Integrating network analysis and interregional trade to study the spatial impact of transport infrastructure using production functions

Résumé

The production function approach is used to analyze the role of transport infrastructure on GDP. Different regression techniques are employed to estimate its impact on production and in combination with other inputs, particularly employment and private capital stock. One problem with the traditional approach is that it incorporates as explanatory variable the public capital stock of transport infrastructure as a whole, i.e., the existing gross capital stock as calculated by statistical offices, which does not take into account the real use that economic agents make of the available transport networks (network effects). To overcome this drawback we introduce a new concept of transport infrastructure capital stock reflecting the real benefit obtained by the regions when accessing markets by using their own infrastructure (internal stock), as well as those of neighboring and non-neighboring regions (imported stock). While the former contributes directly to the production of a region, the latter can be associated to a spillover effect from importing transport infrastructure when exporting goods. Both, the internal and imported infrastructure stocks are calculated by network analysis performed in a GIS, which enables us to account exclusively for the public capital stock actually used in trading flows. Production functions incorporating the internal and imported capital stocks are estimated using panel data econometric techniques . The magnitude and significance of these variables on regional as well as provincial GDP is determined and discussed for the Spanish case and the 1980-2007 period