Photo by Diego Jimenez on Unsplash
 

When the government is paying for something, it should try to spend taxpayer money as effectively as possible. But when it comes to building and fixing roads, the US is almost surely not getting as much as it pays for. Zachary Liscow, Will Nober and Cailin Slattery offer some evidence in “Procurement and Infrastructure Costs,” presented at the 13th annual Municipal Finance Conference at the Brookings Institution (July 11, 2024).

The authors note: “The United States spends a large amount on infrastructure costs: state and local governments spent $266 billion on highways alone in 2022. The spending, on a per-project basis, is very high by international standards—over three times as high as other upper- and middle-income countries.”

Why might this be? The authors gather data across all 50 states by surveying those who work at state-level Departments of Transportation (DOTs) as well as the companies that bid to build roads. They create a detailed dataset of project-level costs. They discover some situations in which the public is paying a lot more for roads and highways than others. One factor is that managing the bidding process for infrastructure projects, and overseeing the project as it happens, will work better if the state DOT is of high quality and has sufficient in-house workers.

In the survey, there is broad agreement that state DOTs have become more understaffed and that reliance on consultants drives up costs. Survey respondents attribute a lack of details in project plans to both a lack of time or experience of DOT engineers and the use of consultants. When there is not enough specificity in the plans the risk to the contractor increases, increasing bids. Moreover, whenever the scope of a project changes this initiates a costly and time-consuming renegotiation process. Survey respondents agree that such changes are a major contributor to costs. We confirm that the state DOT workforce has been shrinking with administrative data on public sector employment.

Their survey data allows them to look at what states are widely considered to have higher- or lower-quality Departments of Transportation, and about the use of consultants. They find:

States that flag concerns about consultant costs have higher costs—a one standard deviation increase in reported consultant costs is associated with an almost 20% increase ($70,000) in cost per lane-mile. States where contractors and procurement officials expect more change orders have significantly higher costs: one additional change order correlates with $25,000 in additional cost per lane-mile at the mean. … More directly, we find that states with (perceived) higher quality DOT employees have lower costs. A state with “neither low nor high quality” employees has almost 30% higher costs per mile than one that rates the DOT employees as “moderately high quality”, all else equal. … A one standard deviation increase in DOT employment per capita is correlated with 16% lower costs.

They also find a problem with lack of bids for projects in many states: “A lack of competition for contracts is an oft-cited cost driver from procurement officials. … We show, with external data on the highway construction industry, that concentration in the industry seems to be rising. Most states have experienced a loss of construction firms, and an increase in size of the remaining firms, in the last 10 years.”

In the survey data, we find that states that do outreach to increase the bidder pool have significantly lower costs, highlighting both the importance of competition and the role the DOT can play in order to increase competition. A one standard deviation (12 percentage point) increase in bidder outreach is correlated with a 17.6% decrease in costs. At the mean, this translates to a decrease in costs of $65,000 per lane-mile and $1 million at the project level. We also find that limits on the amount of work that can be subcontracted is positively correlated with costs. Restrictions on subcontracting can decrease competition by limiting the set of potential prime contractors that can complete the project. Lastly, using our project-level cost data, we find that an additional bidder on a project is associated with 8.3% lower costs, or a savings of approximately $30,000 per lane-mile ($460,000 for the average project).

One of my personal frustrations with how legislation is often discussed arises when there is a heavy focus on the total amount spent, which is easy to measure, and much less focus on what is received for what is spent, which is harder to measure. But the intention (level of spending) is not the outcome (actual results). The estimates in this paper suggest that a number of states are overspending by maybe one-third, or perhaps even more, for the roads and highways they receive.

More By This Author: