AN ESSAY DRAWN FROM 2 0 1 0 53844 Silent and lethal How quiet corruption undermines Africa's development efforts 2 0 1 0 Copyright © 2010 by the International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing 2010 This volume is a product of the staff of the International Bank for Reconstruction and Development/The World Bank. The find- ings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dis- semination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. To order Africa Development Indicators 2010, The Little Data Book on Africa 2010, the Africa Development Indicators 2010­Multiple User CD-ROM, please visit www.worldbank.org/publications. To subscribe to Africa Development Indicators Online please visit http://publications.worldbank.org/ADI. For more information about Africa Development Indicators and its companion products, please visit www.worldbank.org/africa. You can email us at ADI@worldbank.org. Cover design by Communications Development Incorporated. Photo credits: front cover, Mark Evans/iStockphoto; back cover, Arne Hoel/World Bank. The map of Africa is provided by the Map Design Unit/World Bank. SKU: 18202 Contents Foreword vii Acknowledgments ix Executive summary xi Silent and lethal: How quiet corruption undermines Africa's development 1 Notes 23 References 25 Indicator tables Users guide 33 Part I. Basic indicators and national and fiscal accounts 1. Basic indicators 1.1 Basic indicators 37 2. National and fiscal accounts 2.1 Gross domestic product, nominal 38 2.2 Gross domestic product, real 39 2.3 Gross domestic product growth 40 2.4 Gross domestic product per capita, real 41 2.5 Gross domestic product per capita growth 42 2.6 Gross national income, nominal 43 2.7 Gross national income, Atlas method 44 2.8 Gross national income per capita, Atlas method 45 2.9 Gross domestic product deflator (local currency series) 46 2.10 Gross domestic product deflator (U.S. dollar series) 47 2.11 Consumer price index 48 2.12 Price indexes 49 2.13 Gross domestic savings 50 2.14 Gross national savings 51 2.15 General government final consumption expenditure 52 2.16 Household final consumption expenditure 53 2.17 Final consumption expenditure plus discrepancy 54 2.18 Final consumption expenditure plus discrepancy per capita 55 2.19 Gross fixed capital formation 56 2.20 Gross general government fixed capital formation 57 2.21 Private sector fixed capital formation 58 2.22 External trade balance (exports minus imports) 59 2.23 Exports of goods and services, nominal 60 2.24 Imports of goods and services, nominal 61 2.25 Exports of goods and services as a share of GDP 62 2.26 Imports of goods and services as a share of GDP 63 Contents iii 2.27 Balance of payments and current account 64 2.28 Exchange rates and purchasing power parity 66 2.29 Agriculture value added 68 2.30 Industry value added 69 2.31 Services plus discrepancy value added 70 2.32 Central government finances, expense, and revenue 71 2.33 Structure of demand 75 Part II. Millennium Development Goals 3. Millennium Development Goals 3.1 Millennium Development Goal 1: eradicate extreme poverty and hunger 76 3.2 Millennium Development Goal 2: achieve universal primary education 79 3.3 Millennium Development Goal 3: promote gender equality and empower women 80 3.4 Millennium Development Goal 4: reduce child mortality 81 3.5 Millennium Development Goal 5: improve maternal health 82 3.6 Millennium Development Goal 6: combat HIV/AIDS, malaria, and other diseases 83 3.7 Millennium Development Goal 7: ensure environmental sustainability 85 3.8 Millennium Development Goal 8: develop a global partnership for development 87 Part III. Development outcomes 4. Private sector development 4.1 Doing Business indicators 89 4.2 Investment climate 92 4.3 Financial sector infrastructure 94 5. Trade and regional integration 5.1 International trade and tariff barriers 96 5.2 Top three exports and share in total exports, 2007 100 5.3 Regional integration, trade blocs 102 6. Infrastructure 6.1 Water and sanitation 104 6.2 Transportation 105 6.3 Information and communication technology 107 6.4 Energy 109 Participating in growth 7. Human development 7.1 Education 111 7.2 Health 113 8. Agriculture, rural development, and environment 8.1 Rural development 117 8.2 Agriculture 119 8.3 Environment 121 8.4 Fossil fuel emissions 124 9. Labor, migration, and population 9.1 Labor force participation 126 9.2 Labor force composition 128 9.3 Unemployment 130 9.3 Migration and population 132 iv Africa Development Indicators 2010 10. HIV/AIDS 10.1 HIV/AIDS 134 11. Malaria 11.1 Malaria 138 12. Capable states and partnership 12.1 Aid and debt relief 139 12.2 Status of Paris Declaration indicators 142 12.3 Capable states 144 12.4 Governance and anticorruption indicators 146 12.5 Country Policy and Institutional Assessment ratings 148 12.6 Polity indicators 152 Technical notes 153 Technical notes references 195 Users guide: Africa Development Indicators 2007 CD-ROM 197 Contents v Foreword "Sunlight is the best disinfectant," Associate disproportionately. Worse still, it can have Justice of the United States Supreme Court long-term consequences. Denied an educa- Louis Brandeis said in 1914, referring to the tion because of absentee teachers, children benefits of openness and transparency in suffer in adulthood with low cognitive skills tackling corruption in the public sector. To- and weak health. The absence of drugs and day, thanks to the efforts of Transparency In- doctors means unwanted deaths from ma- ternational and other organizations, there is laria and other diseases. Receiving diluted considerable "sunlight" on well known types fertilizer that fails to produce results, farm- of corruption--public officials demanding ers choose not to use any fertilizer, leaving and taking bribes for privileged access to con- them in low-productivity agriculture. tracts or exemptions from regulations. On Quiet corruption does not make the head- average, Africa scores poorly on these indica- lines the way bribery scandals do. It has yet to tors, with some exceptions--Botswana, Cape be picked up by Transparency International Verde, and Mauritius have consistently done and other global indexes of corruption. Tack- well, and Liberia has made great strides. ling quite corruption is at least as difficult as This year's Africa Development Indica- tackling grand corruption. It will require a tors essay sheds light on a different type of combination of strong and committed leader- corruption--what the authors call "quiet ship, policies and institutions at the sectoral corruption"--when public servants fail to level, and--most important--increased ac- deliver services or inputs that have been countability and participation by citizens, the paid for by the government. The most prom- demand side of good governance. By high- inent examples are absentee teachers in pub- lighting quiet corruption in this year's Af- lic schools and absentee doctors in primary rica Development Indicators--itself a tool for clinics. Others include drugs being stolen Africans to hold their governments account- from public clinics and sold in the private able--we hope that the essay will do for quiet market as well as subsidized fertilizer being corruption what Justice Brandeis intended diluted before it reaches farmers. with his famous aphorism. Not only is quiet corruption perva- sive and widespread in Africa, but--as Obiageli K. Ezekwesili the essay points out--it hurts the poor Vice President, Africa Region Foreword vii Acknowledgments Africa Development Indicators is a product of collaborated in the update of the live data- the Africa Region of the World Bank. base. Mehdi Akhlaghi collaborated in the pro- Jorge Saba Arbache was the director of duction of The Little Data Book on Africa 2010. this book and its companions--Africa De- Yohannes Kebede and Michael Mendale velopment Indicators Online 2010, Africa prepared the ADI Online data visualization Development Indicators 2010--Multiple platform. User CD-ROM, and The Little Data Book on Jeffrey Lecksell of the World Bank's Map Africa 2010. Rose Mungai led the work on Design Unit coordinated preparation of the data management, consistency checks, and map. compilation and provided overall data qual- The box on measuring the impact of re- ity assurance. The core team included Mpho gional integration in the technical notes was Chinyolo, Francoise Genouille, Jane K. prepared by Paul Brenton. Njuguna, and Christophe Rockmore. Jane Ann Karasanyi and Ken Omondi pro- Njuguna coordinated all stages of produc- vided administrative and logistical support. tion. Richard Crabbe provided useful pro- The team is grateful to the many people who duction support and suggestions. The over- provided useful comments on the publica- all work was carried out under the guidance tion especially Inger Andersen, Paul Brenton, of Shantayanan Devarajan, Chief Economist Aziz Bouzaher, Cecilia M. Briceno-Garmen- of the Africa Region. dia, Donald Bundy, Moulay Driss Zine Ed- Jorge Saba Arbache, James Habyarima- dine El Idrissi, Madhur Gautam, Giuseppe na, and Vasco Molini wrote the essay. Balu Iarossi, Elizabeth Laura Lule, John F. May, Bumb, Michael Morris, Giuseppe Iarossi, Steven Mink, Emmanuel Mungunasi, Vin- Gäel Raballand, Stephen Minck, Ian Grego- cent Palmade, Mona Prasad, Karima Saleh, ry, David Rohrbach, Aad van Geldermalsen, Rachel Sebudde, Giovanni Tanzillo, Chris- and Alies van Geldermalsen provided useful topher Thomas, Franke Toornstra, Marilou inputs to the essay. Jose Luis Irigoyen, Vi- Jane D. Uy, Stephen Vincent, and Yi-Kyoung vien Foster, and Valerie Marie Helene Layrol Lee. Their feedback and suggestions helped kindly prepared box 2. Stephen Mink, Nancy improve this year's edition. Claire Benjamin, Michael Morris, Prasad C. Staff from External Affairs oversaw Mohan, Jose Luis Irigoyen, Vivien Foster printing and dissemination of the book and and Valerie Marie Helene Layrol provided its companions. useful comments on an earlier draft of the Several institutions provided data to essay. Africa Development Indicators. Their contribu- Azita Amjadi, Ramgopal Erabelly, Ab- tion is very much appreciated. dolreza Farivari, Richard Fix, Shelley Lai Fu, Communications Development Incor- Malarvizhi Veerappan, Shahin Outadi, and porated provided design direction, editing, William Prince collaborated in the data pro- and layout, led by Bruce Ross-Larson and duction. Maja Bresslauer, Mahyar Eshragh- Christopher Trott. Elaine Wilson typeset Tabary, Victor Gabor, and Soong Sup Lee the book. Acknowledgments ix Executive summary Silent and lethal: How quiet corruption With this broader definition in mind, undermines Africa's development efforts the familiar form of big-time corruption is The corruption that often captures newspa- just the "tip of the iceberg"; the quiet corrup- per headlines and provokes worldwide public tion, that is the less frequently observed de- disapproval is dominated by loud "big-time viations from expected conduct, is below the corruption," notably administrative and po- surface. In addition to capturing the notion litical corruption at the highest government that quiet corruption is not as visible, the levels. In response to this notoriety, the iceberg analogy provides two additional in- bulk of anti-corruption measures have been sights. First, quiet corruption occurs across tailored to address this type of corruption. a much wider set of transactions directly However, recent examinations of the level affecting a large number of beneficiaries. and quality of service delivery in developing Quiet corruption is present in a large share countries, including the World Development of health-provider­patient or teacher-pupil Report 2004, have highlighted the need to interactions, for example. Second, quiet expand the scope of the standard definition corruption very often has deep long-term of corruption--the abuse of public office for consequences on households, farms, and private gain. While acknowledging the im- firms. The widespread prevalence of big-time portance of big-time corruption in reducing and quiet corruption in Africa significantly funding for service delivery, recent research undermines the impact of investments to has devoted increasing attention to identi- meet the Millennium Development Goals fying corrupt practices downstream at the (MDGs). In the parlance of this essay, the frontline of public service provision. iceberg of corruption is sinking considerable Following this literature, this essay in- efforts to improve the well-being of Africa's troduces the term "quiet corruption" to indi- citizens, particularly the poor who rely pre- cate various types of malpractice of frontline dominantly on publicly provided services. providers (teachers, doctors, inspectors, and It is important to raise awareness of the other government representatives) that do profile of quiet corruption because this mal- not involve monetary exchange. These be- practice has non-negligible long-term con- haviors include both potentially observable sequences. This essay elaborates both the deviations, such as absenteeism, but also direct consequences, such as the limitation hard to observe deviations from expected of the productivity potential of households, conduct, such as a lower level of effort than firms, and farms, and the indirect conse- expected or the deliberate bending of rules quences, such as distrust of public institu- for personal advantage. For example, recent tions and the notion that frontline provider findings indicate that primary school teach- malpractice is inevitable and omnipresent. ers in a number of African countries are not As an example of direct consequences, in school 15 to 25 percent of the time (absen- we might think how poor service delivery teeism), but, in addition, a considerable frac- caused by absenteeism or low effort on the tion of those in school are not found teaching job might hamper a child's development, (low effort). Frontline provider deviations with potential permanent effects on adult from expected behavior that meet these re- educational attainment, cognitive skills, and quirements broaden the scope of corruption. underlying health. As an indirect effect we Executive summary xi might think of the withdrawal of children and capacity of the national anti-corruption from school because of beliefs about the low units to pursue operationally effective re- quality of education, which shifts the alloca- sponses at the sector level, and adequate tion of time and resources away from human policies and institutions. An equally impor- capital formation toward home production tant second pillar is increasing transparency or labor market participation. in policy formulation and implementation This essay further shows how quiet cor- that empowers citizens to raise the account- ruption manifests itself differently according ability of service providers--bolstering the to the nature of service delivery. It focuses "demand side" for good governance. Finally, on four key sectors (education, health care, successful implementation of anti-corrup- agriculture, and the private sector) whose tion reforms also requires that the prefer- progress and success are crucial for poverty ences and interests of all those involved be eradication and more generally achieving aligned with achieving the objectives of the the MDGs. In presenting examples and out- reform. This often involves better working lining the long-term consequences of quiet conditions. corruption in these sectors, this essay con- Given the complexity of the task, the tends that one of the main reasons Africa is fight against quiet corruption requires tai- lagging behind is the poor service delivery loring policies to country circumstances, that is a consequence of quiet corruption. recognizing that priorities and responses The good news is that quiet corruption may vary depending on different country can be tackled. Whenever a government's conditions. This essay outlines a research determination to deal with quiet corruption agenda to identify interventions to address has increased, for example, by increasing quiet corruption. Experimenting with vari- the availability of information on finances, ous ways to empower beneficiaries and con- inputs and expected outputs, then measur- tinuing the ongoing efforts to tackle