With the ongoing national focus on infrastructure, one of the best papers I’ve read in several years is “Economic Perspectives on Infrastructure Investment” by Harvard’s Edward Glaeser and Massachusetts Institute of Technology’s James Poterba. The economists’ report was prepared for the Aspen Economic Strategy Group and released on July 14, 2021. What distinguishes the paper from so much that is being written about the need for increased infrastructure investment in the United States is its comparison of how engineers frame the problem and how economists do so. (And I write this as an MIT-trained engineer who has spent much of his career trying to explain economic reasoning to transportation engineers.)
Glaeser and Poterba explain in some detail that infrastructure “needs” as defined by groups like the American Society of Civil Engineers and a series of reports by McKinsey are lacking in assessing the benefit/cost ratio of proposed investments. Instead, these reports on infrastructure either compile what amount to engineering wish lists or create ratios of infrastructure spending as a fraction of gross domestic product (GDP) which tell us little about whether these are economically worthwhile projects to invest in. They also point out the potential conflict of interest when engineering consulting firms do the preliminary work estimating a project’s costs and usage—and then gain lucrative contracts to manage the actual projects. This system often results in huge cost overruns (e.g., the Big Dig in Boston and California’s high-speed rail project approved by voters in 2008).
From Glaeser and Poterba’s economic analysis perspective, they argue that, at least, all large infrastructure projects should be vetted via a rigorous benefit/cost analysis (BCA) to decide if they are worth being funded or financed. They suggest several possible ways to require this, including having Congress mandate that federally-funded highway and transit projects must pass a benefit/cost analysis screening in order to proceed; requiring state departments of transportations (DOTs) and metropolitan planning organizations (MPOs) to do that for every project, or shifting large projects to a national infrastructure bank subject to a benefit/cost analysis mandate. The problem with each of these is politics. Many elected officials—federal, state, and local—love to select infrastructure projects based largely on their political benefits rather than their benefits exceeding their costs.
Another long-standing problem in U.S. infrastructure is deferred maintenance. And, again, the problem is mostly political. Elected legislators tend to prioritize new projects with ribbon-cutting opportunities over adequate ongoing maintenance, which may be barely visible to voters. The result is that roads and bridges wear out prematurely, need repeated repaving and other unplanned maintenance, and often need to be rebuilt sooner than they would have otherwise. This is penny-wise and pound-foolish. Consequently, and unfortunately, despite lots of talk about “fix it first” provisions in federal or state infrastructure legislation, this hardly ever seems to happen.
The authors also discuss the many cost-increasing provisions of federal law that raise U.S. transportation infrastructure costs above those of other developed countries. These include “Buy American” provisions and other protectionist measures that increase material costs, Davis-Bacon “prevailing wage” laws and other regulations that increase labor costs, and the well-known practice under design-bid-build contracting of bidding low to get the contract and making it up with numerous cost-increasing change orders.
Another key point on which they elaborate is that the United States should make greater use of pricing and user fees for major transportation projects. This is not only a better way to finance most large projects, but if the pricing is allowed to vary in proportion to demand, it can produce greater user benefits and, in some cases, reduce the total number of lanes needed, which would be an important cost-saving.
Glaeser and Poterba discuss transportation public-private partnerships (P3s) as a useful alternative to business as usual, but they don’t fully explain how valuable the design/build/finance/operate/maintain approach can be in addressing the seemingly intractable politicization of U.S. transportation funding and management that they discuss.
Here is how P3 projects procured as design/build/finance/operate/maintain (DBFOM) address key problems:
Deferred maintenance: Under DBFOM the project is designed to minimize not the initial construction cost but the life-cycle cost because long-term, ongoing maintenance is built into the long-term concession agreement. This amounts to ensuring better long-term stewardship of the expensive and valuable asset.
Benefit/cost analysis: Especially under the revenue-risk version of DBFOM, the project cannot be financed unless the projected revenues are enough to service the bonds issued to finance the project and produce a return on the developer/operator’s equity investment. For an availability payment DBFOM that does not include toll revenue, the state DOT’s willingness to commit to a 35-year stream of payments so the project can be financed means the DOT must complete a benefit/cost analysis to ascertain that the project makes financial sense.
Willingness to charge users: In a number of cases, a megaproject (such as the original express toll lanes on the I-495 Beltway in northern Virginia) was developed based on variable tolling as the means of financing, while the state DOT had deferred adding new lanes due to lack of gas-tax funds. This was an early example of what some have called “outsourcing of political will” to implement variable pricing that the state itself was unwilling to do.
Cost overruns: DBFOM projects are not immune to cost overruns, but the key question is who bears the risk of this happening. To my dismay, the authors dismiss this by writing, “large public entities are typically better able to bear risks than most private firms.” But the large public entity does not bear those cost overruns: taxpayers do, whereas in DBFOM projects the investors willingly take on megaproject cost overrun risks. This premise has been demonstrated by the fact of several bankruptcy filings of toll-financed megaprojects in both Australia and the United States, not mentioned by the authors.
In their summary of the case for (limited) use of long-term P3s, Glaeser and Poterba cite as potential benefits reduced costs, specialized expertise, and improved quality, for all of which there is evidence. But they also state that the case for such public-private partnerships depends on “their being able to borrow at better rates than a city government or bear risk better.”
Thanks to the Transportation Infrastructure Finance and Innovation Act (TIFIA) and tax-exempt private-activity bonds, borrowing cost differences are minimal, but the report’s authors don’t mention those important tools, and as noted above, they devote little attention to the very real risk transfers that are possible (and should be identified in competent value-for-money analyses).
Despite minor shortcomings on P3s, Glaeser and Poterba’s overall economic assessment of how poorly the United States addresses major infrastructure is excellent and much-needed. In this brief review, I have only scratched the surface of the valuable insights and examples set forth in their paper.
A version of this column first appeared in Public Works Financing.