Identifying Nonrational Behavior in Recreation Demand Models

Abstract

In the context of an estimated RPL (Random Parameters Logit) choice model of recreational demand for the game reserves of the KwaZulu-Natal province in South Africa, a test on the existence of an underlying rational preference structure is presented. The proposed procedure allows to identify nonrational behavior in the sample. Using two different data sets extracted from the original data base, we show that by improving the data, apparent nonrationality can be eliminated and welfare analysis may be safely conducted.