We Don’t Need This Much Infrastructure Diminishing Returns & Increasing Political Demands on Infrastructure

Editor’s Note: This report was originally published in the New York University Journal of Law and Liberty[1] and has been lightly edited for style.

Infrastructure is the most celebrated type of government spending. Politicians and commentators across the political spectrum portray it as one of the core functions of government. Most claim that the U.S. has abdicated its responsibility to maintain and improve its infrastructure.

This article argues that the U.S. creates and maintains too much infrastructure. There is widespread acceptance that infrastructure in the U.S. costs too much on a project-by-project basis, due to mandates for everything from environmental planning to interest group contracting goals. There is less widespread understanding that American infrastructure projects today bring lower returns than before, regardless of the cost of individual projects, and many are bringing negative net returns. These low returns would counsel for a vast reduction in government infrastructure spending.

One reason for the reduced value of infrastructure is that investment is going to older systems with long-term diminishing returns, such as railroads, canals, highways, airports, electric transmission, and water distribution—systems which are over a century old. The age of these systems is not due to a refusal to invest in new infrastructure, since it is not obvious that there are equivalent new technological frontier infrastructure projects that require significant public investment. The other reason for declining returns is that infrastructure projects are increasingly selected for reasons unrelated to citizen needs, which means more “roads to nowhere,” unused transit projects, and transmission lines that are not reducing electricity prices.

This article argues that the issues of diminishing returns to infrastructure, increasingly politicized projects, and increasingly burdensome demands on individual projects are related. When infrastructure brings clear benefits to broad numbers of citizens, politicians are rewarded for the increase in well-being that comes from the projects. When returns are abstract or questionable, the main political beneficiaries are the project builders and interest groups that can be enlisted through direct payments. As broad-based returns diminish, political demands for projects and political demands on projects increase.[2]

The cascading problems that emerge from diminishing returns should restrain our hopes for increased infrastructure investment. They should also chasten hopes for a renewed “supply side progressivism,” whatever the virtue of its goals. Both the right and the left have looked at reducing regulatory barriers to infrastructure as part of a move towards increased infrastructure spending. But progressives have focused on barriers to infrastructure projects that further environmental goals, such as electric transmission lines and transit projects. These are the projects that already bring the lowest direct returns to citizens and instead rely on abstract issues like global temperature reduction to justify them. Without direct benefits to citizens, it is difficult if not impossible to create a coalition for such projects without politicized selection of projects as well as cost-increasing mandates and side payments.

The main solution to the problem of decreasing infrastructure returns is to cut back government spending on infrastructure, especially spending coming from the federal government that funds projects with questionable beneficiaries. New capital spending provided by the federal government is funding the lowest valued projects with the most additional requirements. Maintenance support by the federal government is even less plausibly related to core federal functions and only serves to increase normal maintenance costs and support low-return projects. Cutting back federal grants would create more room for the state and local governments that represent the immediate beneficiaries of projects to decide when and if a new project is worthwhile. It would also increase incentives for governments to involve the private sector more in funding and delivering infrastructure. By focusing government funds only on core public projects with clear beneficiaries and allowing the private sector to fund and build as much as possible, the U.S. can both improve our infrastructure and reduce politicized mandates.

There has been much clamor about the inability of the U.S. to invest in infrastructure, often by self-interested groups such as the American Society of Civil Engineers. If one looked at infrastructure as a portion of all government spending one could imagine a continually declining interest. State and local infrastructure and capital investment has gone from almost a fourth of spending in 1970 to around 15% today.[3] As a percentage of federal spending, core infrastructure spending has declined from about 5% in the mid-1950s to about 3% today.[4]

But the relative decline in infrastructure spending is the result of an increase in all other forms of government spending. If one looks at core infrastructure investment, understood as transportation and water projects, it has been almost stable at around 2.5% of Gross Domestic Product for the past 60 years, after a slight jump to 3% of GDP following the interstate highway act of 1956.[5] A stable percent of GDP means real spending on infrastructure has been going up for decades. It is not obvious that real spending on century-old infrastructure should be a stable part of GDP and should be increasing in real terms.

