Discrete/Continuous Contingent Valuation of Private Hunting Access in Kansas

In a recent paper in this Journal, Halstead et al. (1991, Journal of Environmental Management 33, 79-89) discuss the use to Tobit analysis in the evaluation of open-ended contingent valuation surveys with zero bids. This paper generalizes their analysis of the willingness to pay of a sample of 568 Kansas hunters for private hunting access. Because of declining real budgets for public wildlife provisions and limited hunting opportunities for non-landowners, considerable attention has been focused on the potential for user access fees to supply private hunting opportunities. As we note in this paper, Tobit analysis techniques implicitly require the use/non-use censoring rule reflected in zero bids to be generated by the same process as that underlying the continuous level of non-zero bids. An alternative empirical approach that allows a distinction between the use/non-use rule and the process generating the continuous bids of the sub-sample of users is described and applied to a consideration of willingness to pay for private hunting access. The restrictions implicit in the Tobit model are tested empirically and strongly rejected in favor of the alternative model. The empirical analysis demonstrates that variables may influence the discrete use/non-use decision implicit in zero bids in a manner that is significantly different from their effect on continuous, positive bid levels. This distinction suggests that zero bids may reveal statistically significant differences in the tastes and preferences of users and non-users. The results suggest that imposing the restrictions implicit in the Tobit model may induce specification biases and misleading inferences.