Horizontal and Vertical Conflict: Experimental Evidence

Abstract

We experimentally explore the connections between horizontal conflict of interests (citizens have heterogeneous preferences over collective decisions) and vertical conflict of interests (agents in charge of implementing collective decisions earn political rents). We identify two sets of models that incorporate both types of conflicts: electoral models with endogenous rents, and common-agency models. We adapt these models to a laboratory setting and test their main theoretical predictions. In both cases we find support for the proposition that more intense horizontal conflict leads to higher rents. Our findings have important implications. At the macro level, they help explaining the persistence of corruption in very unequal societies. At the micro level, our findings suggest that anti-corruption programs should allocate more resources (e.g., inspectors and auditors) to areas with intense horizontal conflicts.