Permit markets to control pollution have long been popular witheconomists. In recent years they have been embraced increasingly bypolicymakers as well. Conventional wisdom holds that a permit market mustbe competitive to be successful. In the case of the U.S. SO2 allowance market,avoiding market power was deemed sufficiently important that the lawcreated a single national market for allowances. Thus, any significantcontrol over the spatial distribution of emissions was sacrificed. I argue thatthis prioritization was misguided. I develop a spatial framework thatexplicitly accounts for both costs and damages in a set of regions betweenwhich a single pollutant can travel. I show that the welfare losses due tospatial misallocation of emissions are likely to be much larger than anypotential losses due to market power in the smaller regional markets.Moreover, I argue that a small number of traders is unlikely to be a problem,for two reasons. First, they will be on opposite sides of the permit market, sobilateral monopoly is more apt than the usual monopoly or monopsonyanalogy. Second, because they are large but few in number, such firms arelikely to achieve a bargaining outcome that leads to the least-costdistribution of emissions.