flow to their highest valued use, in practice that is not necessarily so if the transfers confer external benefits or costs on third parties.

Such external effects are not rare. In water, for example, transfers from one use to another can affect the quality, quantity, and timing of supply for other downstream users³⁸ (Livingston, 1998). In air pollution control, transfers can affect the spatial distribution of pollution, and that can trigger environmental justice concerns (Tietenberg, 1995b).³⁹ In fisheries quota could be transferred to holders with more damaging gear, or a higher propensity for bycatch. In all cases “leakage” provides another possible external effect. Leakage occurs when pressure on the regulated resource is diverted to an unregulated, or lesser regulated, resource, as when fishermen move their boats to another fishery or polluters move their polluting factory to a country with lower environmental standards.

Western U.S. water markets attempt to solve the externality problem by giving any affected party a chance to intervene in the transfer proceeding (Colby, 1995). In the case of a third-party intervention, the transferring parties bear the burden of establishing the absence of damage to third parties. Although this is probably an effective way to internalize the externality, it raises transaction costs significantly and has resulted in many fewer transfers than would have been the case otherwise (Livingston, 1998). Technology is now making an entrance in water markets (the Water Links electronic water exchange in California, for example) to lower transaction costs (Organization for Economic Co-operation and Development, 1999).

One strategy used in U.S. air pollution control policy to resolve the spatial externality problem is regulatory tiering. Regulatory tiering implies applying more than one regulatory regime at a time. Sulfur oxide pollution in the United States is controlled both by the regulations designed to achieve local ambient air quality standards as well as by the sulfur allowance trading program. All transactions have to satisfy both programs. Thus trading is not restricted by spatial considerations (national trades are possible), but the use of acquired allowances is subject to local regulations protecting human health via the ambient standards. The second regulatory tier protects against the harmful spatial clustering of emissions (by disallowing any specific trades that would violate the standards), while the first tier allows unrestricted trading of allowances. Because the reductions in sulfur are so large and most local ambient standards are not likely to be jeopardized by trades, few trades have been affected by this provision. Yet its very existence serves to allay fears that local air quality could be in jeopardy.

Standard theory suggests that a fully value-maximizing tradable permit system must have full temporal fungibility, implying that allowances can be both borrowed and banked (Kling and Rubin, 1997; Rubin, 1996). Banking allows a