African Development Successes
August 3-5, 2011
Photo credit Epimacus Mikuza
Forty research projects have been commissioned. This third full research conference of the Africa Project included seventeen final presentations, described below. Special appreciation is extended to the Bank of Tanzania, which co-hosted the meeting.
Akresh, de Walque, and Kazianga conducted a unique randomized experiment to estimate the impact of alternative cash transfer delivery mechanisms on household demand for health services in rural Burkina Faso. The two-year pilot program randomly distributed cash transfers that were either conditional or unconditional, and were given to either mothers or fathers. Under the conditional cash transfer schemes, families were required to obtain regular preventative care at local health clinics (growth monitoring and vaccinations) for all children under the age of five. There were no such requirements under the unconditional programs. The authors find that conditional cash transfers have a significant impact (relative to unconditional cash transfers) on reducing the probability that a child was ill during the preceding month. Conditional on being sick, the child was more likely to visit the health clinic and thus was sick for fewer days. Whether transfers were given to mothers or fathers, there was a similar benefit in terms of improved child health.
Comin explores why, despite all of its potential, so few tourists travel to Tanzania. He begins by noting the abnormally high prices of upscale hotels in Tanzania, especially in the safari areas. After conducting a survey of hotel managers, he concludes that those factors that may affect the cost of supplying upscale hotel services are not sufficient to explain these abnormally high prices. He therefore infers that a more likely explanation is high markups, and interviews with hotel managers support this conclusion. Comin next explores the consequences of high prices of upscale hotels for the total number of foreign visitors to Tanzania. He concludes first that cross-country differences in demand are large. Second, once we control for these differences, the differences in upscale hotel prices explain a significant share of cross country differences in demand. Third, cross-country differences in demand are very persistent. Finally, Comin speculates about these drivers of cross-country differences in demand and their policy implications. Having emphasized the role of word-of-mouth, learning-by-doing, and pecuniary externalities in driving differences in demand, he argues that there may be room for the government to induce lower hotel prices and to try to independently increase the foreign perception of the country's attractiveness.
McMillan, Masters, and Kazianga use historical census data from Burkina Faso to characterize local demographic pressures, including population shocks associated with forced repatriation of migrants from Cote d'Ivoire and internal migration associated with disease eradication in river valleys. They combine those data with a new survey of village elders, designed to document change over time and differences across villages in local public goods provision, market infrastructure, and property rights. They hypothesize that higher local population densities are associated with more collective services and a transition from open-access to regulated land use. Controlling for year and province fixed effects, they find that population shocks associated with proximity to Cote d'Ivoire and to river valleys are closely correlated with increased public services, infrastructure, religious facilities and markets. For land rights, they find only a common time trend across all villages.
Kremer, Lee, Robinson, and Rostapshova take advantage of the characteristics of the retail industry in rural Kenya to estimate the rate of return to inventories in a set of retail firms. Using administrative data on whether firms purchased enough to take advantage of quantity discounts from wholesalers, these researchers estimate that the median shop has rates of return of at least 100 percent or more per year. With a second approach, they measure the return to increased investment by surveying shops on a regular basis about "stockouts" (that is, lost sales due to insufficient inventory). Although the average returns are much lower from this approach (likely because firms fear losing customers due to lost goodwill), the authors do not find that marginal rates of return are equal across shops. After examining alternative explanations for these high and heterogeneous marginal returns to capital, they find that credit constraints are not the only explanation. Their results suggest significant departures from the notion of firms as purely rational profit makers.
Do legal reforms intended to modify or displace traditional customs in sub-Saharan Africa improve welfare and economic outcomes? Kutsoati and Morck study the impact of Ghana's Intestate Succession Law (PNDC Law111, 1986) which allows nuclear families to retain most of the family's assets if a spouse dies intestate and the 1998 Children's Act (Act 560) which, in part, mandates the pensions office to pay 60 percent of the unexpired pensions of retirees to minor child(ren). Their survey shows that although most people die intestate and many profess to know Law 111, it is rarely implemented. However, it seems that mere knowledge of the law has provided an incentive for couples to build up family assets jointly, irrespective of education level. Surprisingly, this is least evident among people whose spouses have a matrilineal lineage tradition - the very people the reforms are focused on advancing. Widows with knowledge of the law also reported better (or at least not worse) economic welfare after their husbands' deaths. The researchers also find limited evidence consistent with Act 560 being beneficial for decedents' nuclear families, especially those of decedents with matrilineal lineage traditions. However, this analysis also confirms the tenacity of African traditional rules. Without intensive awareness campaigns and legal aid for those unable to access the law, legal reforms alone may prove ineffective.
