Transaction Costs, Resource Allocation and Liability Rules--A Comment

IN his article on "The Problem of Social Cost" Professor Coase argued that (assuming no transaction costs) the same allocation of resources will come about regardless of which of two joint cost causers is initially charged with the cost, in other words regardless of liability rules.' Various writers--including me-accepted that conclusion for the short run, but had doubts about its validity in the long run situation. The argument was that even if transactions brought about the same short run allocation, liability rules would affect the relative wealth of the two joint cost causing activities, and in the long run this would affect the relative number of firms and hence the relative output of the activities.2 Further thought has convinced me that if one assumes no transaction costs-including no costs of excluding from the benefits the free loaders, that is, those who would gain from a bargain but who are unwilling to pay to bring it about--and if one assumes, as one must, rationality and no legal impediments to bargaining, Coase's analysis must hold for the long run as well as the short run. The reason is simply that (on the given assumptions) the same type of transactions which cured the short run misallocation would also occur to cure the long run ones. For example, if we assume that the cost of factory smoke which destroys neighboring farmers' wheat can be avoided more cheaply by a smoke control device than by growing a smoke resistant wheat, then, even if the loss is left on the farmers they will, under the assumptions made, pay the factory to install the smoke control device. This would, in the short run, result in more factories relative to farmers and lower relative farm output than if the liability rule had been reversed. But if, as a result of this liability rule, farm output is too low relative to factory output those who lose from this "misallocation" would have every reason