48384 ResearchDigest WorldBank Volume 2 l NumbeR 1 l fall 2007 "Fiscal Space"--or Fiscal Policy under Constraints? l Roberto Perotti The notion of fiscal space is not be moved from the second to the In ThIs Issue particularly insightful. The key first. In the end it all boils down to "fiscal Space"--or fiscal Policy under question is what constitutes making an assessment of fiscal poli- Constraints? ... page 1 an optimal fiscal policy under cies based on their economic effects The concept of "fiscal space" may have constraints and their social welfare impact. simply confused the debate on fiscal policy T But wasn't this the problem all constraints he debate on the constraints along? The relationship between foCuS: Climate Change and agriculture in facing fiscal policy started in fiscal policy and macroeconomic africa ... page 2 the aftermath of the financial stability has always been central to African farmers are at the forefront of crises of the late 1990s. Many de- the debate on the effects of fiscal vulnerability to climate change. Some are veloping countries were perceived policy in developing countries. The already adapting as sacrificing their long-run growth notions of fiscal space and its close potential so as to show some sign cousin, macroeconomic space, suf- How Well Do Institutional Theories explain of fiscal discipline. The drive toward fer from loose definitions and so can firms' Perceptions of Property fiscal discipline, demanded by the be a source of confusion. Even when Rights? ... page 3 market and international organiza- correctly defined, they are not par- Why have some countries developed strong tions and later willingly pursued ticularly insightful. The key question protection of property rights--and others by many of these countries, could is what constitutes an optimal fiscal haven't? be ascribed to various reasons (not policy under constraints--the main all independent): fears of unsus- ones being the intertemporal gov- Disability, Poverty, and Schooling in tainability, credit constraints, or ernment budget constraint, credit Developing Countries ... page 4 precautionary savings. There was constraints, and political constraints. Disability is associated with long-run strong pressure on international or- The facts are uncontroversial: In poverty--because it often means greater ganizations to help relax these con- developing countries fiscal policy difficulty in getting an education straints and to provide a framework is typically procyclical--the budget Campaign Contributions and Preferential for thinking about the issue. Thus deficit is positively correlated with access to finance ... page 5 the concept of "fiscal space," which, measures of economic activity such How do firms receive political favors? In according to a recent survey paper as the output gap. By contrast, in emerging markets one important channel is by Perotti, has confused rather than OECD countries there is a zero or preferential access to finance helped the debate. negative correlation. These patterns For Perotti, this notion of fiscal of correlation seem to be due mostly micro-Decomposing the macro space is essentially a restatement to government spending (though Determinants of Human of the intertemporal government a sound assessment of the cyclical Development ... page 6 budget constraint. We know that to properties of tax rates is not pos- Do growth and income distribution explain increase some type of government sible, since we lack good information gains in school enrollment--or are other spending today, we need to reduce on tax rates, except inflation, for factors more important? other types of spending now or in developing countries). The positive the future, increase current or future correlation between output gap and Teacher Compensation and Decentralization in India ... page 7 revenues, or inflate away existing deficit or spending is particularly nominal debt. We also know that if pronounced during downturns. Both What's behind many problems in India's public education sector? Its system of teacher one type of spending has a higher compensation social marginal value than another and the same cost, resources should (continued on page 8) WorldBankResearch Digest FOCUS Climate Change and Agriculture in Africa Climate change will have a response function, the authors simu- ately recognize that the climate has significant impact on African late climate change scenarios, both changed. uniform across Africa and location Using responses to open-ended agriculture. How farmers adapt will specific. The results suggest that not survey questions, Maddison deter- matter all countries are equally vulnerable to mines how well farmers in Africa are A climate change--with predicted vul- able to detect climate change and how frican farmers and agricultural nerability depending on temperature they have adapted to whatever climate systems are at the forefront of and precipitation changes, whether a change they believe has occurred. vulnerability to climate change. country is already hot and dry, and the While significant numbers of farmers Yet until recently there has been little extent to which farms are irrigated. believe that temperatures have already or no comprehensive understand- Maddison, Manley, and Kuruku- increased and that precipitation has ing of vulnerability to climate change lasuriya conduct a similar analysis, declined, those with the greatest ex- across Africa--critical to guide con- except that they use farmers' own per- perience