THE DEMAND FOR INTERCITY PASSENGER TRANSPORTATION

TRANSPORTATION DEMAND MODELS TEND TO BE DEVELOPED FOR SPECIFIC PURPOSES AND ARE UNRELIABLE WHEN USED MORE GENERALLY. THEY ARE INVARIABLY MODE SPECIFIC AND CONSEQUENTLY PROVIDE NO WAY OF LOOKING AT TOTAL TRAVEL WHEN ONE MODE IS CHANGED SIGNIFICANTLY OR A NEW MODE INTRODUCED. THIS PAPER REEXAMINES SOME DEMAND MODELS THAT HAVE BEEN ESTIMATED AND ARE CURRENTLY THE BASIS OF POLICY. THESE ARE GENERALIZED AND RE-ESTIMATED. THE LIMITATIONS OF THE MODELS ARE THEN USED TO DEVELOP A MORE GENERAL CLASS OF DEMAND MODELS, SIMILAR TO THE ABSTRACT MODE MODELS OF BAUMOL AND QUANDT. WHILE THE BAUMOL-QUANDT ABSTRACT MODE MODEL IS A MAJOR THEORETICAL CONTRIBUTION TO DEMAND FORECASTING, IT ADDS LITTLE TO FORECASTING ABILITY. THE DATA DO NOT ALLOW ESTIMATION OF GENERAL MODELS AND FOR THIS REASON THE MODELS MUST BE USED WITH CARE. /AUTHOR/