THEORY AND METHOD IN LAND USE AND TRAVEL FORECASTING

THIS PAPER REVIEWS THE THEORY OF DEMAND AND ITS TRANSLATION INTO CURRENT METHOD IN URBAN TRANSPORTATION PLANNING, NAMELY, THE CONVENTIONAL SEQUENCE OF URBAN TRAVEL-FORECASTING MODELS. THE CURRENT MODELS ARE EXAMINED FROM THE PERSPECTIVES OF APPROPRIATE STRUCTURE, USEFULNESS IN PRACTICE, AND RELEVANCE TO EMERGING VALUES. THE MODELS APPEAR TO FAITHFULLY REFLECT THE UNDERSTANDING OF LAND USE LOCATION AND TRAVEL BEHAVIOR AND OF THE INFORMATION REQUIREMENTS OF AN EARLIER PERIOD IN URBAN TRANSPORTATION PLANNING. A TRANSPORTATION-RELATED GENERAL EQUILIBRIUM LAND USE MODEL IS DERIVED, BASED ON A CAUSAL THEORY OF TRAVEL, NAMELY, THE THEORY OF URBAN PERSON TRAVEL AS A DERIVED DEMAND. THIS LONG-RUN, ACTIVITY-DISTRIBUTION, GENERAL MODEL IS USED TO EXAMINE THE NEW SET OF (SHORT- RUN) TRAVEL-DEMAND MODELS EMPLOYING DIRECT AND CROSS RELATIONS AND THEN THE CONVENTIONAL SEQUENCE OF TRAFFIC MODELS: TRIP GENERATION, TRIP DISTRIBUTION, AND MODAL SPLIT. THE SIMPLIFYING ASSUMPTIONS REQUIRED FOR THESE MODELS ARE EXPLICITYLY EXAMINED FOR THEIR STRUCTURAL (CAUSAL) AND STATISTICAL IMPLICATIONS. IT IS CONCLUDED THAT SEPARATE MODELING OF SHORT-RUN TRAVEL DEMAND FROM LONG-RUN ACTIVITY LOCATION INTRODUCES STRUCTURAL AND STATISTICAL PROBLEMS WHOSE IMPLICATIONS REQUIRE FURTHER RESEARCH. HOWEVER, THE STRUCTURAL AND SPECIFICATION ERRORS REVEALED IN THE CURRENT CONVENTIONAL MODELS ARE SUCH THAT THEY ARE OF DOUBTFUL VALIDITY AND PRODUCE POSSIBLY MISLEADING TRAVEL FORECASTS. SUCH FORECASTS ARE IN DANGER OF BEING BYPASSED IN CURRENT URBAN TRANSPORTATION CONTROVERSIES, AND CONSIDERATION OF USER TRAVEL COSTS MAY BE BYPASSED WITH THEM.