Integrating Revealed and Stated Preference Data to Improve the Estimation of Baseline Risk Fish Ingestion

Quantifying baseline regulatory risk at sites with fish consumption advisories requires estimating what fish consumption would be if the relevant fish consumption advisories were not in force. (Not in force means that no warning sign is present on site and/or no published materials warn anglers about risks from consuming site fish and crab, but all other site conditions remain as they are currently.) Examining what anglers’ fishing and consumption decisions would be in the absence of fish consumption advisories involves modeling angler behavior; therefore, accounting for the effect of the fish consumption advisory in the baseline risk assessment requires an angler behavior model that can examine angler choice under counterfactual advisory scenarios. This manuscript presents the results of an angling model evaluating the array of advisories that are most relevant for the policy outcome needed to inform estimates of baseline-risk fish ingestion. The model integrates revealed preference data presented in Bingham et al. (2011) with SP data collected during the 2013 New Jersey Outdoor Recreation Survey. The experimental design developed for the SP component was designed specifically to collect data on anglers’ stated preferences regarding fish consumption advisories that could be integrated with the existing revealed preference data presented in Bingham et al. (2011). The integration of the SP data improves the model specification, producing statistically significant coefficients on the policy variables that directly inform the evaluation of changes in anglers’ simulated trip-taking behaviors between current and baseline-risk conditions. The model results provide the angler preference function necessary to develop the simulations of changes in angler behavior and consumption presented in Kinnell and Bingham (2014).