48382 ResearchDigest WorldBank VOLUME 1 NUMBER 3 SPRING 2007 Making Finance Work for Africa Patrick Honohan and Thorsten Beck African finance operates in an The authors show that their analysis IN THIS ISSUE extreme environment in terms of leaves only 60 of the 320 basis points Making Finance Work for Africa ... page 1 scale, informality, governance, and of difference unexplained. The authors' analysis shows that How to transform African economies? shocks financial depth is lower in Africa than Start by creating the means to build the E confidence of lenders ntrepreneurs in Africa mention elsewhere not only in raw terms but access to and cost of finance as also when controlling for other fac- FOCUS barriers to their firms' growth tors, such as inflation, per capita in- Water in the Middle East: more often than do those in any other come, and other plausible explanatory How to Manage Scarcity ... page 2 part of the world. If only for that rea- variables. (For some African countries, How the region meets its water son, improving the situation of the such as Mauritius and the Seychelles, challenges matters--for the cost and financial sector should be a priority which score well on depth measures, for who bears it for African policymakers. conditional underperformance is only Indeed, as a new study by Hono- marginally significant.) One associ- Fungibility and the han and Beck finds, conventional ated factor here is the higher recorded "Flypaper Effect" of Aid ... page 3 measures of financial depth and ef- African deposits in international Does donor financing "stick" to the ficiency show African countries falling banks abroad. right sector? A new way of looking at considerably behind those in other Credit-based depth is further be- the fungibility of aid regions. For example, the mean net low international comparators than interest margin in African banks, at are deposit-based measures of depth. Do Girls Gain from Migration- Induced Male Absence? ... page 4 800 basis points, is about 320 basis Consistent with this, the authors points higher than the world average. find that liquidity in African banking How does male migration affect Interestingly, the authors find that systems is high and growing. Interest- households' investment in girls' health and schooling? the gap is even larger if they compare ingly, the shallowest banking systems a matched sample of affiliates of the in the region are the most liquid. The Distributional Effects of WTO same foreign-owned banks in Africa The authors suggest that this pat- Agricultural Reforms in Rich and and elsewhere. tern is consistent with a correlation Poor Countries ... page 5 To explain the wider margins in Af- between the degree of confidence One way to help reduce poverty in rica, the authors take a close look at of depositors in the banks (or in developing countries: require more bank-level correlates of the margins, macroeconomic stability) and that from them in trade liberalization such as operating costs, profitability, of the banks in their borrowers. They and loan loss provisions. They find discuss policy measures that can aid Clientelism, Credibility, and the Policy that all these elements contribute to economic agents in building confi- Choices of Young Democracies ... page 6 the wider margins. The higher profit- dence, including improvements in Can countries count on competitive ability may be associated with high credit information and in the legal elections to make governments more perceived risks. But it is also likely to infrastructure for assuring contract responsive to broad public concerns? be explained in part by less competi- enforcement. Not necessarily tion, as documented by the much Trends in recent financial sector Entry Regulation as a Barrier higher concentration in the typical data for Africa reveal a distinct turn- to Entrepreneurship ... page 7 African banking system. A deeper in- around since the mid-1990s. Private Among the costs of high regulatory terpretation traces the wider margins credit and other depth indicators barriers to entry--fewer new firms to scale and to deficiencies in the have been increasing steadily as a and smaller, slower-growing old ones information and legal infrastructure. (continued on page 8) 2 WorldBankResearch Digest FOCUS toward more efficient courses of action. and potential transformation of agricul- But can they be implemented political- ture. These will change interest group Water in the ly? A new report by a World Bank team alliances. Educated women may lobby looks at ways to navigate through the for more investment in sanitation. And Middle East: region's changing political economy. a transformed agriculture changes both Unlike earlier studies, the report treats the ability of farmers to organize and How to Manage water as a development challenge rath- lobby and the services they request. er than a sectoral one, analyzing the High-value farmers in the Arab Repub- Scarcity impacts of nonwater policies on water. lic of Egypt and Morocco, for example, An international index of the quali- are lobbying for reliable rather than Julia Bucknall and team ty of policies and institutions for man- low-priced irrigation services. aging freshwater shows that the region The report also looks at