Empirical identification of perceived congestion

This paper investigates the empirical identification of perceived congestion and mitigating behavior using observational data. Congestion effects are identified using a procedure based upon a nonlinear function of the choice attributes combined with an equilibrium condition on the sorting behavior of participants. Results suggest that congestion effects can be identified, under certain assumptions, using only revealed preference data, and that ignoring these effects underestimates the strength of preferences for other attributes. The model is applied to data on rock climbers, and the resulting estimates used to simulate the reopening of a currently closed section of a popular NY rock-climbing area and the re-sorting of climbers that would result.