in farming are more likely to certed and focused policy and invest- ceptions of the value of their land (as notice climate change. ment interventions needed to protect in a land sale). The results reinforce Among adaptations made in agriculture, which is central to many the findings of the analysis based on response to climate change, planting African economies. climate variables. Using additional different varieties of the same crop A recent study, synthesized variables, this second analysis finds and changing planting dates are in Climate Change and Agriculture in that distance to markets is an impor- important everywhere. But stratifying Africa, undertook a comprehensive tant factor in farmers' ability to adapt the data by the precise perception quantitative analysis and assessment to climate change. Country differences of climate change provides greater of potential economic impacts of stemming from institutions and regu- insights. Where temperature changes future climate change on agriculture-- lations also affect adaptability. are perceived, farmers plant different and the value of measures to adapt Two approaches are used to mea- varieties, shift from farming to to climate change--for different sure adaptation. In the first, based on nonfarming activities, increase water zones, regions, countries, and farm observed household behavior, Kuru- conservation, and use shading and types in Africa. Four papers address kulasuriya and Mendelsohn examine sheltering techniques. Where changes different aspects of the climate whether the choice of crops is affected in precipitation (especially in the change­agriculture nexus, using by climate in Africa. Not surprisingly, timing of the rains) are perceived, a survey of nearly 10,000 farmers crop choice appears to be very climate varying the planting date appears to be sampled across West Africa (Burkina sensitive. The crop mixes selected by Faso, Cameroon, Ghana, Niger, farmers differ, depending on whether Senegal), southern Africa (South the region is cooler, moderately warm, (continued on page 7) Africa, Zambia, Zimbabwe), East or hot--and whether conditions are Africa (Ethiopia, Kenya), and North dry, medium wet, or wet. Africa (Arab Republic of Egypt). The findings in this paper suggest The analytical framework is based that as temperatures warm, farmers Ariel Dinar, Rashid Hassan, Robert Mendelsohn, on the presumption that there is will shift toward more heat-tolerant James Benhin, and others. Forthcoming. Climate a relationship between the value crops. And as precipitation increases Change and Agriculture in Africa: Impact of land (if available), or the net or decreases, they will shift toward Assessment and Adaptation Strategies. revenue from growing crops, and water-loving or drought-tolerant crops. London: EarthScan. climate, while controlling for water Moreover, while the analysis examines Pradeep Kurukulasuriya and Robert Mendelsohn. 2007. "Crop Selection: Adapting to Climate availability (runoff), soil quality, and choices only among current crops, Change in Africa." Policy Research Working Paper socioeconomic variables. farmers may well have more choices 4307. World Bank, Washington, D.C. Kurukulasuriya and Mendelsohn, where new varieties, more suited to ------. 2007. "A Ricardian Analysis of the in a Ricardian analysis of the impact higher temperatures and lower or high- Impact of Climate Change on African Cropland." of climate change, find that net rev- er rainfall, are developed. Policy Research Working Paper 4305. World enues fall as precipitation falls or as Is the observed behavior of farm- Bank, Washington, D.C. temperatures rise across all surveyed ers in adapting to the present climate David Maddison. 2007. "The Perception of and farms. Examining rainfed and irrigated across Africa optimal? Maddison raises Adaptation to Climate Change in Africa." Policy farms separately, they also find that doubts about whether farmers know Research Working Paper 4308. World Bank, Washington, D.C. rainfed farms are especially climate immediately what constitutes the David Maddison, Marita Manley, and Pradeep sensitive. Irrigated farms, located in best response to climate change when Kurukulasuriya. 2007. "The Impact of Climate relatively cooler parts of Africa, have the agricultural practices it requires Change on African Agriculture: A Ricardian Ap- an immediate positive response to are outside their range of experience. proach." Policy Research Working Paper 4306. warming. Using a climate sensitivity Nor can they be expected to immedi- World Bank, Washington, D.C. WorldBankResearch Digest How Well Do Institutional Theories Explain Firms' Perceptions of Property Rights? classes of firms will have their property rights better protected than others. In l Meghana Ayyagari, Asli Demirgüç-Kunt, and Vojislav Maksimovic oligarchic societies, for example, large incumbent firms would be expected What explains differences in the two most influential legal tradi- to have greater protection of property property rights protection? Legal tions, British common law and French rights than smaller firms. But these origin may be less important than civil law, this theory holds that legal firm-level differences are mostly down- believed, and firm characteristics traditions differ in the priority they at- played by the institutional theories more so tach to protecting the rights of private that focus on country-level differences. investors against the state. As coloni- By considering property