big- able improvements in service delivery have time corruption will go a long way toward been possible. Although there is no "one size this goal. Indeed, although combating loud fits all" recommendation that applies to ev- and visible forms of corruption is necessary, ery sector, this essay advocates the need for fighting quiet corruption is critical if govern- strong and highly motivated leadership in ments want to reduce poverty and promote the fight against corruption, commitment sustainable growth. xii Africa Development Indicators 2010 Silent and lethal: How quiet corruption undermines Africa's development Corruption captures newspaper headlines education providers, agriculture extension and provokes public disapproval. In addition, services, drug inspectors, and the police (see the abuse of public office for private gain-- Razafindrakoto and Roubaud 2006). the most common definition of corruption-- World Development Report 2004 (World has attracted increasing attention by scholars Bank 2003), which examined service deliv- and policy makers interested in economic de- ery, recasts the problem of corruption from velopment. Specifically, corruption and poor a different perspective. While acknowledg- governance help explain why increased fund- ing the importance of big-time corruption in ing allocations, such as those aimed at meet- reducing funding for service delivery, World ing the United Nations Millennium Develop- Development Report 2004 and subsequent ment Goals (MDGs), have not necessarily research have devoted increasing attention translated into improvements in human de- to analyzing corrupt practices downstream, velopment indicators, particularly in Africa.1 at the frontline of public service provision Despite considerable funding increases, the (Reinnika and Svensson 2006). This new region is largely lagging behind in meeting focus has produced two results. First, it has the MDG of reducing child mortality (the enabled the identification of malpractice in- number of children dying before age 5 per volving small monetary transactions, gen- 1,000 live births). Substantial increases in erally referred to as "petty corruption" (de gross enrollment in primary education in Sardan 1999), for example, under-the-table recent years have not been matched by im- payments for services received (Transpar- provements in learning outcomes. Africa's ency International 2005, 2006) or bribes private investment rate is still around 15 to tax collectors and low ranking public of- percent, much lower than in most develop- ficials. Second, the concept of corruption has ing countries. Agricultural productivity is not been gradually extended to practices that do increasing fast enough; the yield per hectare not necessarily involve monetary transac- is still less than half that in other developing tions, such as teacher absenteeism (Patrinos regions. Cutting across all these problems is and Kagia 2007). Furthermore, new survey Africa's fundamental problem, namely weak tools, such as the Public Expenditure Track- governance and associated corruption.2 ing Survey (PETS) and Quantitative Service Until recently, the debate about corrup- Delivery Survey (QSDS), have enabled re- tion and development3 has been dominated searchers to track resources and monitor the by the identification and measurement of attendance of frontline providers. These re- "big-time corruption" (de Sardan 1999), no- search and survey results have improved the tably administrative and political corruption understanding of a broad range of miscon- at the highest levels of government.4 This fo- duct and contributed to reshaping the policy cus has produced measures of governance debate about corruption. weakness and corruption suitable for cross- Following the recent findings on front- country comparisons of political corruption. line provider misconduct, this essay focuses But these measures are not reliable when on behaviors that are difficult to observe it comes to measuring less visible forms of and quantify, but whose impact on service corruption, such as those faced by common delivery and regulation has adverse long- citizens as they interact with health and term effects on households. We introduce Silent and lethal: How quiet corruption undermines Africa's development 1 the term "quiet corruption" to indicate vari- money--either political level thefts or small ous types of malpractice of frontline provid- but frequent bribes--is less salient or "noisy," ers (teachers, doctors, inspectors, and other and consequently less likely to attract public government officials at the front lines of ser- attention. Despite its low visibility, quiet vice provision) that do not involve monetary corruption is ubiquitous. And it is associated exchange. These behaviors include not only with harmful long-term consequences, par- potentially observable deviations, such as ticularly for the poor who are more exposed absenteeism, but also hard to observe devia- to adverse shocks and more reliant on gov- tions from expected conduct, such as a lower ernment services to satisfy their most basic level of effort than expected or the deliberate needs. bending of rules for personal advantage. For Two examples illustrate the magnitude example, education service delivery requires of the consequences of quiet corruption. teachers to be present in school as well as to First, among the reasons for low fertilizer us- deliver classroom instruction required by age among African farmers is the poor qual- the curriculum. Similarly, a building inspec- ity of fertilizers on the market. Despite the tor can turn up to inspect the structural in- capability of manufacturers to produce good tegrity of a new shopping mall but choose to fertilizers, poor controls at the producer and exert little effort in executing the task. wholesaler levels resulted in 43 percent of Quiet corruption, as opposed to cor- the analyzed fertilizers sold in West Africa ruption that involves an exchange of in the 1990s lacking the expected nutrients, meaning that they were basically ineffective Figure 1 Big-time and petty corruption are the "tip of the iceberg" (IDFC 1995). It is likely that poor farmers' experiences with low-quality fertilizers dis- courage fertilizer adoption. Second, a survey of malaria fatalities in rural Tanzania reported that nearly four out of five children who died of malaria sought medical attention from modern health fa- cilities (de Savigny et al. 2008). A range of Big-time and petty manifestations of quiet corruption, includ- ing the absence of diagnostic equipment, corruption drug pilfering, provider absenteeism, and very low levels of diagnostic effort, all con- tributed to this dire statistic (Das and Leon- ard 2009). The concept of quiet corruption is cap- tured in Figure 1. The familiar forms of corruption--both big time and petty--are just the "tip of the iceberg"; the less frequent- Quiet ly observed deviation from expected conduct is quiet corruption. In addition to capturing corruption the notion that quiet corruption is not very visible, the iceberg analogy provides two additional insights. First, quiet corruption occurs across a much wider set of transac- tions affecting a large number of beneficia- ries directly. Quiet corruption is arguably present in a large share of doctor­patient or teacher-pupil interactions, for example. Second, quiet corruption plausibly has deep long-term consequences on households, farms, and firms. Comparing the long-term consequences of different forms of corrup- tion is a hazardous undertaking. In addition 2 Africa Development Indicators 2010 to being affected by the same country level an environment conducive to misconduct by characteristics, all three forms of corruption frontline service providers. are related. The quiet corruption of low-level The scheme of insiders profiting from bi- officials may very well have been "justified" ased rules in the system is mirrored in society. in their minds by the misbehavior of their Police exert their influence to extract benefits superiors involved in big-time corruption. from the disorganized mass of road users; Likewise, by reducing available resources doctors do not show up in public facilities and and compromising the monitoring and en- instead provide services privately; teachers forcement of conduct, big-time corruption do not show up in classes since they have a encourages low-level civil servants to en- second job and their impunity is guaranteed gage in opportunistic behavior. An instance by their superiors in exchange for other fa- of teacher absence could be the result of a vors, and so forth. It follows that corruption poor working environment occasioned by becomes an unavoidable element of daily life big-time corruption or other factors beyond for many citizens and it diffuses throughout the teacher's and education managers' con- the economy; more big-time corruption be- trol. However, it can also be categorized as gets corruption at the frontlines of service quiet corruption--the abuse of public office delivery, which in turn supports big-time by the teacher. The long-term consequences corruption, creating formidable challenges to of this and other instances of absence com- governance and accountability interventions. pound the effects of both big-time and quiet For a number of key public services, the corruption. cumulative nature of human development Corruption is embedded in the politi- implies that poor service delivery experi- cal economy of Africa. A number of studies enced during the early stages of life can have describe the interaction between various long-term consequences. The direct long- forms of corruption and how it is intrinsi- term effects of quiet corruption begin with cally linked to the way power is exercised. 5 poor service delivery during early childhood, In particular, when a social unit is highly which is then amplified by subsequent poor diverse ethnically 6 --as is the case in many service provision throughout childhood. For post-independence African countries--there example, a mother who is a victim of quiet is likely to be suspicion and division among corruption--poor quality antenatal care-- members, making the process of agreeing to might give birth to an underweight child, rules for governance extremely difficult. In who will likely suffer a series of health set- this context, small groups (elites) that are backs during childhood that potentially highly homogeneous are more likely to pre- magnify the immediate effects of the poor vail and impose rules that bias the system antenatal care. This amplification process is in their favor. The enforcement of these bi- also driven by families' increasingly negative ased rules requires either coercion or "addi- expectations of service delivery systems, tional resources" to ensure the cooperation leading to even fewer health service visits of members of other groups who will try to and the use of poorer quality alternatives. In avoid such biased rules. the African context, alternative health ser- In many African states, the coercion op- vices are often nonexistent, of low quality, tion is not feasible because state power is or too costly for the typical household. The limited. In contrast, the option to purchase family's decision to exit the system leads to the cooperation of other groups tends to a worsening of the public sector and can ul- be the most viable. Ruling elites in regimes timately result in the collapse of service de- with limited legitimacy thus regard corrup- livery.8 For example, as McPake et al. (2000) tion purely in terms of its political func- document, the poor quality of health care tionality as a source of patronage resources services in Uganda created a downward spi- to maintain and strengthen the system of ral of underutilization of public health facili- political power.7 The more these elites are ties. Lower demand for services led to even able to privatize state resources, the more lower staff attendance and to shorter open- they can distribute favors and create a base ing hours of health care facilities. of consensus for their privileged position. Considering the pervasiveness of cor- Thus a strategy to control the state creates ruption and that the different types are Silent and lethal: How quiet corruption undermines Africa's development 3 intertwined with the functioning of politi- While these long-term consequences are cal and social systems in many developing very hard to quantify because of the absence countries, it is clear that focusing only on of data that trace out the effects of contem- the monetary forms of corruption misses poraneous misconduct on future outcomes the majority of solutions. Hence, this essay and because of the multiplicity of other fac- attempts to provide a framework to under- tors that may contribute to them, combining stand the implications of the entire "iceberg" evidence from both developed and develop- of misconduct that shapes the level and ing countries provides a sense of the mag- quality of services and regulation in devel- nitude of resulting damage to development. oping countries. The essay outlines evidence The long-term consequences are divided into of quiet corruption and discusses both di- direct consequences, such as the limitation rect and indirect long-term consequences on of the productivity potential of households, households, businesses, and farms. firms, and farms, and indirect consequenc- The framework in Figure 2 describes the es, such as distrust of public institutions mechanisms through which quiet corrup- and the notion that corruption is inevitable tion affects delivery of frontline services, and omnipresent. These two components are such as medical treatment or in-classroom explained in more detail below. instruction, and the provision of business One direct effect of quiet corruption is regulations, such as trading licenses. The the loss of production as a result of the lower three arrows linking quiet corruption and quality of inputs. For example, research on service delivery represent "pathways of in- corruption in the health-care sector rarely fluence." These are denoted as (1) low effort documents how the effect of poor service due to absenteeism, (2) low effort on the job, delivery that might hamper a child's devel- and (3) resource leakage. opment has permanent effects on adult edu- Low effort due to absenteeism refers to cational attainment, cognitive skills, and frontline provider behaviors that restrict the underlying health. The absenteeism of doc- amount of time they are available. Absentee- tors or nurses, for example, might contrib- ism implies that providers work less time at ute to the non-detection of iron deficiency the public facility than contracted for, with (Ramakrishnan et al. 1999) or deficiency of little or no repercussions on their earnings. other micronutrients in a pregnant mother's The second arrow takes into account the ex- diet. The lack of timely intervention affects tent to which frontline providers shirk their the development of the fetus and stunts the duties while on the job. Finally, the third ar- child's full growth. The consequences of this row refers to providers' involvement in the poor health treatment may manifest during leakage of key inputs, such as drugs and adolescence and adulthood and could affect medicines, in the case of health-care work- the individual's productivity (Barker et al. ers, or books and other instructional materi- 1995; Smith 2009). als in the case of teachers. An indirect effect of quiet corruption Despite the difficulties in observing at- operates through changing the beliefs tendance and job effort, the lack of transpar- and expectations of service beneficiaries. ency and accountability, and the weaknesses As a result of this transformation, agents of monitoring and enforcement inherent in may decide to allocate their time in more public service organizations in developing remunerative activities in the short run countries, this essay argues that quiet cor- at the expense of capital accumulation ruption is likely to be equally insidious as and investment in activities that produce big-time corruption. The right-hand portion larger gains only in the long run. A typical of Figure 2 illustrates the linkage between case is the non-investment in the human poor service delivery