Combining Actual and Contingent Behavior to Estimate the Value of Sports Fishing in the Lagoon of Venice

When conducting a travel cost method (TCM) study, researchers often collect observations on actual trips to the natural resource of interest and on intended trips under specified but hypothetical circumstances\u2014usually, changes in prices and/or site quality. Actual and hypothetical trips are combined to estimate single-site TCM demand function for trips. We estimate severalmodels to test whether it is acceptable to pool hypothetical and actual trip data, focusing on respondent heterogeneity in the contingent behavior questions. We apply thesemodels to aTCMstudy about therecreationaluse of the LagoonofVenice for sports fishing. The Lagoon of Venice is a complex natural resource and ecological system that in recent decades has experienced overfishing, water pollution, sediment deposition and the consequences of the introduction of an exotic species of clam. Local policies are being considered that would eventually result, among other things, in restoring fish stocks, thus benefiting recreational anglers. In addition, the European Union's Water Framework Directive includes recreational benefits among the benefits of reaching \u201cgood\u201d ecological and chemical status in rivers, lakes, and coastal waters. In April\u2013July 2002, we conducted a mail survey of anglers with valid licenses fishing on the Lagoon of Venice to gather data on their fishing trips, behaviors and expenditures over the previous year. We also asked questions about trips that would be undertaken under hypothetical changes in the price of a trip and/or in the catch rate. Our models suggest actual and contingent behaviors are driven by the same demand function, and can be pooled for estimation purposes. We use this estimated demand function, and its shift when the catch rate is improved, to compute angler surplus at the current catch rate and the change in surplus accruing from a 50% improvement in the catch rate. For the average angler in our sample, the former is about \u20ac1700/year, while the latter is about \u20ac2800