We are concerned with the definition of choice set used in Random Utility Models of recreation demand. In particular, we are concerned with the spatial boundaries used to define choice sets. In this paper, using a model of day-trip fishing in Maine, we examine the sensitivity of parameter and welfare estimates to changes in the spatial boundary. We fine that there exists some threshold distance beyond which adding more sites to the choice set has negligible effects on the estimation results.
Available at: http://works.bepress.com/george_parsons/22/