendly environment.

28.
In the 1980s and 1990s, most of the developing world moved toward implementing market-friendly reforms. Governments replaced state intervention with private initiatives, public ownership with private ownership, and protectionist policies with competition from foreign producers and investors. The motivation for reforms, and the scope and pace of reform varied widely. In China, reforms were mobilized by the gradual realization that the agricultural sector was not reaching its full potential under state control. Agricultural reforms were followed by the introduction of market mechanisms throughout the economy. In Mexico, the debt crisis of the 1980s led to the push for wide-ranging economic reforms. In Eastern Europe and the former Soviet Union, political transition precipitated dramatic progress toward market reform, with mixed results.

29.
The late 1990s were characterized by several economic crises, beginning with Mexico in 1995, in the emerging economies of East Asia in 1997, andin Brazil andRussia in 1998. These examples underscored the importance of underpinning market liberalization with regulatory and institutional strengthening. Furthermore, in the countries of the former Soviet Union, pernicious corruption subverted major reforms. Today, improving governance is a major item on the reform agenda of many countries, including many MICs.

30.
Policy performance across MICs varies widely as measured by the World Bank's annual Country Policy and Institutional Assessment (CPIA) 5 . The MICs seem to perform best in terms of their macroeconomic policies but there is substantial variation in policy performance across all policy categories. While policy performance improves on average with higher income-per-capita-GDP per capita for countries with CPIA scores in the 4-5 range was about twice that of those in the 1-2 range-examples of good and bad performers may be found in all IBRD income categories. For example, Tunisia earns a slightly higher overall CPIA score than Brazil, a country with more than twice its percapita income.

31.
Countries with market-friendly policies such as openness to trade, disciplined mo netary and fiscal policy, well-developed financial markets, a competitive environment, and transparency and accountability of public sector, achieve on average better long-run economic performance than others (Table 3). GDP growth tends to be higher in countries with sound structural policies and a better-managed public sector. Export growth and gross domestic investment in countries with CPIA scores of 4-5 outperform those in the 1-3 group in all four categories of policies. The availability of private capital flows is less clearly linked to CPIA scores, although countries in the 4-5 range receive more private capital flows per capita. Consistent with the CPIA, countries with high scores are more likely to receive a rating and gain a higher rating if selected.

32.
There is room for improvement in the policy environment of all MICs. Performance is strongest in Economic Management where a majority of the countries achieve good scores in the 4-5 range and only about one-fifth of the countries receive unsatisfactory scores. Progress on Structural Policies is less consistent, with only a minority of countries achieving good ratings. Weaknesses in policy frameworks are most marked in Policies for Social Inclusion/Equity and Public Sector Management and Institutions. This underlines the need for institutional development in most MICs.

33.
Even in MICs with more developed policy environments, there are still gaps to be filled. Chile, for example, the highest rated country according to the CPIA for 2000 receives only a modest score for its environmental policies. Uruguay, another well-rated country also receives only a modest score for its banking sector policies.

34.
The needs for policy reform vary considerably across the MICs. Many lowerincome MICs have still to complete "first-generation" reforms-fiscal adjustment and rationalization of public expenditures and basic market-opening reforms such as trade liberalization, removal of arbitrary government intervention in domestic product markets, financial deregulation, liberalization of foreign investment, and privatization of state enterprises. Removal of market impediments and distortions remains a major agenda items for these countries. At the same time, they need to embark on a more complex set of reforms to strengthen the institutional underpinnings of markets, such as the legal and regulatory framework and the capacity and efficiency of market institutions. One important lesson of recent development experience is that these "second-generation" reforms should not lag seriously behind the market-opening reforms, as they are essential to the efficient and stable functioning of the markets liberalized under the "firstgeneration" reforms.

35.
In the upper-middle income countries there has been greater progress on the firstgeneration reforms, but there is a substantial unfinished agenda of second-generation reforms to strengthen the institutional framework for markets. These include the development of sound and competitive financial systems, capital market development, adoption of international standards in financial transparency and corporate governance, legal and regulatory environment for private transactions (property rights, insolvency laws, etc.), competition policies and related laws and institutions, regulatory framework for private provision of public services, and reform of public governance and institutions. Elements of these reforms are part of the policy agenda of all upper-middle income countries. Strengthening the regulatory and institutiona l framework for the financial system is a particularly important priority in several of these countries, this being crucial not only to fostering efficient private-sector-led growth but also to increasing economic resilience and preventing external shocks from turning into major crises. There are also important second-generation reform issues in the social sectors. In countries that have made greater progress in reducing poverty and expanding access to education and health services, the social policy agenda focuses more on reform of the social security system (efficient social safety nets to cushion the impact of shocks, social risk management instruments, pension system reform) and institutional reforms to improve the quality of education and health services.

