As we noted on the project overview page, Policymakers, think-tank academics, academics, and journalists are increasingly interested in explaining the following pattern: large infrastructure projects in English-speaking countries are generally more expensive per kilometre than are equivalent projects in the advanced industrial democracies of continental Europe and East Asia. Policymakers, journalists, and academics in the United States have observed that the cost of building of mile of subway in an American city is significantly greater than the cost of building an equivalent mile of subway in a European city. Researchers have also noted that the unit costs of building such projects tends to be higher in “Anglosphere” that is predominantly English-speaking countries than in other advanced economies. Researchers have shown that unit construction costs are generally much higher in English-speaking countries for reasons that remain murky despite various academic studies. Resolving this intellectual puzzle is very important in light of the need to provide low-carbon alternatives to car transport and the controversies about over budget projects such as the Queen Elizabeth Line in London and the infamously costly Second Avenue subway.
Our project uses historical data to illuminate this puzzle: we compare unit construction costs for underground projects from 1863 to the present to determine when and why the aforementioned discrepancies in unit construction costs first emerged. Our research is designed to be useful to policymakers who want to build high quality low carbon transport cheaply. We are bringing some historical data into the conversation about this problem.
Other researchers, including those associated with New York University’s Marron Institute, have established that the cost to build a mile of underground rail infrastructure is much higher in English-speaking countries than in other advanced economies. The journalists who have recently written about this question have contrasted the recently completed Copenhagen underground, which was relatively inexpensive, with the far higher costs of the Crossrail project and the infamously delayed Second Avenue Subway project in New York (Vox, 2021). When the construction costs per mile of large numbers of projects are compared, one find that even after we control for geological and other factors, average costs per mile are higher in English-speaking countries than in other democracies that are at equivalent levels of economic development. The United States is notorious for bloated infrastructure costs and delays, although the situation in the United Kingdom is broadly similar.
As a 2010 report by Her Majesty’s Treasury observed, infrastructure costs in the UK are an outlier in Europe. That report also concluded that most of the differential in costs related to legal and political proceedings before construction actually begins. Recent research by scholars at the NYU Marron Institute of Urban Management suggests that the cost of building metro stations in English-speaking countries is not the culprit and it is the greater costs of producing a mile of tunnel that is the main reason for the observed pattern, namely that it costs more to build public transport infrastructure in English-speaking countries. Eric Goldwyn and Alon Levy, the researchers associated with the Marron Institute’s study of comparative transit costs, who are the world’s leading experts on this puzzle, freely admit that they do not yet have a complete explanation for this pattern (Transit Costs Project, 2021). The Marron Institute researchers are great scholars and we are enormously indebted to their prior work. In fact, the spreadsheet with historical subway project costs we created is modelled on the one for modern subways that they created. We have absolutely nothing against the smart guys over the Marron Institute, but by their own admission they can’t really explain this pattern. At least not just by using the relatively recent data in their spreadsheet. That’s where our historical research skills come into play
The existing research literature, does however, make clear is that many of the explanations typically offered by newspaper columnists are not workable. For instance, it is untenable to blame the existence of unions for the costs overruns in US and UK infrastructure projects when the workforces on French and Danish projects are also unionised. In fact, French unions are notoriously militant. What we have is an unresolved puzzle: for some reason, the more like the US the culture and legal system a country it is, the more expensive it is to build underground public transport. Explaining this pattern is obviously important in light of climate change and the importance of good and cheap infrastructure. At a recent webinar hosted by Matt Yglesias, the journalist who runs the (excellent and nerdy) Slow Boring substack, the participants made it clear that this important puzzle remains unsolved.
That’s where we as economic-historical researchers come in! To date, nobody has attempted to do historical research aimed at resolving this question. The admirable cost per mile database constructed by the researchers at NYU only includes relatively recent projects. For instance, the oldest UK project listed in it is the Jubilee Line, which was begun in 1993. That gap in their data is a problem, because the world’s first underground passenger rail line, which was built in London in the 1860s and which is now called the Circle Line. By looking at construction costs from 1863 onwards have evolved in different countries, we are be able to shed light on the historical reasons for the pattern observed above: that it costs much more to build subway lines in the US and other English-speaking countries than in continental Europe.
The Past Speaks 