public ac- does fairly well relative to a compa- countability, generally lower in the Improving public accountability rable set of developing countries. Yet countries of the Middle East and North can turn political opportunities into many nonwater policies inadvertently Africa than in other developing coun- better water outcomes distort use of this scarce resource. tries. Using an internationally compa- Energy subsidies depress the cost of rable index of public accountability, T he Middle East and North Africa water services and encourage waste- it finds that countries in the region region faces big and rapidly in- ful use, and price supports for staple where authorities are more account- creasing water problems (figure crops encourage irrigation of thirsty, able provide better water services than 1). Per capita resources, only a sixth low-value crops. Subsidized credit those where authorities are less so. It of the global average, are set to de- encourages drilling of wells. And the concludes that measures to improve cline by half by 2050. Overpumping of region's overwhelming unemployment public accountability--such as releas- groundwater causes net depletion of problem leaves rural populations de- ing information about water services to national assets valued at the equiva- pendent on agriculture for livelihoods. the public, involving users in decision- lent of up to 2 percent of GDP every The water problem cannot be man- making, and setting clear performance year. aged by water ministries alone. And standards with consequences for fail- The situation is set to get worse. water reformers can have an important ing to meet them--can turn potential Climate change models predict a de- impact if they also work on reforming political opportunities into actual im- crease in rainfall of at least 20 percent. policies in other sectors. provements in water outcomes. Environmental degradation of water The report suggests that would-be The report is optimistic that the already costs 1­2.5 percent of GDP water reformers analyze the changing region will meet its water challenges. every year, and population growth will dynamics of the political economy to But whether the changes can be made increase the water quality challenge. seek opportunities for reform--and without massive cost, borne dispro- The price for taxpayers is high--one in that changes in social, technical, en- portionately by the poor, will depend four dollars of capital budgets goes to vironmental, and economic forces are on what reforms countries implement, water in some countries--and expect- indeed creating such opportunities. in the water sector and elsewhere. ed to increase. The problems result The region is experiencing, or on the Julia Bucknall and team. 2007. Making the from institutional as well as resource cusp of, some major transformations-- Most of Scarcity: Accountability for Better constraints. Institutions in the region including rapid urbanization, growth of Water Management in the Middle East and are not yet able to adapt to changing tourism, increasing education of girls, North Africa. Washington, D.C.: World Bank. water supplies and changing demands from the population. Who will be affected by this ac- Figure 1. Share of Total Renewable Water Resources Withdrawn, by Region celerating challenge? Urban residents will see increasingly frequent service Middle East and North Africa outages. Even now water is delivered South Asia only twice a month in Amman, Jordan, Western Europe and only once a month in Taiz, Re- East Asia and Pacific public of Yemen. Deteriorating water (including Japan and Koreas) quality will have an impact on human North America health and ecosystems. But the great- Europe and Central Asia est impact is likely to be in agricul- Australia and New Zealand ture, which uses around 80 percent of the region's water. Transforming Sub-Saharan Africa agricultural production systems while Latin America and the Caribbean protecting the poor will be a key part 0 10 20 30 40 50 60 70 80 of managing the water challenge. Percent Excellent national strategies exist Source: Compiled from the Food and Agriculture Organization's AQUASTAT data for 1998­2002. to guide the region's water ministries Note: The figure shows the sum of withdrawals across all countries in a region, divided by the sum of all the renewable water available in each country. WorldBankResearch Digest 3 Fungibility and the rule out the alternative explanation "Flypaper Effect" of Aid that the results are due to model mis- specification. These impact evaluation methods Dominique van de Walle and Ren Mu that the money "sticks," like flies on fly- confirm the conclusion drawn from a paper. Theoretical contributions have casual inspection of the data. The im- Some donor aid for a road project shown how such effects can be gener- pact estimates, controlling for selective ated. Others have argued that what placement, suggest a sectoral flypaper was diverted to recipients' own appear to be flypaper effects are in fact effect: some of the project funds were priorities--but it did stick to the due to biases in evaluation methods. diverted from rehabilitation to the road sector The Vietnam project stipulated that building of new roads, but they did stay the