The Wider Economic Impacts of High-speed Railways in France: A Demand-oriented Analysis
Koning, IFSTTAR - East Paris University, Olarte-Bacares, IFSTTAR- East Paris University
We use a Spatial Durbin Model to assess the wider economic impacts brought by high-speed railways. We test this model on a panel of 231 French cities observed over 1982-2010.
Various studies (Graham, 2009, Feddersen and Ahlfeldt, 2011) have recently proposed empirical methods to assess the wider economic impacts (measured in terms of per capita GDP or employment growth) brought by investments in high-speed railways (HSR). Next to travelers’ and environmental benefits, new HSR infrastructures may also stimulate local development by increasing “effective densities” or “potential accessibility”. Adding such economic impacts to standard Costs-Benefits Analyses is of prime interest for policy makers since new HSR projects nowadays present low Internal Rate of Returns, in France notably (Bonnafous, 2014) where the network has been largely expanded since 1982.
Whereas most research has focused on the “supply-side” effects brought by HSR, our approach is more “demand-side” oriented. Thus we believe that HSR can be seen as a medium that disconnects the geography of production and the geography of consumption (Davezies, 2008). Thanks to the new mobility services proposed by HSR, individuals can differentiate their places of jobs and residences. Also, HSR can stimulate tourism and recreational activities. With that respect, cities that do not present any strong productive advantages may also benefit from HSR’s connections: the people present on these territories thus arrive with their income and, through a “multiplier effect”, their consumptions would increase local demand (and level of activity) according to a classical Keynesian scheme.
This article tests such a conjecture on the French case. To do so, we modify at the margin the standard “economic base” modeling framework which stipulates that local activities depend on the strength of export sectors (Krikelas, 1992). Here, the level of jobs in a given city is a function of export-oriented activities as well as of neighboring incomes that are accessible thanks to HSR. Second, we test this model on a panel of 231 French urban units observed over 1982-2010. Our empirical strategy relies on a Spatial Durbin Model (SDM, Elhorst, 2011). As a consequence, our spatial weight matrixes use the information on travel times between pairs of cities connected by HSR, but also by standard rails, highways and roadways networks. The SDM enables us to estimate the marginal propensity to consume locally as a function of accessibility to other territories.
In addition to these benchmark estimates, we test whether the marginal propensities to consume locally are homogeneous across territories. For that purpose, we construct a “presentiality index” that describes the economic specialization of cities towards the presence of individuals (Terrier, 2009). This index (based on amenities, meteorological, socioeconomic data…) allows us to categorize our units of observation into three kinds of territories (presential, intermediary, productive cities). Finally, we try to address endogeneity issues linked to simultaneity and reverse causality between HSR provision and local development (Duranton and Turner, 2009). Thus we substitute contemporaneous travel time matrixes by historic information on travel times necessary to connect pairs of cities in the 19th century. These econometric estimates (in progress) could help us to predict the growth of jobs for French cities planned to be served by HSR in a next future.
Association for European Transport