JOHN P. O'CONNOR [*] Most river basins have experienced development of water projects to promote flood control, recreation, and hydropower and agricultural production. Though the projects helped establish stable economies, there have been adverse impacts to the natural environment and wildlife that reside in these river basins. One of the key policy tools for habitat restoration is management of instream flows. Alternative water banking policies for restoring habitat are compared using a spatial optimization model to identify which policies work best to augment instream flows for habitat restoration. Results indicate that policy goals dictate what type of water bank is preferred. (JEL 025, 018, D61) I. INTRODUCTION During the past century, there has been substantial development of water projects on most of the major river basins in the western United States. The projects reshape basin hydrology to provide flood control, production of hydropower, navigable waterways and to augment stream flows for agricultural production during seasonal dry periods. As projects became operational, water users were granted rights to use the water under the prior appropriations doctrine, where the priority of the right was determined by when the grower initially diverted the water. Consequently, the allocation of water is conducted according to legal rather than economic principles, resulting in a large disparity between the value of water in its many existing and potential uses. Though the projects helped establish stable agricultural-based economies and provided low-cost power to encourage development, there have been adverse effects to the natural environment. Changing the timing of stream flow has resulted in impacts to water flow for fish populations, wetland and wildlife habitat, ecosystem health, and recreation. The impacts to salmon species have been especially pronounced, as the construction of water projects has both impeded fish passage and altered stream flows critical to the fishes life cycle. In Idaho's Snake River there are three salmon stocks listed as threatened or endangered under the Endangered Species Act. To mitigate these impacts and aid in the recovery of endangered salmon species, the National Marine Fisheries Service (NMFS, 1995) recommended the establishment of minimum stream flow targets for the Snake River. The Bureau of Reclamation (BOR) and the Army Corps of Engineers have begun studying a variety of methods for achieving instream flow targets, including water banking. Water transfers have been proposed as a way to reduce the allocative inefficiencies of the prior appropriations doctrine (Burness and Quirk, 1979; Colby, 1990; Howe et al., 1986). Water transfers have received the most attention during drought periods to move water within agriculture (Dinar and Letey, 1991) or to urban uses (Taylor and Young, 1995; Michelsen and Young, 1993). However, recent attention has been given to transferring water for environmental uses (Fadali and Shaw, 1998; Weinberg et al., 1993). Most studies have focused on the implementation of water markets; however, water banks are more common in practice (Howitt, 1994; MacDonnell et al., 1994, 1-3). Though the distinction between these institutions is not precise, markets generally facilitate the permanent transfer of water rights and banks facilitate the temporary transfer of the right to use storage water. As long as water banking is recognized as a beneficial use of water, banks may be more attractive to many water right holders because th ey are able to maintain control over future use of the water and opportunities that may arise from having that right. This removes the all-or-nothing nature of a permanent transfer. Water banks serve as institutional intermediaries that help reduce the cost of transferring water between users. These costs fall under two categories--the cost of making water transfers and the third-party impacts associated with transfers, both of which can be substantial. …