The Senate has begun deliberations over a bipartisan plan to provide $550 billion in new spending for a wide range of infrastructure projects, including roads, bridges, public transit, broadband, and the electrical grid. The good news is that lawmakers avoided raising taxes to cover the cost of the new spending and instead used some reasonable fees and asset sales. The bad news is that half of the offsets come from unused, debt-financed COVID-19 relief funds and the economic return on many of these investments is questionable.

Let’s first review some of the budgetary offsets and then assess the economic value of the various infrastructure projects.

Infrastructure Spending Pay-fors

In order to maximize the economic return from infrastructure spending, the Congressional Budget Office (CBO) has advised that lawmakers offset any new spending with reductions in less valuable federal spending. Unfortunately, lawmakers have mostly chosen to ignore this advice and instead are proposing a menu of non-tax revenue raisers, budget maneuvers, and unspent COVID-19 relief funds.

The plan’s only “spending cuts” are $3 billion in savings from reducing Medicare costs on “discarded medications from large, single-use drug vials.” Surely lawmakers could have found more than $3 billion in Medicare savings by reducing waste and fraud. The General Accountability Office (GAO) reports improper Medicare payments totaled $43 billion in fiscal year 2020.

More than half of the pay-fors are reallocated funds from the various COVID-19 relief bills, which means that much of the new infrastructure plan is debt-financed, not offset. This includes $205 billion of repurposed “certain unused COVID relief dollars,” “$53 billion from certain states returning unused enhanced federal UI supplement” funds, and an unspecified amount from recovering “fraudulently-paid benefits from enhanced federal UI supplement.”

Reallocating these unspent COVID-19 funds is a bipartisan admission of how much of the COVID-19 relief aid was poorly targeted or wasted, as in the case of fraudulently-paid funds. The GAO has indicated that there is as much as $1 trillion in unspent, and debt-financed, COVID-19 relief funds still on the books.   

The plan contains $161 billion in non-tax revenue raisers from user fees and asset sales, including:

• $20 billion from the sale of future spectrum auctions
• $67 billion from the proceeds of the February 2021 c-band auction
• $28 billion from applying information reporting requirements to cryptocurrency
• $21 billion from extending fees on government-sponsored enterprises (GSEs)
• $13 billion from reinstating certain Superfund fees
• $6 billion from extending customs user fees
• $6 billion in sales from the Strategic Petroleum Reserve

While the reporting requirements for cryptocurrency are not technically a tax increase, they are intended to reduce tax avoidance on transactions of cryptocurrency. Extending fees on GSEs (Fannie Mae and Freddie Mac) is fine, but lawmakers would do better to privatize these quasi-government entities and make them full taxpaying companies.

The bipartisan infrastructure plan is not the first such plan to use the proceeds from spectrum auctions and sales from the Strategic Petroleum Reserve (SPR) as payfors. However, lawmakers could have been bolder. Since the U.S. is now an exporter of oil, the SPR is a relic of the 1970s self-imposed oil shocks and could be sold off entirely.

Asset sales are a legitimate means of raising revenues and should be used more often and expanded to include the vast inventory of federal government assets. This should include federal lands, buildings, equipment, and federally run utilities such as the Tennessee Valley Authority and Power Marketing Administration.

The plan also assumes $56 billion in new tax revenues generated by the economic growth resulting from the new infrastructure spending. This amount, which reportedly came from CBO, is roughly in line with the Tax Foundation’s estimate of a similarly sized spending bill. CBO assumes that public investments have an economic return of 5 percent, which Tax Foundation economists also assumed.

However, considering that most of the new spending is for repairs to existing infrastructure and subsidies to loss-making enterprises such as Amtrak and mass transit systems, one should question whether a 5 percent return on these investments is too much. Compared to the productivity enhancing benefits of building a new highway in a third-world country without paved roads, repairing a bridge in Manhattan has much less of an effect on productivity. While some of these investments are needed, they are unlikely to spark an economic renaissance as supporters claim.  

$550 Billion in New Infrastructure Spending over 5 Years

Over the next five years, the bipartisan plan packs more than $100 billion per year in new spending in a dozen different programmatic areas. These include:

• $110 billion for roads, bridges, and major projects
• $66 billion for passenger and freight rail
• $11 billion for highway and pedestrian safety programs
• $39.2 billion for public transit
• $65 billion for broadband
• $17.3 billion for ports and waterways
• $25 billion for airports
• $55 billion for water infrastructure
• $73 billion for power and the grid
• $46 billion for cybersecurity and “resiliency” programs
• $7.5 billion for electric vehicle programs and charging stations
• $7.5 billion for electric school buses and ferries

To be sure, many of the roads, bridges, and highways in America are in need of repair. Ideally, these projects should be funded by increasing the federal gas tax, which would maintain the user-pays principle that has guided highway funding for decades. Debt-financing these projects by using unspent COVID-19 funds undermines the user-pays principle.

We would also like to see lawmakers take steps to reduce the cost of these projects by repealing the Davis-Bacon Act, which inflates the cost of federal projects by requiring “prevailing wages” on construction contracts. The Davis-Bacon Act was passed in 1931 ostensibly to prevent black workers, mostly from the South, from competing with the Northern construction trades. It is an anachronism and should be repealed. CBO says repeal could save as much as $1.5 billion in outlays per year.  

The White House is hailing the plan as “the largest federal investment in passenger rail since the creation of Amtrak,” but lawmakers should be asking instead, “why is the government still running a passenger rail company?”  Amtrak was created in 1970 and has yet to run a profit despite billions in annual subsidies. Lawmakers should be looking to privatize Amtrak, not throw good money after bad.

The White House is also boasting that the plan ensures every American has access to reliable high-speed internet “just as the federal government made a historic effort to provide electricity to every American nearly one hundred years ago.” The red flag with this initiative is that those Depression-era programs—such as the Rural Electrification Administration (REA), Tennessee Valley Authority (TVA), and Power Marketing Administration—are still with us today despite the fact their original mission was completed decades ago.

A much cheaper solution is to give poor people broadband vouchers—the internet version of food stamps—to allow them to access private broadband service. The government should not get into the business of providing broadband service. Otherwise, this broadband initiative will become the REA of the 21^st century.

Similarly, lawmakers should be doing more to encourage private investment in airports, ports, and waterways, not spending billions of taxpayer dollars. As my colleague Alex Muresianu has written, the best way of encouraging private investment in infrastructure is by making bonus expensing permanent and improving the expensing of buildings and structures.

Conclusion

Lawmakers should be applauded for finding common ground on such an important issue as infrastructure spending, and doing so without raising harmful taxes, such as the corporate tax rate. However, roughly half of the package is effectively debt-financed with unspent funds from the various COVID-19 relief bills. This is a telling sign about how much of the spending approved in those bills was unnecessary.

While much of the other offsets in the package are from reasonable asset sales and user fees, the bipartisan plan undermines the long-standing user-pays principle that has traditionally funded roads and highway spending. Doing so makes spending on roads and highways no different than any other federal transfer program.

Finally, the package is full of missed opportunities to cut costs and turn tax-subsidized entities such as Amtrak and the TVA into private, tax-paying entities. And lawmakers should be wary of launching a new broadband program such that members of Congress 100 years from now wonder why such an obsolete program was started in the first place.