Pacific Economic Review, 5: 3 2000) pp. 429±446 PARALLELISM IN THE LAB AND THE FIELD: TESTING ROBUSTNESS OF THE MINIMUM-CONTRIBUTING-SET MECHANISM STEVEN STEWART University of Tennessee, USA MICHAEL MCKEE  University of New Mexico and Georgia State University, USA ROBERT P. BERRENS University of New Mexico, USA ALOK K. BOHARA University of New Mexico, USA DAVID BROOKSHIRE University of New Mexico, USA Abstract. Efforts to ``test'' public-goods provision mechanisms in field settings encounter a fundamental obstacle: investigators cannot determine whether the aggregate valuation of the public good exceeds the cost. Experimental laboratory settings can fix the provision of the public good to be efficient. This allows investigation of the performance of the mechanism under potential field settings. This paper reports the results of a set of laboratory experiments designed to test the robustness of the minimum-contributing-set MCS) mechanism to field conditions. The reported results support further use and investigation of the MCS mechanism for the provision of step-level public goods. 1. INTRODUCTION In many instances it is necessary or desirable to fund a collective good from voluntary contributions. It would be valuable to have a working mechanism that would generate successful provision of such goods when it is efficient to do so. In his recent survey of experimental research on the voluntary provision of public goods, Ledyard 1995) identifies several factors that contribute to group success in the voluntary provision of public goods. One factor he touches upon only briefly is the production process for providing the public good. Hirshliefer 1983) has suggested that the provision or production rule is pivotal to whether free-riding is sufficiently pervasive to prevent private provision entirely. One class of production processes is characterized by discrete step-level) public goods that can be provided only if a threshold sum of money is collected. Because of the binary nature of the production function, this class is usually referred to as a ``provision-point'' mechanism. Provision-point mechanisms have been found to be fairly effective in voluntary provision of public goods in a few field applications. Bagnoli and # Blackwell Publishers Ltd 2000. 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.  Address for correspondence: University of New Mexico, Albuquerque, NM 87131, USA. We wish to thank Carol Silva for discussions at the initial stages of this research, seminar participants at the University of South Carolina especially McKinley Blackburn and Glenn Harrison), two anonymous reviewers, and Stuart Mestelman for comments that led to substantial improvements in the exposition and the empirical analysis. Financial support for this research was provided by grant R824679 David Brookshire ± Project Administrator) from the US Environmental Protection Agency.