Since infrastructure typically involves financing public goods where private returns are uncertain, it is difficult to demonstrate its economic value. But if one focuses on the main benefits of new infrastructure, especially its value to direct users, it is clear that increasing spending is bringing lower returns than in the past, either on a project-by-project or a dollar-by-dollar basis, and that the total returns are often negative.

Highways and roads are the largest proportion of core infrastructure spending, but many studies show declining returns on these projects. One estimate found that highway investment contributed about a third of total annual productivity growth in the 1950s and early ‘60s, during the height of the interstate boom, but this kept declining until it contributed only 7% of growth in the 1980s. By that point, investments in highways were returning less than returns to private capital.[6] A paper written by Chad Shirley and Clifford Winston likewise found declining economic returns from highway investments until in the 1980s and 1990s they returned less than 5%.[7]

The returns to transit investment have been consistently lower than those for roads and have been declining at an ever faster rate.[8] The number of passenger trips on public transit remained largely constant from 1960 until 2019, even though the number of vehicle miles traveled by transit more than doubled and the cost per passenger trip increased more than five times.[9] After Covid, transit ridership dropped 25% below its previous level, even as driving and air travel returned in full force.[10] Transit now represents less than 2% of passenger miles traveled but receives almost 20% of government passenger transport funding. And this percentage actually undercounts transit’s share because support for highways and air service helps move freight as well as people, and transit buses use the highways and roads as well.[11] Farebox recovery ratios, or the percent of operating costs paid directly by riders and users of transit systems, dropped from over 90% in the middle of the 20th century to about a third today.[12] By any measure, building increased transit capacity that carries fewer passengers at higher costs demonstrates a rapidly declining return on investment.

Although it does not receive the press of other infrastructure,“water infrastructure,” mainly projects for sewers and consumable water, makes up a third of all U.S. core infrastructure spending.[13] In terms of the main service provided by this infrastructure—giving water to and removing water from households and businesses—little has changed in the past century. Yet the total spent on such water utilities, ignoring other water projects such as ports and transportation, has been stable at about 0.5% of GDP for decades and is now well over $100 billion a year.[14]

What have Americans received for this increased water spending? It has not saved Americans from more expensive water or led to substantially more consumable water. The Bureau of Labor Statistics includes a combination of water, sewer, and trash collection in an inflation measure, but water and sewer make up the bulk of that. By this measure, real prices for U.S. consumers of water and sanitation services have increased by about 50% since 2000.[15] This is an impressive failure of policy. Most of us drink the same way and flush the same way as our parents and, for that matter, our great-grandparents did, but we are paying more and more for it. The real costs for a gallon of water coming from the tap or going down the drain have gone up even though we’ve managed to limit our usage. Total American public water use did go up from 1950 to the 1980s, but it has been mostly stable since then and has gone down from its peak.[16]

Once one moves beyond core infrastructure into broader network infrastructure areas, such as electricity and internet networks, one finds the same problems with lower returns. The basic mechanics of electricity transmission and distribution are now over a century old, and the basic service provided by electricity, i.e., to get electrons to consumers on demand, has not changed during that period. After decades of declining real electricity prices in the 20th century, however, the real price per kilowatt hour actually increased slightly from the early 1970s to the early 2010s.[17] In some parts of California, which have engaged in the most electricity infrastructure spending, real prices to consumers have nearly doubled since the late 1970s.[18] Overall, real spending on electricity has more than doubled since 1970, to over $400 billion per year, and this is almost exactly in line with the increase in total electricity usage over this period. Like water usage, overall electricity usage has been stable in the most recent years.[19] Electricity is an area that has absorbed massive amounts of research, innovation, and investment, and yet consumers have seen almost no gains and, in recent years, have seen little increase in the actual product delivered.