work entail rehabilitating earth in the road sector. Diversion was less T he fungibility of aid--in which roads and not building new roads. In- likely in communes that had been do- donor aid earmarked for a proj- dependent administrative data show ing less rehabilitation before the proj- ect substitutes for rather than that an average of 4.6 kilometers per ect. Spending on rehabilitation plus supplements local spending intended commune were rehabilitated under building accords reasonably closely to for that purpose--has important im- the project. Supervision and other the total project allocation. plications for its effectiveness. And evidence confirm that the project was Thus project communes ignored if aid ostensibly tied to a specific implemented as planned. Yet a casual the donor's stipulation that they not intervention simply displaces local look at the data suggests only half the build new roads. Instead, they im- resources, it becomes difficult to de- expected difference between project posed their own priorities and diverted termine exactly what its development and nonproject com- some of the funding impact has been. munes in increments of to priority road build- How much aid fungibility is there? road rehabilitated dur- Displacement ing. The authors also There is little consensus. Most econo- ing the project period. occurred in find that the quality mists think that fungibility is the It also suggests that the of rehabilitated roads norm, while most aid donors behave project communes built that less improved in the project as though there were none. The issue significantly more new communes. This again is routinely ignored in project work. roads than the nonpro- rehabilitation entailed a switch from Aid fungibility has typically been ject communes did. took place the donor's preferred examined using cross-country regres- Looking casually at technology of rehabilitat- sions on macro aggregates over time. data can be misleading. than the aid ing earth roads. In a new study van de Walle and Mu For example, if the proj- Yet project aid did take a different approach, using impact ect was implemented in had (ostensibly) stick to the road sector. evaluation methods to examine the is- communes with different financed. This is shown both by sue in the context of project aid. They characteristics than the the estimated kilometers estimate the impact of a World Bank­ nonproject comparators, However, the of rehabilitated and built financed rural road rehabilitation simple mean differences roads and by supporting project in Vietnam on the kilometers will be wrong. Accurate aid stuck to the evidence that the project of roads actually rehabilitated. The impact assessments of sector, with the did not lead to the con- authors ask, did the aid for this proj- rural roads have been struction of other basic ect end up funding what the donor rare because the neces- donor getting infrastructure. intended? sary data are usually Impact evaluations From a theoretical perspective the lacking. The authors' more built have become more so- expected outcomes are ambiguous. study is one of the first roads phisticated in recent Economic theory predicts that if proj- to use a panel data set years, worrying about ect aid is less than or equal to what of communes in project selection, impact hetero- recipients intended to spend, they will and nonproject areas before and after geneity, and attrition bias. Yet evalu- spend as planned and treat the aid as a project to rigorously examine road ations still rarely establish that an in- general budget support. The aid will impact and fungibility. tervention actually funded what it was be fully fungible, and there will be no The authors first ascertain that full intended to fund and was supplemen- difference in impact between project fungibility is possible, since most of tal to local spending. One conclusion areas and the nonproject areas used the project communes would have from the analysis is that ascertaining for comparison. spent more on road rehabilitation this should be a first step in evaluating Against that is plentiful empiri- without the project than they in fact the impacts of development projects. cal evidence for "flypaper effects"--in received from the project. They then which grants to local governments use a difference-in-difference estima- Dominique van de Walle and Ren Mu. Forthcom- stimulate much higher local spend- tor coupled with propensity score ing. "Fungibility and the Flypaper Effect of Project ing of the type intended than would matching and weighting methods to Aid: Micro Evidence for Vietnam." Journal of an increase in community income, so determine the impact of the aid and Development Economics. 