Edwards analyzes some of the most important aspects of Tanzania's economic development since independence in 1961. His discussion emphasizes the role of foreign assistance throughout the last half century, and focuses on macroeconomic policies and overall economic performance. In particular, he analyzes the role played by foreign donors - and by foreign advisors, for that matter - both in the collapse of the economy in the early 1980s and its subsequent recovery. He emphasizes both technical and political economy issues related to imbalances, disequilibria, adjustment, and reform. Because of the emphasis on foreign aid and macroeconomics, this paper plays special attention to three important episodes: the exchange rate crisis of the late 1970s and early 1980s which culminated in the IMF Stand-by Program and the maxi-devaluation of 1986; the serious impasse between donors and the Tanzanian authorities in the mid 1990s; and the reform program implemented during the second half of the 1990s and during the first decade of the 2000s.
Kalemli-Ozcan and Sorensen use establishment-level survey data to quantify the extent of capital misallocation within ten African countries. Interest rates vary dramatically across firms, which is reflected in high variation in the marginal product of capital. The implication is that countries could produce significantly more if capital were allocated optimally. Firms with less access to finance have higher levels of return on capital. Non-exporting firms have less access to capital; there is some evidence that this is also the case for small and young firms. The authors show that country variation in the strength of property rights and legal systems help to explain the variation in the extent of misallocation across African countries.
Interventions designed to prevent mother-to-child transmission (PMTCT) of HIV can reduce the cumulative probability of transmission from an HIV-positive woman to her child by as much as 40 percentage points. Wilson examines the fertility responses to scaling-up PMTCT in Zambia, a country where approximately 15 percent of adults age 15-49 are HIV positive. His results suggest that the local introduction of PMTCT generated a small reduction in average pregnancy rates. The fertility response was greater among women in their main childbearing years and women who had completed secondary school. PMTCT also substantially increased breastfeeding rates.
The majority of people in rural Africa do not have bank accounts. Dupas, Green, Keats, and Robinson combine survey and experimental evidence from rural Kenya to document supply and demand factors that help to explain such low levels of financial inclusion. In their experiment, they waived the fixed cost of opening an account. They find that 18 percent of people actively used their account. Many of those who did not use the account reported that it was because they did not fully trust the bank, or because withdrawal fees were prohibitive. Available credit options also were unattractive to the great majority of individuals in this sample. These results suggest that while simply expanding access to banking services (by, for instance, lowering financial barriers to ownership) will benefit a minority, broader success may be unobtainable unless the quality of services is simultaneously improved. On the demand side, many questions remain. Many people who report that formal banks are their preferred saving tool still do not use formal accounts.
After reviewing traditional HIV prevention strategies, as well as what is known from prior literature on incentives, de Walque, Dow, Medlin, and Nathan discuss the theoretical pathways by which incentive programs could work to reduce risky sexual behavior and prevent HIV infection. They turn to a specific example in detail, elaborating on one such project in Tanzania (the RESPECT Project, of which the authors are the principal investigators) which has implemented a novel randomized trial of outcome-based incentives to reduce risky sexual behavior. RESPECT was a one-year intervention in which participants were tested for curable sexually transmitted infections (STIs) at four-month intervals, and received cash rewards if STIs were negative. This paper focuses on the lessons learned from the RESPECT trial and its implications for future such efforts.
Caldeira, Foucault, and Rota-Graziosi analyze the average and distributional effects of decentralization on the access to some poverty-related public services in Benin. Analyzing panel data from local governments, called communes, for 2006 and 2007 they find that decentralization has a positive overall effect on access to basic services. However, this effect appears to be non-monotonic, following an inverted U-shaped curve. Moreover, it varies according to communes' wealth: it is positive for sufficiently wealthy communes and may be negative for the poorest ones. The authors also point out that decentralization may increase intra-jurisdictional inequalities, especially within the poorest areas. Decentralization in Benin successfully contributes to the reduction in poverty by improving the average access to poverty-related services. However, the devil is in the details, as decentralization seems to increase inequality in terms of access within and between communes.
Using datasets on household surveys and bank penetration at the district-level in 2006 and 2009, Allen, Carletti, Cull, Qian, Senbet, and Valenzuela explore the impact of Equity Bank - a leading private commercial bank - as well as other banks, on the use of banking services in Kenya. They find that the presence of domestic private banks (including Equity), government- or government-influenced banks, or foreign banks has a positive impact on financial access at the district level. In particular, Equity Bank has a first-order effect on enhancing households' access to bank accounts and bank credit. This effect is stronger for Kenyans with low income, without their own permanent house, with less education, and without a salaried job. These results imply that Equity Bank's business model - providing financial services to the population sectors ignored by traditional commercial banks, and generating sustainable profits in the process - can be an important solution to the financial access problem that has hindered the development of the financial sector in many African countries.
