Temporal correlation in recreation demand models with limited data

Typical single-site recreation demand models assume the decision to take a recreation trip on a particular choice occasion is independent of all past trip decisions. This paper develops a model of recreation behavior that allows for temporal correlation between choice occasions, but can be estimated on standard recreation demand data (i.e. seasonal summaries of behavior). Nested within the temporally correlated model is the Poisson count data model often employed in single-site recreation demand studies.