The economic impact of transport infrastructure: a review of project-level vs. aggregate-level evidence

Abstract

This article aims to assess how well the economic impact of transport infrastructure is understood. Analyses of large samples of individual investment projects based on data in the so-called Flyvbjerg database have documented systematic construction cost overruns and first-year traffic demand shortfalls. They have interpreted these findings as being indicative of negative social welfare consequences of typical transport infrastructure investment projects, but that interpretation has been challenged on methodological grounds. The evidence base of ex post social cost–benefit analyses is growing and it, too, has challenged the suggestion that transport infrastructure projects destroy social welfare. Overall, it seems fair to conclude that our understanding of the project-level social welfare consequences of transport infrastructure is improving but remains far from conclusive. In contrast, aggregate-level quantitative meta-analyses provide robust results of a significant and positive relationship between transport infrastructure and economic activity at the level of regions or countries, especially in the long run and at higher levels of geographical aggregation. These aggregate-level results imply that project-level analyses should indeed consider the entire life cycle of projects, not just construction and the first year of operation, and that they should also acknowledge the presence of network effects as well as wider economic impacts – as difficult to measure and controversial as they are.