To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.jeem.2006.08.001 Byline: Charles D. Kolstad (a)(b) Keywords: Climate change; International environmental agreements; Voluntary provision of public goods; Treaties; Uncertainty; Learning Abstract: This paper addresses the subject of self-enforcing international environmental agreements (IEAs). The standard model of IEAs is adapted to include uncertainty in environmental costs and benefits, as well as learning about these costs and benefits. The paper investigates the extent to which the size of the coalition changes as a result of learning and systematic uncertainty (also known as model uncertainty). Results are that systematic uncertainty by itself decreases the size of an IEA. Learning has the further effect of either increasing or decreasing the size of an IEA, depending on parameters of the problem. Author Affiliation: (a) Department of Economics and Bren School of Environmental Science & Management, University of California, Santa Barbara, CA, USA (b) University Fellow, Resources for the Future, Washington, DC, USA Article History: Received 15 September 2005 Article Note: (footnote) [star] This paper has benefited from comment from two anonymous referees, Brian Copeland, Ita Falk, Michael Finus, Don Fullerton, Andy King, Eric Orts, Rohini Somanathan, Eswaran Somanathan, Samudra Vijay, David Zilberman and from seminar participants at Dartmouth College, Harvard University, North Carolina State University, Yale University, the Delhi School of Economics, the Indian Institute of Management (Calcutta), the Indian Statistical Institute (Delhi), the Madras School of Economics, and the Universities of Central Florida, Connecticut, Wyoming, and Leuven (Belgium). I am particularly grateful to Alistair Ulph of the University of Manchester for many discussions early on and throughout the development of this paper. In fact, the model presented her is derived from a model in Ulph and that intellectual debt should be recognized. Furthermore, Prof. Ulph perceptively pointed out a flaw in a result which no longer appears in this paper.