rights protec- P zation spread around the world, British tion at the firm level, the paper sheds rotection of property rights is a colonizers brought with them a legal light on the likely importance of firm key factor in the efficient opera- tradition that stressed private property characteristics such as size, private tion of contracts and the devel- rights protection, while French coloniz- ownership, incorporation, and industry opment of financial institutions. The ers spread a legal tradition less condu- classification. critical question is why some coun- cive to such protection. The results show that the institu- tries have managed to develop strong Second, the endowment view em- tional variables explain about 50 per- protection of property rights while phasizes the role of geography and cent of the cross-country variation in others have not. the disease environment in shaping property rights protection, suggesting A substantial body of theoretical the institutional environment and the that current research on institutions work tries to explain the historical de- underlying property rights. In this view does indeed address first-order ef- terminants of these differences. There it is not the identity of the colonizer fects. A country's legal origin predicts is also a growing body of empirical but the colonization strategy that de- property rights protection better than work that assesses the relative contri- termined the extent of property rights its religion, ethnic fractionalization, or bution of different historical determi- protection. According to this view, natural endowments. nants in the cross-country variation of countries that are closer to the equa- But this result depends critically on property rights protection. The most tor, with a more tropical climate, were sample selection. Removing formerly influential work has emphasized the more inhospitable to European set- socialist economies from the sample importance of legal traditions, initial tlers and therefore fostered extractive significantly reduces the explanatory endowments, ethnic fractionalization, institutions rather than institutions power of legal origin. In this smaller and culture and religion. that protect property rights. sample ethnic fractionalization and A paper by Ayyagari, Demirgüç- Third, political theories predict that endowments are more important in Kunt, and Maksimovic takes a more governments become more interven- explaining the variation in property micro approach, using firm-level tionist as the ethnic heterogeneity of rights protection. Since formerly so- survey data to study how well firm- countries increases, with the groups cialist economies have much more in level and institutional factors explain in power implementing policies that common than legal tradition, these entrepreneurs' perceptions of prop- expropriate as much as possible from two variables are probably capturing erty rights protection. The authors other ethnic groups. Thus the ethnic country characteristics that are not use survey responses from more than diversity view would predict that coun- necessarily caused by differences in 7,500 firms in 80 countries to compare tries with greater ethnic fractionaliza- legal systems. the different theories and to examine tion are less likely to protect property Focusing on firm-level differences, the relative influence of firm effects rights. the authors find that such variables compared with country effects. The Fourth, many scholars argue that as size, ownership structure, and analysis is based on variance decom- religion shapes national views on pro- organizational form are comparable position, which allows the authors tection of property rights. For example, in importance to institutional factors to focus directly on the relative im- they argue that the Catholic religion in explaining variations in property portance of the different effects in fostered authoritarian societies with rights protection--and in some cases explaining property rights without any powerful bonds between church and even more important. These results assumptions on causality or structural state, limiting protection of private suggest that firm-level differences, analysis. property rights. Thus the fourth view, which have received little attention The paper evaluates four potential culture and religion, predicts that differ- in the literature, are also significant historical determinants of property ences in religion and in the systems in explaining perceptions of property rights protection. First, the law and of beliefs and culture that stem from rights protection. finance view predicts that historically such differences can explain differenc- Meghana Ayyagari, Asli Demirgüç-Kunt, and determined differences in legal tra- es in property rights protection across Vojislav Maksimovic. Forthcoming. "How Well Do ditions help explain differences in countries. Institutional Theories Explain Firms' Perceptions of protection of property rights today. Finally, implicit in several of these Property Rights?" Review of Financial Focusing on the differences between theories is the prediction that some Studies. WorldBankResearch Digest Disability, Poverty, and Schooling in in poorer households--though much Developing Countries of this association comes from the fact that adults with disabilities have lower educational attainment, which leads l Deon Filmer to lower economic status. able to use more than one limb; para- Given this finding, it is particularly Children with