4 WorldBankResearch Digest Do Girls Gain the extent that girls gain more from affirming the potential intergenera- such a change in preferences, male tional benefits of averting nutritional from Migration- migration could well amplify gender and other health shocks in early differences in the gains from migra- childhood. Induced Male tion. Migration also has a significant In two recent papers Mansuri and positive effect on all schooling Absence? examines the extent to which the decisions and a strong dampening expected gains from migration are effect on child labor market activity. Ghazala Mansuri borne out in rural Pakistan, where This suggests that the opportunity migration for work is substantial-- cost of time spent by children in with more than one in four families school may be substantial, at least Girls have bigger health and reporting at least one migrant--and for some households. education gains than boys in migration is largely temporary and Most important, migrant house- households with migrant males-- legal. The papers focus on house- holds make much larger investments except in those headed by women hold investments in child health and in the schooling of girls, leading to a schooling, arguably the two most im- significant reduction in gender gaps R ecent debates on "feasible portant contributors to inequalities in school enrollment, retention, and globalization" have focused of opportunity. attainment. In contrast, there are no on the importance of open- Work migration is undertaken gender differences in labor market ing up international labor markets exclusively by men, and female head- activity. Thus the smaller relative to low-skilled guest workers from ship is almost exclusively a result of gains for boys do not appear to re- developing countries. Liberalizing temporary migration-induced male flect higher labor demands. labor markets in this way, it is ar- absence. This makes the context par- A comparison of siblings corrobo- gued, would lead to large gains in ticularly useful for examining gender rates these results. income and could help reduce in- differences in human capital invest- Interestingly, though, while female equalities of wealth and opportunity ment not only between households headship has no impact on school both within and across countries. with and without migrants, but also enrollment for either boys or girls, Key to this is the expectation that between migrant households, by fe- adolescent girls in female-headed remittances from migrants will fuel male headship. migrant households are much more private investment in human capi- The key empirical problem in as- likely to drop out of school and con- tal in their communities of origin. sessing the impact of migration on sequently have lower school attain- Where gender gaps in human capital household behavior is that house- ment. The opposite is true for boys. indicators are large, remittances are hold characteristics that influence They do significantly better in such also expected to reduce gender in- the decision to migrate are also likely households. Is this because female equalities. to affect other household outcomes, headship has a differential impact Migration can affect household including child schooling, labor mar- on the work burden of children? The behavior through a number of com- ket activity, early childhood growth, data do not support this. While both peting channels, however, creating and any observed gender disparities boys and girls work substantially new constraints that could dampen in all of these. more in female-headed households, this potential enhanced investment. The papers use two strategies to there are no gender differences in Most significant are migration- deal with this potential endogeneity work burden. Nor does female head- induced changes in household struc- problem: instrumenting for migra- ship have an impact on the woman's ture. Children in migrant households tion, and comparing siblings within own labor supply. often have less adult supervision migrant households by exploiting This suggests that migration- and are required to spend more time the fact that many schooling and induced female headship may cre- on household production, the care of health decisions are time sensitive. ate constraints that dampen at least younger siblings, or other domestic Finally, the analysis obtains the im- some of the gains from migration. chores. pact of female headship by compar- In particular, the evidence suggests Conversely, where migration is ing female-headed migrant house- greater restrictions on the mobility of undertaken mainly by men, and holds with male-headed ones. adolescent girls, perhaps due to the there are important gender differ- Using two measures of early child greater social vulnerability of female- ences in preferences relating to child growth (weight-for-age and height- headed households. welfare, migration-induced "male for-age z-scores), instrumental vari- absence" could change the balance able estimates show significantly Ghazala Mansuri. 2006. "Migration, School At- of household preferences in a differ- better growth outcomes among chil- tainment, and Child Labor: Evidence from Rural Pakistan." Policy Research Working Paper 3945. ent direction. A substantial body of dren in migrant households. More- World Bank, Washington, D.C. research has shown that investments over, the effects are much larger for ------. 