Identifying Oil Extraction Effects in Developing Countries: Evidence from Ghana

Abstract

Over the last few decades, the role of natural resources to the development of economies has gained considerable attention. Empirical studies establishing the link between natural resources and development have largely resorted to aggregated country level studies with most finding little to no development in developing economies. The underlying reasons for such findings are the absence of infrastructure, stable government and presence of corruption. In recent years, there has been a switch in examining natural resource effect. Studies have focused on moving away from macro-level analysis to understanding how extraction of the resource lead to development at the micro-level of countries. This focus allows for heterogeneous study at the lowest administrative level. However, the current literature is largely centered on developed economies with little attention on developing countries, like Ghana that discovered one of the largest oil reserve in West Africa in over a decade in 2007 and started extraction in 2010. This thesis comprises of three chapters and each examines the effects of oil extraction in Ghana and various economic decisions that resulted from it. The thesis uses data from the Ghana Living Standard Survey which is a nationwide survey conducted by the Ghana Statistical Service in collaboration with the World Bank. The first chapter examines how oil extraction made it possible for the government of Ghana to implement a new wage policy -the Single Spine Pay Policyfor public sector workers. By adopting a novel unconditional quantile estimation within a difference-in-differences framework, the chapter examines the significance and effectiveness of this policy in addressing wage disparities and more importantly productivity of public sector workers. The study finds that the policy largely affected public sector workers at the lower tail of the earnings distribution. Female workers in the education and health sub-sectors and male workers in the administration sub-sector were the largest beneficiaries. However, there was a reduction in the level of productivity for public sector workers largely from the beneficiaries of the gains in earnings. The second chapter examines spillover effects of oil extraction on income, employment and migration in Ghana. To capture the spillover effects, individuals living in districts closer to the oil extraction area are used as treated group and those living further away (about 250km) are used as control group. The study employs a difference-in-differences strategy and finds that there is, on average, a positive spillover effect on the income of individuals living closest to the oil extraction area. These effects decrease for income and migration but increase for employment, the further away an individual is from the oil extraction area. Moreover, the spillover effects are heterogeneous and vary by gender and sector. The positive effect on income observed is largely for men and workers in the agricultural and retail sectors. Migrants are mostly women and workers in retail and other services. There is a significant reduction in agriculture and service sectors employment but an increase in the manufacturing sector. The findings suggest that oil extraction in the south of Ghana deepens the economic disparities that exist between the north and south of Ghana. The third chapter examines the effect of oil extraction on the well being of households. Existing empirical literature examining impact of natural resources at the micro level, undertake a partial analysis of the well being with most focused on average estimation with no consideration for changes along the distribution of these outcomes. This study employs an unconditional quantile technique using the Recentered In uence Function (RIF) in a difference-in-differences framework. Examining oil effect along the distribution of expenditure, the study uses households in close proximity to the oil extraction area as treated group and households further away as control group. The study finds that oil extraction has a positive effect at low expenditure levels or on poor households, but negative effects at high expenditure or on rich households. These results are broadly consistent with microeconomic theory predictions given a downward sloping oil extraction effect on households' expenditure share on food and an upward slope on non-food expenditure share.