There is an increasing demand for public spending on internet infrastructure, meaning more high-speed cable or fiber-optic wires. Since the private sector has largely solved the problem of broadband access, however, more public spending is bringing few returns. According to the Federal Communications Commission, 96% of the U.S. population has access to a broadband network if they care to use it.[20] Most of the public focus on internet investment has been directed at connecting rural citizens, but there is not a large “digital divide” between urban and rural residents, even on the choice of whether to pay for broadband internet. 85% of households in the largest metro areas are currently paying for and using broadband internet service compared to 75% in the most rural areas. This divide is even less stark if one looks at smaller metros and normal rural areas.[21] Compare this divide to the divide in rural and urban electricity consumption during the Great Depression and the changes wrought by the rural electrification movement. In 1930, before rural electrification, almost 90% of nonfarmers had electricity, but only about 10% of farmers did. Within a few decades, virtually all farmers had electricity.[22] The returns to public spending on broadband infrastructure in a nation that is almost entirely built-out are almost nil, especially when considering that the accessible broadband available to almost all Americans is sufficient for high-definition video streaming and there are ever more options for wireless and satellite internet.

Many argue that even if the U.S. has not skimped on new infrastructure, they have refused to fund infrastructure maintenance. John Oliver on his popular HBO show did a segment on maintenance, which, like much of the show, would be very funny if one agrees with the premise it’s based on.[23] There is, in fact, no evidence that our infrastructure is less well-maintained then in the past. Although many cite the fact that about 7% of U.S. highway bridges are structurally deficient or in poor shape, that is actually a drop of more than half since 2000.[24] The share of major roads in poor condition has been close to stable over the same period, but the number of roads in “good” condition has improved from 46% to 54%.[25] According to the Federal Aviation Administration, 98% of airport runways are in excellent to fair condition, far above the federal minimum of 93%.[26] Maintenance and operation spending has increased faster than capital spending in recent years, and, for the first time since we have had records, maintenance spending is a larger proportion of infrastructure spending than new investment.[27]

The increased maintenance spending has, in fact, been a result of the ever-increasing additions to the infrastructure capital stock, which increases maintenance needs as well as natural depreciation. This explains why state and local “net” infrastructure capital spending, spending after maintenance and depreciation, as a percent of GDP has gone down since the 1950s.

The fact that we have fewer “net” additions to infrastructure, but that these proportionally fewer new projects have lower returns, considered on a project rather than a cost basis, may be surprising. Yet there is a literature on the productivity of public capital, and it consistently finds, to quote one study, that “the larger the stock and the better its quality, the lower will be the impact of additions to this stock.”[28] The diminishing returns on U.S. infrastructure have certainly set in, and for many projects returns are far below cost. One might think the diminished pool of new projects would mean the remaining ones are more carefully selected and thus provide greater benefits. By any reasonable measure, this is not the case.

Unlike in the private sector, diminishing returns in the public sector do not lead to less investment. They merely change the nature of the projects that receive approval. If broad-based benefits to citizens are less important in the political calculus for infrastructure spending, the “political return” in terms of rewarding certain groups must go up to compensate. By many measures, we know that the public sector is investing more in lower return projects that cater to narrower interest groups than in the past, and this is further exacerbating the natural decline in returns from built-out systems.

The decline in returns from new highways has been partially a result of their politicization. Originally, maps designed by the federal Bureau of Public Roads (collected in “the Yellow Book” in 1955) formed the outlines of the interstate highway system. The highways were carefully placed to maximize traffic flow.[29] A study by later Nobel Prize-winning economist Daniel McFadden found that in the 1950s and ‘60s, the California Division of Highways decided routes based on a quantifiable cost-benefit analysis.[30] Such cost-benefit analyses of road projects was typical during this era.

For the past four decades, however, highway spending has become more related to government heft. The last federal act building out the interstate highway system in 1987 approved the massive “Big Dig” tunnel in Boston, even though the government’s own estimates found a benefit-to-cost ratio of 0.3 to 0.5—and that was before further massive cost-increases.[31] All Roads Lead to Congress, an exceptionally detailed review of the passage of the 2005 highway bill, showed that almost the entire debate on the bill revolved around whether to raise the minimum percentage of gas taxes returned to “donor” states, those that provided more gas taxes to the federal government than they received, above the previous standard of 90.5%. The increased distribution of funds based on state size and gas taxes meant the funds were not directed at the urban areas with the most congestion or the most demand for new or expanded roads.[32] A Federal Highway Administration cost-benefit analysis confirmed that rural highway and rural interstate investment in particular should be cut by about half to meet cost-benefit metrics.[33]

