The relationship between transportation demand and supply: Granger-Causality test using time-series data

Abstract

Transport demand and supply are deemed to determine each other in a cyclic manner. The major idea has been that the demand is usually the preceding one. However, in urban cases, usually the land use variables in place of supply interfere this process. Cleansing the land use variables, the regional/national level variable pairs of demand and supply are employed to analyze the cause-effect mechanism. For objectivity, the Granger-causality test (GCT) is used to understand the relationship between transportation demand and supply. The Analyses were made at four dimensions; (a)whether the nexus is one-directional or bi-directional, (b)its significance level, (c)whether demand or supply is the preceding, (d)whether the effects are short-term or long-term. Using the Turkish statistics, the GCT results showed that, in the short/medium run, overwhelmingly the supply variables preceded (mostly in railway mode), mostly unidirectional (one-way causality) manner, however, in the long-run almost no relationship was found. In other transportation modes, no significant relationship is observed. Finally, bi-directional relations were usually observed in suburban rail. The investments then should be made according to known demand. Usually, the effects of supply (especially of railways and roadways) could rather fade away in the long-run. Still, no general statement can be made for the demand/supply causality especially in terms of which one is preceding and of the direction of causality. The chaotic nature of the process reigns over with the changing conditions.