At the end of their pernicious civil wars, Liberia and Sierra Leone faced critical challenges relating to government revenues: resuscitating collapsed revenues; ensuring that natural resource revenues promote peace and development; and eliminating highly distortionary forms of taxation. Dessy examines the two countries' efforts to address these challenges. The empirical evidence here highlights the disparities between Liberia and Sierra Leone with respect to the state capacity to rehabilitate its revenue in the aftermath a civil war. A political economic model of state revenue rehabilitation then links state performance to the choice of revenue rehabilitation strategy. It also formalizes the impediments to state adoption of a Stiglitz - like patience strategy in the rehabilitation of public revenue after a civil war. An international civil society could substitute for the lack of a pro-active national civil society in a resource-rich post-conflict society. But unless such a device becomes part of the peace-consolidation process, the abundance of natural resources is bound to be a curse that creates a vicious cycle of civil wars and inefficient resource extraction.
Casaburi, Kremer, and Mullainathan use a newly collected plot-level panel database compiled from the administrative records of a large Kenyan contract farming scheme to study productivity and land allocation patterns between 1987 and 2006. They find that a negative rainfall shock reduces the amount of land allocated to cane in the subsequent replant cycle by 4 to 5 percent. However, smaller and more productive plots are less affected. These estimates suggest that plots in the bottom 25th percentile of the plot size distribution and in the top 10th percentile of the productivity distribution expand following a shock. Combined with strong evidence of decreasing returns to land, the result suggests a potential "cleansing effect of recessions". The researchers then compare the observed land allocation across producers with the one that maximizes expected total output. They find that optimization implies both a redistribution of land toward small producers and an increase in variance of plot sizes. The observed misallocation implies an average loss of about 5-15 percent of total yearly production. Finally, they use a simulation to assess the dynamic impact of rainfall shocks. Shocks have a static negative effect on yields but a potentially positive dynamic effect via reallocation toward more productive and smaller producers. The authors show that a one-time aggregate rainfall shock that hits all of the plots reduces contemporaneous output by 15 percent but raises output in the subsequent period by 2 percent.
In 2007 the Nigerian federal government began implementing a conditional grants scheme (CGS) that has made funds available to state governments to pursue the Millennium Development Goals, with a special focus on projects addressing the key health-related goals of reducing child mortality rates and the prevalence of disease, and improving maternal care. Has the program been effective in improving public health outcomes? To evaluate the impact of the program Baggott, Hainmueller, and Hiscox have examined evidence on health outcomes gathered in two large-scale surveys of Nigerian households conducted in 2006 and 2009. The preliminary findings indicate that the initial CGS program, administered by state governments, has not had large positive effects on health outcomes in Nigeria. By some measures, health outcomes actually appear to have changed for the worse in local government areas (LGAs) participating.
Since independence, a quiet revolution has taken place in maize production in the Sahel, with Mali increasing production more than ten-fold and yields going up about 2 percent a year. Foltz, Aldana, and Laris use farm-level panel data from southern Mali's maize growing regions to demonstrate this success in agricultural production and technological change. They analyze the determinants of production to separate increases in input use from technological change. The estimates show that farmers' adoption of increased fertilizer use, rather than the adoption of improvements in seeds and management, has driven much of the productivity growth. They also find strong evidence of observed and unobserved heterogeneity, which affects both the choice of fertilizer amounts and the marginal returns to fertilizer use. The results demonstrate the key changes behind this silent maize revolution and point to the importance of taking into account farmer heterogeneity in estimating productivity and returns to fertilizer.
In the study of political systems it is often asked how power is allocated and shared across multiple heterogeneous groups. This is a particularly relevant question in the case of autocracies, which often lack formal rules for the allocation of power. Rainer and Trebbi present novel evidence on the power-sharing dynamics of national political elites in a panel of African countries, most of which were autocracies for the vast portion of the period being analyzed. Using a new dataset on the ethnicity of cabinet ministers since a country's independence, they show that political power is allocated across ethnic groups in proportion to their population shares, which is in line with what is observed in parliamentary democracies. They also show that the ruling ethnic group is over-represented relative to other groups in the government by a large margin, which is not different from standard "formateur premia" of democracies. Movements in and out of autocracy in the country do not reduce disproportionality in the representation of ethnicities in the cabinet. As opposed to the view of a ruling ethnic elite which monolithically controls power in Africa, this paper shows that power is proportionally divided between ethnicities, implying that within-ethnicity frictions may be more central in explaining political failure than previously assessed.