Elite Capture of Local Institutions: Evidence from a Field Experiment in Afghanistan

Abstract: This paper looks at the effect of local governance decision rules on elite capture of community resources. In an experiment implemented across 250 villages in Afghanistan we study the impact of two randomly-assigned methods for electing local development councils and of two randomly-assigned methods for selecting local development projects on the choice of development projects. For project selection, we contrast outcomes of consultation meetings—at which villagers discuss and debate alternative projects—with secret-ballot referenda. For elections, we contrast outcomes of cluster elections, in which a male and female representative are elected from each section of the village separately, against at-large elections, in which council representatives are elected across the entire village. The results indicate that the choice of projects is generally responsive to the preferences of non-elite male villagers. Members of the village elite have substantial influence on the choice of projects only in villages that combine at-large elections of the council with the consultation meeting procedure for project selection.