Stated Preference Methods and Models: Analyzing Recreational Angling in New England Groundfisheries

Abstract

Policy analysis of nonmarket goods requires accurate knowledge about the behavior of economic agents. This dissertation explores several facets of behavior models in recreational angling for three New England groundfish species.

Stated preference methods are used frequently for nonmarket applications because data are scarce, but survey design can affect the results of behavior models via changes in respondents' cognitive processes. Methodological biases due to task complexity, measured by survey length, number of alternatives, and the degree of information overlap are observed in discrete choice experiment questionnaires, evidenced by differences in estimated model parameters and error variances. Additionally, ignoring task complexity increases mean marginal willingness-to-pay estimates. Information processing and decision heuristics should be considered in survey design and accounted for in estimated models.

Empirical specifications for utility models of recreational angling are also explored because numerous variants are employed in analyzing stated preference data. Inclusion of responses from different survey subpopulations affect estimated utility function parameters and mean marginal willingness-to-pay values. Utility models that are nonlinear in catch are as statistically robust as their linear counterparts but allow for diminishing marginal utility in fish, which is more consistent with recreational angling behavior. Failure to account for sources of heterogeneity such as angler avidity, species familiarity, and demographic information affect behavioral interpretations considerably.

Recreational fisheries are commonly managed using bag (creel) and minimum size restrictions. Many surveys include regulations as attributes in choice experiments, but models of angler behavior should not contain regulatory variables explicitly because they rarely factor into angler participation decisions directly. Because catch is random, regulations affect angler decisions indirectly by changing the underlying distributions for keep and release. A framework for understanding the effect of regulations on angler behavior given the stochastic nature of catch is developed. Short-run and long-run fishery implications are evaluated using a bioeconomic simulation.