Day-to-day dynamics of urban commuter departure time and route switching decisions: Joint model estimation

This chapter focuses on the day-to-day dynamic decisions of departure time and route by actual commuters in the Dallas, Texas urban area. A model previously developed based on "laboratory like" experiments provides the framework for analysing this phenomenon. The underlying behavioural mechanism is a boundedly-rational search to achieve an acceptable commute. Joint Indifference Bands (IBs) for route and departure time acceptability decisions are formulated, and specifications are developed for both the systematic and random components. Ln particular, econometric issues associated with specifying the random error structure are addressed for parameter estimation purposes. Insights into the effects of attributes are obtained through the analysis of the model's performance and estimated parameter values. A general probit model form is used for the dynamic switching model, allowing the introduction of state dependence and serial correlation in the model specification. The Monte Carlo simulation approach is followed to obtain maximum likelihood parameter estimates of the multinomial probit choice probability function. The data used for estimation were obtained from a field survey in Dallas, Texas conducted over a period of 10 working days, yielding observations on consecutive day-to-day decisions of commuters.