Single versus Multiple Prize Contests to Finance Public Goods: Theory and Experimental Evidence

This paper investigates single and multiple prize contests as incentive mechanisms for the private provision of public goods, under the assumptions of income heterogeneity and incomplete information about income levels. We compare experimentally a one-prize contest with a three-prize contest in a case where theory predicts that several prizes maximise revenues. We find that, contrary to the theoretical predictions, total contributions are significantly higher in the one-prize contest. In both treatments contributions converge towards theoretical predictions over successive rounds, but the effects of repetition are different: convergence is fast in the one-prize treatment, while gradual and with some undershooting in the three-prize treatment. Focusing on individual income types, the better performance of the single-prize contest is largely explained by the contributions of high income individuals: a single larger prize provides a more effective incentive for richer individuals than three smaller prizes.