Taxation, Corruption, and Engagement With the Formal State: Experimental Evidence From the D.R. Congo

Abstract

Few countries achieve peace and prosperity without a capable and accountable government. Political scientists and historians have documented the consolidation of modern nation states in Europe, and the emergence of accountable governance when citizens resist taxation. Although theoretically appealing, the presumed link between state building and citizen engagement in politics has escaped testing with modern quantitative methods due to the non-randomness of taxation and other state-building initiatives. I fill this gap by partnering with governments to embed randomized evaluations into their tax programs. In the first chapter, I examine the first field experiment to randomize tax collection: a door-to-door property tax campaign in Kananga, D.R. Congo. Consistent with the classic ‘tax-participation hypothesis,’ individuals living in neighborhoods assigned to the tax program are nearly five percentage points more likely to engage in costly participation. The second chapter investigates the determinants of tax compliance in this tax campaign. Despite the prominence of pecuniary motives in models of tax compliance, I find that the perceived legitimacy of the state is a stronger predictor of citizen tax payment. Moreover, the campaign improves citizens’ perceptions of the provincial government, indicating a positive feedback loop in tax collection and state legitimacy. In the third chapter, I examine citizens’ decisions to participate in corruption. This co-authored chapter explores the elasticity of citizen bribe payment at the roadway tolls in Kananga with respect to price. We argue that citizens’ limited responsiveness to monetary incentives to obtain a valid receipt reflects the fact that bribes accelerate interactions at the toll. The dissertation thus brings experimental evidence to bear on long-theorized links between state building and citizen political engagement.