Densities Rather than Shares: Improving the Measurement of Congestion in Recreation Demand Models

While economists have successfully addressed the endogeneity of congestion in revealed preference approaches, little attention has been devoted to measuring congestion. In fact, most applications measure congestion as a simple count of users or, at best, consider the shares of users, defined as the proportion of people visiting a site during the course of the season. However, only when all sites have a similar size or length and the total number of visitors is constant over time, can these measures be used as a proxy for congestion. In general, these assumptions are unrealistic. In this paper we examine site density measures rather than participation shares as a measure of congestion. We show that endogeneity problems can be addressed using a control function approach and demonstrate that the density of users leads to more statistically significant coefficients of congestion. Copyright Springer Science+Business Media Dordrecht 2015