Power, Compensation and Corruption: Theory and Evidence

Abstract

Using a multi-period model of corruption in a heterogeneous society, we show that societal corruption and output crucially depend on the interplay among the productivity distribution, the societal compensation structure, and the distribution of power. The power distribution influences the curvature of the output-maximizing societal compensation structure in any period and the relation between societal corruption and inequality. Societal corruption varies in a U-shaped manner with the bias of the power distribution. Corruption attains its minimum in a society characterized by relatively higher levels of “petty” corruption compared with “high-level” corruption. Societies with differing power distributions can experience dramatically divergent evolutions of their corruption levels and output. The positive implications of our theory for the relations among corruption, compensation, power, inequality, and growth are consistent with prior empirical evidence. Our theory also leads to the testable prediction that societal corruption varies in a U-shaped manner with the relative level of public sector wages. We find significant support for this prediction in our empirical analysis. From a policy standpoint, our results suggest that high level corruption should be targeted by legal systems in developed economies while petty corruption is relatively more pernicious in developing economies, and that highly productive economies should have relatively powerful public sectors.