Connect with us

Hi, what are you looking for?

Editor's Pick

Comparing the Cost of IRA Energy Tax Credits to Expensing

Adam N. Michel

The House Republican debt ceiling package repeals many of the Inflation Reduction Act (IRA) energy tax credits. Revised estimates of the tax credits confirm that their costs will be at least two times higher than originally thought.

Putting aside the countless problems with each of the specific tax credits, it’s worth comparing the size of the special interest IRA tax subsidies to other more beneficial tax changes Congress could have made. For example, the energy credits could end up costing almost twice as much as making full expensing permanent, the most pro‐​growth investment incentive in the 2017 Tax Cuts and Jobs Act.

Costs Revised Up

When passed in August 2022, congressional scorekeepers estimated that the energy tax credits in the IRA would cost $271 billion dollars over ten years. Since then, Goldman Sachs, Credit Suisse, The Brookings Institution, and The Mercatus Center have all published estimates showing that the cost of the IRA tax credits could be two to three times larger than initially projected. Goldman and Brookings put the estimated ten‐​year cost at over $1 trillion.

In an updated preliminary score to repeal most of the IRA tax provisions, the JCT now estimates that repealing the tax credits would raise $570 billion, doubling their original estimate. Notably, the score does not include the cost of some of the subsidies for electric vehicles. Goldman estimates the total for all EV credits could cost as much as $390 billion over the decade, which is more than 27 times larger than the originally estimated cost.

The new higher cost estimates are the result of different assumptions about how the credit rules will be written by the Treasury Department and demand for incentive‐​eligible products. Because most of the credits are uncapped, the costs can increase significantly if eligibility rules are interpreted broadly and if consumer demand for the subsidized products is higher than anticipated. Accurately projecting these variables is challenging, if not impossible.

For proponents of the credits, the ballooning cost could be interpreted as a good thing, signifying additional adoption of government‐​favored technologies. However, it could also be the result of gamesmanship, where industry has found a way (through lobbying Treasury or other creative means) to access incentives that Congress did not intend. For example, companies are encouraging customers to lease instead of buy new electric vehicles to sidestep restrictions on tax credits if the car is purchased by an individual.

The high fiscal cost of industry‐​specific tax credits may come at the expense of extending or expanding neutral tax policies that benefit all Americans. Neutral tax policies create broader economic benefits, including for IRA‐​targeted industries.

A Lost Opportunity

Under current law, effective tax rates on U.S. investment are scheduled to increase as immediate expensing for new investments phases out and other changes take place. For example, the effective marginal tax rate on capital income from business equipment is set to increase by more than 200 percent between 2021 and 2027, according to the Congressional Budget Office. If Congress does not stop these tax increases, it will make all types of new investment more expensive and businesses will do less investing, including the types of investment necessary for green energy production targeted by the IRA.

Two significant tax changes start to happen in 2022 and 2023. Beginning in 2022, research and development spending—including related wages and other associated costs—cannot be deducted from revenues immediately. Instead, the costs must be amortized over five years. By spreading these deductions out over time, the most innovative U.S. businesses will face higher tax rates because they will only be able to recoup as little as 83 percent of their real research costs (investment deductions are eroded by inflation and time). Similar changes start in 2023 for the remainder of business investments as 100 percent bonus depreciation or “full expensing” begins to phase out. Costs that were immediately deductible following the 2017 reforms now must be spread out over as many as twenty years.

As I explained in detail recently, expensing makes it easier to invest in tools, equipment, and research that allow Americans to produce new goods and services and earn higher incomes for their work. Expensing also creates an incentive for businesses to re‐​shore their operations without the economic costs of more heavy‐​handed industrial policies or industry‐​specific tax credits. By lowering the cost of new investments made in the United States, expensing is a powerful incentive that boosts overall levels of investment and economic growth. In general, expensing offers the most economic benefits for the budgetary cost of the policy.

Making R&D expensing and 100 percent bonus depreciation permanent would lower revenues by about $582 billion over a decade, according to Tax Foundation estimates. The lost revenue is about $150 billion less after accounting for the economic benefits of more investment and the resulting larger economy.

The IRA energy credits could cost between $300 billion and $760 billion more than originally estimated when Congress passed the law. In other words, Congress could have stopped the limits on R&D spending and expensing for $170 billion less than the upper‐​bound unanticipated cost increases of the IRA credits. The updated score from JCT shows that the IRA credits would all but fully cover the lost revenue from full expensing reforms.

Figure 1 shows the original and updated JCT estimates of the IRA energy credits, the Goldman IRA cost estimate, and the Tax Foundation cost estimates for R&D and expensing. Compared to the Goldman estimate, expiring technology‐​neutral expensing rules reduce revenue by a bit more than half the cost of the energy credits.

The IRA reveals Congress’ preference for the heavier hand of industry‐​specific subsidies over the proven incentives of neutral policy changes that encourage investment of all types. Lest you think this is just a preference among Democrat lawmakers, the bipartisan CHIPS and Science Act similarly authorized about $280 billion in industry‐​specific subsidies. Despite R&D being a key component of advanced chip manufacturing, Congress chose not to include the legislative fix to the 2022 R&D amortization requirement that disincentivizes all U.S. R&D spending.

Congress chose to subsidize specific politically popular technologies and leave in place scheduled tax increases on all other forms of domestic investment. The type of targeted subsidies used in the IRA have historically failed to create the desired economic transformations as policymakers don’t have the knowledge and expertise to direct sustainable innovation. Because the IRA credits are more akin to direct government spending, and expensing is a broad‐​based tax cut that removes current disincentives to invest, comparing the two is a little apples to oranges. However, when deficits are high, and Congress faces tough fiscal tradeoffs, they are ultimately making decisions between subsidies, such as those in the IRA, or broad‐​based tax cuts that benefit everyone.

Repealing the IRA provisions would provide hundreds of billions of dollars in savings that could go toward deficit reduction and maintaining a pro‐​growth tax code.

You May Also Like

Editor's Pick

Because this series is about the New Deal’s contributions to economic recovery, it’s essential that we recognize the difference, as Roosevelt himself did, between...

Editor's Pick

Neal McCluskey The 1990 book Politics, Markets, and America’s Schools, by John Chubb and Terry Moe and published by the Brookings Institution, had a major...

Tech News

Photo by Tim Stevens for The Verge Lucid, the EV startup behind the Air sedan that competes with the Tesla Model S, has announced...

Tech News

Illustration: Alex Castro / The Verge Google’s dark web reporting feature — a service that allows users to run one-off scans to see if...