36.
The capacity of MICs to manage the policy agenda varies considerably. Capacity constraints are of course more serious in the lower-income MICs. The more advanced MICs possess greater capacity for formulating and implementing reforms, but even they face gaps in capacity in managing elements of the more complex institutional reform agenda, such as modernization of the banking system and the capital market, corporate legal reform, competition policies, social security reform, and environmental regulations. Developing the needed administrative capabilities remains an important challenge.

37.
The MICs vary in two important dimensions in terms of their external environment: (i) their access to external private and official capital flows and the volatility of such flows; and (ii) their degree of integration into the international economy through trade and their vulnerability to terms of trade shocks. Despite the increased internationalization of finance, the majority of MICs receive very little private capital. Access to international capital markets is often limited, and even when there is access, it is normally not at the maturities appropriate for long-term development programs and projects. And volatility in private capital flows is a problem even for those MICs with broader access. While the MICs have become more integrated in international trade, progress is uneven, and trade-related reform remains an important agenda item in many MICs. Also, for a number of MICs, volatility in the terms of trade continues to be a serious problem.

38.
Total capital flows to the MICs rose from 2.6 percent of their GDP in the 1980s, a level depressed by the debt crisis during the decade, to nearly 4 percent of GDP in the 1990s. Private capital flows to MICs surged to $171 billion (in nominal terms) annually in the 1990s from an average of only $36 billion during the 1980s. As a share of GDP, Source: Global Development Finance private flows more than doubled in the last two decades. In this period, low-income countries also received higher amounts of capital flows as a share of their GDP, but unlike the MICs, they received mostly official flows (Figures 9 and 10).

39.
While the MICs received around 95 percent of private capital flows to the developing countries during the 1990s, they accounted for only about 11 percent of global private flows in the same period. This is significantly lower than their share in world GDP or exports (about one-fifth), an indicator of their lack of creditworthiness compared to high-income countries. Within MICs, access to international capital markets is highly skewed. Indeed, in 1998, only ten MICs received more than two-thirds-and twenty countries received more than 80 percent-of all capital flows to developing countries. 6 The distribution was even more skewed with respect to portfolio capital flows, which provide a better measure of a country's integration into international financial markets. In 1998, the ten top recipients among the MICs accounted for about 80 percent of all equity flows and 90 percent of all bond flows to developing countries. Typically, it is the richest countries among the MICs that have better access to capital markets, and these countries increased their share of total flows in the 1990s (Figure 11). The majority of MICs received little capital flows, and the share of the poorest group within MICs has declined in recent years.

40.
The number of MICs rated by Standard and Poor's doubled between 1995 and 2000 from 23 to 52 countries. But only a handful of these countries are rated investment grade. The majority of the MICs still have little access to international capital markets, or if they do have access, they have to pay a high price for raising finance.

The interest spread (over US dollar London Interbank
Offer Rate (LIBOR)) paid by a MIC borrower is often significantly higher than even the least creditworthy borrowers (issuers of 'junk bonds') in the high income countries. Figure  12 shows, for 1996-2000, the Emerging Market Bond Index spread indicating the borrowing costs for 11 emerging market (MIC) countries, the high yield spread paid by the least creditworthy US corporate borrowers, and the swap spread incurred by the most creditworthy borrowers in the US (i.e., the spread between AAA and AA-rated issues). Evidently, borrowing costs of emerging market issuers are both the highest and the most volatile.

42.
Creditworthiness varies considerably within the MICs. This is evident from Figure 13 showing sovereign credit ratings by Moody's and spreads as of September 2000. While borrowing costs (spreads) are expectedly higher for less creditworthy countries, similarly rated countries often have to pay different spreads. Borrowing costs rise dramatically as country creditworthiness deteriorates. Thus, capital becomes scarcer as the need for capital becomes more intense. Access to capital also becomes more uncertain as market perceptions get worse ( Figure 14). When credit ratings are downgraded, not only do borrowing costs go up, but also the volatility of spreads rises dramatically.

43.
Even the most creditworthy among the MICs are sub ject to disruption of capital flows and financial crisis from time to time. The 1980s were marked by a series of debt crises. The 1990s saw a series of currency and financial crises: Mexico in 1994/95, East Asia (Indonesia, Korea, Malaysia and Thailand) in 1997/98, Russia in 1998, and Brazil in 1999. Private non-foreign-direct-investment (FDI) flows to these countries suffered sharp reversals during the crises (Figure 15).

44.
The volatility of capital flows varies depending on the composition of capital flows. FDI tends to be more stable than portfolio flows. Indeed, in the aftermath of the crisis, FDI flows increased in the East Asia crisis countries and Brazil as depressed asset prices encouraged mergers and acquisitions. Non-FDI flows, on the other hand, declined sharply. Within non-FDI flows, the reversal of flows mirrored the initial composition of the flows. Prior to their crises, Mexico and Brazil received comparatively larger amounts of portfolio (bond and equity) flows than East Asia which received larger amounts of bank lending. Following the crisis, Mexico and Brazil suffered a collapse of portfolio flows, whereas the East Asian crisis countries suffered an acute reversal of bank lending, especially short-term debt. Russia seems to have suffered on all fronts including FDI, portfolio flows and bank lending.

