Implications of realization uncertainty on random utility models: the case of lottery rationed hunting

摘要

Conditional multinomial logit (MNL) models are often used to estimate the value of non-market resources under the assumption that consumer choices will be realized with certainty. The conditional MNL may not be appropriate when choices made may notbe realized with certainty, such as when rationing mechanisms are used as a means of managing non-market resources. The case examined in this paper is that of moose hunting in Newfoundland, Canada, where a priority pool lottery rationing system is used to allocate moose hunting licenses. Two models that address this issue, developed by Boxall (1995) and Rouwendal (1989), are compared to estimate welfare measures for moose hunting in Newfoundland. The results from these two models differ significantly from the results of a simple conditional MNL model.