An exploration of the robustness of alternative laboratory methodologies: Matching funds and the provision of public goods

Abstract

The voluntary provision of a pure public good is studied in the presence of an anonymous external donor. New data generated using experimental procedures employing both extra-credit and cash incentives, as well as asynchronous access to real-time decision rounds lasting several days, are compared to previous data generated using traditional cash-only, synchronous-access laboratory procedures. The effect on resource allocations to the public good of introducing external matching funds is examined in two different settings, lump-sum matching and one-to-one matching. The new data confirm the robustness of results previously reported in Baker et al. (2009) to the change in laboratory procedures and incentives. The new data are then used to extend the parameter space in which the two matching mechanisms are studied, including: varying within-round information regarding the current level of public-good allocations and varying group size from four to twenty group members. Allocations in lump-sum matching are no worse, and sometimes better, than one-to-one matching in these new treatments.