45.
Private flows tend to respond pro-cyclically to economic shocks. Thus, an adverse shock (e.g., a decline in commodity prices) can lead to a reversal of private flows and exacerbate the impact of the shock. Short-term debt is potentially the most reversible component of capital flows. It is not only pro-cyclical to economic shocks, but it also tends to respond more to negative shocks than to positive shocks. Because of such procyclical behavior, reversal of short-term debt exacerbated the Asian crisis.

46.
Volatility in capital flows can have severe economic and social consequences for the affected countries, as illustrated by the East Asian crisis. The East Asian experience also showed that, given the increasing global financial integration, crisis in one emerging economy can quickly spread to others and threaten international financial stability.

47.
The economies of MICs became more outward-oriented in the 1990s. The trade share of GDP in these countries grew annually at an average rate of 3 percent in 1988-98, a rate similar to that of high-income countries. Trade expansion was, however, very uneven across the MICs and concentrated in specific regions ( Figure 16). Eight of the thirteen countries with the fastest increase in openness were in Latin America, all with annual growth in their GDP trade shares of above 5.3 percent. Eight of the thirteen countries with the slowest increase in openness were in Africa and the Middle East, each with annual growth in trade shares of below 0.2 percent and some with negative growth.

48.
While the MICs have made substantial progress in reducing trade barriers and in general have less restrictive trade regimes than low-income countries (Figure 17), trade-related reforms remain an important agenda item in many of these countries. First, almost one-third of MICs still have high trade restrictions ( Figure 17).
Second, in countries that have substantially reduced trade barriers (tariffs and non-tariff barriers), there is a sizable unfinished agenda of complementary reforms to maximize the benefits from trade liberalization and enhance its impact on growth. These complementary, "second-generation" reforms include, for example, improvements in domestic competition policies, strengthening of trade-related institutions, modernization of product standards systems and conformity assessment infrastructure, reform of traderelated finance, and promotion of trade in services.

49.
Volatility in terms of trade (TOT) remains important for a number of MICs. The MICs faced on average a TOT decline of 0.55 percent annually during 1988-98 (Table 4). While this is less than the 1 percent annual decline faced by low-income countries, it contrasts with the stable TOT faced by high-income countries. The MICs lost on average 0.4 percent of GDP annually as a result of TOT decline. The average TOT deterioration in the MICs was relatively large in this period, since the oil price was exceptionally low in 1998 and the MICs include several oil exporters (e.g., Venezuela, Mexico, Algeria, Nigeria, Gabon, Trinidad and Tobago), although several MICs in Central and Latin America and Southeast Asia experienced TOT gains. In Paraguay, Guatemala, Panama and Argentina real income grew on average 0.35 percentage points faster than real output, and in Indonesia, Malaysia, Thailand and the Philippines the average gain was around 0.5 percentage points. The difference in standard deviation between the country groupings is remarkable. The TOT and the equivalent gains and losses in terms of GDP are much more volatile in MICs than in high-income countries.

50.
The MICs vary from very small countries to huge federal systems, and from countries that are still poor to much better-off transition economies and relatively sophisticated emerging market economies. As the sections above show, the problems and development challenges faced by the MICs vary enormously, and a 'one-size-fits-all' approach is unworkable. It is nevertheless possible to classify the MICs into sub-groups that broadly share common characteristics. These sub-groups could be useful as a basis for further research and for guiding the policies of international institutions. The diversity among the MICs calls for an appropriately differentiated approach on the part of these institutions in defining their roles in these countries and formulating responsive support strategies.