Highway acts have changed into general transportation acts or general spending acts. Although the original goal of federal highway investment was to direct gas-tax revenue to interstate travel, which states would have an incentive to underfund, in recent acts the federal government has funded ever-narrower benefits. The 2005 act was loaded with 5,634 earmarks (as opposed to just 121 in the highway act 18 years earlier), such as the infamous “Bridge to Nowhere.”[34] Modern transportation acts have funded bikeways, sidewalks, nature trails, historic paving materials, and transportation museums that have no plausible relationship to interstate, or often even local, travel.[35]

The fact that ever-increasing proportions of transportation funding is spent on transit when it is carrying an ever-smaller proportion of travel is further evidence of the politicization of transport. Despite declining transit use, in the 15 years to 2021, federal transit spending more than doubled, far above the 70% overall increase in transportation spending.[36] The transit projects are also more politicized now than in the past. Consider one of the biggest transit projects in modern years: the $9 billion spent after 1990 building out a metro system in Los Angeles. The returns to this project were questionable at the outset, considering low local demand, but they were exacerbated when Congress, at the behest of some local interests and through local congressman Henry Waxman, banned funding for the line through parts of the Wilshire Boulevard neighborhood, which contained by some measures the most dense urban corridor West of the Mississippi, and redirected it to lower density areas.[37] A Federal Reserve study showed that after the build-out of the subway, the annual costs, including capital and operating costs, were still almost double the annual benefits at the most generous estimate and about eight times at the least generous.[38] The fact that much of modern transit spending goes to lower-density rural areas is more evidence of politicization. For instance, for small fixed route transit systems in Indiana, transit costs around $18 per passenger trip just for operating expenses. These systems recover only about 5% of that operating cost back in fares, while the rest is subsidized. No one bothers to calculate plausible returns to capital investments that cannot even bring in 1/20th of the operating expenses from customers.[39] These politicized projects make the diminished returns to politicized highways look rosy by comparison.

Airport infrastructure should be largely self-funded, since the airlines that use airports can pay for them.[40] But today we have increasing public investment in airports—although much of that investment is not going to facilitate travel. While in the 1960s about 75% of airport investment involved runways or taxiways, and only 25% was spent on terminals and similar structures, by the early 1970s, that was reversed.[41] Today the Airports Council estimates that only about 25% of total airport investment involves airfields. The rest involves terminal spending and other projects (such as expanding retail) that have little to do with transportation per se and cater more to outside interests.[42] The federal Airport Improvement Program, a program that spends more than $3 billion a year, funds the airports least in need. Although 88% of all enplanements are at large airports, these larger airports only receive 25% of annual funding. A recent Government Accountability Office report also found that grants to smaller airports grew faster than those to the large ones.[43] Similarly, the Essential Air Service Program provides subsidies of about $300 million a year, and up to more than $200 per passenger, to keep airlines traveling to these small, already overfunded airports.[44] Most of the direct airfield investment today targets maintenance, as opposed to new runways which are rarely built anymore in America. A major new airport has not been created in almost 30 years. New airport investment is simply going to less valuable projects that bring lower returns than in the past.

The returns to water infrastructure have likewise declined rapidly due partially to political demands for less worthwhile projects.[45] The most obvious non-price benefit of modern water infrastructure is that our water—both our drinking water and our surface waters such as rivers and lakes—is cleaner. But demands by environmental groups have led to the funding of increasingly marginal environmental water projects that bring few direct health or recreational benefits to citizens.

