Fiscal Policy

The Case against the Senate Infrastructure Agreement

Traffic on the 405 Freeway in Los Angeles, Calif., in 2011. (Eric Thayer/Reuters)
The infrastructure proposal would be both wasteful and ineffective.

The Senate is poised to pass a $550 billion Infrastructure Investment and Jobs Act (IIJA) that would pile on hundreds of billions in new debt without fixing America’s broken infrastructure system. There are three main reasons for responsible lawmakers to oppose this bill:

The Cost
The case for adding $550 billion in federal infrastructure spending is weak. First and foremost, Washington has already borrowed a staggering $6 trillion since the pandemic began, and is projected to borrow $12 trillion more over the next decade just to finance current programs. On top of that, President Biden and congressional Democrats have proposed a $3.5 trillion reconciliation bill plus an 8.4 percent increase in discretionary spending that would cost $1 trillion over the decade. Congress is highly unlikely to allow temporary policies such as the expanded child tax credit to expire, and those extensions would also cost $1 trillion over the decade. Overall, the national debt held by the public would leap from $17 trillion before the pandemic to $40 trillion a decade from now.

How dangerous is a $40 trillion national debt? Consider that Washington overwhelmingly relies on short-term borrowing that must be rolled over into the current interest rate every few years. Every time that the average interest rate rises one percentage point, the interest cost on $40 trillion of debt increases by $400 billion per year, or $4 trillion over the decade. That is more than double the cost of the 2017 tax cuts. Over 30 years, each one percentage point interest-rate increase would cost as much as creating a new Defense Department. Washington’s long-term solvency would thus depend on interest rates remaining below 4 percent forever. Feeling lucky?

In that context, adding another $550 billion for infrastructure digs the budget hole deeper, and proves once again that Congress has lost the ability to say no. And while some suggest that $550 billion is not much money in today’s economy, it would comprise one of the largest non-emergency spending bills of the past 50 years.

If Congress is truly desperate to upgrade America’s infrastructure, lawmakers should first ask themselves why the hundreds of billions of dollars that federal, state, and local governments already spend annually on infrastructure is not getting the job done (more on that later). Then they should incentivize the nation’s governors to apply towards infrastructure the $350 billion they recently received to shore up non-existent budget deficits. After all, infrastructure is primarily a state and local activity, and the perfect use of a one-time cash windfall for states. It makes no sense for Washington to go deep into debt for infrastructure while states are sitting on enormous federal handouts that they do not need.

Fake Offsets
Several lawmakers pledged to support an infrastructure bill only if it does not add to the national debt. And despite assurances that the IIJA is paid for, the Congressional Budget Office has determined that the Senate bill would add $256 billion in deficits over the decade. The true shortfall rises to $350 billion after including the bill’s $90 billion bailout of the highway and transit trust funds, which are heading towards insolvency.

Ultimately, GOP lawmakers demanded $550 billion in spending offsets to pay for the bill, but were not willing to take the political heat of actually reducing $550 billion in benefits for any constituencies. So they took the classic Washington route of relying on gimmicks and fake offsets.

For example, lawmakers claimed $50 billion in “savings” from having the Department of Health and Human Services propose an expensive regulation and then “canceling” it through 2025. This is like a family considering a $10,000 vacation, then changing their mind and treating that $10,000 as new income to spend elsewhere. The provision to sell oil from the Strategic Petroleum Reserve would save a few billion dollars only until Congress repurchased the oil later at a potentially higher price. A pension-smoothing provision allowing some companies to underfund their pension plans now (which raises their taxable income and taxes paid today) and replenish those funds later would similarly kick the can down the road. The one-year extension of a series of annual spending cuts known as the “mandatory sequester” that are currently scheduled to last until 2030 would be more credible if Congress had not recently canceled the 2020 and 2021 sequester cuts.