today and the direct capital of children because of beliefs about and indirect long-term consequences of big- the low quality of education, which shifts time corruption and quiet corruption. Be- the allocation of time and resources away cause of its nature, quiet corruption can af- from education toward home production fect incentives and distort the allocation of or labor market participation. Another ex- resources at the individual, household, firm, ample, as mentioned above, is the lack of and farm levels. adoption of fertilizers and other productive 4 Africa Development Indicators 2010 inputs by farmers who have had bad prior socioeconomic characteristics and political experiences. power of the main clientele. The implication Finally, the notion that corruption is is that reform strategies should differ de- generally ubiquitous and inevitable implies pending on the nature of service. There will that it is an "accumulating process": the more be differences across countries as well, in corrupt the system, the more it produces accord with the levels of accountability and a downward spiral of malpractice (de Sar- transparency and the systems of monitor- dan 1999). Within a corrupt environment, ing, enforceability of rules and procedures, people adjust their strategies accordingly and punishment of corruption. As a conse- and contribute to the general acceptance of quence, there is no recipe on how to prevent the phenomenon, thus making it routine. If and fight quiet corruption that is valid for all professional standards are substituted with sectors and countries. The aim of this essay a pure "fend for yourself" attitude at every is not to arrive at specific recommendations level (Lindelow, Serneels, and Lemma 2005), but rather to stimulate debate around this the system falls into a vicious cycle in which critical development topic, expecting that every misconduct is tolerated and the struc- it will increase interest and efforts that are ture of incentives becomes biased against much needed to combat quiet corruption. those who adhere to the standards. While quiet corruption is indeed pres- Some sectors are more vulnerable to ent in all sectors, the next sections present quiet corruption than others; the main de- evidence and discuss the consequences of terminants are the level of transparency quiet corruption in education, health, the and accountability in the sector, the asym- private sector, and agriculture. This selec- metry of information, and the discretion tion is based on the importance of these sec- and monopoly power of service providers, tors for Africa's development as well as the all of which create incentives for miscon- existing evidence on quiet corruption. For duct. The manifestation will also differ from each sector, the presence of quiet corrup- rural to urban areas and will depend on the tion in the typology presented in Figure 2 Figure 2 The functioning of quiet corruption and its long-term consequences Big-time and petty corruption Low effort due to absenteeism Lower productivity Poor service delivery Long-term Lower human capital and regulations direct Quiet consequences corruption Low effort on the job Markets concentration Long-term indirect Poor resources consequences allocation Resource Disbelief in leakage public institutions Silent and lethal: How quiet corruption undermines Africa's development 5 is documented. Furthermore, and to the For purposes of this essay, the frequency of extent possible, the direct and indirect long- documented deviations represents an upper term consequences for economic agents are bound of the prevalence of quiet corruption. presented. Following the framework of Figure 2, the long-term consequences are divided into di- Education rect and indirect effects. The education sector prepares youth for productive engagement in the social, politi- Teacher capture cal, and economic realms as adults. In Africa A considerable body of evidence documents education accounts for a large fraction of the capture of service delivery systems by key government expenditure, with a large share actors in the service delivery chain (Mizala of public resources accruing to teachers. and Romaguera 2004 and others). "Capture" Teacher remuneration accounts for nearly refers to a situation in which key actors are three-quarters of recurrent expenditure in able to alter the rules, such as the conditions education in developing countries (Bruns, of service or the allocation of expenditure in Mingat, and Rakatomalala 2003). Quiet cor- the sector, to their advantage and to the detri- ruption in education, therefore, is not only ment of service beneficiaries and the society costly in terms of the direct loss of consider- at large. In the case of the education system, able scarce public resources, but more impor- teachers are a key group of actors that have tantly in terms of its long-term consequenc- exerted considerable influence over both the es for the human capital base. Given the long allocation of resources within the system, but term consequences of adults with lower skills more importantly, the rules that define their and poor attitude, quiet corruption in educa- conditions of service. Much of this power is tion undermines the serious efforts being in- exercised as a result of the influence wielded vested in the eradication of poverty and im- by teacher unions or through the direct in- provement in the competitiveness of African volvement of current or retired teachers in economies (World Bank 2009). local and national politics. This section presents three different This "teacher power" could represent an forms of quiet corruption that have been important constraint on the extent to which identified in the literature. First, the issue of levels of learning can be improved in devel- frontline provider capture of the education oping countries. Two examples demonstrate system: teachers modify the rules and in- the effects of teacher capture on the learn- fluence the allocation of education budgets. ing levels of pupils. In 1998, the Bolivian Second, evidence for low levels of teacher government introduced a policy to ascertain effort in the form of attendance and effort the quality of teachers through a "teach- on the job is discussed. Finally, evidence of ing sufficiency examination." Participation the extent of the leakage of non-salary cash in the test was voluntary and teachers who flows and instructional materials in the edu- passed the exams received a wage increase cation sector is presented. Short-term im- relative to the traditional wage scale. In ad- pacts of each of these forms of quiet corrup- dition, head teachers had to pass this test in tion are linked to long-term effects through order to continue in their role as principals. the cumulative nature of skills acquisition The first round of implementation revealed and evidence from cohort studies in devel- very low levels of teacher quality: 60 percent oped countries. of the teachers who participated failed the Identifying quiet corruption in educa- test and only a very small fraction received tion, as in any other sector, is not straight- a wage increase. The teacher union rejected forward. Much of the evidence presented the results of the test, claiming that the in- below does not unambiguously categorize vitation to participate and the assignment of any observed deviation from expected be- grades were problematic. A series of demon- havior as quiet corruption. For example, it strations and hunger strikes calling for the is difficult to establish the extent to which elimination of the examinations followed. teacher absenteeism or low levels of school The government capitulated and in the inspection reflect either a poor working second year of implementation more than environment or the abuse of public office. 18,500 teachers received wage increases. 6 Africa Development Indicators 2010 The policy was discontinued and replaced by Table 1 Percentage of grade 6 students receiving extra lessons a range of largely non-performance-based incentives (Mizala and Romaguera 2004). Percentage of grade 6 students receiving extra lessons A 2007 proposal by the ministry of edu- SACMEQ I SACMEQ II cation in Uganda to improve management at Country 1995 2000 the school level through performance con- Mauritius 77.5 86.6 tracts with head teachers met a similar fate. Kenya 68.6 87.7 In the proposed contracts, head teachers Zanzibar 46.1 55.9 would sign an agreement with the local gov- Zambia 44.8 55.1 ernment outlining a series of goals to be met Namibia 34.7 44.7 over a two-year period. Failure to meet these Malawi 22.1 79.7 goals could lead to demotions or transfers. Total 49.0 68.3 Even though the policy explicitly stipulated that head teachers would be the main archi- Source: Paviot, Heinsohn, and Korkman (2008). tects of the performance targets, the teacher union successfully opposed the policy on the basis that the penalties included in the con- duration of absence during a given time pe- tracts were excessive and unfair. riod (usually 1­4 weeks). For example, an Other rules that are subverted for teach- important UNICEF multi-country survey of er benefit include the implicit or explicit 14 developing countries conducted in 1995 sanctioning of additional instruction out- and reported in Postlethwaite (1998) reports side regular school hours. The legitimacy high levels of head-teacher reported absen- of this practice in a number of countries is teeism. Among the African countries, Tan- particularly pernicious when public teach- zania, Uganda, and Zambia were the worst ers selectively cover material during regular performers. More than half the teachers in school hours and other material during their Tanzania and Uganda were absent at least private tutoring sessions (Jayachandran one day in the previous week and about a 2008; Dang and Rogers 2008). The extent quarter of teachers were absent for two or to which this extracurricular instruction more days. In Zambia, a quarter of teachers might be occurring is suggested by the high were absent for two or more days. Using a and increasing prevalence of extra tuition, similar methodology, Das et al. (2004) report which is shown in Table 1 using data from an average absence duration of two days per the Southern and Eastern Africa Consor- month in Zambian primary schools in 2002. tium for Monitoring Educational Quality Concerns about the quality of head- (SACMEQ). While these data are generated teacher or teacher reports of absenteeism from pupil reports that are unreliable for es- have motivated the use of direct observation tablishing the existence or level of extra tu- of teacher attendance (Table 2).9 In this ap- ition fees, they suggest a considerable degree proach, which relies on unannounced visits, of discrimination: households that cannot a teacher is reported as being absent if he afford extra lessons receive less and/or lower or she cannot be found by the enumeration quality instruction than is stipulated by the team at a time that he or she is scheduled curriculum. to be in school. This methodology has been widely used in western Kenya and a large Low levels of teacher effort multi-country study that included Uganda Teacher effort is an important input into (Chaudhury et al. 2006). Kremer et al. (2004) learning (Park and Hannum 2002; Hanushek, report that 20 percent of teachers in rural Kain, and Rivkin 2005). Perhaps the most western Kenyan primary schools could not important form of quiet corruption in edu- be found during school hours. In Uganda, cation is the low levels of teacher effort that two waves of surveys using this methodolo- arise from teacher absence and low effort gy found teacher absence rates of 27 percent while in school. Evidence on the extent of in 2002 and 20 percent in 2007.10 teacher absence has improved greatly over The above studies document considerably the last decade. Early evidence comes from higher levels of absence among head teach- head-teacher or teacher self-reports of the ers and other senior teachers.12 Whether Silent and lethal: How quiet corruption undermines Africa's development 7 Table 2 Estimates of teacher developed countries. While this measure absenteeism indicates lower levels of instruction, it as- % teachers Days absent sumes that teacher attendance patterns and absent per month (direct (teacher in-class effort and quality are similar across Country observation) self report) developed and developing countries. Higher Uganda (2003)a 27 levels of teacher attendance and the use of Uganda (2007)b 20 substitute teachers in developed countries Kenya (2003)c 20 further deepen the instruction time gap be- Zambia (2007)d 20 tween developed and developing countries. Burkina Faso (1995­8)e 2.2 Additional evidence from direct obser- Cameroon (1995­8) e 1.8 vation studies suggests that in-class behav- Cote d'Ivoire (1995­8) e 1.3 ior of teachers differs from their developed Madagascar (1995­8) e 2.5 country colleagues. While there are prob- Senegal (1995­8) e 4.7 lems of interpretation,13 direct observation Zambia (2002)f 2.0 surveys suggest that even among those teachers who are found present in school, Sources: a. Chaudhury et al. (2006); b. Habyarimana (2007); c. Glewwe, Kremer, and Moulin (2009)11; d. Halsey, Rogers, and Vegas (2009); in-class effort is low. For instance in western e. Postlethwaite (1998); f. Das et al. (2004). Kenya, Glewwe, Kremer, and Moulin (2009) document that 12 percent of teachers were found to be outside the classroom when they this represents low effort is difficult to say should have been teaching. An even higher because a range of duties may draw princi- fraction is estimated in Uganda where nearly pals away from school (to attend meetings or one-third of teachers found were not in the request or collect resources). Taking the ex- classroom during learning periods (Habya- ample of Uganda, head teachers were twice rimana 2007). as likely to be absent as regular teachers Teacher influence and capture could (Habyarimana 2007). In addition, assum- explain much of the low levels of teacher ing that the reported reasons for absence are effort documented. For example, the multi- credible, only half of the absences are offi- country study finds that teacher absence is cially sanctioned. Perhaps even more conse- not concentrated among a few "ghost" teach- quential is evidence from the multi-country ers, but rather is driven by the behavior of a study that documents higher regular teacher large share of teachers. Punishment of poor absence when the head teacher is absent. attendance or tardiness on the job is very Chaudhury et al. (2006) also collected infor- rare. In many cases, warning letters or re- mation about the likelihood that teachers ports of teacher misconduct do not trigger were warned or fired for absence. The results any sanctions. from India are indicative of the low levels of sanctions: only one teacher in 3,000 schools Leakage of resources had been fired for absence despite high ab- Schools combine instructional materials and senteeism rates. The level and quality of in- teacher and pupil interaction to produce cog- struction provided by teachers is not only nitive skills. This essay documents the leak- a function of their training and attendance age of two key inputs: instructional materials patterns, but also of their behavior while in and school inspection. A teacher with few or school. Measuring the effort level of teachers no instructional materials will find it harder in school and the extent to which it qualifies to impart the necessary skills to her charges. as quiet corruption is challenging. Indeed, In addition, school inspection ensures that a number of studies have tried to quantify the right pedagogical strategies are being this effect with limited success. Some studies implemented and instructional materials are rely on adding up the number of class hours well deployed.14 using the school's timetable. For example, The clearest example of the extent of Postlethwaite (1998) reports that students leakage of instructional resources comes in his developing country sample received from two PETS surveys in Uganda in the only about 80 percent of the total timetable- 1990s. The first study revealed that an av- derived annual instruction time as those in erage of only 13 percent of the resources 8 Africa