Transportation infrastructure and local growth: Evidence from Italy
Transportation infrastructure and local growth: Historical evidence from the South of Italy
Large transportation infrastructure projects are considered a promising investment to spur economic growth in lagging areas by many policymakers. This column presents historical evidence that questions this assumption. It studies the most important Italian infrastructure project in the aftermath of WWII: the 440km freeway connecting the Southern regions of Italy. It finds, that while the freeway caused a significant reorganization of both economic activity and population from places far from the freeway to locations close to it, there is no evidence that it had any long-run effect on economic growth of the Southern region as a whole.
While the policy debate at the EU level focuses on setting up the Recovery Fund, national Governments are increasingly engaged in planning how to spend the EU money. In Italy, the Government organized a seven-day consultation with relevant stakeholders (Stati Generali) to discuss how to use EU funds to tackle the long-standing problems of the economy. A widely popular proposal is to expand the national transportation network through new large-scale projects, such as a bridge connecting Sicily with the rest of the peninsula. Ostensibly, the idea seems to fit the needs of the South of Italy, which lags behind the rest of the country and is far from the main economic hubs.
Does transportation infrastructure cause growth?
This proposal rests on the assumption that investing in transportation infrastructure spurs local economic growth. This belief is strong enough to bring together coalitions broad enough to support large and expensive projects in underdeveloped areas. It is not hard to persuade people that the transportation – growth nexus is right: the world is full of anecdotal evidence about rich regions that are highly connected with the rest of the world. However, as pointed out by Duranton and Venables (2018 p.16), “it does not follow from these observations that all well-connected places are rich or that improving connectivity necessarily brings development.” In fact, the more rigorous empirical evidence suggests prudence. A recent NBER working paper by Duranton et al. (2020) reviews the literature assessing the economic effects of transportation investment on economic activity. They conclude that the literature finds strong evidence for the influence of transportation infrastructure on the geographical distribution of economic activity. However, there is “little compelling evidence about transportation infrastructure creating economic growth” (Duranton et al. 2020, p.2).
The Italian A3 freeway
The use of large-scale transportation investments to push underdeveloped areas is not a new issue in policymakers’ agenda. The Italian A3 freeway is a leading example. Planned in the late 1950s, during the years of the ‘economic miracle’, it aimed at connecting one of the least developed regions in the South (Calabria) to the thriving areas of the peninsula. The plaque installed at the start of the freeway made this aim clear: “After XXI centuries the road that Rome opened to unite the people of the South reopened today on the ancient footsteps from Salerno to Reggio Calabria to continue and complete the great route of traffic and work between the North and the South of Italy.”
The prospect of boosting local development did not go unnoticed among politicians from Calabrian constituencies, and quickly turned into a political economy diatribe. The initial feasibility plan proposed three possible routes: along the Tyrrhenian coast, along the Ionic coast, or through the centre of the region (see Figure 1). The median lane was chosen so that the freeway would pass the town of Cosenza (D’Antone 2008) which was the hometown and constituency of two very powerful politicians: the socialist Giacomo Mancini and the Christian democratic Riccardo Misasi. Mancini, in particular, who was a former MP and regional secretary of the Socialist party, was a prominent member of the national administration. The freeway was mostly built between 1962 and 1974, although part of it was completed only later on. The final freeway featured frequent interruptions and roadworks, which might have diminished its role in bringing connectivity.
Figure 1 Map of the A3 freeway
Note: The black line is the actual freeway, while the blue and red lines are the alternative routes that were considered.
In our paper (Ciani et al. 2020), we study the long-term economic effects (over four decades) of the freeway on local development. We use the three planned routes to identify treated and control units. As the Calabrian terrain is extremely rugged and includes mountainous sites, we restrict control units to those with a similar degree of accessibility as the treated ones. This builds on the logic that those remote areas would have been excluded from the transition to a modern economy between 1951 and 2001 anyway. Although the selection of the final route was far from random, we exploit the fact that it was driven by the political desire to make it pass through the town of Cosenza. Hence, following an ‘inconsequential units’ approach (Redding and Turner 2015), we exclude the local labor market of Cosenza, and compare municipalities that ended up being close to the freeway (only because they were on the route passing through Cosenza) with other municipalities along the alternative routes. In a difference-in-difference framework, we contrast employment and population growth in municipalities along the freeway with the growth in municipalities on the discarded paths.
Re-organization versus growth
We find that the freeway led to stronger growth in the municipalities closer to it. The difference in the long-run is sizable. Moving from being close to a freeway entrance (the first percentile of distance, 2 km) to further away (at the 75th percentiles, 60 km) leads to a 17 percentage point difference in private employment between 1961 and 2001. The effect on the number of jobs is stronger than the effect on population size (8 percentage points), hence the proximity to the freeway increases the employment rate. The differential growth started in the tradable sector and spread to non-tradable employment later in time. These results are robust to a series of sensitivity checks including the modernization of the overall road network and the presence of criminal organizations across municipalities.
These divergent trends might point to a positive growth effect of the freeway itself but they could also be the result of the relocation of economic activity from more distant locations to closer to the freeway. One way to understand whether this is the case is to exploit a third area, unaffected by the construction of the freeway. By using data from municipalities located in Southern regions without highways (Sardinia or the Southern part of Apulia) as comparison units we fail to find evidence that the freeway helped Calabria to converge towards more developed areas of Italy. Those municipalities closer to the A3 did better than the other Calabrian municipalities, but still worse than the ‘third area’ counterparts, suggesting that most of the effects are driven by the relocation of economic activity within the region. We also experimented with a synthetic control approach on macro-level data (regional level data) and again fail to find any positive impact of the transportation infrastructure on regional growth as a whole.
Among policymakers, investing in transportation infrastructure is considered a promising way to revitalize growth. The consensus is stronger in countries, like Italy, where lagging regions are distant from the centres of economic activity and alternative policies to stimulate catching up resulted in massive failures (Accetturo and de Blasio 2019). Rigorous empirical evidence, however, suggests that transport investments might not be a magic wand for economic growth. The case of the freeway Salerno-Reggio Calabria illustrates that new transportation infrastructure can produce relevant re-allocation effects across municipalities, without boosting regional economic development. In this case, benefits and losses associated with the infrastructure mostly compensated within the region, without prompting additional growth.
Authors’ note: The views expressed in this column are those of the authors only. They do not necessarily represent the views of the Institutions they are affiliated to.