2006. "Migration, Sex Bias, and Child in children tend to increase where girls. And estimation on samples dis- Growth in Rural Pakistan." Policy Research mothers exercise greater control over aggregated by age indicate that this Working Paper 3946. World Bank, Washington, the use of household resources. To advantage is sustained as girls age, D.C. WorldBankResearch Digest 5 The Distributional Effects of WTO are generally rather small, even for the abolition of rich countries' agricultural Agricultural Reforms in Rich and trade barriers. Table 1 reports on four illustrative economies. Poor Countries Brazil and Thailand are big gainers. The increased demand for their agri- Thomas Hertel, Roman Keeney, Maros households by their main income cultural exports boosts rural incomes, Ivanic, and L. Alan Winters source. They then calculate for each and because agriculture is relatively stratum the poverty elasticities at the important for unskilled workers, their poverty line ($1 or $2 a day), showing wages also rise more generally. Be- The mix of high levels of support, the sensitivity of the number of poor cause the poor are clustered around specialization, and great wealth is to changes in the prices of consumer the poverty line, the income benefits toxic for agricultural reform goods, in taxes, and in each of 10 shift substantial numbers above it. sources of income. The global model Bangladesh and Zambia miss out, T he modern apology for preserv- estimates how these prices and factor especially in the Doha scenario. Part ing rich countries' agricultural incomes change with liberalization, and of the reason is that Doha includes the protection is that it supports from these two sets of information the abolition of rich countries' export sub- poor farmers in the North and that authors calculate the poverty effects. sidies--which does little or nothing for liberalization would benefit only rich In developed countries the average poverty in most developing countries landowners in the South. Both asser- farm household earns most of its in- because the poor either are isolated or tions contain grains of the truth, but a come from nonfarm activities. Thus in are net purchasers of the commodities new article by Hertel, Keeney, Ivanic, Japan, for example, Doha liberalization affected. In addition, Bangladesh loses and Winters shows that the predomi- might cut farm earnings by 16 percent because it is an importer of cotton, and nant effects are the very opposite. but boost farmers' nonfarm earnings Zambia because its farmers are so iso- The authors explore two liberaliza- by 0.6 percent. The net effect is an in- lated that they receive relatively little tion scenarios: first, a prediction based come loss of just 1.4 percent, because increase in demand and so poor that on the WTO's Hong Kong Ministerial agriculture provides only 12 percent what they do get is insufficient to pull (December 2005) of what the Doha of total earnings for Japanese farm them above $1 a day. Round might bring in agriculture (75 households on average. The authors also compare the "pov- percent cuts in tariffs, 60­75 percent In the United States most farms erty friendliness" of different types of cuts in domestic support, and the abo- earn little from farming, but large trade liberalization. Agricultural lib- lition of export subsidies), and second, farms producing sensitive products eralization has larger poverty effects complete liberalization. They translate are highly specialized, with the rich- than nonagricultural liberalization. And these general rules into changes in est deriving up to 90 percent of their developing countries' own agricultural policies on more than 5,000 products income from agriculture. Full agricul- liberalization always reduces poverty, in before reaggregating them into a man- tural trade liberalization would cut most cases more so than rich countries' ageable number of categories. the total income of the wealthiest rice reforms. It both reduces the prices of The authors convert the policy farmers by about 19 percent and that staple foods for the poor and creates changes into resulting medium-run of cotton farmers by 13 percent. This market opportunities because develop- (three- to five-year) changes in prices, combination of high levels of support, ing countries trade a good deal with outputs, and factor rewards using a specialization, and great wealth is one another. That Doha requires very global simulation model (GTAP-AGR). toxic for agricultural reform. little trade liberalization by developing The model contains detailed treat- In the developing economies, rich- countries is a clear source of weakness ments of farm input and production country agricultural trade liberalization in its poverty-reducing credentials. decisions, labor and land markets, boosts real returns in agriculture--for and consumption patterns. land and farm workers--and cuts them Thomas Hertel, Roman Keeney, Maros Ivanic, and Finally, the authors calculate the elsewhere. How this translates into L. Alan Winters. Forthcoming. "Distributional effects on farm incomes in developed national poverty depends on a range Effects of WTO Agricultural Reforms in Rich and countries and on poverty among all of factors. The effects, though varied, Poor Countries." Economic Policy. households in 15 developing coun- tries on which they have good data on the sources of income. Among devel- Table 1. Effects of Rich-Country Liberalization of Agriculture on National Poverty oped countries they give the United States special attention, identifying Full liberalization Doha liberalization Country Percent Thousands Percent Thousands farm households specializing in four Bangladesh ­0.06* ­27* 0.00* 0* heavily supported sensitive crops-- Brazil ­1.88 ­431 ­0.73 ­167 rice, sugar, cotton, and dairy--by their Thailand ­7.10 ­84 ­1.43 ­17 Zambia 0.13 8 0.03 2 place in the wealth distribution. For each developing country the * Result cannot be distinguished from zero change at the 95 percent confidence level. Note: The table shows percentage and absolute changes in the population below a poverty line of $1 a day, with a negative value authors distinguish seven strata of indicating a reduction in poverty. 