Development Indicators 2010 intended for schools were reaching them. the consequences of these deviations is Poor information flows about the size of the straightforward. The acquisition of skills and capitation grant and the timing of resource competencies is a cumulative process: cogni- flows provided local education authorities tive achievement today defines how much a with the cover to divert these resources. child will learn tomorrow. Therefore, quiet These findings motivated a government-led corruption today that leads to contempora- intervention to increase the transparency neous lower levels of learning has long-term of grants disbursement. In addition to pros- effects. Some of these long-term effects oper- ecuting offenders, the newspaper and radio ate through decisions that households make. campaign led to large increases in school lev- For example a household might decide that a el funding (Reinnika and Svensson 2005). child who is not learning very much could be In the Zambia PETS of 2002, Das et al. better utilized to look after cows. Evidence of (2004) found that resources meant for reno- this cumulative learning process comes from vation were more likely to go to schools in cohort studies in developed countries that the middle of the wealth distribution, which demonstrate the strong correlation between suggests a degree of collusion between head cognitive skills while young and competen- teachers of middle-tier schools and local ed- cies and earnings in adult life (for example, ucation authorities. Other PETS carried out Case and Paxson 2008). For each of the three in Africa all point to considerable leakages in forms of quiet corruption in education iden- non-salary funding (Gauthier 2006). tified above: teacher capture, low effort, and Finally, the quality of instruction and leakage of resources, there is evidence of a pace of learning is supposed to be monitored negative impact on learning. These short- by a variety of standards officials who typi- term impacts translate into long-term con- cally work in the local education authority. sequences through the cumulative nature of Regular inspection provides crucial infor- skills acquisition and the dynamic decisions mation about challenges and successes that of households. In short, quiet corruption has can be used to generate improvements in arguably grave consequences for the future service delivery.15 While low levels of school competencies of Africa's youth. inspection can be the result of a poor work- Households make dynamic decisions ing environment in which willing officials about whether to enroll/continue a child in lack the means to conduct their duties, the school and how much to invest in time and prima facie evidence is examined as an up- resources based on complementary inputs per bound of the extent of deviations from provided by teachers and schools, and par- norms. It is important to bear in mind that ticularly, based on perceptions of the child's countries have different norms of inspec- learning. Each of the three forms of quiet tion frequency that determine the extent corruption impinges on household decisions to which observed rates of inspection dif- and consequently on the competencies and fer from stipulated rates. This essay instead attainment of children. Teacher instruction presents evidence from a number of sources time is a crucial input in the production of that simply report the fraction of schools competencies and skills that are crucial in a that have been visited by an inspector in the wide range of market and non-market activi- year since the survey. Postlethwaite (1998) ties. And the level and quality of teacher in- suggests that in Madagascar, Togo, Ugan- struction time is affected by all three forms da, and Tanzania, more than 70 percent of of quiet corruption. students were in schools that had not been The capture of the education system af- inspected in the previous year. More recent fects attainment and long-run skills acquisi- evidence from a number of surveys (Uganda tion in several ways. First, as the examples Unit Cost Study and PETS in Zambia and above demonstrate, capture supports lower elsewhere) suggest similarly low levels of levels of teacher quality and managerial ef- inspection. fort at the school level. While evidence of the link between teacher quality and learn- Long-term consequences for education ing outcomes is thin, a study in Israel found While it is difficult to attribute all of the de- that teacher training is associated with viations above to misbehavior, establishing learning gains (Angrist and Lavy 2001). Silent and lethal: How quiet corruption undermines Africa's development 9 Second, to varying degrees, capture supports (2004), leading to even lower learning gains. deeper levels of quiet corruption: low levels Second, as mentioned earlier, households of teacher effort, leakage of instructional make human capital investment decisions materials, and lower levels of inspection. on the basis of current and expected learn- The effects of teacher absence, leakage ing achievements. A child who is struggling of instructional materials, or lower moni- in a school system characterized by quiet toring of learning are well documented. For corruption is more likely to drop out or be instance, using data from Zambian primary removed from school, leading to permanent- schools, Das and others (2007) found that ly low levels of skills and competencies. Evi- an increase in absence duration of one day dence from long-term cohort studies (Case, per month reduces test scores by about 4­8 Lubotsky, and Paxson 2002) confirm the percent of the average annual gains in Eng- long-lasting nature of the adverse effects of lish and mathematics. In addition, Kremer, learning deficiencies at an early age on pro- Miguel, and Thornton (2004) found that test ductivity in later years and complete the link score gains in response to a girls' scholarship between quiet corruption and direct long- program in western Kenya are in part the term consequences. result of increases in teacher attendance of nearly 6 percentage points. Duflo, Hanna, Health and Ryan (2008) showed test score gains of As in the education sector, quiet corruption in 0.2 standard deviations that corresponded the health care sector is widespread in Africa. to a halving of teacher absence in non-for- However, as in the other sectors discussed in mal schools in India. this essay, it is generally very difficult to as- Several studies document the positive certain the intent and therefore culpability of link between more resources and short-term providers. In laying out the evidence for each learning gains in Africa. Using the results of these behaviors in health care and their at- of a newspaper campaign that increased tendant consequences, we advise the reader the amount of funding reaching schools in to keep this caveat in mind. Uganda, Bjorkman (2006) reported gains in national test scores attributable to the Low levels of provider effort increase in capitation grant flows. In addi- As in the education sector, the quality and tion, evidence from Zambia suggests that level of health services depend on the qual- increases in unanticipated funding increase ity of providers, the frequency of attendance, the test scores of grade 6 pupils (Das et al. and effort levels while at work. Evidence on 2004). Evidence from western Kenya sug- quality and effort levels is only just beginning gests that the randomized provision of to emerge and is briefly discussed below. On textbooks increased learning only for the the other hand, evidence on health provider best students (Glewwe, Kremer, and Moulin attendance is considerable and suggests a 2009). In contrast, a recent study in Brazil very discouraging situation. Differences in showed that higher levels of resource leak- measurement methodology notwithstand- age at the municipality level are associated ing, reports of health provider absence are with lower learning gains of pupils (Ferraz, very high. PETS in Mozambique and Chad Finan, and Moreira 2009). (Gauthier 2006) document a rate of absen- While test score outcomes are typically teeism in public facilities of around 20 per- measured for 10- to 14-year-old students, cent. A direct observation survey in Uganda the deleterious effects of quiet corruption recorded a rate of 37 percent in the first extend throughout a child's adolescent and round (2002) that went down 4 percentage adult life. Two particular channels amplify points in the second survey in 2003. Buck- the long-term consequences. First, low lev- ing this trend is Cameroon, where estimated els of learning occasioned by quiet corrup- absence rate is only 5.6 percent. tion produce a poor learning environment in In another survey based on direct ques- the next year that leads to further teacher tions to African, Asian, and Latin American and student absenteeism. This dynamic is physicians who had obtained a master's of reinforced by the links between student and public health degree in Europe between 1976 teacher effort documented by Kremer et al. and 1996 (Macq and Van Lerbeghe 2000), 10 Africa Development Indicators 2010 doctors comprised the category most absent on studies carried out in India, Indonesia, among health care personnel, declaring only Mexico, Paraguay, and Tanzania, estimates 73 percent of their time serving the public, of doctor competence and practice paint a that is, in their official job capacity. The rest disturbing picture. Restricting the focus to of the time was divided between a second the evidence from Tanzania, Leonard and job, generally in the private health care sec- Masatu (2007) found that health provider tor, or in teaching or other activities often competence is considerably poorer in rural unrelated to the core activity. This is likely a areas. However, more appropriate to our def- conservative estimate, because respondents inition of quiet corruption, the gap between have incentives to underreport the effective what providers know and what they do, that time spent in other activities. is, provider effort, is particularly worse in A study undertaken in Uganda between government facilities. 1994 and 1997 involving health workers, community members, and the Health Unit Leakage of resources Management Committee documents an Estimating the degree of resource leakage in environment where so-called "coping strat- health care is very challenging. In many coun- egies," activities not directly related to the tries, governments do not state how much job position, are predominant (McPake et al. they have allocated for various health care in- 2000).16 Health workers, besides openly ad- puts, depriving the analysts of a benchmark mitting to extra-legal user fees for services against which to assess receipts. and selling drugs, also declare that their However, for some resources, leakage greatest source of income is agriculture, thus can be measured as the difference between implicitly acknowledging a high rate of ab- stipulated resource flows (typically non-sala- senteeism. A comparison of working hours ry budgetary resources) and actual amounts declared with observed working durations received. Leakage amounts in this category over the course of a month reveal a striking vary from about 38 percent in Kenya to 99 gap: effective hours worked are in most cases percent in Chad (Table 3). While the differ- one-third or less than what is declared. This ences do not singularly represent leakage is consistent with very low utilization of close to the frontline of service delivery, these facilities, which are open for only two the magnitude of the leakage gives a sense to three hours in the morning. of the size and importance of this form of Absenteeism thus gives way to a vicious quiet corruption. cycle of low service utilization by the public, In addition to leakage of non-salary which then further reinforces the poor at- cash flows, the leakage of health goods is tendance of health workers.17 For example, pervasive. A qualitative survey of 50 health Banerjee, Deaton, and Duflo (2004) found workers in Mozambique and Cape Verde that in Rajasthan, India, over the course of concluded that this practice is widespread, 18 months, nurses, who were assigned to particularly among doctors (Ferrinho et al. staff the clinic on a regular basis, were only 2004). The study also documents an "institu- to be found in the facilities 12 percent of the tionalization" of this phenomenon: Mozam- time. The record of absences, and therefore bican health workers report the existence of closure of facilities, followed no pattern, informal contracts between private clinics meaning that patients' likelihood of finding a provider was unpredictable, thereby dis- Table 3 Leakage of resources couraging patients from using the facilities in health care or keeping appointments. % of cash/inkind Resource Country (year) resources leaked category In addition to high absence rates of Kenya (2004) 38 Non-salary budget health-care workers, a number of studies suggest that quality and effort on the job Tanzania (1999) 41 Non-salary budget is very low. Using direct clinic observation Uganda (2000) 70 Drugs and supplies and vignettes, Das and Hammer (2005) and Ghana (2000) 80 Non-salary budget Leonard and Masatu (2005) provide a sense Chad (2004) 99 Non-salary budget of the magnitude of the problems of low Source: Gauthier (2006). quality and effort in health care.18 Drawing Silent and lethal: How quiet corruption undermines Africa's development 11 and public hospitals to ensure a steady sup- in antifreeze (Cohen et al. 2007). Another ply of certain medicines. Results from a sur- study in South Asia (Newton et al. 2001) re- vey of 90 Mozambican health care workers ported that 38 percent of the anti-malarial, corroborate the findings above (Schwallbach artesunate-based products sold on the mar- et al. 2000). ket contained a lower-than-standard quan- tity of active ingredient, drastically reducing Weak regulation of drugs their efficacy. Akunyili (2005) found that Quiet corruption in the regulation of phar- during the 1990s, Nigeria was flooded with maceuticals is rampant and deadly. The effi- counterfeit drugs that, according to some cacy of medication is dependent on the care- studies, accounted for more than 50 percent ful regulation of standards in the production, of drugs sold in drugstores. While specific distribution, and prescription of pharmaceu- statistics are not available on deaths or seri- ticals. Inadequate or weakly implemented ous illnesses caused by fake medicines, an- quality controls can lead to distribution of ecdotal evidence suggests a connection be- poor quality and often counterfeited drugs, tween drug efficacy and number of fatalities. which result in severe health consequences A further consequence of weak regulation including death of the consumers. Cohen et was the total ban of Nigerian-made pharma- al. (2007) document the existence of sev- ceuticals imposed by neighboring countries. eral areas where quiet corruption along the value chain from production to consumption Long-term consequences for health care of pharmaceuticals compromises long-term Even if drugs are stolen and do not reach fa- health. cilities, they may be reaching the target popu- This high vulnerability to corruption de- lation through different channels. Although rives from the specific features of the health the distribution may be inequitable due to sector. Information between consumer and pricing out particular population segments producer is highly asymmetric. The typi- and could possibly be less effective because cal consumer cannot verify in advance the medicines are dispensed by non-trained per- quality of the medication and has to rely on sonnel, this outcome does not necessarily information provided either by the pharma- imply a dramatic worsening of health con- ceutical producer or the health-care provid- ditions in the population. We sidestep this er. Second, the consumer's inability to verify issue by elucidating the link between quiet quality necessitates government regulation. corruption and contemporaneous and long- The great latitude in regulating pharmaceu- term health outcomes and the long-lasting tical quality is sometimes abused by regula- beliefs of health service users. tors, either directly as a result of low effort