Why Do We Need Infrastructure Policy?

My title at IFP is "Senior Infrastructure Fellow," which means that in addition to publishing this newsletter, I do work related to infrastructure policy. At a high level, that work is figuring out which things we should be building, and how we can make building those things easier. But it’s worth diving a little deeper, to understand what physical infrastructure is, and why it's worth having a policy arm devoted to it.

Broadly, infrastructure falls into one of four categories:

• Energy — Structures that manipulate energy, changing its form and moving it to where it's needed. This covers obvious things like power plants, transmission lines, and other electrical infrastructure, but also things like oil and gas pipelines.

• Transportation — Structures that enable the movement of people and goods. Roads, bridges, railroads, mass transit lines, and so on.

• Water — Structures that control the movement of water. This includes things like public drinking water (water mains, aqueducts, water treatment plants), flood control (dams, dikes, levees, retention ponds), irrigation, and sewage.

• Information — Structures that enable communication and/or the movement of data. Telephone lines, fiber optic and undersea cables, and so on.

More generally, infrastructure is largely physical structures and works that make it possible to move things around, whether that's atoms, energy, or bits.

But why is infrastructure policy important?

One of the government's most important functions is to solve coordination problems and address market failures — to prevent undesirable outcomes that occur from markets acting alone, and to enable mutually beneficial transactions that in the absence of some coordination mechanism might not exist. For instance, consider contracts, and what they might look like in the absence of some authority that will enforce them. If I sign a contract with a builder to build a house for me, and give them an upfront payment so they can buy materials and equipment, they might be incentivized to simply run off with my payment rather than doing the work — after all, they have my money right now. And if I know this, I will be reluctant to hand over any money to the builder. The result is that I don’t get my house built, and the builder doesn’t get any work — everyone loses! A government that enforces contract provisions between parties and penalizes breaking them makes this sort of thing more costly, which makes it much less risky for me to hand over my up-front payment. By making it costly to break contracts, people will be much more likely to sign them, engaging in mutually beneficial transactions that otherwise might not occur.

Different sectors of the economy can benefit from different levels of this sort of government intervention. In some (perhaps most) cases, government intervention can be relatively minor, or may not be needed at all beyond enforcement of basic rule of law. But the nature of infrastructure makes it extremely susceptible to a variety of market and coordination failures:

• Public goods — Many infrastructure projects have "public good" elements, in that they are non-rivalrous (they don't get used up) and non-excludable (you can't prevent people from using them). Flood control structures, for instance, such as levees and dams, fall into this category. A levee doesn't get "used up" the more people it protects, and it's impractical to prevent people from getting the benefits of it.

  Road networks and mass transit are not strictly public goods (since they can be both rivalrous and excludable), but they also provide significant "spillover benefits" to people other than those who use them, which can't be easily charged for. As Alain Bertraud notes in “Order Without Design,” "...once a road network has been built, it is impractical to allocate and to recover its cost from beneficiaries, since not only roads users but also landowners benefit from better accessibility and hence increased land values.”

  In both cases, because this sort of infrastructure creates significant benefits that can't easily be charged for, markets and private actors will under-provide it. As Bertraud notes, “Government has to substitute design for markets to ensure an adequate supply of all public goods, including roads. An adequate supply of urban roads is particularly important, as roads provide the indispensable mobility that allows labor markets to function and cities to exist.”

• Monopolies — Infrastructure projects also often have monopoly elements, or other restrictions that prevent market-based competition from occurring. Infrastructure, particularly large horizontal infrastructure such as roads, is often limited by geography or occupies physical space in such a way that it's not reasonable to have more than one of something. It would be impractical, for instance, to have many competing urban road networks, or electrical grids, or storm sewers. And even if there were enough physical space to build a second network, it usually isn’t profitable to do so, as the new entrant wouldn’t be able to charge sufficiently high prices to recoup their upfront investments.  Infrastructure will thus often be a “natural monopoly."

  In the absence of competition, there's once again the risk of under-providing service (as well as a reduction of consumer surplus). Under traditional economic theory, a monopoly will be incentivized to restrict output to try to maximize their profit, which both transfers money from the consumer to the producer and results in a "deadweight loss" — there are transactions that would have benefits for both buyer and seller that don’t occur so the seller can maintain a higher price. Government policy can try to correct this by using things like rate caps or allowable returns on investment. The electrical grid, for instance, was largely built by private companies, but they quickly began to be regulated by state utility commissions, which set rates based on an allowable rate of return.

• Holdups and holdouts — Because much of it is built for the purpose of moving things around, infrastructure projects tend to be long, horizontal construction projects that cross over many parcels of land. Building them thus requires getting the permission of many landowners — in some cases, such as with long-distance transmission lines, there can be thousands of landowners. These sorts of projects are at risk of being "held up" by landowners refusing to sell their land or grant permission for construction, or only agreeing to if paid a very high price. Because delay is costly, and companies may have made up-front investments, this gives a small number of landowners the ability to inflict very high costs on a project, potentially making it non-viable. As with our contract example from earlier, if builders know this, they may be reluctant to start infrastructure projects in the first place, once again resulting in less infrastructure than we would like. 

  Historically, building infrastructure has thus required government tools like eminent domain so that holdouts can’t prevent projects from getting built. With Los Angeles’ road improvement in the early 20th century, for instance, even though the plan was widely popular it still required laws that “made it easier to condemn property, allowed the city and county governments to pool funds, and let the city ‘overrule majority protests’ which prevented ‘intransigent homeowners’ from stopping improvements.”

In every case, if left purely up to the free market, we can expect less infrastructure to be built than we would like, perhaps much less. And of course, extensive land-use regulation in the US means that in practice we don’t have a free market but a highly restricted one, which only makes the problem worse. With good policy, however, we can potentially correct these problems. Policies that make it easier to build things can greatly enable infrastructure construction. The transcontinental railroad, for instance, was made possible by the Pacific Railway Act of 1862, which provided government loans and land grants to railroad companies. Policies that make it harder to build, on the other hand, will greatly impede the construction of infrastructure, and make the infrastructure that we do get much more expensive. When Washington DC was building its metro, for instance, a ruling that required each segment of the line to have its own environmental impact statement delayed the project by a year and added an estimated $120 million to the cost ($826 million in 2023 dollars). And the “regulatory morass” of getting transmission lines built has prevented the construction of cross-country High Voltage Direct Current (HVDC) transmission lines.

Since the 1970s, the US has steadily made things harder and harder to build. Not only has this made infrastructure like roads much more expensive, but it's likely prevented many projects from existing that otherwise might. This is bad, because it means wasted money and foregone benefits, but it's also bad because solving these sorts of problems is fundamentally why government exists. A government that can't is a government that isn't doing its job. We need ways of addressing the things that are making it hard to build infrastructure or resulting in less of it than we want, so we can continue to reap the benefits of it.