Working Paper 145 - Assessing the Returns to Education in the Gambia
|Publishing Date||16/02/2012 15:41|
Description The importance of education in development is a perennial topic in economics especially in the context of sub-Saharan Africa’s development experience. The connection is not surprising since the region stands out both in its low level of schooling and its low average rate of economic growth. In the macroeconomic growth literature, evidence shows that education is positively associated with economic growth, a result that accords well with many previous studies. Micro-level research on private rates of returns to education has shown disparate estimates in sub-Saharan Africa in the private benefits to education. Our work focuses on private returns to education in The Gambia, a small country in West Africa with very low levels of schooling. Like other countries in the region, it also has achieved little economic growth since independence in 1965. It is therefore not surprising that the country is not on schedule to achieve one of the Millennium Development Goals: universal primary education by 2015. This work adds to the large literature that provides a range of estimates on the private rate of returns to education in Africa. What has been found to date is that there is a great deal of heterogeneity in estimates of returns to education in Africa. It could be the case that there are indeed very large differences between countries in the rates of returns to education since there has been very little replication of estimates within a single country. However, part of the difference in estimates may also be due to the use of improved econometric techniques among recent papers. Some of these new approaches have addressed issues such as ability bias and selection - problems that were not always addressed in many earlier papers. Another possibility is that differences in estimation strategies can also produce different results since the estimates may be specific to only a subset of the population in a given country. Specifically, the estimates from using an instrumental variable approach may not be comparable across different studies that employ different instruments since such an estimation strategy produces the local average treatment effects. Typical estimates using instrumental variables, in which the most common instrument measures access to schooling, provide measures of the returns to schooling for those who would have continued in school but did not have access to schooling. Given that in the African context there is great variation across countries, ethnic groups, religions, and the proclivity of parents to send their children to school even when it is available and affordable, one should also expect great variation in estimates of returns based on that population. This work contributes to the literature by providing the first estimates of the private rate of returns to education for The Gambia and among its regions. Our estimates rely on the exploitation of the exogenous variations in the availability of schools across the country at the district level and its interaction with year of birth of individuals to control for ability bias. In addition, we use exogenous rainfall shocks to control for selection bias. Like many instrument variables, ours are not perfect. We discuss the possible violation of the exclusion restriction and provide further robustness checks to mitigate them. Our study uses three nationally representative household surveys from 1992, 1998 and 2003 that provide a very high coverage rate for the overall population of The Gambia. The results show high and significant private rate of returns to education for individuals in the wage sector. We estimate private rate of returns to education of 23% overall, using instrumental variable estimation. In addition, we found that returns were higher for men and urban residents. The results also demonstrate large significant differences in the rate of returns to education across regions, with poorer regions registering higher rate of returns. It is worth stressing that all sub-regions and sub-groups experience high rate of returns to education. What these finding suggests is that barriers to schooling in terms of both direct and indirect costs are substantial in The Gambia. Therefore, government policies should be geared towards lowering the cost of education in the country.