Stable international environmental agreements with a stock pollutant, uncertainty and learning

In this paper I address the question of how uncertainty about damage costs and the possibility of resolving that uncertainty in the future affects the incentives for countries to join an international environmental agreement. I use a two-period model with a stock pollutant where the number of countries generating pollution can be arbitrarily large. The stability concept employed is such that size of the stable IEA can be anywhere between 2 and the grand coalition of all countries depending on parameter values. The dynamic structure allows two different membership rules for an IEA: fixed ( countries commit at the outset to be members for both periods) or variable ( countries decide each period whether to join). I show that with fixed membership learning results in at least as high membership and global welfare as no learning ( unless both the expected value and variance of damage costs are high). With variable membership, learning leads to higher membership ( in the second period) but lower global welfare than no learning. For most parameter values variable membership results in higher global welfare than fixed membership.