disabilities are less lyzed lower limbs only; paralyzed all worrisome that youth with disabilities likely to acquire the education they four limbs." In the Jamaica survey, by are almost always substantially less need to earn high incomes and avoid contrast, there is one category, defined likely to participate in schooling. Chil- poverty simply as a household member having dren with disabilities are less likely a "physical or mental disability." The to start school and in some countries T Mongolia survey describes sight and have lower transition rates, resulting he target of universal education hearing problems as seeing or hearing in lower schooling attainment. The remains elusive: worldwide, "with difficulty," while others typically deficit in school participation associ- around 100 million children of characterize such problems as "blind" ated with disability is typically larger primary school age are not in school. and "deaf." than those associated with other char- Children with disabilities face par- Despite these limitations the data acteristics, such as gender, rural resi- ticular hurdles in attending, and are revealing. Consistent with similar dence, or economic status (figure 1). completing, school in developing surveys, the 14 surveys identify around This analysis suggests that disabil- countries. 1­2 percent of the population as hav- ity is associated with long-run poverty: While there has been policy ing a disability. Data for Cambodia, children with disabilities are less likely discussion about interventions to with two surveys and varying defini- to acquire the human capital that will increase access to schooling for chil- tions, suggest that the percentage allow them to earn higher incomes. dren with disabilities, there has been is not always sensitive to the exact In all countries the schooling gap be- little systematic empirical analysis definition: different definitions can tween children with a disability and on which to base this policy. A large give similar prevalence rates, and those without one starts at grade 1. part of the reason for this is the lack vice versa. In addition, other aspects That suggests that efforts to boost the of appropriate and comparable data. of the surveys, such as the training enrollment of children with disabili- A new study by Filmer aims to start of enumerators or the use to which ties are needed at the earliest grades. filling some gaps in knowledge using interviewees expect the survey to be And the finding that the disability existing Demographic and Health Sur- put, might affect estimated prevalence deficit widens from grade to grade veys, Multiple Indicator Cluster Sur- rates. in countries that have achieved high veys, and other Integrated Household The surveys provide little evidence enrollment among children without a Surveys. The 14 surveys--the earliest to suggest that youth with disabilities disability suggests that special effort from 1992 and the most recent from are generally more or less likely to live may be needed to keep youth with 2004--provide data on the prevalence in richer or poorer households. Adults disabilities in school--once they have of disability and its association with with disabilities, however, typically live started attending--in all countries. poverty and schooling in 1 transition The findings of the analysis should and 12 developing economies. be treated as tentative. The goal was Defining disability is complicated-- bolivia to exploit existing data to help orient and controversial. Purely medical def- burundi policy, but the clearest message is initions are giving way to definitions Cambodia (DHS) that better data are needed. Establish- that incorporate continuous measures Cambodia (SeS) ing clear and consistent measures of of the activities that people can un- Chad disability for household surveys and dertake, the extent of participation in Colombia national censuses would be a start. A society and social and civic life, and India recent review suggests that questions the role of adaptive technologies. The Indonesia focusing on functionality, limited to a definitions of disability in the data Jamaica core set of activities, and allowing for sets used in the analysis are most mongolia variation in the degree of functional closely consistent with a focus on im- mozambique limitation (rather than simply the pairment--such as having a missing Romania presence or absence of a limitation) limb or limited or no sight. South africa should be preferred. But there is much variation across Zambia Implementing these questions the surveys. In the Cambodia Socio- in samples with sufficient observa- -80 -60 -40 -20 0 20 Economic Survey, for example, the tions to allow detailed analysis will be disability question includes a detailed Disability female Rural Poorest quintile relative to richest needed to build the quantitative evi- list of potential cases: "amputation of one limb; amputation of more than one limb; unable to use one limb; un- (continued on page 5) WorldBankResearch Digest Campaign Contributions and the four years after the election. This Preferential Access to Finance suggests that contributing firms in- creased in value because they gained preferential access to finance from l Stijn Claessens, Erik Feijen, and Luc Laeven banks. Claessens, Feijen, and Laeven at- Preferential access to finance is tional development and distortions tempt to quantify the overall cost of an important channel for political in real and financial markets, political