In the early 20th century, fixes to water and sewer systems represented some of the greatest public investments in human history. From 1902 to 1929, deaths from waterborne diseases fell by almost 90%, and the percent of all urban deaths accounted for by waterborne diseases dropped to about one-seventh its previous level, to just 1.4% of urban deaths.[46] The investment in sand filters, chlorination, and other innovations brought immense and obvious gains.[47] Despite ever-increasing investment in water cleanliness in recent decades, however, drinking water has long posed little threat to Americans. The number of drinking water disease outbreaks, including diseases associated with both biological and chemical irritants, has been fairly stable since the 1990s, at around 10 to 20 a year. Waterborne diseases from both drinking and surface water now represent less than 10% of the threat of food-related illnesses.[48] Yet annual investment in drinking water quality increased from $20 billion in 1970 to around $70 billion in the 2010s.[49] Billions of dollars a year are dedicated to reducing an increasingly infinitesimal threat.[50]

Starting in the mid-20th century, environmental groups and the broader public became more concerned with the quality of surface water for recreation, fishing, and the like. Annual investment in surface water pollution control doubled from the early 1970s to the 2010s, to around $80 billion a year. In the early part of that period, there were indeed significant improvements in surface water quality. The percentage of monitored sites rated “not fishable” dropped from near 30% to 15%. But most of that decline happened before the mid-1990s. A recent estimate shows that modern surface water quality investment has costs that outweigh its benefits.[51]

Today, groups such as the U.S. Water Alliance and environmental lawyers have successfully pushed for more subsidies and more mandated investment in water infrastructure, often in increasingly marginal areas. An increasing number of the violations of water quality standards, and concomitant requirements to rectify them, are in rural and low-income areas, meaning that most of the burdens of putative improvement are placed on communities that have little ability to pay and have typically chosen not to.[52] Jefferson County, Alabama, home to the city of Birmingham, faced lawsuits about its water quality and entered a consent decree in the 1990s that forced it to rework its sewer system. Its sewer debt grew from $300 million to over $3 billion by 2004. The average household sewer bill rose over 350%. These increased costs (and some bad financing decisions) led the county in 2011 to descend into what was then the largest municipal bankruptcy in American history. There was little indication that the citizens of Jefferson County noticed the change in their water quality, but they did notice a massive increase in costs.[53]

There has also been an increase in the politicization of the subsidies provided to water projects. Since the 1992 Water Resources Development Act, the Army Corps of Engineers can assist groups in carrying out water projects as part of its Section 219 program, but, as the army notes, the “projects must be specifically named by Congress.”[54] The most recent water act included dozens of new projects explicitly named in the bill, including $35 million for water infrastructure and environmental restoration in each of the three counties in Delaware. (It helped that the chair of the Senate Environment and Public Works Committee, and of course the president, are from Delaware.) The act also provided $103 million for water, wastewater, and stormwater management in Los Angeles, although that city has substantial resources to manage water on its own. Increasing amounts of public water investment has been devoted to local “stormwater management,” which is a particularly low-return effort to control runoff from streets, buildings, and public spaces.[55] From 2022 to 2023, Congress also redirected about $2.3 billion in water funds to earmarks, which proportionally reduced the amounts states received for their professional water plans.[56]

Most of the modern drive for electricity infrastructure involves building extensive transmission systems to far-flung renewable sources. These are demanded by environmental groups and a growing renewable energy industry. The power of these groups is demonstrated by the fact that renewable energy gets about five times the federal subsidies as fossil fuels even though it produces about one-eighth as much energy.[57] Most electricity investment is still done by the private sector, but much private sector renewable investment happens in response to “renewable portfolio standards” or other mandates accompanied with subsidies, whereby state regulators require certain proportions of electricity to come from renewables. Although new renewable projects can sometimes reduce costs to consumers, many cost-increasing renewable projects are mandated due to concerns about the secondary effects of fossil fuel projects.