While there are few micro-level stud- or indirectly through inducements by drug ies that demonstrate a causal link between producers or distributors. quiet corruption in health care and poor For example, pharmaceutical companies health outcomes, a number of cross-country should follow specific protocols defined by regressions suggest a strong relationship. To the World Health Organization in the pro- establish the link between quiet corruption duction and distribution of medication.19 and long-term consequences, a useful start- Failure to comply with mandated procedures ing point is research that has estimated the for handling raw materials and storing, pack- long-run consequences of malaria eradica- aging, and labeling products compromises tion (Cutler et al. 2007), famine, and low the quality of the product. Low effort or cap- birth weight on long-term labor market con- ture of the regulatory authority implies that sequences (Almond et al. 2006), and cohort these regulations are weakly or selectively studies in the United Kingdom and United enforced, which results in the selling of sub- States that examined the effects of low birth standard and sometimes harmful products. weight on cognitive skills and long-term A number of examples highlight the wellbeing--the so-called Barker hypothesis costs of weak regulatory systems. In 1995 (Barker et al. 1995; Barker 1998). in Haiti, 89 people died after using Paraceta- Cross-country research has demonstrat- mol (acetaminophen) cough syrup prepared ed a negative association between country- with diethylene glycol, a toxic chemical used level measures of corruption and health 12 Africa Development Indicators 2010 care indicators. To the extent that country- consequences. Two recent pieces of evidence level corruption is linked to quiet corruption confirm the link between contemporaneous through the "mirror effect" described in the quiet corruption and birth outcomes. Gold- introduction, these results potentially re- stein et al. (2009) found that absence of a flect the effects of quiet corruption. Gupta, nurse responsible for pre- and post-HIV-test Davoodi, and Tiongson (2000) showed that counseling has a great impact on whether corruption indicators are positively associ- prenatal care patients in Kenya are tested ated with child and infant mortality, the for HIV. In addition, they found that women likelihood of an attended birth, immuniza- who are not tested and counseled are more tion coverage, and low birth weight. Closer likely to give birth without a professional to one of the forms of quiet corruption de- attendant, less likely to receive preventive fined above, Rajkumar and Swaroop (2008) mother-to-child transmission medication, found that the effectiveness of public health and less likely to breastfeed their babies.20 spending in reducing child mortality de- They concluded that reducing absenteeism pends crucially on the perception of higher in public health facilities could reduce ver- government integrity. Wagstaff and Claeson tical transmission of HIV by 0.5­1.5 infec- (2004), replicating a Filmer, Hammer, and tions per 1000 live births. Pritchett (2000) study using more recent The second piece of evidence comes from data, found that public spending reduces an intervention inspired by World Develop- under-five child mortality only where gov- ment Report 2004 (World Bank 2003). Bjork- ernance is good, as measured by the World man and Svensson (2007) document the re- Bank's Country Policy and Institutional sults of a report card intervention in Uganda Assessment (CPIA) score. This study specifi- in which beneficiaries were provided with in- cally explored the implications of additional formation on the performance of their public spending for reaching the MDGs, and con- facility in relation to regional and national cludes that more spending in medium and standards. The impact of the report card was low CPIA countries would not reduce child stunning. It raised both the level of health mortality and that per-capita income growth service utilization and provider attendance offers a better investment if mortality de- and, consequently, reduced infant mortality clines are the objective. by one-third, increasing birth weight, and Micro evidence is more specific, enabling improving other health outcomes. a richer description of how quiet corruption The link between these two pieces of evi- in health translates into poor service deliv- dence and long-term consequences is drawn ery and documentation of some of the direct from cohort studies primarily in developed and indirect long-term consequences listed countries that document long-term conse- in Figure 2. Indirect evidence of the link quences of low birth weight and short stat- between service utilization and health out- ure during early childhood with long-term comes comes from a study in Uganda that cognitive and health outcomes. Height at age estimated the effect of banning user fees on three, which is a function of nutrition and utilization and morbidity (Deininger and health during infancy, affects cognitive skills Mpuga 2004). In addition to direct effects in adulthood (Case, Lubotsky, and Paxson on service utilization, quiet corruption alters 2002). Other studies, such as Almond, Chay, the beliefs of households about the efficacy and Lee (2005), demonstrate a link between of treatments obtained in public facilities. mothers' health during gestation and long- Such beliefs reinforce lower service utiliza- term health and labor market outcomes of tion in preference for traditional, sometimes the children (for a review of these and other life-threatening, interventions. studies, see Smith (2009). Assuming that Given the importance of physical and the same mechanisms operating in these de- cognitive development during a child's ges- veloped countries also apply to developing tation and early years, quiet corruption that countries, quiet corruption in health care affects the utilization of key inputs such as that particularly affects early childhood out- prenatal and post-natal care, immunization, comes has large and long-lasting effects on and the treatment of infant and child infec- the competitiveness of an economy and the tions is likely to have far-reaching, long-term well-being of its citizens. Silent and lethal: How quiet corruption undermines Africa's development 13 Private Sector and Agriculture five firms expected to make informal pay- This section describes the long-term conse- ments to obtain government services. quences of quiet corruption in the private While many of these payments tend to sector and agriculture, areas with a high be small, their high frequency makes them potential to contribute to economic growth a considerable cost for firms. For example, and poverty reduction in Africa. While the Svensson (2003) reported that among prevalence of informal payments for ser- Ugandan firms that paid bribes, the average vices is widely documented for the private amount of informal payments was equiva- sector, evidence for quiet corruption in both lent to US $8,280 (a median US $1,820), cor- the private sector and agriculture is sparse. responding to nearly 8 percent of the firms' Although enterprises are the immediate vic- total costs (1 percent at the median). tims of corruption, our contention is that In the last column in Table 4, a large they do not always bear the ultimate burden fraction of firms report that corruption is because they can often pass on any increased a major impediment to firm operations and costs to consumers.21 growth. While it is hard to say how much of A growing body of evidence on the prev- the firms' expectations of having to make in- alence of quiet corruption in the private sec- formal payments are captured by this mea- tor draws from recent surveys of firms in sure or the extent to which other business developing countries. Both the firms' expe- environment factors are more important rience of actual petty corruption and their constraints, this category likely incorpo- perception of corruption as an impediment to rates forms of quiet corruption. The last row firm operations were elicited on the survey. (Spearman Correlation) reports the results While these two sets of questions shed light of the comparison between the corruption on the extent and severity of petty corrup- measures. The Spearman index 22 reveals a tion, the prevalence of quiet corruption has positive correlation between incidence mea- been difficult to document. However, a care- sures and the perceived one. Yet, none of ful examination of the survey data is indica- the associations is statistically significant, tive of the contours of quiet corruption. which suggests that the rankings reflect dis- Table 4 shows results for five corrup- tinct impediments. tion indicators for the sub-Saharan African The discrepancy between perceived cor- countries surveyed. The first four indicators ruption and incidence of corruption has report the likelihood that firms make any attracted a great deal of attention from informal payments to obtain licenses, con- scholars and policy makers.23 Quiet corrup- tracts, or other services. The last indicator tion may help explain the divide. As noted measures the extent to which corruption is by Herrera, Lijane, and Rodriguez, (2008), a major or severe constraint to the firms' op- perceived corruption partially captures the erations. There is considerable variation in invisible element, notably the uncertainty the extent to which firms expect to make in- stemming from engaging in corrupt trans- formal payments. For example, the share of actions. For firms paying bribes, corrup- firms in Cape Verde, Mauritius, and Namibia tion has an immediate cost in the form of expected to make informal payments is con- illegal payments--petty corruption--but siderably lower than that of the Organisation also an additional cost represented by the for Economic Co-operation and Develop- capriciousness of interaction with public ment (OECD) country average (Investment institutions. Climate Assessment 2009). Nevertheless, a Although corruption can be seen as a large percentage of firms expect that they "tax" and might still allow companies to op- will make informal payments to obtain gov- erate normally (Shleifer and Vishny 1993), ernment services and contracts. In nearly the critical difference between a normal tax half the countries (16 of 35), more than 50 and the "corruption tax" is their predictabil- percent of firms reported an expectation of ity. In the first case, firms know the level having to make informal payments to "get and frequency of payments. In the case of things done." In particular, in Burkina Faso, corruption tax, capture by decentralized Cameroon, Democratic Republic of Congo, regulatory officials with considerable discre- Guinea, and Kenya, nearly four out of every tion engenders uncertainty about the level 14 Africa Development Indicators 2010 Table 4 Incidence of corruption and perceived corruption in sub-Saharan Africa countries Perceived Incidence of corruption corruption % of firms % of firms expected to pay % of firms % of firms expected to give % of firms informal payment expected to give expected to give gifts to secure identifying to public officials gifts to get an gifts in meetings a government corruption as a (to get things done) operating license with tax officials contract major constraint Angola (2006) 46.8 10.08 14.84 38.45 36.06 Benin (2004) 57.65 41.25 21.21 75.43 83.85 Botswana (2006) 27.62 3.29 4.47 22.92 22.58 Burkina Faso (2006) 86.96 0 19.51 80.77 53.96 Burundi (2006) 56.46 40.26 22.63 44.36 19.72 Cameroon (2006) 77.6 50.81 65.43 85.23 52.05 Cape Verde (2006) 5.63 0 10.42 14.08 16.33 Congo, Dem. Rep. (2006) 83.79 66.25 64.42 80.54 20.02 Congo, Rep. (2009) 49.21 42.79 37.1 75.18 65.02 Côte d'Ivoire (2009) 30.64 31.8 13.62 32.34 74.99 Ethiopia (2006) 12.42 2.7 4.35 11.8 23.08 Gabon (2009) 26.09 0 22.81 26.61 41.35 Gambia, the (2006) 52.42 23.42 13.56 50.3 9.78 Ghana (2007) 38.77 22.6 18.08 61.23 9.86 Guinea (2006) 84.75 51.87 57.34 74.58 47.66 Guinea-Bissau (2006) 62.72 15.33 22.7 48.41 44.01 Kenya (2007) 79.22 28.75 32.25 71.2 38.35 Lesotho (2009) 13.96 3.34 9.2 26.37 46.71 Liberia (2009) 55.22 49.63 54.42 51.59 31.19 Madagascar (2009) 19.2 18.6 6.79 14.13 42.71 Malawi (2006) 35.65 4.92 15.33 12.26 46.84 Mali (2007) 28.88 24.04 31.08 80.35 15.7 Mauritania (2006) 82.12 33.23 48.23 76.16 17.1 Mauritius (2009) 1.59 0 0.28 8.81 50.72 Mozambique (2007) 14.84 6.87 9.79 31.65 25.36 Namibia (2006) 11.36 0 2.6 8.08 19.14 Niger (2006) 69.7 8.33 17.05 80 58.54 Nigeria (2007) 40.9 40.29 22.85 44.57 24.7 Rwanda (2006) 19.96 4.58 4.9 14.37 4.35 Senegal (2007) 18.12 21.09 18.66 36.32 23.84 Sierra Leone (2009) 18.8 8.71 8.58 33.85 36.87 South Africa (2007) 15.09 0 3.13 33.2 16.87 Tanzania (2006) 49.47 20.05 14.7 42.69 19.73 Uganda (2006) 51.7 12.86 14.53 46.43 23.57 Zambia (2007) 14.33 2.61 4.89 27.39 12.08 Spearman Correlation Index 0.28 0.15 0.23 0.18 with perceived corruption Source of raw data: www.enterpisesurveys.org. The higher the percentage, the higher the incidence of and perceived corruption. No Spearman coefficient is statistically significant. and frequency of informal payments. For operations. By introducing uncertainty into example, an uncertain number of interac- the cost of regulatory and other publicly pro- tions with the revenue authority or electric- vided inputs, the capture of regulatory and ity provider might be required to continue other services increases the gap between Silent and lethal: How quiet corruption undermines Africa's development 15 actual and perceived corruption. The report- market meets the required chemical com- ed gap between perceived and actual cor- position consistent with extension recom- ruption could be even larger, given that the mendations and that packages sold are of existing firms are those that are relatively the right weight. However, even in developed successful in operating in a corrupt environ- countries, where strict laws protect consum- ment. As a result of these selective concerns, ers from adulteration, the verification of Hausmann and Velasco (2005) questioned fraud is a serious problem. This is because it the reliability of firm-based perceptions of is often not easy to trace back at which step corruption. They note that a more telling of the production or sale the adulteration indicator is the underlying industrial struc- occurs. To address this problem, a common ture, because responses to quiet corruption strategy of the national agencies (such as the in the private sector include a higher degree U.S. Department of Agriculture) is to require of informalization and a high market con- some form of certification of dealers and centration of formal firms. to conduct spot checks through accredited laboratories. Weak regulation of agricultural inputs Unfortunately, for many countries in Another example of the capture of the regu- sub-Saharan Africa, exercising this type latory function of government that has grave of control might be out of reach. Many lack consequences for reducing poverty and in- qualified laboratories, skilled staff, and tech- creasing economic growth is the market for nical tools for conducting even simple sur- fertilizers. As with the market for pharma- veys. Furthermore, the modalities of product ceuticals, asymmetric information between commercialization represent an obstacle for producers and farmers necessitates public the controls; for example, while in developed regulation. National standards agencies are countries, fertilizers are sold in bags, in Af- supposed to ensure that fertilizer sold on the rica, in part due to the high cost, retail sellers Box 1 Quiet corruption in a port authority in Nigeria The Lagos port in Nigeria represents an interesting case of a poorly offences may abandon goods in the port, wait for "their" goods to regulated business environment that gives way to quiet corruption be auctioned, and then bypass the import regulation to get their episodes. In 2006, the reform of the Lagos port was praised as one goods at a relatively low price. In the second scenario, an importer of the best in sub-Saharan Africa in the last decade. Within a few makes a false declaration including an undervaluation of declared months of operation under private ownership, productivity had