How enforcement institutions affect impersonal exchange

In an experiment we study anonymous market interactions in the presence of “judges”. Our transactions resemble anonymous credit transactions where “lenders” can give loans and “borrowers” can repay them. When borrowers default, judges are free to enforce repayment but are themselves paid differently in each of three treatments. First, paying judges according to lenders’ votes maximizes surplus and the equality of earnings. In contrast, paying judges according to borrowers’ votes triggers insufficient enforcement, destroying the exchange opportunities and producing the lowest surplus and the most unequal distribution of earnings. Lastly, judges paid the average earnings of borrowers and lenders achieve results close to those based on lender voting. We employ a steps-of-reasoning argument to interpret the performances of different institutions. When voting and enforcement rights are allocated to different classes of actors, the difficulty of their task changes, and arguably as a consequence they focus on high or low surplus equilibria.