There are some local air pollution reductions associated with renewable sources, but here too the U.S. is dealing with diminishing returns. For instance, due to the a long-term federal and state regulatory focus on reducing electrical plant airborne emissions, the percent of nitrogen oxide, one of the major regulated air pollutants, that comes from utilities in New England, for example, is only 4%—less than that from commercial marine and shipping activities.[58] The strongest argument for grid expansion to renewables is that it helps reduce global warming. But many of these subsidies and investments cannot pass a “social cost of carbon” test for cost efficiency, even if using global as opposed to U.S., or especially local citizen, impacts of climate change, since subsidies to energy can encourage increased use of energy overall that offsets some of the reduced emissions.[59] Even if such projects do pass the social cost of carbon test, it is difficult to convince affected citizens of such projects of the gains. For those concerned about global warming, alternatives, including direct regulation of polluters, carbon taxes, and other measures, would often be more efficient than forcing a build-out of renewable energy infrastructure, but those measures have a much smaller constituency than subsidies to individual projects that can drive up prices.[60]

Electricity investment finds itself in the same situation as most other modern infrastructure investments. The tangible benefits to citizens to building out older infrastructure are minimal and the arguments for such infrastructure are abstract. Few citizens would even be aware that an electricity transmission project changed global emissions. Without broad-based gains, politicians need to award projects based on the demands of narrower interest groups. Politicians must also increase the mandates on such projects to buy off more interest groups and let them share in the putative gains.

In the book Regulatory Takings, William Fischel notes a standard trajectory for new infrastructure systems and the expansiveness of compensation for landholders and other parties affected by them. In the early years of railroads or highways, when returns to new projects were highest, legislatures, judges, and juries provided minimal recompense to those who lost property due to the eminent domain takings of the projects, or they reduced the compensation to property owners based on the supposed gains they received from the infrastructure itself. The public was eager to see new projects completed and was unwilling to let individuals stand in the way. After the systems were built-out and the returns decreased, the compensation to property owners and others affected by the infrastructure increased.[61] Lower infrastructure returns for older infrastructure systems led to higher eminent domain costs, further lowering the returns of that infrastructure.

One could propose a more general tendency across all infrastructure projects. As the infrastructure returns to the general population go down, the tendency to increase side payments and coalition buy-ins for projects goes up. This explains the combination of decreased returns, worse projects, and increased costs. Eminent domain is one aspect of this, but one can see the trajectory in many areas. As already discussed, the 1970s and early ‘80s was an inflection point in terms of returns to highway and many other infrastructure projects. It was also an inflection point in terms of their increased costs. Highways are the best-studied and most easily comparable subjects of infrastructure research, and a recent study showed that cost per mile of highway projects began exploding in the 1970s and early ‘80s.[62]

Since the 1970s, politicians have found numerous ways to increase the political coalition for infrastructure spending by increasing side payments for certain interest groups. One example has been creating quotas or goals for the race and sex of the contractors and subcontractors who build infrastructure projects, which has brought identity groups into the infrastructure-funding coalition. Minority contracting goals were first mandated in U.S. law in the late 1970s, when the federal government required 10% of federally funded transportation projects’ contracts go to firms owned by “disadvantaged” individuals, i.e., racial minorities. Beyond minorities, there now are requirements to give 3% of federal contracts to “HUBZones” or disadvantaged areas, 5% to women-owned businesses, 3% percent to disabled veterans, and so forth. Such requirements have received broad support from both parties.[63] The costs of these contractor goals are not insubstantial. One study showed that after California banned racial and sexual preferences, the cost of California highway projects relative to federal projects fell by over 5%.[64] Another study found that a 10-percentage-point increase in disadvantaged business contracting goals increased costs by 25% on average.[65] Spread these increased costs across U.S. infrastructure spending, and the funds required to meet these identity group goals absorb a significant proportion of all infrastructure funds. Yet they have also played a critical role in maintaining a political coalition for infrastructure spending despite decreasing general returns.[66]

Many people are aware of the increasing cost and length of National Environmental Policy Act environmental impact statements, first instituted in the early 1970s.[67] Such impact studies have brought environmental groups, who were once inveterate foes of building, into the infrastructure process, by making sure that the projects either met their demands with funds for environmental remediation, or involved environmental consultants directly in the projects. There are many other groups that benefit from infrastructure consultancies. Section 106 of the National Historic Preservation Act requires federal agencies that build or fund state and local projects to consider a project’s impact on historic properties. Some have called it the “archaeologist full employment act,” since it leads to extensive reports and often extensive digs on any property that might contain historic objects.[68] One sees similarly increasing mandates beginning in the 1970s and ‘80s on everything from Buy America requirements to recycling goals.[69]

