Count regression models for recreation demand: an application to Clear Lake

Abstract

<p>An important objective for policy-makers is how to allocate resources for the enjoyment of its citizens. Outdoor recreation is a very popular hobby for a lot of people. The sites they travel to for recreational purposes are public sites such as the Clear Lake, located in central Iowa. The users of the lake often care about the quality of the water. It is the goal of the researcher to determine how much they are willing to pay in order to preserve or improve the water quality. The researcher must decide on not only the theoretical methodology, but the appropriate statistical model. The focus of this thesis is using count data models to estimate individuals' willingness to pay. A common count data model is the Poisson model, however it is restrictive and often alternative models must be used. This thesis introduces a new count data model to the literature: The Conway-Maxwell-Poisson regression model. Using the data gathered by individual users at Clear Lake, I contrast this model with a popular alternate to the Poisson model, the negative binomial model.</p>