6 WorldBankResearch Digest Clientelism, Credibility, and the Policy Choices of Young Democracies Philip Keefer When elected governments fail to believe the broad policy promises good provision, higher private good promote broad social welfare, the of political competitors, electoral provision, and greater corruption. solution may be reforms to increase competition forces them to bid up Nor does it appear that unobserved the performance promises they make effects, unrelated to the credibility political credibility to voters. When promises are not arguments, simultaneously deter- A n important development puz- credible, voters can rely only on past mine government policy choices and zle is that among poor coun- performance and clientelist promises democratic age. tries elected governments do not (promises from political competitors These results point to the impor- make dramatically different policy that are credible to narrow groups, tance of reforms that increase politi- decisions than unelected governments. or individuals, in society). Neither cal credibility. Credibility depends In a new article Keefer shows that in provides a strong basis for electoral most on actions by politicians (such poor democracies measures of gov- accountability. as investing in the programmatic rep- ernance, schooling, infrastructure, Reliance on past performance utation of their political parties); do- government spending as a fraction can lead voters to retain even poorly nors can do little to influence these. of national income, and the business performing incumbents, since vot- But it also requires that citizens have climate are generally about the same ers cannot be sure that noncredible the information to verify whether pol- as or worse than those in poor non- challengers will perform better. Reli- iticians have fulfilled their promises. democracies. ance on credible clientelist promises, Policies to boost citizen information This finding is inconsistent with as Keefer and Vlaicu show in a 2005 about policy outcomes, and a cor- the common assumption that com- paper, drives politicians to underpro- responding relaxation of government petitive elections should make poli- vide public goods and to overprovide controls on information, are there- ticians notably more responsive to targeted goods that benefit narrow fore key and already have a place in the concerns of the general public. constituencies that happen to be- the donor toolkit. Understanding why the assumption lieve the politicians' promises. This The analysis also suggests reform fails to hold--why policymakers in is the pattern of policy performance strategies in the absence of political some democracies are not more sen- observed in young democracies: they credibility. For example, reforms that sitive to broad social welfare--is key are exceptionally accountable for pri- are difficult for citizens to observe to the design of successful develop- vate good provision, but exception- and are broad in their effects (such as ment strategies in those countries. ally unaccountable for public good civil service reform) are more likely to Keefer presents new evidence provision. succeed if they are linked to specific showing that the cleavage in demo- The association between demo- outcomes that citizens can moni- cratic performance also runs be- cratic age and this pattern of policies tor (textbook delivery, road mainte- tween young and old democracies. is robust to many alternative expla- nance) and more strictly supervised. Countries with fewer continuous nations. In describing the political Some reforms that appear to be years of competitive elections are market imperfections that interfere economically inferior may also be more reluctant to provide public with electoral accountability, Keefer preferable if they help to build cred- goods (education, access to informa- and Khemani, in a 2005 article, also ibility while satisfying the incentives tion, high-quality bureaucracies, the point to imperfectly informed voters of politicians to give targeted ben- rule of law), more enthusiastic about and social polarization, which may efits (increasing capitation grants for private goods (jobs in the public sec- particularly afflict young democra- students rather than increasing fund- tor, pork barrel infrastructure), and cies. Young democracies may also ing for schools generally; increas- less restrained in the exercise of rent exhibit systematically different politi- ing redistributive grants to the poor seeking and corruption. While there cal and electoral institutions or face rather than increasing the quality of is a substantial overlap between greater exposure to political violence. services to