extending preferential bank credit favors for firms in emerging markets connections are likely to have greater to contributing firms. A welfare loss like Brazil value than in the more developed would arise if the rate of return on the I countries studied to date. And fourth, investment financed by the credit is n many countries, especially de- much of this value could come from lower than that on resources invested veloping ones, close links between preferential access to finance, since elsewhere. They estimate this invest- politicians and business are asso- Brazil has among the highest interest ment distortion cost by comparing ciated with rent-seeking behavior and rates and lowest degrees of financial the return on investment generated distortions in the institutional envi- intermediation in the world. by contributing firms with that of ronment. The result: lower and less Claessens, Feijen, and Laeven in- noncontributing firms. This is a lower- efficient growth and more unequal troduce two methodological improve- bound measure, since there are likely distribution of income and wealth. ments over other studies. First, they to be many other costs associated Evidence of these detrimental ef- improve on the relatively rough mea- with the investment distortion that the fects exists at the very general level, sures of political connections often authors do not capture. They estimate as in the negative relationships be- used in the non-U.S. literature. Using the cost to be at least 0.2 percent of tween corruption and growth and a novel data set of firm- and candi- GDP a year--a significant amount. general development. But empirical date-level campaign contributions, Finance may not be the only chan- evidence on the specific mechanisms they determine whether a connection nel through which firms benefit from and channels through which business exists between a specific politician political connections. But the results gains favors from politicians is by and a specific firm--and, if so, mea- obtained by the authors support the nature more scarce. The lack of such sure the intensity of the connection notion that it is an important one. evidence hinders the identification of rather than only indicating its pres- More generally, their findings provide concrete policy recommendations for ence. Second, they mitigate the omit- new evidence on the cost of politi- remedying the effects. ted variables problem that plagues cal connections in emerging markets In a new paper Claessens, Feijen, the literature by using panel data to such as Brazil. This corroborates other and Laeven attempt to identify a exploit the variation over time in po- evidence that the rents from corrup- specific channel by studying the link litical connections. tion are particularly large where there between campaign contributions The authors find that better con- are government-imposed distortions, and business favors. Studies on the nected firms (that is, those that give with negative welfare effects. And it relationship between campaign con- relatively more campaign contribu- implies that the operation of corpora- tributions and policy outcomes have tions) have significantly higher stock tions in such environments, including generally been inconclusive because market returns around the time their financing and financial structure, of an inability to distinguish between election results are announced and depends on their relationships with the matching of politicians' voting that the intensity of the connection politicians. dispositions with firms' preferences matters. Contributions to candidates and politicians' incentives to provide for the federal congress who win the Stijn Claessens, Erik Feijen, and Luc Laeven. Forthcoming. "Political Connections and Preferen- contributors with specific favors. The election have an even larger positive tial Access to Finance: The Role of Campaign Con- authors overcome this problem by impact on stock returns than contri- tributions." Journal of Financial Economics. looking at the relationships between butions to candidates who lose the firm-specific campaign contributions, election. This empirical evidence (continued from page 4) firm valuation, and firm-specific pref- strongly supports the notion that, in erential access to finance around the an environment with many distortions, dence base for empirically grounded times of the 1998 and 2002 elections political connections can increase firm policies. An important complement to in Brazil. value. that analysis would be evaluations of Brazil is well suited for such a The authors next investigate the impact of alternative interventions study. First, it is one of the few coun- whether the channel for the increase to increase the enrollment of children tries that require each candidate to in value runs through preferential ac- with disabilities. register campaign contributions and cess to bank finance. They find that justify campaign expenditures. Sec- for firms that made contributions to Deon Filmer. Forthcoming. "Disability, Poverty, ond, Brazil is known for strong rela- (elected) federal deputies, the shares and Schooling in Developing Countries: Results tions between politicians and firms. of financing obtained through bank from 14 Household Surveys." World Bank Eco- Third, given Brazil's level of institu- loans increased substantially during nomic Review. WorldBankResearch Digest Micro-Decomposing the Macro Determinants of Human Development and (observed) nonincome factors to schooling