The impending multibillion-dollar infrastructure deal between the Biden administration and the Senate Republicans has been hailed as welcome bipartisan cooperation that augurs well for the revitalization of this nation’s aging infrastructure—and for improving the lives of many Americans. This narrative plays so well because the term “infrastructure” now carries a seal of approval. Private expenditures may be suspect as greedy, but not public expenditures.

The great appeal of the term “infrastructure” is that it helps bridge the divide between classical liberal and modern progressive attitudes toward government spending. The most thoughtful students of laissez-faire economics do not limit permissible government actions to the prevention of force and fraud and the enforcement of private agreements. They also believe that government intervention is often needed to provide those collective facilities—such as roads, railways, bridges, tunnels, communications, and pipelines—that cannot be put together solely through coordinated private investment.

Many of these operations are long and skinny and thus presuppose the ability to assemble land from multiple owners, who in the absence of a threat of condemnation could hold out for higher prices that doom all collective projects.

The application of state power changes the bargaining equation by removing that blocking position of individual owners while providing just compensation for those whose property has been taken. Ideally, this one-two combination can make the state a superior player to, or essential partner with, private industry. Once those roads, rails, pipes, and wires are acquired by the state, they are often subject to traditional common carrier obligations of universal service at reasonable rates, whether the state retains operational control over these facilities or transfers them to private parties. The use of state power for infrastructure is thus tied to a credible theory of market failure.

Progressives who sense the implicit appeal of this arrangement seek to use it for as many activities as possible, all to expand the role of government power. It has been common, for example, in recent times to hear extensive defenses of the public funding of “human infrastructure,” on the ground that investing in the education of young children, community college students, and job trainees pays handsome dividends by increasing the level of human capital. From there, it becomes a quick leap to defend public investment in housing, Internet upgrades, research and development, and much more.

The progressive position offers no limiting principle for public expenditures, and the infrastructure exception quickly overwhelms the classical liberal presumption in favor of private investment. The abiding worry with the progressive position is that people are much more willing to spend other people’s money than their own, so that a dollar of private investment in education is far more likely to yield a decent return than free government funds paid to those who ask for them.

Yet these grubby matters of incentive don’t cut much ice with progressives, who argue that market failure requires state funding of education in order to create a public good.  But it is a mistake to think that the primary beneficiary of education is the public at large, and not the particular students who receive training. That is why, notwithstanding the heavy tax burdens to fund public education, a thriving market for private education continues to exist even as the level of public education in many communities continues to deteriorate, given the union monopoly.

In response, it is often said that public education is needed for those families who cannot pay for their children’s education, to which there are two answers. First, the prices currently charged could be much reduced if the regulatory burdens on these private options were curtailed. Second, the remaining financial shortfalls could be met by targeted student vouchers, the use of which could separate state funding of education from state control over education.

It is, therefore, a mixed blessing that all “human infrastructure” expenditures were removed from the compromise bill, as a separate bill devoted exclusively to that purpose may emerge. But even properly designated infrastructure projects raise further questions. How much money should be devoted to infrastructure, and how much to private investment? Clearly, at some point, there are diminishing marginal returns to both; so, ideally, the last dollar spent on infrastructure should generate the same as the last dollar spent on private investment. And which infrastructure projects should receive funding, and how should that money be spent?

On these issues, the current breakdown of the items that have made it this far through the process reveals some disappointing news. From the classical liberal point of view, the government should be highly reluctant to engage in infrastructure projects when private investment is adequate. That proposition immediately calls into question the bill’s broadband expenditures of $65 billion, water storage for $50 billion, power infrastructure for $73 billion, and electric vehicles for $15 billion. These four questionable programs alone total $203 billion, or roughly 37 percent of the $550 billion base figure.

All these items have already attracted extensive private investment, so why distort the incentives by subsidizing some firms and competing unfairly with others? To be sure, there may be shortfalls in these investments, but the last thing to do is throw huge sums of money at any problem without first ensuring that the regulatory apparatus is not choking off private investment. All too often a state permit system operates as a barrier to entry. The sensible approach is to ask why these barriers are in place, and then to remove those that make no sense.

A good example is the senseless limitation on telemedicine across state lines. State licensing laws restrict the scope of the national market in health care, a field that covers close to 20 percent of the economy, where efficiencies are desperately needed. If a license given in one state allows a physician to treat persons from out of state—so long as the patients enter the physician’s home state—what possible reason is there to fear that fraud or incompetence will flourish when telemedicine crosses state lines? It would be far better to remove the barrier than to expand some health care subsidy to expand access to care, which is what a modern infrastructure bill is sure to do. Broader access will be of greatest benefit to underserved individuals with limited income, who are less likely to have easy access to medical care in their home jurisdiction.

Yet at no point does the current infrastructure bill ask whether deregulation can achieve superior outcomes over devising yet another quasi-entitlement program. This missed tradeoff was one of the most serious design flaws of the Affordable Care Act, which also chose to stifle interstate competition in insurance markets on the view that larger subsidies are preferable to lower market prices.

The problems in the bill run deeper. A government should think of itself as a trustee for the entire public, including citizens in low tax states, not just for its preferred clientele.  The minimum condition for meeting this constraint is to pay no more than competitive prices and wages for goods and labor acquired by government. Any additional payment is a disguised gift of public funds to private parties. But today, this simple proposition is far from the norm in dealing with government expenditures. Senate Majority Leader Chuck Schumer has gloated that “New York has not seen such a federal infusion of infrastructure dollars” that will “rebuild and revive” state infrastructure, through the creation of new union jobs no less—all to the tune of close to $11 billion. The bill also rewards states that have the worst infrastructure systems by imposing taxes on other states that have not run their infrastructure into the ground. In effect, the infrastructure bill ignores one of the major constraints of a sound theory of public expenditures: to avoid counterproductive cross-subsidies that could not take hold in competitive markets.

Unfortunately, today’s politics have no guardrails against penalizing well-run states and propping up badly run ones. And so it is that New York, with one of the worst-run public transportation systems in the land, is rewarded nationally for its rich union contracts and its incompetent management structures. In a sensible constitutional order, the courts would check massive cross-subsidies that are imposed by differential regulation—e.g., on zoning or landmark preservation laws—and look over public expenditures to improve overall social performance. The entrenched cross-subsidies in the current bill are anathema to the classical liberal mindset, however appealing they are to progressives.

Fortunately, the Republicans were able to bleed some of the worst subsidies out of the compromise bill. But by failing to take a tough position on infrastructure spending more generally, they opened the door to more government spending in the future. Indeed, when Republicans regain power, they will most likely use this playbook to help their own constituencies. The current bipartisan consensus is a recipe for long-term decline in the economic and institutional design of public finance programs in the United States.

© 2021 by the Board of Trustees of Leland Stanford Junior University.