This dissertation studies whether politicians in rentier states are more likely to use public sector employment as a redistribution mechanism than in non-rentier states, in exchange for political support. I find that larger resource rents are associated with more public employment, conditional on being a democracy with low productivity and high inequality. In this context, larger resource rents can also change public employment composition, in favor of blue-collar workers and political appointees. Moreover, being resource-wealthy is associated with contextual factors that discourage public service reforms meant to curve down excessive or non-meritocratic public employment growth.

These results are consistent with theory and existing evidence that public jobs can be used by political elites for clientelistic redistribution - that is, in exchange of political support. When productivity is low and inequality high, it can be convenient for politicians to provide more public employment (versus public goods or other private transfers) because it increases their chances of staying in power. Natural resource rents further enhance this tendency, by raising the stakes of keeping office and granting a larger envelope for redistribution.

My findings are relevant because they call for realism in reforms aimed at making the public services of resource rich countries leaner or more meritocratic, by highlighting the potential incompatibility of these policies with political incentives. Furthermore, my work contributes to the literatures on the political economy of the natural resource curse, clientelism and public employment, and public service reforms.

In the first essay, I test existing theories of redistributive politics to consider whether natural resource rents affect politicians’ willingness to redistribute income through public employment. Using panel data for 138 countries over 23 years, I find that the relationship between resource rents and public employment size is contingent on the political regime type. In democracies, resource wealth is generally associated with larger public employment, as this strategy provides electoral advantages. The opposite holds for the average autocracy, since more resource rents increase autocrats’ ability to repress (hence reducing the need to please the broader population).

In the second essay, I explore the effect of natural resource revenues on municipal public employment in Peru. I exploit the variation in exogenous mining revenue shocks across municipalities, due to a legal reform in 2004 which sharply increased the mining revenues transferred to mineral-producing municipalities. Post reform, producing municipalities significantly increasedpublic employment, providing mainly temporary contracts to (predominantly) blue collar workers and political appointees. Consistent with theory, this employment growth composition suggests that it was partly driven by redistribution concerns and likely clientelistic.

In the third essay, I survey the (relevant) theoretical and empirical literature to explain why resource rich countries are less likely to implement meritocratic reforms of their public services (and thus reduce clientelistic employment). The review shows that public service reforms, while generally politically unattractive, are particularly challenging in resource-wealthy countries. This is because resource-wealth is associated with contextual factors that further discourage such reforms, such as lower productivity, being less democratic and more prone to violent conflict, and having less programmatic political parties and deeper political budget cycles.