attainments. This holds l Sylvie Lambert, Martin Ravallion, nationally as well as within urban and and Dominique van de Walle rural areas, for majority and minor- ity ethnic groups in Vietnam, and for literate and illiterate groups in Neither economic growth nor aggregating the empirical micro rela- Morocco. The decomposition cannot income redistribution was tionships and so throwing light on the tell us what drives these structural important in driving enrollment macroeconomics of human develop- changes, since they are economywide gains in Morocco and Vietnam. The ment. The analytical complication over factors. But it suggests that there were explanations lie elsewhere other decompositions (in the Blinder- substantial public policy efforts to Oaxaca tradition) is that to convinc- increase enrollments and that these W ingly estimate the decomposition the were successful; increases in the over- hat are the proximate deter- authors must use a flexible, nonpara- all economic returns to schooling may minants of aggregate human metric representation of the economic also have played a part. development outcomes? gradient in human development out- The results look very different What is the role of economic growth comes across households. when the cross-sectional disparities in and of such factors as changes in The paper shows that changes in schooling between socioeconomic or income inequality or public policy aggregate human development out- geographic groups are studied. Struc- reform? comes can be additively decomposed ture becomes generally secondary There have been two approaches into four components: a pure growth to differences in mean incomes and to examining these questions in the effect associated with differences in nonincome factors and, though less past. Cross-country comparisons mean income; a redistribution effect often, to intergroup differences in the have found that attainments in basic attributed to differences in the dis- distribution of income. health and education tend to be high- tribution of income; nonincome fac- Differences in secondary-level er in higher-income countries and, tors (such as parental education and enrollments between Vietnam's ur- less obviously, that economic growth distance to school); and a structural ban and rural areas are due largely tends to be accompanied by improve- component reflecting any differences to differences in mean consumption. ments in social indicators. But the in the model parameters--the human In Morocco, by contrast, nonincome implications for development policy development returns to income or factors are dominant in explaining the remain unclear. Is the explanation nonincome characteristics. urban-rural differences in enrollments that higher average income allows a The decomposition method is used at both primary and secondary levels. society to buy goods and services that to study changes over time as well as Disparity in mean consumption is promote health and schooling? Or is the intergroup disparities in school also a major factor in the education it that there is a correlation between enrollments for boys and for girls inequality between Vietnam's ethnic average incomes and other country in Morocco and Vietnam during the minority and majority groups, though characteristics, such as better public 1990s. Enrollments have risen in both the effect of this economic inequal- services? countries but are appreciably higher in ity has shifted over time from primary Another approach has examined Vietnam. Real consumption per per- to secondary schooling. Nonincome the micro determinants of human son rose appreciably over the period factors also play a big part here, in- development attainments. This lit- in Vietnam (5.8 percent a year) but creasing the gap in primary school erature too has pointed to the role fell in Morocco. Inequality remained enrollments in both 1993 and 1998 but of income and its distribution, but it stable in Morocco while it rose in reducing ethnic group differences in has also emphasized a wide range of Vietnam. secondary enrollments in 1998. nonincome characteristics, includ- Taken as a whole, the results for Finally, income differences also ing those of households, providers, the two countries do not suggest explain a sizable share of the enroll- and geographic location. But this that aggregate economic growth or ment gap between children with and literature has not effectively explored changes in the distribution of income without literate fathers in Morocco, the implications for the aggregate have been important determinants of though here too nonincome factors relationship between human develop- improvements in aggregate education are dominant. ment attainments in a country and attainments. What is probably most economic growth, the distribution of surprising in these findings is how income, and nonincome factors. little education benefit is attributed to A new paper by Lambert, Raval- the robust economic growth observed Sylvie Lambert, Martin Ravallion, and Domi- lion, and van de Walle aims to help in Vietnam over this period. nique van de Walle. 