The bipartisan agreement press release claims that the IIJA is paid for by including items that are not part of the CBO score. For instance, lawmakers take credit for lower-than-expected spending on a few past stimulus programs and for the revenue from spectrum sales that occurred this past February. These savings are already taking place independent of the infrastructure bill, and thus do not reduce the debt impact of the bill. Furthermore, the lower-than-expected cost of some federal programs is more than offset by the higher-than-expected cost of other programs. There is no pot of savings waiting to be spent.

Lawmakers have also promised large “dynamic scoring” revenues from the economic growth resulting from this bill. However, a new CBO analysis suggests that — due to the lack of legitimate spending offsets — this bill’s macroeconomic effects would save Washington just $6 billion over the decade, as a negligible rise in economic growth would be largely offset by higher interest costs. Still, this is slightly better than an early version of infrastructure legislation that economists at the University of Pennsylvania calculated would reduce long-term economic growth, wages, and capital stock.

Not only is infrastructure spending not guaranteed to provide long-term growth, it is also among the least effective short-term stimulus options. This is because federal infrastructure policies take several years to be implemented (remember the not-so “shovel-ready” 2009 stimulus provisions?), and are often offset by state and local infrastructure spending cutbacks. Additionally, highway expansion, high-speed rail, airport infrastructure, and new utilities are often needed most in regions with growing populations and economic activity — and these regions typically have the lowest unemployment rates and stimulus needs.

Lack of Infrastructure Reforms
Lawmakers demanding $550 billion in new spending to address America’s “neglected infrastructure” should instead be asking what happened to the nearly $500 billion that government at all levels already spend on infrastructure every year.

As I’ve written here before, America’s infrastructure planning and construction processes are among the slowest and most bureaucratic in the entire world. The inflation-adjusted cost of interstate construction spending per mile quadrupled from 1960 through 1990, and has continued to grow since then. Labor costs are higher in part because the Davis-Bacon Act raises wage costs by as much as 22 percent, and America requires many more workers to do the same building work as Europe. Government-mandated Project Labor Agreements can raise construction costs by as much as 30 percent. U.S. subway systems are by far the most expensive to build in the world, and in New York City cost quadruple the world average.

Environmental Impact Statements require on average seven years to complete — with some projects taking 17 years or even 25 years. These statements take no more than one to two years in Canada and 3.5 years in the European Union. And because no ground can be broken until all appeals are litigated, project opponents can effectively veto infrastructure projects by filing endless environmental-impact lawsuits.

Overall, a wide-ranging CBO analysis found that federal investments deliver half the returns of private-sector investments. CBO added that federal investments financed by taxes or debt — rather than spending offsets — could even have negative returns. Anyone who denies that America’s infrastructure process is broken should examine Boston’s Big Dig, or California’s doomed high-speed-rail boondoggle.

The bipartisan infrastructure agreement simply throws $550 billion more at this dysfunctional system. In doing so, it ensures that more money will be misallocated, misspent, delayed, and wasted. As long as Congress continues to measure progress by spending levels rather than outcomes, America’s infrastructure will not materially improve.

It is also worth noting that — despite ineffective spending — $550 billion would have been enough of a “moon shot” to substantially improve any one area of infrastructure, such as high-speed rail, highway congestion, bridge and dam safety, or the electrical grid. Instead, lawmakers sprinkled modest budget increases across approximately 20 infrastructure needs. This was enough to please many different interest groups, but not enough to truly “fix” any of these needs. Someone who leaves the United States today and returns a decade from now will not likely notice anything notably different in America’s infrastructure resulting from this bill.

America’s infrastructure could certainly use some upgrades. Yet the Infrastructure Investment and Jobs Act simply tosses more funding across several infrastructure areas without addressing the failures or waste in the current system. Lawmakers should instead overhaul the nearly $500 billion in existing annual infrastructure spending, and then use modest matching funds to incentivize governors to put much of their $350 billion federal windfall into infrastructure reform. If Congress truly wants to think creatively, they could even revive legislation to shift more of the highway program back to the states. There is no emergency forcing Congress to pass a bad bill quickly. It’s time for lawmakers to roll up their sleeves and build a more effective system, rather than just sign ever-bigger checks.


The Latest