risen goods and decides when caught to abandon the consignment in at the container terminals. Chronic delays for berthing space had order to obtain the goods through auction, which, in any case, is nearly vanished, leading shipping lines to reduce their congestion cheaper than full payment of import duties with penalty fees for surcharge. However, the benefits of this reform did not last long. In false declaration and incidental port charges. In both cases, the February 2009, the Nigerian Ports Authority (NPA) announced a tem- importer needs to be sure that during the auction process, his porary but immediate suspension of ship entry to enable terminals cargo will be assigned to him and not to another, which is where to clear "alarming" backlogs. In addition, for vessels already heading collusion with the Port Authority plays an important role. The result into Lagos, the NPA considered diverting them elsewhere. of the auction has to be known in advance; otherwise the importer How could the situation deteriorate from the post-reform high would not abandon the cargo. to this point in less than three years? Raballand and Mjekiqi (2009) These cases present all the characteristics of quiet cor- attribute it to a customs circular. On June 12, 2008, Customs man- ruption. In an environment where regulations provide several agement issued a circular (Customs Circular No. 026/2008) to dis- loopholes, reckless businessmen with the connivance of public allow the clearance of goods that featured discrepancies such as authorities manage to avoid clearance costs or to import pro- lack of appropriate import clearance documents and false declara- hibited goods. However, the mechanism used, abandonment of tion. This circular, in fact, modified the behavior of some importers/ cargo that is recovered later via a public auction, has conse- customs brokers; priority clearance in favor of goods that were quences, less visible in the short run, that go beyond the direct easily cleared was given, while the others were abandoned in the revenue loss of clearance evasion. As the Nigerian case shows, port. After the publication of this circular, the amount of uncleared the long-run effect is the port congestion and delays in clear- and abandoned cargo started to grow and congestion increased. ance that completely eliminated the benefits of the 2006 reform There are two possible situations that explain cargo abandon- with obvious consequences on the competitiveness of Nigerian ment. Importers of prohibited goods or those with other related producers. 16 Africa Development Indicators 2010 usually open the bags and sell small amounts IFDC (2007) in sub-Saharan Africa, the (about 1­2 kg). This exposes the products to situation doesn't seem to have improved various forms of adulteration, such as addi- substantially. A survey conducted in 2007 tion of sand or substitution of cheaper and (IDFC 2007) documents the share of sam- unsuitable fertilizer that consumers cannot ples of poor-quality fertilizer sold in 10 Af- easily detect. In addition, a particular type rican countries. Column 3 of Table 5 shows of fertilizer adulteration is the addition of the percentage of fertilizers not satisfying heavy metals; varying amounts of arsenic, quality standards. As the table illustrates, cadmium, chromium, lead, and nickel have a considerable share of widely used fertiliz- been found in fertilizer materials in sub- ers, such as NPK or urea, are of insufficient Saharan Africa. These contaminants are dif- quality: they either show a high moisture or ficult to detect but can cause serious harm if low nutrient content or simply are wrongly they get into the food chain. labeled. For typical African farmers, the cost Although there is little evidence of fertil- of fertilizer and improved seeds accounts for izer adulteration in sub-Saharan Africa, there a large share of their resources. At best, the is empirical support for nutrient deficiencies use of defective inputs does not have any ef- in sold fertilizers. A survey of wholesalers fect on yields, and at worst it degrades the by the International Fertilizer Development soil. Voortmann (2009) documents that Center (IFDC 1995) on fertilizer quality in there are even cases of poor application that West Africa found that, of the 80 fertilizers cause declining yields. analyzed, 43 percent lacked the appropriate Similar lax regulation has been reported nutrients. Of the 685 bags sampled, only 58 for other key inputs such as improved seeds. percent were within a negligible range of the In this regard, survey evidence is non-exis- indicated weight. In addition, 20 percent of tent and anecdotes from experts and other the bags sampled did not have information actors are the only window to misconduct on the type and concentration of nutrients. in this area. For example, the lack of control Furthermore, in only 7 percent of the cases enabled a dishonest company in Zimbabwe did the labels contain the complete address to buy sorghum grain from a late maturing of the responsible party. To compound the variety and sell the same as an early matur- farmers' problem, the IFDC report suggests ing variety. The use the late maturing variety that frequent cases of deliberate adultera- did not provide any grain unless the season tion occur at the retail level where sellers was unusually long and completely jeopar- can easily add deleterious or harmful ingre- dized the harvest of many farmers. dients to increase the weight and sell under- While there is no reliable empirical weight items or even completely misbranded evidence, misconduct that undermines the products. quality/suitability of agricultural inputs Episodes of mislabeled fertilizers sold likely has important implications for agri- at retail level were documented in Mali dur- cultural productivity. Some recent empirical ing the 1990s, when the country started to evidence is suggestive. Initial randomized import low-cost, but poor quality, stocks from Nigeria (Morris et al. 2007) and more Table 5 The prevalence of substandard fertilizers in West Africa recently in Zimbabwe (Djurfeldt et al. 2005). A recent survey on fertilizers in Kenya (GDS Product Total samples Deficient samples Percentage 2005) shows that adulteration and sales of Urea 50 4 8 counterfeit products are isolated events. Ammonium sulphate 7 2 28.6 Nonetheless, among the products sold on Calcium ammonium nitrate (CAN) 9 3 33.3 the market, the survey documents a wide Triple super phosphate (TSP) 4 0 0 fluctuation in the nitrogen and phosphorus Muriate of potash (MOP) 2 0 0 concentration, often not reported on the la- Diammonium phosphates (DAP) 19 1 5.3 bels. Furthermore, about 3 to 5 percent of Nitrogen phosphorus potassium (NPK) 54 19 35.2 fertilizers are deliberately mislabeled in or- Total 145 29 20 der to sell inferior quality fertilizers. On the basis of anecdotal evidence and Note: One sample of ammonium nitrate is excluded. Source: IFDC (2007). a more recent analysis conducted by the Silent and lethal: How quiet corruption undermines Africa's development 17 experiments to measure the impact of fertil- by producers. Auriol, Flauchelm, and Straub izer and improved seeds on yields of maize (2009) showed that, in sectors producing in Western Kenya found no impact as a re- goods for more corrupt public institutions, sult of defective seed or fertilizer. The most the formal sector is dominated by a few large recent results of randomized evaluations in firms. These big formal players recoup the the same area, which show that higher than additional costs resulting from bribery by suggested levels of fertilizer use are the marking up prices and passing these costs on most cost-effective, could possibly reflect to consumers. Furthermore, thanks to their ineffective enforcement of standards (Duflo, network of relationships with politicians, Kremer, and Robinson 2008). these big firms are able to obtain favorable access to inputs such as credit (Khwaja and Long-term consequences for the Mian 2005; Li et al. 2008) or obtain a system private sector and agriculture of regulations biased against new entrants Quiet corruption modifies the structure of that de facto preserves their dominant posi- incentives for entrepreneurs and farmers tion, or both. Hence, the long-run indirect to conduct business, which permanently consequences of quiet corruption are a less alters their current and future investment dynamic economy in which consumers face decisions. Moreover, as discussed below, the higher prices. negative effects go beyond the single perfor- Evidence for the impact of corruption on mance of individual entrepreneurs or farm- increasing the degree of informalization of ers. In the case of firms, quiet corruption acts markets comes from the most recent Invest- as an additional fixed cost that pushes many ment Climate Assessment data (2006­2009) companies out of the market or to the infor- (World Bank 2009).25 In the latest round of mal sector, leaving the most lucrative activi- interviews, entrepreneurs were asked to re- ties to a few large firms that are well connect- port the existence of informal competitors ed with public authorities.24 A recent paper in their markets by responding yes or no on Paraguay illustrates the effects of capture to the question, "Does this establishment of the regulatory function of government compete against unregistered or informal Box 2 Quiet corruption in public utilities In state-owned utilities providing power, telephone, and water Distribution losses take place when utilities fail to adequately services, quiet corruption takes a variety of forms including over- maintain distribution networks and, in addition, tolerate clandes- manning, undercollection of bills, and distribution losses. Recent tine connections, which amount to theft of scarce energy and estimates suggest that these forms of quiet corruption cost Africa water resources. African power utilities typically lose 23 percent some US $5.7 billion a year on aggregate, or just short of 1 percent of their energy in distribution losses. Similarly, African water utili- of GDP (Foster and Briceno-Garmendia 2009). ties typically lose 35 percent of their water in distribution losses, Over-manning takes place when state-owned enterprises retain nearly twice the 20 percent benchmark. These losses amount to more employees than is strictly necessary to discharge their func- US $1.8 billion a year, or 0.3 percent of GDP. tions, often because of political pressure to provide jobs for mem- bers of certain interest groups. Over-manning is found to be par- ticularly material in the case of state-owned telephone incumbents, Table Distribution losses, undercollection, and which amount to US $1.5 billion a year, or 0.2 percent of GDP. over-manning costs as a percentage of GDP in sub-Saharan Africa's energy, information These enterprises have on average only 94 connections per em- and communications technologies (ICT), and ployee, compared to developing-country benchmarks of 420 con- water and sanitation services (WSS) sectors nections per employee, an over-employment ratio of 600 percent. Energy ICT WSS Total Undercollection of bills is a result of lack of effort on the part Distribution losses 0.2 -- 0.1 0.3 of revenue collection officers or their petty corruption in collusion Undercollection 0.3 -- 0.1 0.4 with consumers and is most frequently due to non-payment of bills Over-manning 0.0 0.2 0.0 0.2 by government departments. This problem is prevalent in power Total 0.5 0.2 0.2 0.9 and water utilities, where non-payment can be found across the income spectrum and carries an overall cost of US $2.4 billion a Source: Foster and Briceno-Garmendia (2009). year, or 0.4 percent of GDP. 18 Africa Development Indicators 2010 firms?" Table 6 presents the association be- Table 6 Correlation between perceived market informalization and incidence of perceived corruption tween the degree of informalization of the market with perceived and experienced cor- % of firms expected to give gifts % of firms identifying corruption in meetings with tax officials as a major constraint ruption. To improve comparisons, the sam- Total ­0.1011** ­0.1256** ple is disaggregated by manufacturing and Manufacturing ­0.0955** ­0.0935** nonmanufacturing status and by the num- Non-manufacturing ­0.1100** ­0.1319** ber of regular employees. A negative correla- No. of employees: 1­6 ­0.1368** ­0.0822** tion indicates that as corruption increases, No. of employees: 7­9 ­0.0894** ­0.1937** so does the degree of informalization (infor- No. of employees: 10­15 ­0.1060** ­0.2136** malization is a yes/no question; yes= 0 and No. of employees: 16­35 ­0.1354** ­0.1412** no=1)). No. of employees: over 35 ­0.1079** ­0.0291 Table 6 shows that informalization is correlated with both perceived and expe- **Significant at 5% rienced corruption, suggesting that cor- Source of raw data: ICA (2009). ruption effectively acts as an entry barrier: many firms unable to afford extra costs are forced to remain in the informal market. prices covering 11 routes and 7 countries.27 This barrier applies in particular to firms Their analysis showed that prices in the with fewer than 35 employees and the retail main corridors in Africa were higher than in and services sectors where rates of informal- other regions. Furthermore, most of these ity are higher. prices were not supported by the underlying While there is evidence of market con- structure of costs, since the transport sec- centration in Africa, only a few cases docu- tor is labor-intensive and wages in Africa ment the effect of corruption on market are relatively low. For example, in 2007, in structure.26 Fafchamps (2004) finds that a China, the average transport price was 5 US small group of well-connected traders cap- cents per ton-kilometer, while in Africa in tures the most lucrative markets, leaving the the Durban-Lusaka corridor it was 6 cents, 8 remainder to small, inefficient firms that are cents for the Mombasa-Kampala route, and unable to scale up and challenge these trad- 11 cents for the Douala-Ndjaména corridor. ers' dominant position. In addition, the so- Given that the underlying costs of providing cial network these big firms create plays an these services are not higher than in China, important role by limiting competition from the only plausible explanation for the price outsiders. Ramachandran, Shah, and Tata differences is market power. Teravaninthorn (2007) are more direct in describing this and Raballand specifically considered the capture as a "capacity of lobbying"; although cost of corruption and found two crucial generally less efficient, these large, formal results. The "corruption tax" in the form of firms are better connected, thus protecting levies that policemen and custom officials their high profit margins by resisting exter- charge is significant in West Africa. Such nal competition. It follows that, although costs account for about 20 to 27 percent of African markets were broadly liberalized in variable operating costs in some corridors, the 1980s and 1990s, a few enterprises with although this "tax" is almost insignificant in high market share were able to retain their Eastern and Southern Africa (1 percent). Yet, market power by investing resources in their this is only the visible part of corruption, relationships with the government. Thus the measurable component. Much more im- quiet corruption acts as a constraint on the portant is the untraceable part represented competitiveness of African manufacturing by the extra costs transport companies face and the growth and poverty reduction ben- when dealing with allocation of freight. It efits of private-sector development. goes without saying that often these "extra In addition to reinforcing a non-dynamic costs" are not charged in exchange for any private sector, quiet corruption in the trans- service but are explicitly imposed to create port sector further diminishes the prospects an entry barrier to potential competitors. In- of African manufacturing. This is shown by siders, often well connected with the ruling Teravaninthorn and Raballand (2008), who regimes, thus compensate for these costs by constructed a dataset of transport costs and imposing very high markups on prices and Silent and lethal: How quiet corruption undermines Africa's development 19 enjoy all the benefits of a monopoly. This may even damage crops. Hence, in the long ultimately enables truckers and freight for- run, farmers that find no increase in yields warders to pass on much of these additional or find decreases may be driven to reduce or costs to final users. even completely avoid using fertilizers, turn- While empirical evidence of the trans- ing to a low-input type of agriculture. This port case of quiet corruption described type of agriculture, as widely documented in above is difficult to find, a striking example literature, is bound to produce low yield and of transport market reform in Rwanda after is more exposed to natural shocks. 