2007. "A Micro-Decomposi- bridge this gap. It develops and Most of the changes over time are tion Analysis of the Macroeconomic Determinants implements a micro-based decom- attributable to changes in the struc- of Human Development." Policy Research Working position method for consistently ture of the model linking incomes Paper 4358. World Bank, Washington, D.C. WorldBankResearch Digest Teacher Compensation and But simply shifting the existing Decentralization in India system--with its lack of performance orientation, lack of external account- l Lant Pritchett and Rinku Murgai ability, and strictures on compensa- tion--to the PRIs is unlikely to lead to The system of teacher compensation, · The pay of public sector teachers improvements. That said, decentral- combining high pay and zero has very little variance even potentially ization to PRIs, if done well, has the accountability, is at the heart of related to performance--much less so potential to break the political impetus problems in India's public education than the pay of either private sector behind business as usual by combin- teachers or other private sector sala- ing a reallocation of functions across P ried workers. tiers of government (states and PRIs) ublicly produced elementary ed- Each of these elements of the with the freedom for PRIs to develop ucation in India faces enormous system of compensation reinforces the systems of compensation aligned with problems. Although enrollments lack of accountability. There is nothing the realities of public employment and have increased, a recent survey of rural in the system to attract people well the particularities of the practice of areas in all districts found shockingly matched to teaching, to retain the teaching. low levels of learning achievement, best and most committed teachers, With the development of a new confirming the accumulating piece- or to motivate good teachers (or, for cadre of teachers under district con- meal evidence. There are many other that matter, to prevent good teachers trol, newly hired teachers can be indicators of distress: high levels of from becoming disillusioned, cynical, launched into a new system and sail dissatisfaction with teachers among and embittered and yet stay until out of today's perfect storm--and pub- parents and students, a massive and they reach the age of 60). Moreover, licly produced schooling will be able to ongoing shift into private schooling, the institutional context of basic compete with private alternatives. and the unhappiness of the public sec- schooling--all the other relationships tor teachers themselves. In a new pa- of accountability--is also weak. Lant Pritchett and Rinku Murgai. Forthcoming. per Pritchett and Murgai argue that the This system of compensation plays "Teacher Compensation: Can Decentralization to system of teacher compensation in the a large part in producing today's "per- Local Bodies Take India from the Perfect Storm public sector is at the heart of many of fect storm" in public schooling: the through Troubled Waters to Clear Sailing?" India these problems. learning achievement of students is Policy Forum. Washington, D.C.: Brookings A system of compensation for any low, the absenteeism of teachers is Institution. Also available at http://lpritchett.org/ research_education.htm. performance-oriented organization high, the treatment of students by should attract, retain, and motivate teachers is often abysmal, recourse workers who, on a day-to-day basis, to private instruction is rampant, pursue the goals of the organization. parents and students are dissatisfied All four elements of a system of com- with government schools, and people FOCUS pensation--durability of the employ- are voting with their feet and their Climate Change and ment relationship, structure of pay pocketbooks for the private sector. Agriculture in Africa across states of the world, assignment Perhaps worst of all, the potentially (continued from page 2) of workers to tasks, and cash relative good teachers in the public system to benefits--should work together to- are disenchanted and overburdened an important response. There is also ward this goal. and feel disrespected by parents and evidence that adaptation measures While there are many variations management. Any reform of teacher are linked to baseline climate and that across states in India, it is not unfair compensation needs to be pro-teacher adaptation occurs mainly on sites that to describe the system of teacher because the present system is dramati- are already marginal (hot and dry). compensation as combining high pay cally anti-teacher. The propensity of farmers to adapt and zero accountability. The paper Most observers agree that there differs across locations, and under- documents four facts about the sys- is no possibility of significant reform standing the underlying factors would tem of teacher compensation: of teacher compensation in the ex- require further analysis. Yet only those · There is little or no ability to isting system. But the devolution of who perceive climate change under- separate teachers from service--for education to the local governing bod- take adaptation--and the perception any cause. ies known as Panchayati Raj institu- of climate change appears to hinge on · The average pay in public sector tions (PRIs) provides an opportunity farmers' experience and the availability teaching is very high relative to the to sail out of this perfect storm--to of affordable technologies, manage- pay in alternative employment (both completely restructure the system of ment practices, and extension advice private teaching and other private sec- compensation to be consistent with related to climate change. While the tor jobs). an accountable, performance-oriented policy options for promoting greater · The degree of overpayment is public sector. Decentralization to PRIs awareness of climate change