How to Tell What’s Infrastructure and What Isn’t

With his proposed American Jobs Plan, President Joe Biden is aiming to preside over the greatest expansion of infrastructure in decades — not only the physical stuff but also the definition of the term itself. In doing so, however, he risks undermining public support for very programs he intends to promote.

The reason for the expansive definition seems clear enough: Americans across the political spectrum say that infrastructure is a wise use of government funds. By grouping more government projects under the banner of “infrastructure spending,” the Biden administration is hoping to increase public support for its plan.

But only a fraction of the administration’s proposal would be devoted to the type of shovel-ready public works that are typically denoted as infrastructure — roads, rails, airports, water systems and so forth. Instead, most of the money would go toward other forms of public-sector investment. That’s not the same thing, and the difference is more than semantic.

An almost-defining feature of modern capitalist economies is something called a spillover effect: the way one type of investment increases productivity and the demand for more investment. In essence, spillover effects make investments more attractive than they otherwise would be.

The construction of a new factory, for example, creates demand for homes for new workers. The dense clustering of homes fosters retail establishments, which draw in new residents both by employing more workers and by improving the local quality of life.

Infrastructure is investment that amplifies or increases spillover effects. It needs not be publicly funded. The canals and eventually railways that laced together England during the Industrial Revolution, for instance, were largely funded by aristocrats.

Nonetheless, financial markets sometimes find it difficult to raise the funds for expensive but clearly productive projects. It was obvious in the early 19th century, for example, that a canal connecting a port on the east coast of North America to the fertile soils of the Midwest would radically increase the productivity of both regions. Every attempt met with bankruptcy, however — until New York State provided the upfront funding for the Erie Canal.

The canal was so successful that the entire cost (funded by tolls, not taxpayers) was recouped in the first year. Such gains were possible because the canal unlocked the spillover effects of linking the East Coast and the Midwest.

Almost ever since, infrastructure has enjoyed a good reputation with voters.

The most popular categories of public infrastructure spending all have this similar feature: The returns from spillover effects are so large that costs can be recovered from user fees.

Other kinds of public infrastructure, such as mass transit, have a mixed reputation largely because their costs typically can’t be recouped in fees. But these kinds of investments have another kind of spillover effect: They spur publicly beneficial private-sector investment. Huge private-sector investments in central business districts, which are crucial to the prosperity of entire metropolitan regions, can be seen as spillover effects of spending on public transit.

How do the elements of Biden’s proposal stack up in terms of spillover effects? Repairing America’s aging roads and bridges clearly makes the grade, provided the funds are well allocated. Paying for this portion of the proposal with an increase in the gas tax would enhance accountability. Increased funding for scientific research has similar potential, though it would be nearly impossible to judge whether it’s been put to its best use.

Spending on rural broadband is harder to defend on classic infrastructural rationale. State and local governments are reluctant to subsidize such investments precisely because the enhanced spillover effects are mild. Such programs are better viewed as a kind of transfer spending to rural communities rather than as a meaningful expansion of infrastructure. The spending may well be justifiable — especially as high-speed internet becomes a basic necessity — but the argument is more about ensuring that children have access to the modern economy no matter where they live.

The “Everything Is Infrastructure” argument is weakest when it comes to spending on care-giving. These aren’t one-time projects, but ongoing spending on things such as child and elder care. The administration’s defenders argue that the spending will nonetheless enhance the productivity of family members who had been forced to quit work in order to provide that care.

This misses a painful but crucial point. Suppose, instead of providing more services, the government simply gave an equivalent amount in cash to those families struggling with care-giving costs.

Some families would choose to use those funds to help defray the costs of private care, perhaps allowing a family member to work more outside the home. Many, however, would still find it cost-effective to keep one member at home. For these families, getting government money is preferable to getting a government service.

Government spending on care not only fails to provide spillover benefits, but it’s not even worth its cost to every family who receives it. This kind of program is better described not as infrastructure but as a type of transfer spending — from the well off to the less well off.

Again, it may be that this is a transfer worth making. But it’s a harder sell, because some voters are skeptical of the efficiency of such programs while others are flat-out opposed them. The Biden administration is clearly hoping that it can make some of these programs more popular by including them as part of something it calls an “infrastructure package.” But it could just as easily work the other way around — making infrastructure spending less popular.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Karl W. Smith is a Bloomberg Opinion columnist. He was formerly vice president for federal policy at the Tax Foundation and assistant professor of economics at the University of North Carolina. He is also co-founder of the economics blog Modeled Behavior.

©2021 Bloomberg L.P.

BQ Install

Bloomberg Quint

Add BloombergQuint App to Home screen.