Quantifying the Incidence of a Global Oil Price Shock in Nigeria : A Roy Model with Non-Homothetic Preferences and Endogenous Transfers

In this paper, I explore the distributional effects of a negative terms-of-trade shock to an oil-exporting country. Specifically, I begin by developing a computable Roy model of the Nigerian economy with non-homothetic preferences, after-tax income effects, and government endogenous transfers through oil subsidies. I then use this model to simulate the impact of the global oil price shock that hits the Nigerian economy in 2016. In doing so, I focus on differential impacts across workers due to their different skills, locations, and comparative advantages. I find that the oil shock led to a reverse Dutch disease characterized by increases of real output, relative wage and labor in the manufacturing and agricultural sectors. Further, the combination of the shock and subsidies removal affected poor and rich income groups differently, in accordance with their labor specialization patterns across sectors and differences in the consumption bundles they consumed. More precisely, I show that the shock causes aggregate welfare losses that are concentrated in high-income groups, as they tend to specialize in the crude oil and service sectors, the most hit by the shock. Next, using a series of government budget-balancing tax policy counterfactuals, I first show that the removal of the subsidies was the best response to the shock regarding the welfare outcomes of low-income workers. Second, I show that while removing the subsidies, the government can use fiscally neutral progressive tax redistribution schemes to lessen the extent of the welfare losses, particularly on the poor populations. Finally, I show that not accounting for heterogeneity in the expenditure behavior would introduce biases in the magnitude of the welfare changes