Are Transportation Cost Overruns Deliberate?
Congress should be worried about how the funds are spent, because the current process rewards efforts to exaggerate the benefits and understate the costs of transportation projects.
As Congress honed its recently announced deal on a long-term transportation bill, the debate centered on funding sources beyond the fuel tax. But going forward, Congress should be worried about how the funds are spent, because the current process rewards efforts to exaggerate the benefits and understate the costs of transportation projects. California's high-speed-rail project is a prime example: The cost estimate has more than doubled, to $68 billion, since voters approved funding for a share of the project.
These biased estimates result in the building of projects with questionable economic merit. Reform is needed.
To be sure, accurately estimating the cost of a new highway or rail line is technically demanding. It requires projections of labor and material costs, and analysts build in contingency funds for the unexpected problems that are bound to pop up. User-demand estimates require forecasting demographic trends and economic growth, along with the impact of thousands of business and residential location decisions.
These estimates are bound to be off the mark in one direction or the other — but it appears that proposals systematically underestimate costs and overestimate use.
Bent Flyvbjerg, Mette Skamris Holm, and Søren Buhl examined cost estimates for 258 large public transportation projects totaling $90 billion in countries around the world. They found the average cost overrun to be almost 28 percent. Rail projects were the worst, with an average cost overrun of nearly 45 percent. Other researchers have found estimates of project benefits, measured in terms of the number of people who will use the project, to also be consistently overly optimistic. This too is especially true for rail lines.
The fact that projections are consistently off the mark in the direction that makes the projects look good suggests that these errors may be deliberate attempts to cook the books. It is possible that elected officials pressure analysts to make projects look better than they really are for a number of reasons. Politicians look like they are doing something about transportation problems when they get behind transit proposals. Also, construction companies and their workers benefit from transportation spending, and in return these special interests provide government officials with political support.
These biases distort project selection. At any given time, there are many potential projects to fund. Limited transportation funds should go to the projects that provide the highest return. The record suggests this is not happening.
Before the president signs any long-term transportation bill, project-evaluation reform is needed. As a condition of receiving federal funds, a project's cost and benefit forecasts should be compared with the actual costs and benefits of similar completed projects. This would provide elected officials and taxpayers a means to determine whether the estimates seem reasonable. This process should be transparent, and the information should be made public.
Also, the estimates should be provided under a range of assumptions. What would happen to projects costs if inflation were to rise from 2 to 5 percent during construction? What would ridership on a new fixed-rail transit line be if economic-growth estimates were cut in half?
Finally, an analyst's salary at state and local transportation agencies should be tied to the accuracy of their estimates, and the track record of a transportation agency's estimates should be made public by posting results on its webpage.
While political pressures would remain, these reforms would provide stronger incentives to do good work. Now that lawmakers have agreed on the broad strokes of transportation funding, they must work to enact these improvements to the process.