the poor). young and poor democracies, the Theoretically, though, these com- Demand-side interventions to differences between younger and peting explanations either are em- accelerate reforms should focus on older democracies persist even when bedded in the credibility argument the civil society groups that have the controlling for income per capita. (information) or predict patterns most comprehensive strategies for The best explanation for this par- of performance other than that ob- bridging the credibility gap between ticular pattern of policy performance served among young democracies. politicians and citizens. is the inability of political competi- Empirically, none of these variables Philip Keefer. Forthcoming. "Clientelism, Credibil- tors to make policy promises that are affects the significant association ity and the Policy Choices of Young Democracies." credible to all citizens. When citizens of democratic age with lower public American Journal of Political Science. WorldBankResearch Digest 7 Entry Regulation as a Barrier to Entrepreneurship Leora Klapper, Luc Laeven, and Raghuram Rajan Regulatory barriers against Interestingly, the costs matter most in industries in countries with costly bu- entrepreneurship hamper the richer countries and in countries that reaucratic barriers. Thus costly entry creation of new firms, discourage are not corrupt, where the regulations regulations are a form of protection on the books are more likely to be en- that has the most deleterious effect the entry of small firms, and affect forced. on the performance of seasoned in- even older firms, which grow more These findings suggest an explana- cumbents. slowly and to a smaller size tion for the low level of incorporation Here, a comparison between Italy, in Italy: the average direct cost of com- with high entry barriers, and the Unit- E ntrepreneurship is important plying with the bureaucratic regula- ed Kingdom, with low ones, is par- for the continued dynamism tions for registering a new corporation ticularly telling. Across all industries, of the modern economy and in Italy is 20 percent of firms start out larger contributes to economic growth. Yet per capita GNP--twice when young in Italy, but many countries put in place regu- the average for other By then grow more slowly-- lations that make it more difficult G-7 countries in Europe. so that by age 10, firms in to start a new firm. In a new paper The authors next hampering the United Kingdom are Klapper, Laeven, and Rajan look at study the effect of bu- the creation about twice as large as the cost of meeting the regulatory reaucratic entry regula- those in Italy. This sug- requirements for setting up a limited tions on the average of new firms, gests that Italy has small liability company and study the ef- size of entrant firms. firms not because there fect of such entry regulations on the Since the high entry costly entry is too much entry but be- creation of new firms, the average costs are largely fixed, regulations cause there is too little. size of firms that finally are able to they should be reflected To check whether incorporate, and the dynamism of in a larger average serve as entry regulations proxy incumbent firms. size of entrants into for other aspects of the The authors start by investigating high-entry industries a form of business environment the cross-country picture of the incor- in countries with high protection. that are likely to have an poration of new firms, using a com- costs. The authors find impact on entry--such as prehensive database of corporations that this is indeed the The absence financial development, across a large number of developed case. That means that labor regulation, and and transition economies in Europe. such regulations not of the protection of intellectual Some facts are striking. Italy, for ex- only discourage small disciplining property--the authors in- ample, with its myriad small corpora- firms from setting up, clude these environmen- tions, might be expected to have tre- they also force others effect of tal variables interacted mendous incorporation of new firms. to grow without the with the characteristics In fact, new corporations (those that protection of limited li- competition of the industry they are are one or two years old) represent ability until they reach a from new most likely to influence. only 3.8 percent of the total in Italy. scale at which they can The results show that In contrast, the average for France, afford the cost of incor- firms has these aspects of the Germany, and the United Kingdom is poration. business environment 13.5 percent. If entry regulations real adverse do matter, but primarily To study whether entry costs af- indiscriminately screen effects for the rate of incorpora- fect the extent of incorporation, the out small young firms, tion, not for the size of authors examine whether the share of constraints on their entrants or the productiv- new corporations is lower in an indus- emergence should mute the disciplin- ity growth of incumbents. Particularly try with a higher "natural" propensity ary effects of competition, with older noteworthy is that the harmful effects for entry when the country has higher firms more likely to be lazy and less of onerous entry regulations persist costs of complying with bureaucratic capable of enhancing productivity. despite the inclusion of these other requirements for incorporation. They Testing whether entry regulations af- interactions. find that the rate at which new corpo- fect the productivity growth of older rations are created in "naturally high- incumbent firms, the authors find Leora Klapper, Luc Laeven, and Raghuram entry" industries is relatively lower that value added per employee for Rajan. 