Abstract
This paper examines the modeling of consumer decision making under uncertainty about congestion and the relative roles of situational versus antecedent variables. The primary hypothesis in that a failure to allow for the effects of unrealized prior expectations concerning congestion can impart a downward bias to estimates of individual benefits from relieving congestion. It is argued that the appropriate model of decision making at congested sites is a recursive system when users are (1) unable to predict congestion levels with certainty and (2) can only partially adjust the length of the stay to the optimal length once actual congestion becomes known.