The Effect of Fuel Prices on Demands for Road and Rail Travel - an Application to the French Case
DELSAUT Marianne, CETE De L'Ouest
This study is in line with the current reflections about the change in travel behaviour due to the increase in fuel price. This may encourage car users to car-pool, to shift from road to rail, to work from home, etc. This paper analyses the fuel price sensitivity of French traffic. Crucially, it develops the use of innovative econometric models. Partial adjustment models have been selected from the literature review to represent traffic between 1990 and 2010. The road traffic model includes explanatory variables such as the fuel price, GDP, the length of the motorway network, and a lagged variable of traffic. Furthermore, a rail demand model including fuel price as an explanatory variable is processed. After this, the short-term and long-term fuel price elasticities for transport demand are measured as a means of testing the national transport policies to be implemented, especially in the fields of fuel taxes and modal shift. Therefore, the use of elasticity estimates in traffic forecasting models is discussed. As a result, this study falls within the scope of a project promoting the sustainable development of regional territories.
This study is in line with the current reflections about the change in travel behaviour due to the increase in fuel price. Fuel cost 0,89 Euros2005 per litre in 1990, increasing to 1,22 Euros2005 per litre in 2008. This rise may encourage car users to car-pool, to shift from road to rail, to work from home or to reconsider their workplace and housing arrangements in order to reduce their journey distance to work. One way to estimate future travel behaviour is to build econometric models of transport demand. The resulting estimates of price elasticities are key-results to prepare for the future transport demand. Numerous studies have been undertaken in order to quantify the impact of a rise in price on the demand. However, they mainly focus on US and European cases. This paper analyses the fuel price sensitivity of French traffic. Crucially, it develops the use of innovative econometric models. Moreover, in France, most traffic forecasting models do not consider the impact of fuel price sensitivity on traffic for two main reasons. Firstly, further studies are needed to set the standard values of elasticity for use in traffic forecasting models. Secondly, it is still unclear at what stage of the four-stage models it would be appropriate to consider fuel price elasticities. That is why there is a need to reassess the relationship between traffic and fuel price in relation to the modelling tools.
A literature review throughout the US and Europe as well as a meta-analysis carried out in the French context allow a comparison with the results of the current study. The key-articles show a fall in road traffic and a rise in rail traffic resulting from an increase in price. Furthermore, they argue that these effects are more compounded in the long run.
We selected from the literature review the partial adjustment models (PAM) to represent traffic. Such models allow the determination of price elasticities in both the short and long term without requiring a large amount of data - as opposed to static models and error correction models. The national road traffic model is computed between 1990 and 2010 and includes explanatory variables such as the fuel price, GDP, the length of the motorway network and a lagged variable of traffic. The influence of competition between road and rail or air transports is disregarded as road transport remains by far predominant on a national scale according to the household surveys. According to the micro-economic theory, the influence of the variable 'fuel price' is expected to be negative as an increase in fuel price is likely to lead to a fall in consumer surplus, which implies a lower demand in transport. The price elasticity of demand is expected to be lower in the short term than in the long considering the time required to adjust behaviours.
Moreover, the study also focuses on the long distance trips made on the motorway network. Thus, the analysis of the elasticities becomes more accurate. Indeed, long distance trips are more likely to be made for leisure purposes and short distance trips for commuting. Then, the analysis of long distance trips allows a reflection on the difference in fuel price sensibility according to the purpose of travel.
This paper improves the knowledge about transport demand, particularly in the evaluation of fuel price elasticity of road demand in France. Furthermore, we process a rail demand model also based on partial adjustment models. As opposed to the case of roads, the rail demand is more likely to be subjected to road competition on a national scale. Consequently, the model is going to include the fuel price as an explanatory variable.
A quantitative estimation of the elasticities of demand allows a test of the national transport policies to be implemented, especially in the fields of fuel taxes and modal shift. Moreover, these estimates could be introduced in order to specify more accurately the traffic forecasting models. As a result, this study falls within the scope of a project promoting the sustainable development of territories.
Association for European Transport