The effects of fiscal windfalls on corrruption and selection into politics: evidence from the Peruvian commodity boom (2003-2014)

Abstract

Some resource-dependent countries experience substantial increases in government revenue during natural resource booms. Recent theoretical and empirical contributions (Robinson et al. 2006, Brollo et al. 2013) have argued that such abundance of windfalls can have a negative effect on the functioning of local institutions. In particular, abundance of windfalls could increase the incidence of corruption because their presence aggravates the moral hazard problem that exists between citizens and their elected authorities. This is so because an increased budget means that officials can appropiate rents illegally without compromising their obligations with the electorate, thus distorting the inferences that citizens make about their authorities’ competence. Furthermore, such agency problem could be aggravated by the self-selection of relatively worse candidates entering politics (assuming rents are more valuable for the relatively less skilled), because incumbent mayors (including corrupt ones) would face weaker competition. Exploiting substantial time and spatial variation in the amount of mining-related transfers received by the districts of Peru during the 2000s commodity boom, I implement a difference-in-differences strategy to put the mentioned theory to test. I find evidence supporting the hypothesis of a non-monotonic effect of windfalls on the incidence of corruption, and a negative effect of windfalls on the quality of candidates drawn to challenge incumbents and compete for public office. I do not find strong evidence of the theorized relationship between the two phenomena, but I propose further research pathways to improve our understanding of the results.