2006. "Entry Regulation as a Barrier to in countries with higher entry costs, older incumbents grows relatively Entrepreneurship." Journal of Financial Eco- suggesting that these costs matter. more slowly in naturally high-entry nomics 82 (3): 591­629. 8 WorldBankResearch Digest (continued from page 1) Recent Policy Research Working Papers share of GDP--and the increase is fact that it operates in an environment broad. Indeed, four of every five coun- that is extreme in four key dimensions: WPS4168 Mobility and Earnings in Ethiopia's tries for which data are available have scale, informality, governance, and Urban Labor Markets, 1994­2004 Arne Bigsten, Taye Mengistae, and had an increase in financial depth shocks. The authors examine alterna- Abebe Shimeles since 2000. tive policy solutions that have been WPS4162 An Analysis of Crop Choice: Adapting to Climate Change in Latin American Farms Similarly, organized securities mar- suggested to deal with these and Niggol Seo and Robert Mendelsohn kets have shown a sharp increase in distinguish between two approaches: WPS4158 Walking Up the Down Escalator: Public Investment and Fiscal Stability capitalization and listings in the past modernist and activist. William Easterly, Timothy Irwin, and few years. In the Lagos stock exchange Each approach has its advocates Luis Serven alone about $3 billion in new capital and its strengths, though each also WPS4155 How Good a Map? Putting Small Area Estimation to the Test was raised in 2006 to meet the new can overreach. Modernist solutions Gabriel Demombynes, Chris Elbers, banking capital requirements. And sometimes overreach by unthinkingly Jean O. Lanjouw, and Peter Lanjouw WPS4149 The Worldwide Governance Indicators listed bonds have increased in number attempting to transplant models from Project: Answering the Critics and maturity. advanced economies into unprepared Daniel Kaufmann, Aart Kraay, and Massimo Mastruzzi The region's stock exchanges are environments (an example is the ad- WPS4146 International Financial Integration through still very small (except for Johannes- vanced version of the Basel 2 capital Equity Markets: Which Firms from Which Countries Go Global? burg's). But a comparison of their accord for banks). Activism is surely Stijn Claessens and Sergio L. Schmukler performance with that of other devel- necessary to help ensure direct access WPS4140 Corporate Governance and Regulation: Can There Be Too Much of a Good Thing? oping country exchanges using the for all to the needed financial services, Valentina G. Bruno and Stijn Claessens World Bank's financial sector devel- but weak governance is its Achilles' WPS4138 Migration and Mental Health: Evidence from a Natural Experiment opment indicators (http://www.fsdi.org) heel. The authors review how these Steven Stillman, David McKenzie, and shows that African markets are close tensions can be resolved in instances John Gibson to the average in size (relative to the ranging from microfinance regulation WPS4131 Will Markets Direct Investments Under the Kyoto Protocol? economy), access, and stability. Only to regional cooperation on currency Donald F. Larson and Gunnar Breustedt in efficiency are African markets well unions. WPS4127 Poverty and Environmental Impacts of Electricity Price Reforms in Montenegro behind the rest--a deficiency that Patricia Silva, Irina Klytchnikova, and again can be associated with gaps in Dragana Radevic Patrick Honohan and Thorsten Beck. 2007. Mak- the information infrastructure. ing Finance Work for Africa. Washington, Papers can be downloaded at http://econ.worldbank.org The shortcomings of African fi- D.C.: World Bank. To download the World Bank Research E-Newsletter, go to Data nance can be attributed largely to the & Research at http://www.worldbank.org The World Bank Research Digest is a quarterly publica- The Research Digest is financed by the Bank's Editorial Committee: Jean-Jacques Dethier (man- tion aimed at disseminating findings of World Bank Research Committee and managed by DECRS, the aging editor), Asli Demirgüç-Kunt, and Alan Gelb. research. The views and interpretations in the articles research support unit of the Development Economics Editor: Alison Strong; research assistance: Thi are those of the authors and do not necessarily repre- Vice Presidency (DEC). The Research Digest is not Trang Linh Phu; production: Evelyn Alfaro-Bloch. sent the views of the World Bank, its Executive Direc- copyrighted and may be reproduced with appropriate Information or free subscriptions: send email to tors, or the countries they represent. source attribution. researchdigest@worldbank.org or visit http://econ.worldbank.org/research_digest The World Bank 1818 H Street, NW Washington, DC 20433, USA Printed on Recycled Paper