Addressing the “Big Dig Syndrome”

Transportation projects have credibility problem

The transportation infrastructure industry has a major credibility problem. Its track record on mega-projects is terrible. The costs are often grossly under-estimated, and traffic is all too often over-estimated. A poster child for the problem is Boston’s infamous Big Dig (Central Artery/Tunnel) project, originally pegged at $3 billion and now finally completed for nearly $15 billion. Many recent rail projects have similar, well-documented histories. It’s going to be harder and harder to get public and political support for much-needed mega-projects unless we can come up with better-performing delivery models.

That is the challenge taken up by Danish academic Bent Flyvbjerg and colleagues in Megaprojects and Risk (Cambridge University Press, 2003), which I highly recommend. They document the global nature of the problem, analyze its causes, and offer very useful ideas on doing better.

First, they cite studies showing that this is hardly a new problem, nor is it unique to a few countries. One of the most comprehensive studies (Aalburg University) covers 258 highway and rail projects ($90 billion worth) in 20 countries. Nearly all (90%) suffered cost overruns, with the average rail project costing 45% more than projected, the average highway project 20% more. Traffic forecasts were also far from accurate, with rail projects generating an average of 39% less traffic than forecast (though highway projects averaged a 9% under-estimate of traffic).

Flyvbjerg concludes that the “cost estimates used in public debates, media coverage, and decision-making for transport infrastructure are highly, systematically, and significantly deceptive. So are the cost-benefit analyses.” And most controversially, that this “cannot be explained by error and seem[s] to be best explained by strategic misrepresentation, namely lying.”

Those are harsh words, but many other analysts have reached similar conclusions. Flyvbjerg goes on to explain why this comes about. First, he cites two MIT researchers’ conclusion that “the incentives to produce optimistic estimates of viability are very strong and the disincentives weak.” And the reason for that is a lack of accountability of the parties involved, not a lack of technical skills or insufficient data.

Another key insight is that “risk is simply disregarded in feasibility studies . . . by assuming what the World Bank calls the EGAP principle: Everything Goes According to Plan.” But in mega-projects like the Big Dig, the Channel Tunnel, or the Los Angeles Red Line, things seldom go according to plan, and nobody should expect that they would.

Asking why risk is disregarded leads Flyvbjerg to question the conventional approach to project development, in which government is the project promoter and financier, and private firms are only too happy to do the best-case feasibility studies, produce the designs, and take on construction contracts fattened by numerous change orders. That is one form of “public-private partnership,” but it is not one that serves the public interest. It leads directly to the kinds of results documented in the Aalburg University study.

A much better delivery model is a public-private partnership that more appropriately “allocates risks to parties who have an incentive to reduce the negative impacts.” The current system puts the big-deal risks-of cost over-runs and of miserable traffic-onto the shoulders of hapless taxpayers. Not only is this inherently undesirable, but a system set up in this way “is likely to increase the total risks and costs of a project.” If somebody else is picking up the tab, you don’t have strong incentives to anticipate the kinds of things that will lead to problems and costly change orders.

Far better to put commercial-type risks, such as construction cost risk and traffic risk, on the shoulders of sophisticated investors. But to bring that about requires that there be true risk capital involved in a mega-project. Indeed, one of Flyvbjerg’s strongest conclusions is that the decision to proceed with such a project should be based on “the willingness of private financiers to participate in the project without a sovereign guarantee.” By putting their own capital at risk, such investors will be seriously involved in monitoring how the project is done, to mitigate the inherent risks.

One recommended model is what Europeans call the long-term concession or build-operate-transfer (BOT) model, under which the selected private consortium gains a long-term ownership interest in the project, sufficiently long that they have a reasonable likelihood of making a return on the investment (and also have strong incentives to build it right in the first place, to minimize life-cycle as opposed to up-front costs). But this is not a brief for the inherent superiority of the private sector. Flyvbjerg also supports a state-owned enterprise model, provided that it, too, is financed with at least one-third genuine risk capital from private investors. The point is to create accountability and risk-management, which the conventional government-led model simply does not provide.

The United States lags much of Europe and Australia in making use of true risk capital and long-term ownership models for transportation infrastructure. But unless we make the case for the superior performance of these models, and start to make them the norm for mega-projects, it’s going to be more and more difficult to persuade opinion leaders and policy makers that billion-dollar projects should be undertaken at all. Why saddle taxpayers with mega-risks? We need to make the case that mega-projects can solve real transportation problems. But we also need to shift their inherent risks to the shoulders of global investors, where they belong.

Robert W. Poole Jr. is director of transportation studies and founder of the Reason Foundation.