SEEMINGLY UNRELATED STATIC AND DYNAMIC URBAN TRAVEL DEMANDS.

THIS PAPER EXAMINES THE USEFULNESS OF GROUPING TRIP DEMAND EQUATIONS FOR ESTIMATION PURPOSES USING AS EXAMPLES TWO OF THE MARKETS SERVED BY THE MONTREAL URBAN COMMUNITY TRANSIT COMMISSION. THE EQUATIONS ARE ESTIMATED BOTH UNDER THE ASSUMPTION THAT CONSUMERS ADJUST IMMEDIATELY TO CHANGES IN FARE AND SERVICE FREQUENCY AND UNDER THE ASSUMPTION THAT THEIR RESPONSE VARIES OVER TIME. SINGLE-EQUATION ESTIMATES OBTAINED BY COMBINING BOX-JENKINS RTH ORDER AUTOREGRESSIVE PROCESSES WITH THE COCHRANE-ORCUTT MAXIMUM LIKELIHOOD ITERATIVE TECHNIQUE FOR NON-LINEAR EQUATIONS ARE COMPARED TO ESTIMATES OBTAINED BY USING A MAXIMUM LIKELIHOOD ITERATE OF PARKS' SEEMINGLY UNRELATED PROCEDURE GENERALIZED TO HIGHER ORDER AUTOREGRESSIVE SCHEMES. FORECASTING EXPERIENCE WITH BOTH ESTIMATION METHODS IS ALSO SUMMARIZED.(Author/publisher).