Many of these mandates are placed by the federal government on state and local governments that receive federal funding and carry out the actual infrastructure projects. One side effect of the federal government acting mainly as a grant-funder is that it cares little about the actual impact of the grants on the community that receives them and cares more about general mandates placed across projects nationally. The main interest groups that lobby the federal government for mandates are not looking at individual project returns; they are looking at how generalized grants can be leveraged for the benefit of groups such as minority contractors or archaeologists. Along with increased interest group participation in the selection of projects, these heightened requirements have reduced the returns of projects by widening the number of special interest beneficiaries and increasing costs.

The 2021 Infrastructure Investment and Jobs Act, also known as the Bipartisan Infrastructure Law, is a culmination of the recent trajectory of low returns, worse projects, and increased mandates. It demonstrates that the fundamental tendencies of American infrastructure spending will continue. Beyond carrying forward the existing and poorly designed transportation funding formulas, it added over $550 billion in new spending over five years that deviated even further from projects that could bring plausible returns. As much as $55 billion was allocated towards water infrastructure, mainly appropriations for drinking water and clean surface water, a field that has already been bringing negative returns for years.[70] Transit, which carries less than 2% of U.S. passenger miles, got almost $40 billion, or about a fourth of combined personal passenger spending. Amtrak and passenger rail, which carry an infinitesimal proportion of passenger miles, received over $50 billion in advanced appropriations.[71] Broadband, already built out, got $65 billion.[72] After a decade in which Congress had banned explicit earmarks, the debate on the act marked the return in force of earmark spending demands, with over a thousand of what were known as “member designated projects” in the accompanying House report to the bill.[73] The new act seems purposely designed to ensure that the lowest returning projects get the highest proportion of spending.

Much of the infrastructure act involved maintenance spending on already well-maintained projects, or, even less plausibly related to the investment in the act’s title, operating expenses. The Affordable Connectivity Program in the act gave $14.2 billion to subsidize internet demand. It has nothing to do with the supply of infrastructure.[74] By 2019, more than a third of federal transit spending, or over $4 billion, was already subsidizing transit operating expenses. But the combination of the infrastructure bill and other recent acts has increased federal operating subsidies for transit to $22 billion a year—meaning the federal government became the single largest source of transit operating revenue.[75] Much of the increased infrastructure “investment” in recent years is propping up questionable projects that would not be self-sustaining by subsidizing maintenance and even operations.

The Bipartisan Infrastructure Law and some of the Biden administration’s recent actions have also continued the trend of continuously adding new mandates. “Buy America” requirements have been in U.S. infrastructure law since the 1970s, and the Trump administration further tightened them.[76] But the recent infrastructure act created a broader “Build America, Buy America” mandate, which requires iron, steel, manufactured products, and, as a new addition, construction materials on projects to be produced in the United States. The law extended this mandate beyond projects funded by the act itself and even beyond the traditional spheres of transportation and water projects to all federally funded “infrastructure,” including airports, electric transmission, and even individual government buildings.[77]

The Biden administration has been working assiduously to expand the base of interest groups involved in its broadly defined infrastructure push through increased mandates. On Biden’s first day in office, he released an executive order directing that spending on minority contractors go up to 15% of all federal contracts from the current level of around 10%, and has since continuously pushed to increase that amount.[78] Another executive order for the first time required Project Labor Agreements—mandated deals with labor unions—on all large federal construction projects.[79] The administration, through its environmental justice, or Justice 40, initiative, has tried to reserve 40% of all climate and green investments, including broadly defined environmental grants for water, port, and electric vehicle projects, for disadvantaged communities. This almost certainly means placing infrastructure in suboptimal locations, since poorer communities do not usually need electric car charging stations, and port expansions have more to do with international logistics than issues at the local level.[80] Despite efforts in the infrastructure act to limit environmental reviews, the Biden administration later strengthened and extended these reviews through regulation.[81]