1994 sheds light on the potential gains that arise from thoroughly addressing the prob- Final remarks lems quiet corruption can generate. After a It is becoming widely accepted that improv- radical reform of the transport sector that ing service delivery to the poor is both a eliminated entry barriers to the transport widespread political demand and central to market, prices declined by more than 30 the realization of the MDGs. Improving gov- percent in nominal terms and almost 75 per- ernance is integral to achieving these goals. cent in real terms. This result can be largely Where transparency and accountability attributed to elimination of quiet corrup- mechanisms are weak or lacking, poor peo- tion in the transport sector, since no major ple are often marginalized and development investment in infrastructure was carried out outcomes suffer. This essay attempts to un- during this period in Rwanda. veil the iceberg that threatens to sink Africa's As with the health care and education efforts to improve well being and growth by sectors, more investment in the transport documenting a broadened scope of corrup- sector does not necessarily mean that ser- tion beyond the behaviors recently brought vice conditions improve. In fact, the case of to light by innovative survey tools and in- transport reform in Rwanda clearly reveals spired by World Development Report 2004. As that seriously addressing the effect of quiet discussed, some of these behaviors are not corruption, that is the cartelized market readily observed or are difficult to measure, structure of the trucking industry, might but their long-term consequences are often induce effective gains that eclipse any other severe and cannot be ignored by policy mak- potential benefit arising from a pure increase ers, citizens, international institutions, and in expenditures. Teravaninthorn and Rabal- donor organizations. land (2008) noted that, although the condi- This essay outlined a framework to un- tion of African roads is worse than those in derstand the nature and impact of quiet other parts of the world, an investment in corruption, which captures the implicit and their improvement is not necessarily bound less tangible forms of corruption. The quiet to succeed in transport price reduction. corruption approach embraces the recogni- Moving on to the long-term effects of tion that government spending on social quiet corruption on farmers' investment services alone is not sufficient to understand decisions, two main triggers can be iden- the quantity and quality of public services or tified. In the first case, farmers indirectly the determinants of public service delivery pay the cost of the corruption tax that performance. This approach looks at issues emanates from other sectors. As explained that are complementary to the more visible earlier, markets in Africa are generally not forms of corruption and have broad implica- competitive and the market of agricultural tions for strategies and policies that focus inputs is not an exception. To the contrary, on results. these products are often imported (Svensson Examples of the existence and conse- 2003), which leads to even higher informal quences of quiet corruption--such as the payments. The high prices of fertilizer force case of low teacher instruction time that the minority of farmers that purchase fertil- leads to poor competencies and ends with izer to use it in rather small doses. the decision of households to disenroll a As pointed out earlier, lax regulation of child--highlighted how quiet corruption fertilizer quality compounds the affordabil- can lead to substantial long-term impacts ity problem and often leads to a negligible on poverty. The good news is that quiet impact on yields; poor quality fertilizers corruption can be tackled. The report card 20 Africa Development Indicators 2010 example from Uganda--in which beneficia- operationally effective responses at the sec- ries were provided with information from tor level, and good policies and institutions. the second generation of corruption indi- Equally important is more transparency and cators on the performance of their public increasing accountability and participation health facility in relation to regional and na- by citizens, the "demand side" for good gov- tional standards and its impacts on the level ernance. Success will also require the estab- of health service utilization and provider lishment of strategies for addressing weak- attendance--confirms this. nesses in existing governance capacity and Progress in service delivery has been pos- accountability in the delivery of services. sible because of the increasing determination Strengthening enforcement and administra- of governments to deal with corruption, as tive control, management of public finance, well as the increasing availability of infor- government decentralization, systematic mation on finances, inputs, outputs, pricing, dissemination of information about projects and oversight of public service provision by and budgets, and investments in human civil society, which are being used to gener- capital are also essential. Successful imple- ate information on performance and to track mentation of anticorruption reforms will absenteeism, leakage of funds, and informal also require that the preferences of all those user fees. The way forward, however, will re- involved be aligned with achieving the ob- quire the development of a third generation jectives or goals of the reform. This often in- of indicators--ones that measure quality of volves better working conditions. services and the performance of service pro- Of course, given the complexity of the viders. A recent intervention to improve edu- task, the fight against quiet corruption re- cation services in Uttar Pradesh highlights quires tailoring policies to country circum- the features that new projects and programs stances, recognizing that priorities and need to incorporate (Banerjee et al. 2008). By responses may vary depending on the differ- teaching households with very little school- ent country conditions. This essay outlines ing to identify children who are struggling in a research agenda to identify interventions school, the intervention empowered house- to address quiet corruption. Experimenting holds to evaluate service quality. with various ways to empower beneficiaries As quiet corruption manifests different- and continuing the ongoing efforts to tackle ly in each economic sector, there is no "one big-time corruption will go a long way to- size fits all" recommendation that applies ward this goal. Indeed, although combating to each and every sector. But vital to fight- loud corruption is necessary, fighting quiet ing quiet corruption at large are strong and corruption is critical if governments want highly motivated leadership, commitment of to reduce poverty and promote sustainable the national anticorruption units to pursue growth. Silent and lethal: How quiet corruption undermines Africa's development 21 Notes 1. See, for example, Rajkumar and Swaroop (2008) fact, the evidence on the relationship between and Amin, Das, and Goldstein (2009). performance pay and teacher absence is quite mixed. Duflo, Hanna, and Ryan (2008) find a posi- 2. Gupta, Davoodi, and Tiongson (2000) show that cor- tive effect on attendance of a pay-for-inputs per- ruption is associated with higher levels of infant mor- formance contract in non-formal schools in India. tality and school dropout and lower birth weight. On the contrary, Glewwe, Kremer, Moulin, and Zitzewitz (2004) and Muralidharan and Sundara- 3. Early evidence of negative association between man (2006) do not find evidence of a teacher at- political corruption and development comes from tendance response to output-based performance Mauro (1995) and Kaufmann and Wei (1999). For pay. the negative relationship between corruption and the capacity to attract foreign direct investments, 13. Since direct observation requires enumerators to see Wei (2000). Tanzi (1998) reviews some of the physically establish attendance, the presence of an evidence and finds support that corruption is as- outsider in a school could be driving some of these sociated with lower government revenue receipts "in-school effort" measures and therefore casting and also alters the composition of public spend- some doubt over the validity of these estimates. ing away from productive sectors. Other evidence In addition, they could be affected by the fact that comes from Baldacci et al. (2004) and Gupta et al. different pedagogical styles entail different levels (2000). A recent strand of literature has extended of direct teacher-pupil interaction. this analysis to examine the extent to which these relationships are affected by institutional quality or 14. PETS have some important limitations in only be- the level of corruption (Meon and Sekkat 2005; ing able to define leakage unambiguously for fund- Mendez and Sepulveda 2006; Aidt 2009). ing flows with clear rules, such as teacher salaries or capitation grants. Given that some important 4. See Scott (1972) for a broad overview of the vari- resource flows in some education systems are not ous forms of political corruption. rule-based, it is difficult to accurately characterize the extent of leakage. 5. Harsch (1993); Wunsch (2000). 15. There is evidence of an association between the 6. See Olson (1965). frequency of inspection and the level of teacher absenteeism (see Chaudhury 2006). 7. See Harsch (1993). 16. The Health Unit Management Committees were 8. See Hirshmann (1970) in this regard. established with the objective of overseeing the management of the public health facility. It con- 9. Recent Annual School Censuses use a similar sists of public health providers together with mem- head-teacher reported measure of absence. bers of the community. 10. The 2002 data come from Chaudhury et al. 17. Evidence of the perception of absenteeism in par- (2006), while the 2007 estimate is from Habya- ticular in Latin America corroborates these find- rimana (2007). ings. Surveys of hospital nurses' perceptions of the frequency of chronic absenteeism among doc- 11. Two districts in western Kenya. tors reported rates of 98 percent in Costa Rica, 30 percent in Nicaragua, 38 percent in Colombia 12. This result suggests that low teacher remuneration (Giedion, Morales, and Acosta 2001) and 24­31 is not a major determinant of teacher absence. In percent across public and social security hospitals Notes 23 in Argentina (Schargrodsky, Mera, and Weinschel- one indicator occur along with higher magnitudes baum 2001). in the other indicator, the two rankings reflect dis- tinct sources of impediments. 18. Vignettes are hypothetical cases presented to doctors in order to estimate doctor quality. Doctor 23. In this regard it is worth mentioning the contribu- questions, diagnostics, and prescriptions are com- tions from Kaufmann and Kraay (2007), Gelb et al. pared to expert panels or existing protocols. (2007), and Gonzalez et al. (2007). 19. The so-defined "good manufacturing practice." 24. The issues of political lobbying and corruption that are focused on obtaining privileged access to rent- 20. WHO guidelines on breastfeeding have been re- seeking activities and the economic and social laxed to accommodate the poor availability or suit- costs of rent-seeking are broadly treated in the ability of formula. literature. See, for example, Baghawati (1982) and Krusell and Rios-Rull (1996). 21. The ability to pass on corruption-related costs to consumers is subject to, among other factors, the 25. For an interesting case of high market informaliza- demand conditions and market structure. tion caused by corruption, see Auriol et al. (2009) on Paraguay. 22. Spearman's rank correlation coefficient is used to analyze the correspondence between the ranking 26. See Biggs and Srivastava (1996) and Van Biese- defined by incidence of corruption measures and broeck (2005). the ranking obtained by the perceived corruption measure. A positive but insignificant rank correla- 27. Burkina Faso, Cameroon, Chad, Ghana, Kenya, tion suggests that, although higher magnitudes in Uganda, and Zambia. 24 Africa Development Indicators 2010 References Aidt, T. 2009. "Corruption, Institutions, and Economic Programs: Evidence from a Randomized Evaluation Development." Oxford Review of Economic Policy 25 in Education in India," Unpublished manuscript. (2): 271­91. Barker, D. J. P. 1998. Mothers, Babies and Diseases in Akunyili, D. 2005. "Counterfeit and Substandard Drugs, Later Life. London: Churchill Livingstone. Nigeria's Experience: Implications, Challenges, Ac- tions and Recommendations." Paper presented Barker, D. J. P., P. D. Gluckman, K. M. Godfrey, J. E. at World Bank Meeting for Key Interest Groups in Harding, J. A. Owens, and J. S. Robinson. 1995. Health, Washington, DC, March 11. "Fetal Nutrition and Cardiovascular Disease in Adult Life." Lancet 341: 938­41. Almond, D., K. Chay, and D. Lee. 2005. "The Costs of Low Birth Weight." Quarterly Journal of Economics Bhagwati, J. 1982. "Directly Unproductive, Profit Seek- 120 (3): 1031­83. ing (DUP) Activities." Journal of Political Economy 90: 998­1002. Almond, D., L. Edlund, H. Li, and J. Zhang. 2006. "Long-Term Effects of the 1959­1961 China Fam- Biggs, T., and P. Srivastava. 1996. "Structural Aspects ine: Mainland China and Hong Kong." NBER Working of Manufacturing in sub-Saharan Africa: Findings Paper No. 13384. New York: Columbia University. from a Seven Country Enterprise Survey." World Bank Discussion Paper No. 346, Africa Technical Amin, S., J. Das, and M. Goldstein. 2009. Are You Be- Department Series. Washington, DC: World Bank. ing Served? New Tools for Measuring Service Deliv- ery. Washington, DC: World Bank. Björkman, M. 2006. "Does Money Matter for Student Performance? Evidence from a Grant Program in Angrist, J., and V. Lavy. 2001. "Does Teacher Train- Uganda." Innocenzo Gasparini Institute for Economic ing Affect Pupil Learning? Evidence from Matched Research Working Paper No. 326. Milan: Bocconi Comparisons in Jerusalem Public Schools." Journal University. of Labor Economics 19 (2): 343­69. Björkman, M., and J. Svensson. 2007. "Power to Auriol, E., T. Flochelm, and S. Straub. 2009. "`La the People: Evidence from a Randomized Field Patria Contratista': Public Procurement and Rent- Experiment of a Community-Based Monitoring Seeking in Paraguay." Unpublished manuscript. Project in Uganda." World Bank Policy Research Working Paper No. 4268. Washington, DC: World Baldacci, E., B. Clements, S. Gupta, and C. Mulas- Bank. Granados. 2004. "Persistence of Fiscal Adjustments and Expenditure Composition in Low-Income Coun- Bruns, B., A. Mingat, and R. Rakatomalala. 2003. "A tries." In S. Gupta, B. Clements, and G. Inchauste Chance for Every Child: Achieving Universal Primary (eds.), Helping Countries Develop: The Role of Fiscal Education by 2015." Washington, DC: World Bank. Policy, 48­66. Washington, DC: International Mon- etary Fund. Case, A., D. Lubotsky, and C. Paxson. 2002. "Eco- nomic Status and Health in Childhood: The Origins Banerjee, A., A. Deaton, and E. Duflo. 2004. ìWealth, of the Gradient." The American Economic Review 92 Health and Health Services in Rural Rajasthan.î (5): 1308­34. American Economic Review 94 (2): 326ñ30. Case, A., and C. Paxson, 2008. "Stature and Status: Banerjee, A., R. Banerji, E. Duflo, R. Glennerster, Height, Ability, and Labor Market Outcomes." Jour- and S. Khemani. 2008. "Pitfalls of Participatory nal of Political Economy 116 (3): 499­532. References 25 Chaudhury, N., J. S. Hammer, M. Kremer, K. Muralid- Duflo, E., R. Hanna, and S. Ryan. 2008. "Monitoring haran, and F. Halsey Rogers. 