FLYVBJERG

Our survey looked at several hundred projects in more than 20 countries and we discovered that there seemed to be a formula at work in deciding which projects get built: Project approval equals underestimated cost, plus overestimated revenue, plus undervalued environmental impact, plus overvalued economic development effect.

But during project implementation, when fact defeats fiction, the consequences of this formula are huge cost overruns, delays, revenue that doesn’t materialize, crippling debt and often endless litigation to determine who is responsible for the mess. The misinvestments, overspending and financial problems are of Enron-scandal magnitude, though less transparent and harder to fix.

U.S. examples include Boston’s Central Artery/Tunnel, Los Angeles’ subway and Denver’s international airport. In Britain, there’s the Channel tunnel, the Humber Bridge and the Tyne metro system. And the problem is not just limited to transportation infrastructure. Viewing the rebuilding of Iraq as a megaproject, we find the distinctive pattern of misrepresentation, cost overrun and benefit shortfall.

We tested our data against four possible explanations for this misrepresentation–technical, economic, psychological and political–and found that political explanations best fit the data. This has been confirmed in interviews with forecasters. Project promoters and forecasters appear to deliberately "cook" forecasts. "[The] minister leans on his advisors, the advisors lean on the feasibility consultants and the forecasts get ‘talked up,’" says one traffic forecaster.

Our data show that technical problems, inadequate models, over-optimism and insufficient level of detail in initial forecasts aren’t plausible excuses. For the past 70 years, cost overruns have stayed high and largely constant, with nine out of 10 projects experiencing overruns. No one is learning from past mistakes or, more likely, no one wants to learn.

ENDS AND MEANS

Undoubtedly, most project proponents believe their projects will benefit society and that they are thus justified in cooking costs and benefits to get projects built. The ends justify the means, or so the players reason.

The entire structure of incentives is geared toward underestimating costs and overestimating benefits. When a project goes forward, a lot of people profit–engineers, contractors, bankers, landowners and developers. In addition, politicians with a "monument" complex gain satisfaction, administrators receive larger budgets and cities gain from investments and infrastructure that would otherwise go elsewhere.

Some argue that no projects, including our most treasured, would ever be undertaken without misrepresentation. An icon like Sydney’s opera house, with a cost overrun of 1,400%, is often cited.

The argument is easy to refute. Quite simply, many viable projects exist that don’t require misrepresentation. The record shows that architectural treasures like the Bilbao Guggenheim, with highly complex and innovative engineering, can be built on time and budget.

The problem is not that good projects don’t exist. The problem is that the Machiavellian make-believe world of misrepresentation makes it very difficult to decide which projects deserve undertaking. The result is that too many projects proceed that should not, and many projects don’t proceed that probably should because they lose out to projects with better misrepresentation.

One must be skeptical of those involved in promoting and building large public works projects. Their forecasts cannot be trusted and should be examined carefully by independent specialists. Institutional checks and balances also should be developed, but it would be naive to think the misrepresentation can be easily stopped.

Bent Flyvbjerg is principal author of Megaprojects and Risk: An Anatomy of Ambition (Cambridge Univ. Press, 2003), and professor of planning at Aalborg University, Denmark.