Conventional wisdom in the U.S. transportation infrastructure community is that our infrastructure is “crumbling,” and that China shows the way forward by “spending more on infrastructure each year than North America and Western Europe combined.” This has also been a theme of economist and New York Times columnist Paul Krugman in recent years. The comparison is very misleading, since China is an economy still emerging from underdevelopment, not an advanced western nation that already has extensive infrastructure. But there is also the question of whether China is making sound investments.
On the latter point, I was not completely surprised to read the following in a letter to the editor of The Economist (September 3, 2016):
“Over half of the investments in transport infrastructure in China are of such low quality that they destroy economic value instead of generating it—the costs of that spending are larger than the benefits they generate. Unless China shifts to fewer and higher-quality infrastructure investments, the country is heading for an infrastructure-led national financial and economic crisis.”
The letter writer was Bent Flyvbjerg, professor of economics at Oxford’s Said Business School and lead author of the classic volume, Megaprojects and Risk (Cambridge University Press, 2003).
His conclusions about China are derived from a major research project, summarized in “Does Infrastructure Investment Lead to Economic Growth or Economic Fragility? Evidence from China,” by Atif Ansar, Bent Flyvbjerg, Alexander Budzier and Daniel Lunn. I downloaded and read the paper last week, and found it a very well-done piece of empirical research.
Flyvbjerg and his team obtained data on 95 completed transport infrastructure projects in China. They are among the best-documented Chinese projects, since they received some of their financing from either the Asian Development Bank or the World Bank. Among the team’s findings from this dataset are the following:
- 75% of these road and rail projects had cost overruns;
- Actual costs averaged 31% higher than budget, but many were much worse;
- Road projects were generally completed on schedule, but rail projects on average took 25% longer than planned;
- 65% of the projects had “benefit shortfalls,” primarily lower traffic and revenue than forecast, averaging 41%, and some were much worse;
- The remaining one-third were overloaded, by an average of 61%.
The approved plans that authorized the projects to go forward (including ADB or WB financing) included benefit/cost ratios of 1.4 to 1.5—meaning “planners expected the net present benefits to exceed net present costs by about 40-50 percent.” Using data from the 95 completed projects, Flyvbjerg’s team recomputed their actual B/C ratios, finding that 55% of the projects had an actual ratio of less than 1.0. And, “a majority of these value-destroying projects suffered the double whammy of a cost overrun and a benefit shortfall.”
Well, you might think, at least China’s economy will be improved by all this new infrastructure, even if some of the money was wasted. The latter part of the paper addresses the macroeconomic consequences of these poor outcomes. First, the authors cite other work demonstrating that China’s biggest economic growth came prior to most of this infrastructure development. Bannister and Berechman found that “The ‘China miracle’ happened not because it had glittering skyscrapers and modern highways but because bold economic liberalization and institutional reforms . . . created competition and nurtured private entrepreneurship.”
Flyvbjerg and colleagues conclude that massive over-spending and poor outcomes in infrastructure have led to “an accumulation of a destabilizing pile of debt in the economy; unprecedented monetary expansion . . . and subsequent economic fragility.” And also, “perhaps even more damaging, are the opportunities foregone to build the right infrastructure.”
I’ve only skimmed the surface of this very important paper. Everyone interested in sound U.S. infrastructure policy should read it and ponder its findings.
(This article first ran in the October 2016 issue of Surface Transportation Innovations)