2006. "Missing in Works: Getting Teachers to Come to School." Mimeo, Action: Teacher and Health Worker Absence in De- Massachusetts Institute of Technology. veloping Countries." Journal of Economic Perspec- tives 20 (1): 91­116. Duflo E., M. Kremer, and J. Robinson. 2008. "How High are Rates of Return to Fertilizer? Evidence from Field Cohen J. C., M. F. Mrazek., L. Hawkins. 2007. "Cor- Experiments in Kenya." Paper presented at the Ameri- ruption and Pharmaceuticals: Strengthening Good can Economic Association Meeting, New Orleans. Governance to Improve Access." In J. E. Campos and S. Pradhan (eds.), The Many Faces of Corrup- Fafchamps, M. 2004. Market Institutions in sub-Saha- tion. Washington, DC: World Bank. ran Africa. Cambridge, MA: Massachusetts Institute of Technology Press. Cutler, D., W. Fung, M. Kremer, M. Singhal, and T. Vogl. 2007. "Mosquitoes: The Long-Term Effects Ferraz, C., F. Finan, and D. B. Moreira. 2009. "Cor- of Malaria Eradication in India." National Bureau rupting Learning: Evidence from Missing Federal Ed- for Economic Research Working Paper No. 13539. ucation Funds in Brazil." Unpublished manuscript. Cambridge, MA: NBER. Ferrinho, P., M. C. Omar, M. de Jesus Fernandes, P. Dang, H. A., and H. Rogers. 2008. "The Growing Blaise, M. Bugalho, and W. Van Lerberghe. 2004. Phenomenon of Private Tutoring: Does It Deepen "Pilfering for Survival: How Health Workers Use Ac- Human Capital, Widen Inequalities, or Waste Re- cess to Drugs as a Coping Strategy." Human Re- sources?" World Bank Research Observer 23 (2): sources for Health 2 (4): 1­6. 161­200. Filmer, D., J. S. Hammer, and L. H. Pritchett. 2000. Das, J., S. Dercon, J. Habyarimana, and P. Krish- "Weak Links in the Chain: A Diagnosis of Health nan. 2004. "When Can School Inputs Improve Test Policy in Poor Countries." World Bank Research Ob- Scores?" Policy Research Working Paper No. 3217. server 15 (2): 199­224. Washington, DC: World Bank. Foster, V., and C. M. Briceno-Garmendia (eds.). Das J., S. Dercon, J. Habyarimana, and P. Krishnana. 2009. Africa's Infrastructure: A Time for Transforma- 2007. "Teacher Shocks and Student Learning: Evi- tion. Washington, DC: World Bank. dence from Zambia." Journal of Human Resources 42 (4): 820­62. Gauthier, B. 2006. "PETS-QSDS in sub-Saharan Africa: A Stocktaking Study." Study commissioned by the World Das, J., and J. Hammer. 2005. "Which Doctor? Com- Bank as part of the project Measuring Progress in bining Vignettes and Item Response to Measure Public Services Delivery. Washington, DC: World Bank. Clinical Competence." Journal of Development Eco- nomics 78 (2): 348­83. Gauthier, B., and W. Wane. 2006. "Leakage of Public Resources in the Health Sector: An Empirical Investi- Das, J., and K. Leonard. 2009. "Use of Vignettes to gation of Chad." Paper presented at the Center for the Measure the Quality of Health Care." In S. Amin, J. Study of African Economies Conference, Oxford, UK. Das, and M. Goldstein (eds.). Are You Being Served? New Tools for Measuring Service Delivery, 299­312. GDS (Global Development Solutions). 2005. "From Washington, DC: World Bank. Laboratory to the Dining Table: Tracing the Value Chain of Kenyan Maize." Global Development Solu- Deininger, K., and P. Mpuga. 2004. "Economic and tions, LLC. Prepared for World Bank. Welfare Effects of the Abolition of Health User Fees: Evidence from Uganda." Policy Research Working Gelb, A., V. Ramachandran , M. Kedia Shah, and G. Paper No. 3276. Washington, DC: World Bank. Turner. 2007. "What Matters to African Firms? The Relevance of Perceptions Data." Policy Research Work- Djurfeldt, G., H. Holmen, M. Jirström, and R. Lars- ing Paper No. 4446. Washington, DC: World Bank. son. 2005. The African Food Crisis: Lessons from the Asian Green Revolution. Cambridge, MA: CABI Giedion, U., L. G. Morales, and O. L. Acosta. 2001. Publishing. "The Impact of Health Reforms on Irregularities in 26 Africa Development Indicators 2010 Bogotá Hospitals." In R. Di Tella and W. D. Savedoff Causes." Center for International Development, (eds.), Diagnosis Corruption. Washington, DC: Inter- Harvard University. Unpublished manuscript. American Development Bank. Herrera, A. M., L. Lijane, and P. Rodriguez. 2008. Glewwe, P., M. Kremer, S. Moulin, and E. Zitzewitz. "Bribery and the Nature of Corruption." Unpublished 2004. "Retrospective vs. Prospective Analyses of manuscript. School Inputs: The Case of Flip Charts in Kenya." Journal of Development Economics 74 (1): 251­68. Hirschmann, A. O. 1970. Exit, Voice and Loyalty: Re- sponses to Decline in Firms, Organizations and Glewwe, P., M. Kremer, and S. Moulin. 2009. "Many States. Cambridge, MA: Harvard University Press. Children Left Behind? Textbooks and Test Scores in Kenya." American Economic Journal: Applied Eco- IFDC (International Fertilizer Development Center). nomics 1 (1): 112­35. 1995. "The Quality of Fertilizers in West Africa." Miscellaneous Fertilizer Studies No. 13. Muscle Goldstein, M., J. Graff-Zivin, J. Habyarimana, C. Shoals. AL: IDFC. Pop-Eleches, and H. Thirumurthy. 2009. "Health Worker Absence, HIV Testing and Behavioral IFDC (International Fertilizer Development Center). Change: Evidence from Western Kenya." Unpub- 2007. "Assessment of Quality and Truth in Labeling lished manuscript. of Fertilizers in Africa" Fertilizer Quality and Content Report. Muscle Shoals. AL: IDFC. Gonzalez, A., J. E. Lopez-Cordova, and E. Valladares. 2007. "The Incidence of Graft on Developing Coun- Investment Climate Assessment (ICA). 2009. On the try Firms." Policy Research Working Paper No. 4394. website: http://www.enterprisesurveys.org/ Washington, DC: World Bank. Kaufmann, D., and S. Wei. 1999. "Does `Grease Mon- Gupta, S., H. Davoodi, and E. Tiongson. 2000. "Cor- ey' Speed Up the Wheels of Commerce?" National ruption and the Provision of Health Care and Edu- Bureau for Economic Research Working Paper No. cation Services." IMF Working Paper, WP/00/116. 7093. New York: NBER. Washington, DC: International Monetary Fund. Kaufmann, D., and A. Kraay. 2007. "Governance Indi- Gupta, S., L. de Mello, and R. Sharan. 2000. "Corruption cators: Where We Are, Where We Should Be Going." and Military Spending." IMF Working Paper, WP/00/23. Policy Research Working Paper No. 4370. Washing- Washington, DC: International Monetary Fund. ton, DC: World Bank. Habyarimana, J. 2007. "The Determinants of Teacher Khwaja, A. I., and A. Mian. 2005. "Do Lenders Favor Absenteeism: Evidence from Panel Data from Ugan- Politically Connected Firms? Rent Provision in an da." Mimeo, Georgetown University. Emerging Financial Market." Quarterly Journal of Economics 120 (4): 1371­1411. Halsey Rogers, F., and E. Vegas. 2009. "No More Cut- ting Class? Reducing Teacher Absence and Providing Kremer, M., E. Miguel, and R. Thornton. 2004. "In- Incentives for Performance." Policy Research Work- centives to Learn." NBER Working Paper No. 10971. ing Paper No. 4847. Washington, DC: World Bank. Cambridge, MA: NBER. Hanushek, E., J. Kain, and S. Rivkin. 2005. "Teachers, Krusell, P., and J.V. Rios-Rull. 1996. "Vested Interests Schools and Academic Achievement." Econometrica in a Positive Theory of Stagnation and Growth." Re- 73 (2): 417­58. view of Economic Studies 63: 301­329. Harsch, E. 1993. "Accumulators and Democrats: Chal- Jayachandran, S. 2008. "Incentives to Teach Badly? lenging State Corruption in Africa." Journal of Mod- After-School Tutoring in Developing Countries." Un- ern African Studies 31 (1): 31­48. published manuscript. Hausmann, R., and A. Velasco. 2005. "Slow Growth Leive, A., and K. Xu. 2008. "Coping with Out-of-pocket in Latin America: Common Outcomes, Common Health Payments: Empirical Evidence from 15 African References 27 Countries." Bulletin of the World Health Organization Morris, M., V. A. Kelly, R. J. Kopicki, and D. Byerlee. 86 (11): 849­856. 2007. "Fertilizer Use in Africa, Lessons Learned and Good Practice Guide." Washington, DC: World Leonard, K. L., and M. C. Masatu. 2005. "The Use of Bank. Direct Clinician Observation and Vignettes for Health Services Quality Evaluation in Developing Countries." Muralidharan, K., and V. Sundararman. 2006. Social Science and Medicine 61 (9): 1944­51. "Teacher Incentives in Developing Countries: Experi- mental Evidence from India." Mimeo, University of Leonard, K. L., and M. C. Masatu. 2007. "Reexam- California, San Diego. ining the Gap between Medical Ability and Practice Using the Hawthorne Effect." University of Maryland. Newton, P., S. Proux, M. Green, F. Smithuis, J. Unpublished manuscript. Rozendaal, and others. 2001. "Fake Artesunate in Southeast Asia" Lancet 357 (9272): 1948­50. Li, H., L. Meng, Q. Wang, and L. A. Zhou. 2008. "Po- litical Connections, Financing and Firm Performance: Olson, M. 1965. The Logic of Collective Action. Cam- Evidence from 44 Chinese Private Firms." Journal of bridge, MA: Harvard University Press. Development Economics 87 (2): 283­99. Park, A., and E. Hannum. 2002. "Do Teachers Affect Lindelow, M., P. Serneels, and T. Lemma. 2005. Learning in Developing Countries? Evidence from "The Performance of Health Workers in Ethiopia Student-Teacher Data from China." Unpublished Results from Qualitative Research." Policy Re- manuscript. search Working Paper No. 3558. Washington, DC: World Bank. Patrinos, H. A., and R. Kagia. 2007. "Maximizing the Performance of Education Systems: The Case of Macq, J., and W. Van Lerberghe. 2000. "Managing Teacher Absenteeism." In J. E. Campos and S. Prad- Health Services in Developing Countries: Moonlight- han (eds.), The Many Faces of Corruption. Washing- ing to Serve the Public?" In P. Ferrinho and W.Van ton, DC: World Bank. Lerberghe (eds.), Providing Health Care under Ad- verse Conditions: Health Personnel Performance and Paviot, L., N. Heinsohn, and J. Korkman. 2008. "Ex- Individual Coping Strategies. Antwerp: ITG Press. tra Tuition in Southern and Eastern Africa: Coverage, Growth, and Linkages with Pupil Achievement." In- Mauro, P. 1995. "Corruption and Growth." Quarterly ternational Journal of Educational Development 28 Journal of Economics 110: 681­712. (2): 149­60. McPake, B., E. Asiimwe, F. Mwesigye, M. Ofumb, P. Postlethwaite, N. 1998. "The Condition of Primary Streefland, and A. Turinde. 2000. "Coping Strat- Schools in Least Developed Countries." International egies of Health Workers in Uganda." In P. Ferrinho Review of Education 44 (4): 289­317. and W.Van Lerberghe (eds.), Providing Health Care under Adverse Conditions: Health Personnel Perfor- Raballand, G., and E. Mjekiqi. 2009. "Nigeria's Trade mance and Individual Coping Strategies. Antwerp: Policy Facilitates Unofficial Trade but not Manufac- ITG Press. turing." Unpublished manuscript. Mendez, F., and F. Sepulveda. 2006. "Corruption, Rajkumar, A. S., and V. Swaroop. 2008. "Public Growth and Political Regimes: Cross-country Evi- Spending and Outcomes: Does Governance Matter?" dence." European Journal of Political Economy 22 Journal of Development Economics 86: 96­111. (1): 82­98. Ramachandran, V., M. K. Shah, and G. M. Tata. 2007. Meon, P. G., and K. Sekkat. 2005. "Does Corruption "How Does Influence-Peddling Impact Industrial Grease or Sand the Wheels of Growth?" Public Competition? Evidence from Enterprise Surveys in Choice 122: 69­97. Africa." Working Paper No. 127. Washington, DC: Center for Global Development. Mizala, A., and P. Romaguera. 2004. "School and Teacher Performance Incentives: The Latin American Ramakrishnan, U., R. Manjrekar, J. Rivera, T. Gonza- Experience." Mimeo, University of Chile. les-Cossio, and R. Martorell. 1999. "Micronutrients 28 Africa Development Indicators 2010 and Pregnancy Outcomes: A Review of Literature." Svensson, J. 2003. "Who Must Pay Bribes and How Nutrition Research 19 (1): 103­59. Much? Evidence from a Cross-Section of Firms." Quarterly Journal of Economics 142: 207­29. Razafindrakoto, M., and F. Roubaud. 2006. "Are In- ternational Databases on Corruption Reliable? A Tanzi, V. 1998. "Corruption Around the World: Comparison of Expert Opinion Surveys and House- Causes, Scope and Cures." IMF Working Paper hold Surveys in sub-Saharan Africa." Manuscript, WP/98/63. Washington, DC: International Mon- Institut de Recherche pour le Developpement (IRD/ etary Fund. DIAL), Paris, France. Available at http://www.dial. prd.fr/dial_publications/PDF/Doc_travail/2006-17_ Teravaninthorn, S., and G. Raballand. 2008. "Trans- english.pdf port Process and Costs in Africa, A Review of the Main International Corridors." Washington, DC: Reinikka, R., and J. Svensson. 2005. "Fighting Cor- World Bank. ruption to Improve Schooling: Evidence from a Newspaper Campaign in Uganda." Journal of the Transparency International. 2005. Stealing the Future: European Economic Association 3 (2-3): 259­67. Corruption in the Classroom, Ten Real World Experi- ences. Berlin: Transparency International. Reinikka, R., and J. Svensson. 2006. "Using Micro- Surveys to Measure and Explain Corruption." World Transparency International. 2006. "Corruption in Development 34 (2): 359­70. Hospitals." Global Corruption Report 2006. Berlin: Transparency International. De Sardan, O. 1999. "A Moral Economy of Corruption in Africa." Journal of Modern African Studies 37 (1): Van Biesebroeck, J. 2005. "Firm Size Matters: Growth 25-52. and Productivity Growth in African Manufacturing." Economic Development and Cultural Change 53 (3): de Savigny, D., H. Kasale, C. Mbuya, and G. Reid. 545­83. 2008. Fixing Health Systems (2nd ed.). Ottawa, ON (Canada): International Development Research Voortman, R. L. 2009. "Explorations into African Land Centre. Resource Ecology; On the Chemistry Between Soils, Plants and Fertilizers." (Free University Amsterdam Schargrodsky, E., J. Mera, and F. Weinschelbaum. Ph.D. thesis, forthcoming). 2001. "Transparency and Accountability in Argen- tina's Hospitals." In R. Di Tella and W.D. Savedoff Wagstaff, A., and M. Claeson. 2004. The Millennium (eds.), Diagnosis Corruption. Washington, DC: Inter- Development Goals for Health: Rising to the Chal- American Development Bank. lenges. Washington, DC: World Bank. Scott, J. 1972. Comparative Political Corruption. Engle- Wei, S. 2000. "How Taxing Is Corruption on International wood Cliffs, NJ: Prentice-Hall. Investors?" Review of Economics and Statistics 82: 1­11. Shleifer, A., and R. Vishny. 1993. "Corruption." The Quarterly Journal of Economics 108 (3): 599­617. World Bank. 2003. World Development Report 2004: Making Services Work for Poor People. New York: Schwalbach, J., M. Abdula, Y. Adam, and Z. Khan. Oxford University Press. 2000. "Good Samaritan or Exploiter of Illness? Cop- ing Strategies of Mozambican Health Care Provid- World Bank. 2009. African Development Indicators ers." In P. Ferrinho and W. Van Lerberghe (eds.), 2008/2009: Youth and Employment in Africa. The Providing Health Care under Adverse Conditions: Potential, the Problem, the Promise. Washington, Health Personnel Performance and Individual Cop- DC: World Bank. ing Strategies. Antwerp: ITG Press. Wunsch, J. S. 2000. "Refunding the African State Smith, J. 2009. "The Impact of Childhood Health on and Local Self Governance: The Neglected Foun- Adult Labor Market Outcomes." Review of Econom- dation." Journal of Modern African Studies 38 (3): ics and Statistics 91 (3): 478­89. 487­509. References 29 2 0 1 0 Africa Development Indicators 2010 provides the most detailed collection of data on Africa available in one volume. It contains more than 450 macroeconomic, sectoral, and social indicators, covering 53 African countries. Additional data may be found on the companion CD-ROM, covering about 1,600 indicators from 1961 to 2008. · Basic indicators · National and fiscal accounts · External accounts and exchange rates · Millennium Development Goals · Private sector development · Trade and regional integration · Infrastructure · Human development · Agriculture, rural development, and environment · Labor, migration, and population · HIV/AIDS and malaria · Capable states and partnership · Paris Declaration indicators · Governance and polity Designed to provide all those interested in Africa with quick reference and a reliable set of data to monitor development programs and aid flows in the region, this is an invaluable reference tool for analysts and policy makers who want a better understanding of the economic and social developments occurring in Africa. SKU 32381