51.
Based on the foregoing discussion of the profile of the MICs in terms of their development challenges and integration into the international financial system, a threeway classification of these countries is proposed here (Table 5). While the classification is defined in terms of the different levels of international capital market access, it takes into account the other dimensions of the MICs discussed in this paper, including the extent of poverty and the quality of the policy and institutional framework. In general, these three dimensions are correlated: countries with higher capital market access tend to have lower poverty and a higher level of achievement in policy and institutional reform. As elsewhere in this paper, MICs are defined in the classification below to include all IBRD-eligible (including blend) countries: • Low capital access countries (LCACs). Close to one half of the MICs fall into this category. These countries are either not rated in terms of the Institutional Investor Rating (IIR) or have IIRs of less than 35 and have typically very limited access to international capital markets. They have high and pervasive poverty (with a mean incidence of about 40 percent), have relatively weak policy frameworks, and slow growth. The main policy priorities in these countries are typically to structure competitive markets open to international trade and improve the environment for private sector investment and growth. A substantial part of the first-generation reform agenda-to unshackle markets and competition and remove distortionary government interventions-remains unfinished in many of these countries.
• Medium capital access countries (MCACs). About 30 percent of the MICs fall into this category. These are countries with an IIR between 35 and 51. They have significant, though not very stable, access to international capital markets, and also face relatively high spreads. These countries have lower but still high poverty incidence (mean incidence of 30 percent), more satisfactory policy frameworks based on freer markets, and low but positive per capita income growth. In these economies, perceived country risk needs to be reduced to improve access to international capital markets. These countries need to complete their first-generation reform agenda and move at the same time on the secondgeneration institutional reforms.
• High capital access countries (HCACs). A smaller group of MICs, 15 in number, have IIR of above 51. These countries enjoy strong and more stable capital market access and also face much lower spreads in the international financial markets. Average poverty incidence is 15 percent in these countries, they are well advanced in their market reforms, and they enjoy substantial per capita income growth on average. Despite their strong capital market access, these countries are not immune to volatility in capital flows as market perceptions shift, as the recent experience of the three East Asian members of this group-Korea, Malaysia, Thailand-testifies. Achieving greater resilience to shocks is a key objective for these countries, and their remaining reform agenda focuses primarily on second-generation reforms to strengthen the institutional framework for efficient and stable functioning of markets. Excluding China, poverty in this group tends to be less pervasive and more geographically concentrated. However, national poverty levels can rise sharply in the event of negative economic shocks (the East Asian HCACs are again a case in point), necessitating stronger systems of social protection. Notes : Six of the 77 countries are not included in this classification due to incomplete data. ".." denotes data not available.
Source: Staff calculations based on World Bank data.

52.
As noted in Section II, the MICs are of increasing importance in relation to a range of global public goods. The HCAC category, for example, while only 15 in number, accounts for more than 60 percent of all private capital flows to developing countries and is, there fore, important from the standpoint of regional and international financial stability. The LCAC and MCAC categories are particularly important for the global objectives of poverty reduction and protection of the environment and health. More than half of the world's poor live in these countries. Because it includes China, even the HCAC category accounts for about a quarter of the world's poor (Table 6).

53.
Inevitably, the proposed classification is not watertight and exceptions may be found in each group that do not confirm to the norm. China, for example, is an HCAC, but its poverty incidence of 54 percent and CPIA rating are well out of line with others in this group. South Africa is another HCAC where a high poverty incidence of 36 percent-a legacy of the apartheid regime-is well above average for this group. Lower on the scale, Brazil provides the opposite example as an MCAC with a low poverty incidence of 17 percent (though with high regional concentrations of poverty). Algeria is something of an exception among LCACs as it has low poverty (estimated at 15 percent) although its policy performance is well in line with the average for the group.

54.
It should be noted that while average per capita income rises as we move up the capital market access categories, there is substantial variation around the averages. The two extreme cases are those of China and India of course. The average per capita income for the HCAC category is close to $4,000 but that of China is only $780. Similarly, the average per capita income of the MCAC category is about $2,700, but that of India, which belongs to this category, is only $450.

55.
The MICs, often referred to as one group, are in fact characterized by a high level of diversity. All of them continue to face major development challenges, but the nature of these challenges varies considerably. Many countries continue to face pervasive poverty, while in others the national incidence of poverty has been reduced but these gains are still fragile and there remain concentrations of high poverty at the sub-national level. A number of MICs are still at a relatively early stage of policy reform. Others have substantially completed the basic market-opening reforms but face a major agenda of institutional reforms to complement the policy changes already adopted. To varying degrees, the MICs face institutional capacity constraints in managing the reform agenda. Their degree of integration into the global economy also varies greatly. The majority of MICs still have very limited access to private foreign capital. Some enjoy broader access, but volatility in private capital flows is a problem even for these countries.

56.
This assessment suggests that there is a continuing role for international development institutions, such as the World Bank, is assisting the MICs, as a substantial part of the development agenda in these countries remains unfinished. It also suggests that these institutions need to employ an appropriately differentiated approach to assistance to the MICs, given the highly diverse nature of the challenges that these countries face. For instance, in MICs with low access to international capital markets, the traditional role of these institutions as sources of development finance still holds as in the past. But in MICs with high capital market access, the provision of external finance is less of an issue, except in the event of disruption of their market access, and the focus of the international development community needs to be more on assisting policy development and institutional change.

57.
There is also a global public goods rationale to the continued engagement of the international development institutions in the MICs. With the rising role and integration of the MICs in the world economy, developments in these countries have become increasingly important to a range of global public goods. Helping the MICs address financial instability, poverty, environmental degradation and communicable diseases assists with these objectives at the global level.