Unlocking water markets: an experimental approach

Abstract

Despite decades of advocacy by water resource economists, water markets (leases and sales of water rights between willing buyers and sellers) have largely failed to develop in the western US. Although there are a number of explanations for this failure, we explore one potential reason that has received less attention in the economics and policy literature: farmers as sellers may have preferences for different elements of a water market transaction that are not captured in the relative comparison of their profits from farming and their profits from agreeing to a deal. We recruited irrigators with senior water rights in the upper Yakima River Basin in Washington state to participate in a series of experimental auctions. These auctions asked participants to imagine that they owned and operated a 100-acre timothy hay farm with a given level of net revenue (i.e. an induced value). Participants then reacted to series of offers for 1-year leases from hypothetical buyers where several attributes of the lease varied across tasks. These attributes were the type of buyer (the state Department of Ecology, an irrigation district, or a developer), the management of the water bank (run by the state or by a new nonprofit), the lease type (split-season and full-season), and the offer price. Participants were paid in relation to their hypothetical farms‟ earnings. This paper presents results from 7 sessions with irrigators (n=49) and a comparison group of undergraduates (n=38). Our results show that irrigators are more likely to accept split-season than fullseason leases (controlling for differences in farm profits). Compared to the Dept of Ecology, they are more likely to accept a lease from an irrigation district and less likely to accept an offer from a Developer. They do not, however, have strong preferences over the management of the water bank. Although offer price was statistically significant in both groups, we find students were more influenced by offer price. Students had no detectable preferences over lease type, although some preferred a state-run water bank when leasing to a developer. Most notably, we find farmers were far more likely than students to reject offers from buyers even though it would increase their winnings from the experiment. Our results could be used in ongoing water supply policy debates in the Basin to simulate the amount of water that could be freed by water markets.