Corruption, investments and contributions to public goods: Experimental evidence from rural Liberia

Abstract

We analyze how corruption affects incentives to invest or contribute to public goods. We obtain a proxy for corruption among Liberian community leaders by keeping track of a flow of inputs associated with a development intervention, measuring these inputs before and after giving them in custody to the chief. We then use the “gap” between these measurements (“missing inputs”) to explain variation in investment behavior of villagers. Investment behavior is gauged with two simple artefactual field experiments. Our main results are that corruption (i) undermines incentives for voluntary contributions to local public goods and (ii) may reduce private investments of individuals subject to rent-seeking by the chief in real life. We also provide weaker evidence that the impact of corruption on investments and contributions to public goods is heterogeneous: this impact may be gender-specific and appears to vary with accessibility of communities.