What factors explain the allocation of motorways in Portugal?

Transport infrastructure is an expensive investment and is considered to be an important driver of economic growth. Portugal’s road network remained of poor quality until the last quarter of the 20th century. The country had less than 200 km of motorways before joining the EU in 1986, whereas in 2017 it had the fifth and third highest endowments of motorways relative to population and GDP, respectively, in the EU.

This study investigates the factors that influenced the allocation of motorways across municipalities in mainland Portugal between 1981 and 2011.

The analysis, based on Poisson Pseudo-maximum Likelihood models, suggests that population size and population accessibility (i.e., market potential) in 1981 are important determinants of motorway density in 2011. Likewise, physical and geographical variables also help explain the spatial distribution of motorways, as terrain ruggedness, distance to the coast, and distance to the border with Spain are negatively associated with motorway density.

The analysis also considers the influence of the proximity to historical and pre-existing transport networks on the allocation of motorways and finds that municipalities that are closer to the 1800’s (mainly dirt / non-paved) itineraries, the main roads of the 1945’s National Road Plan, and 1981’s train stations appear to have higher motorway densities in 2011; this effect, however, is concentrated in the vast and sparsely populated area of the country that excludes the high-density Portuguese “blue banana” (i.e., the coastal strip between the Metropolitan Area of Lisbon and the north of the country). Interestingly, it is also only in this low-density region that partisan alignment between the municipal and the national levels of government appears to affect the allocation of motorways. While more research on how this difference occurs is needed, a plausible interpretation is that motorways are more of a political asset in more remote or less urbanised areas.

Click here to go to the paper by Bruno T. Rocha, Nuno Afonso, Patrícia C. Melo, and João de Abreu e Silva.