may be higher for public sector teachers at the is certainly no panacea--but it may limited, the perception of climate early stages of a career. well be the last best hope. change is already widespread. WorldBankResearch Digest (continued from page 1) public investment and social spending by cutting productive government Recent Policy Research (particularly targeted programs) fall spending and thus future growth and Working Papers as a percentage of GDP during down- revenues. International organizations turns. should instead have gone against the 4301 India's Journey toward an effective Patent What explains the observed procy- market, they assert, by shifting the fo- System clical behavior of fiscal policy in devel- cus from the short to the long run. Bruce Abramson 4328 Does IDa engage in Defensive lending? oping countries? Three explanations But the alternative is self- Carolin Geginat and Aart Kraay have been put forward. First, and most insurance, which should be seen as 4346 Infant mortality over the business Cycle in obvious, is credit constraint facing the just a form of precautionary savings the Developing World Sarah Baird, Jed Friedman, and Norbert government: it is precisely when ex- against noncatastrophic shocks. How Schady ternal credit is likely to dry up that the can an institutional setup that is more 4351 leakage of Public Resources in the Health Sector: an empirical Investigation of Chad government may need to increase the conducive to such self-insurance be Bernard Gauthier and Waly Wane budget deficit. Second, there may be created? Many stabilization funds and 4353 The long-Run Impact of orphanhood Kathleen Beegle, Joachim De Weerdt, a "hidden deficit": sudden stops could fiscal rules (such as numerical limits and Stefan Dercon be associated with fiscal cuts required on the deficit) have been created in 4357 Structure and Performance of the Services Sector in Transition economies to make room for the costs of bailing developing countries with the dual Ana M. Fernandes out the financial system and other aim of ensuring discipline in fiscal 4362 labor market Policy in Developing Countries: a Selective Review of the literature and contingent liabilities. Finally, several policy and avoiding a procyclical bias Needs for the future politico-economic reasons have been in the presence of credit constraints. Gary S. Fields put forward, including the "starve the It is often argued that such fis- 4367 The Vanishing farms? The Impact of International migration Leviathan" argument of Alberto Ale- cal rules are difficult or impossible Juna Miluka, Gero Carletto, Benjamin Davis, and Alberto Zezz sina and Guido Tabellini. Voters can- to implement. Using the example of 4369 Coordinating Public Debt management with not observe fiscal policy precisely, par- Chile, Perotti shows that this is not fiscal and monetary Policies: an analytical ticularly the liabilities accumulated by necessarily so and that sound fiscal framework Eriko Togo the government. When the economy rules on the cyclically adjusted deficit 4375 The Determinants of Rising Informality in is booming, to prevent the govern- are feasible. The reason that we see brazil: evidence from Gross Worker flows Mariano Bosch, Edwin Goni, and William ment from appropriating tax revenues, so little self-insurance has to do with Maloney voters demand more public goods for political economy: in good times it is 4377 early Identification of at-Risk Youth in latin america: an application of Cluster themselves, imparting a procyclical politically impossible to put aside tax analysis bias to fiscal policy. The implication, revenues while withstanding accusa- Emilie Bagby and Wendy Cunningham one for which there is some empiri- tions of neglecting health, education, 4378 Do Regulation and Institutional Design matter for Infrastructure Sector Performance? cal support, is that the procyclicality and infrastructure. Luis Andres, Jose Luis Guasch, and of fiscal policy should be stronger in Stephane Straub 4379 Designing economic Instruments for the more corrupt democracies. environment in a Decentralized fiscal System Critics of the advice of interna- James Alm and H. Spencer Banzhaf tional organizations in the aftermath Roberto Perotti. 2007. "Fiscal Policy in Developing of the financial crises argue that the Countries: A Framework and Some Questions." fiscal policy that was implemented Working Papers can be downloaded at http://econ.worldbank.org Policy Research Working Paper 4365. World To download the WorldBankResearchE-Newsletter, actually reduced countries' net worth, Bank, Washington, D.C. go to Data & 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 (managing tion aimed at disseminating findings of World bank Research Committee and managed by DeCRS, the editor), asli Demirgüç-Kunt, and alan Gelb. editor: research. The views and interpretations in the articles research support unit of the Development economics alison Strong; research assistance: Thi Trang linh Phu; are those of the authors and do not necessarily repre- Vice Presidency (DeC). The Research Digest is not production: Roula I. Yazigi. Information or free sub- sent the views of the World bank, its executive Direc- copyrighted and may be reproduced with appropriate scriptions: send email to researchdigest@worldbank. tors, or the countries they represent. source attribution. org or visit http://econ.worldbank.org/research_digest The World Bank 1818 H Street, NW Washington, DC 20433, USA Printed on Recycled Paper