The recent infrastructure programs have brought both hope and hardship for the “supply-side progressivism” movement. The infrastructure act and the Inflation Reduction Act focused precisely on those environmental projects that the progressive movement was championing, but they also brought the increased mandates they feared, because, as suggested, the recent acts focused on those projects with the lowest direct returns to affected citizens, such as electric transmission or rail transit. There is little possibility of getting these low-return projects approved by politicians if there aren’t side-payments for political allies since there is little consumer or user constituency for them. Whatever the general coalitional issues for supply-side progressivism, and they are substantial, the projects that the movement focuses on are not promising in terms of streamlining mandates and permitting reform.[82] Reducing political mandates for projects that are already heavily subsidized and bringing negative returns would not so much open investment as move the subsidized benefits from one group to another.

The biggest question for the government is not “why can’t it build things?” but “why is it building things that people don’t need?” One reason could simply be naturally diminishing returns, in that the classic infrastructure goods that we need—roads, transit, airports, and so forth—are already well supplied in much of the country. But if the issue was just diminishing returns the government might respond by stopping the low-return projects or lowering mandates on them. The government’s response has been the opposite.

There is a need for new infrastructure projects in America. New housing requires new roads, pipes, and electric distribution to make it livable. New factories, warehouses, and data centers need new internet and electric networks to service and power them. Pipelines, probably the most protested and least loved of all current infrastructure, are needed due to the increase in natural gas and oil production. (Pipelines also happen to be more environmentally sound than moving gas or oil by truck or railroad.) Many of these projects do not require new public investment so much as the removal of roadblocks against private investment. There is much benefit to removing these regulations, and, occasionally, providing appropriate public support for areas in which the private sector returns are questionable. Yet much of America’s public investment in transit projects, clean water projects, electricity projects, and internet projects can and should be stopped. At the very least, the government, most especially the federal government, should stop subsidizing them. Infrastructure should be funded as much as possible by user fees on the direct beneficiaries of the projects. Where possible, the government should allow the private sector to take the lead in building them out.

Even with a retreat from federal and other government subsidies, there are substantial possibilities for privatized infrastructure investment in the United States. Despite the well-demonstrated benefits, privatized services appear to be a fairly stable proportion of all local government services, where most actual service provision takes place, since the early 1990s.[83] Why did the privatization movement stall out? Some studies show spending constraints and budget crunches encourage privatization.[84] The flood of federal money, especially for maintenance and operation, has prevented those spending constraints from emerging, deterred the privatization of projects, and increased the costs on those that remain. Similarly, mandates on federal transit projects that protect the collective bargaining of labor unions prevent both increased efficiency and privatization.[85]

The way for the government to build better, or to build back better, is for it to build less. Much of our modern infrastructure investment is not serving a plausible purpose, and the main reason is that we do not need new public infrastructure commensurate with the level of current public spending. Federal grants and mandates are a significant reason for the continuance and growth of unnecessary projects, the politicization of project placement, and the increased demands on such projects. The log-rolling nature of federal subsidies means ever-increasing infrastructure investment is going to low-density areas that have little need for new projects. Ending most federal support, funding more projects at the state and local level, and opening more options for the private sector would both raise returns and reduce the political problems that have made U.S. infrastructure famously wasteful.

Judge Glock is the director of research and a senior fellow at the Manhattan Institute and a contributing editor at City Journal. He was formerly the senior director of policy and research at the Cicero Institute, a nonpartisan think tank based in Austin, and a visiting professor of economics at West Virginia University. He writes about the intersection of economics, finance, and housing, with a perspective informed by his work in economic history.

Glock’s work has been featured in National Affairs, Tax Notes, the Journal of American History, NPR, the New York Times, and the Wall Street Journal, among other places. He is the author of the book The Dead Pledge: The Origins of the Mortgage Market and Federal Bailouts, 1913-1939, published in 2021 by Columbia University Press. He received his Ph.D. in history with a focus on economic history from Rutgers University.

The Manhattan Institute thanks the Milstein Innovation in Infrastructure Project for supporting the publication of this paper.

Please see Endnotes in PDF

Photo: Thomas Northcut / DigitalVision via Getty Images