Over at Resources for the Future, Bob Hahn and Robert Stavins have posted a new discussion paper (here) on the implications of the "Coase Theorem" for the allocation of emissions allowances in a cap-and-trade system.
The Coase Theorem basically says that entitlements to resources will ultimately be allocated by the market (or by some central planner, as Steve Cheung pointed out) to maximize efficiency, regardless of the initial allocation, so long as transaction costs are zero and the other standard assumptions of neoclassical economic theory hold. Hahn and Stavins refer to market reallocations that improve efficiency regardless of initial allocations as the "independence property" of the "Coase Theorem." They note that "[a] number of factors call the independence property into question theoretically, including market power, transaction costs, non-cost-minimizing behavior, and conditional allowance allocations. However, "in practice," they find, "support for the independence property in some, but not all cap-and-trade applications."
I do not doubt their empirical finding for a moment. Markets often do manage to reallocate entitlements to more highly valued uses. But - and this cannot be over-stressed - that has nothing to do with the "Coase Theorem." As Coase himself has noted time and time again, the assumptions behind that theorem, including most importantly the assumption of zero transaction costs, never hold in the real world. When market reallocations improve efficiency, it is not because of the "independence property" of the "Coase Theorem"; it is in spite of the existence of positive transaction costs and other impediments to transacting. Always.
It would be useful for economists (and legal scholars) to bear in mind that the "Coase Theorem" doesn't posit that markets will sometimes reallocate property rights to improve efficiency; it posits that they will never fail to do so under the conditions specified in neoclassical economic theory. Those are two very different propositions, with very different implications. In finding that emissions markets sometimes, but don't always reallocate allowances to maximize efficiency, Hahn and Stavins are only telling us something we should already know: that the "Coase Theorem" does not apply, and has no implications for cap-and-trade regimes in the real world.