A BEHAVIORAL APPROACH TO MODAL SPLIT FORECASTING

MUCH OF THE LITERATURE ON MODAL SPLIT SUFFERS FROM EITHER EXCESS AGGREGATION OR INSUFFICIENT ATTENTION TO BEHAVIORAL CONSIDERATIONS. THIS PAPER FORMULATES A BEHAVIORALLY ORIENTED MODEL OF INDIVIDUAL CHOICE. THE MODEL IS COMPARED TO A NUMBER OF EXISTING MODELS ON THEORETICAL GROUNDS AND IS ALSO TESTED ON 1956 CHICAGO DATA FOR THE JOURNEY TO WORK. THE MODEL IS DESIGNED TO PREDICT THE DIVERSION EFFECTS OF GIVEN CHANGES IN SYSTEM CHARACTERISTICS, E. G. A CHANGE IN BUS FARES OR PARKING COSTS, OR CHANGES IN COMMUTER CHARACTERISTICS. THE MODEL ALSO PROVIDES A NEW WAY OF MEASURING THE VALUE OF TRAVEL TIME. SPECIFIC ESTIMATES OF VALUE OF TRAVEL TIME, AND THE EFFECTS OF RELATIVE SYSTEM IMPROVEMENTS, ARE CALCULATED. ALL OF THE ESTIMATED COEFFICIENTS AND RESULTS ARE IN GOOD ACCORD WITH "A PRIORI" THEORETICAL CONSIDERATIONS. /TR/