Eakinomics: Infrastructure as a Response to COVID-19

Infrastructure seems to be having another one of its moments – if not another Infrastructure Week! As Congress contemplates further policy responses to the economic fallout of the COVID-19 pandemic, infrastructure is increasingly mentioned as among the candidates. Having seen the disaster of “shovel ready” projects in the Obama-era American Recovery and Reinvestment Act (ARRA) and the undisciplined (and failed) efforts in the Trump years, it might be a useful time for a review of key issues.

To begin, what is infrastructure? How can one decide what should, and what should not, be under consideration for infrastructure spending? I think the key is to take a disciplined approach.

First, it should be something that cannot, or will not, be provided by the private sector alone. There is infrastructure that is valuable to production that has some combination of two characteristics: (1) it is jointly consumed; (2) it is not possible to exclude people or firms from using it. (Technically, these kinds of goods have the label “public goods.” See Chapter 4 of Rosen and Gayer.) One truck driving on a highway does not stop another from doing so as well; both “consume” the highway at the same time without diminishing the other. (Notice the same is not true of the diesel fuel they use; that is a private good.) The same is true of airports and the air traffic control system. Or the broadband backbone. Non-excludable means you can’t stop someone from using the infrastructure, making it quite difficult to charge for its use. Neither characteristic is absolute, and they change with technology and market conditions. But the more something is shared in use and difficult to charge for, the better candidate it is for genuinely being public infrastructure.

Second, it should be tied as directly as possible to better economic performance. The prime imperative is to improve efficiency and reduce the costs of the national supply chain. The economy has already weathered severe supply shocks, and one can anticipate more in the pandemic’s future. And responding to the pandemic will raise the costs of doing business. It is important to use policy to offset these changes to the extent possible. Better connectivity across ports, railroads, airports, trucking, and other parts of the economy will improve efficiency. Those impacts are direct and immediate; they are the best candidates for infrastructure spending.

A third criteria is whether they are national in scope. The federal government solves national problems; it should be first and foremost focused on improving national connectivity in transportation and the like. Projects that are regional or metropolitan in scope should have lesser priority and be the responsibility of states and localities.

The basic moral is simple: In the absence of a disciplined test for what constitutes infrastructure, the definition will become indefensibly elastic and the outcome determined by raw politics alone. That won’t support a better recovery.

The second issue is to be clear about what infrastructure spending will and will not accomplish. The biggest sin is to use the s-word. Infrastructure should not be about “stimulus.” The most important aspect of projected recoveries for the U.S. economy is not the growth rate in the latter half of 2020. The most important feature is that even those forecasts that anticipate rapid growth in the 3^rd and 4^th quarters still have the level of economic activity at the end of 2021 below that at the beginning of 2020.

Infrastructure projects are long-lived. They provide long-lived benefits that have nothing to do with the near-term pace of growth and should be planned and implemented independent of it as well. As it turns out, few projects are genuinely “shovel-ready,” and the history of using public-works spending as a tool of “stimulus” is replete with failure.

As noted earlier, not every problem is a federal problem. If your water mains are breaking, potholes expanding, and local congestion rising, the first step is to change the local management. It is a core function of local, county, state, and regional authorities to plan, construct, and maintain their infrastructure.

The genuine federal role in infrastructure is quite limited. Indeed, an important step would be to identify the uniquely federal role — national connectivity of economic activity, as mentioned above — and have the federal government pursue that and only that. It might also be involved in financing other infrastructure projects whose benefits are regional or local, but in doing so it should demand serious analysis that these projects generate quality results.

Infrastructure can contribute to a policy mix that responds effectively to the pandemic recession. But it is not a silver bullet and will only be successful if it is undertaken in a disciplined fashion.