��������

Key Takeaways

• The U.S. risks falling behind the rest of the world in innovation because tax incentives are much more limited than those offered by foreign nations.
• The largest U.S. innovation tax incentive was significantly weakened by a 2022 law change.
• Another tax incentive, the U.S. R&D tax credit, suffers from three design flaws and needs to be reformed.
• Policymakers should act quickly and consider policies that both simplify and expand U.S. innovation tax incentives.

Although research and development are critical to economic growth and global competitiveness, U.S. companies underinvest in R&D (Solow 1957; Romer 1990).  The high-cost, high-risk nature of R&D spending, coupled with the fact that some of the benefits accrue to other companies when the innovation hits the market (Lucking, Bloom, and Van Reenen 2019), helps explain the underinvestment.

The federal government has tried to spur innovation and encourage R&D investment through a series of tax incentives. As shown in the figure below, a survey of corporate executives conducted by co-author Mary Cowx highlights that R&D-related tax incentives are among the most influential factors shaping R&D investment decisions.

Figure 1: Financing Factors and R&D Spending

This figure depicts survey responses to the question: “How much do the following factors affect the amount your company spends on R&D?” (N = 116)  The responses are grouped by whether the factor is “not at all” or "at least a little" important.

But measured against the tax incentives of other countries, the U.S. tax policies are lagging. The figure below presents OECD data on the value of R&D tax subsidies available for large, profitable companies in countries around the world. Twenty years ago, the U.S. provided a similar amount of incentives as other OECD countries. Now, however, the current level of U.S. incentives is roughly 20 percent of the OECD average and less than 10 percent of what China offers. This wedge between the U.S. and China grew in 2022, due to both a substantial reduction of U.S. benefits in a 2022 tax policy change and the expansion of the Chinese “super deduction.” 

Without reforms to restore and enhance R&D-specific tax incentives, the U.S. risks losing ground in global innovation leadership.

This policy brief describes and discusses key concerns with current U.S. innovation tax policies; compares the U.S. approach to other countries’ policies; and urges policymakers to take action to safeguard America’s global leadership in innovation.

Figure 2: Implied R&D Tax Subsidy

This figure depicts the implied R&D subsidy by country and the average implied R&D subsidy computed across all OECD countries. 

Challenges with current U.S. innovation tax incentives

There are two types of innovation tax incentives: input- and output-based incentives. Input-based incentives are tied to amounts spent on investing in innovation, including R&D deductions and credits. Output-based incentives provide lower tax rates on income earned from a firm’s innovation assets — a system popular among several European countries and known as “patent boxes.”

The U.S. primarily has input-based incentives. The two largest U.S. R&D tax incentives are the R&D tax deduction and the R&D tax credit. Together, these benefits provided U.S. companies with more than $100 billion in tax savings in 2021, the most recent year with available data (IRS 2024).  The figure below illustrates the aggregate magnitudes of these U.S. R&D tax incentives for corporations as of 2021, based on data from the Internal Revenue Service: 

Figure 3: Tax Benefits Claimed on U.S. Corporate Income Tax Returns

This figure depicts the aggregate estimated annual tax benefits attributable to the R&D tax deduction, R&D tax credit, and Orphan Drug credit, as reported by the IRS. The amounts reflect the total deductions or credits claimed by taxpayers in 2021.

R&D tax deduction

First enacted in 1954, the R&D tax deduction allows companies to deduct qualifying R&D spending when calculating taxable income.[1]  For example, wages paid to a scientist developing a new drug formula or to an engineer designing advanced semiconductor technology have historically been immediately deductible as R&D costs. This immediate tax deduction lowers the after-tax cost of innovation and provides a simple yet economically important incentive for firms to invest in R&D. In tax year 2021, U.S. corporations deducted more than $327 billion of R&D costs on their tax returns, resulting in $69 billion in tax savings (IRS 2024).

However, recent policy changes significantly weakened this tax incentive. Prior to 2022, U.S. companies could immediately deduct 100 percent of their R&D expenditures in the year incurred.  This deduction significantly reduces the after-tax cost of innovation. However, beginning in 2022, U.S. companies are now required to spread the deduction for domestic R&D expenditures over a five-year period, meaning they can deduct only 10 percent of their R&D costs in the year of the investment.

Delaying tax deductions significantly reduces their value, relative to immediate expensing, due to the time value of money. Our recent study (Cowx, Lester, and Nessa 2025) documents three key findings:

1. Companies were forced to defer more than $59 billion in tax benefits under this new policy, reducing the present value of R&D benefits by half. The figure below illustrates the magnitude of the tax benefits that have been deferred within our sample of large publicly traded corporations. This estimate represents cash held by the government rather than with U.S. businesses, restricting firms’ ability to invest this cash in R&D.
2. Companies cut their R&D investment in response to this new policy. The most research- intensive public companies in our sample cut their R&D by 11.6 percent in the first year alone.
3. Companies also reduce share repurchases and capital investment spending to compensate for the increased tax burdens.

Our research shows that the current policy is bad for domestic innovation. Thus far, congressional efforts to reverse this policy have failed.

Figure 4: Aggregate R&D Deferred Tax Benefits by Industry

This figure depicts aggregate capitalized R&D deferred tax assets by industry (2-digit SIC) in millions, based on hand-collected data from financial statement deferred tax asset disclosures specifically related to the R&D capitalization requirement.  We report the 15 industries with the largest R&D deferred tax assets, with a subtotal for all other industries labeled "Other Industries."

R&D tax credit

The R&D tax credit, officially called the Credit for Increasing Research Activities, was introduced in 1981 to encourage incremental R&D investment. This means that the R&D credit specifically rewards firms for increasing their R&D spending beyond a historical benchmark amount. (This differs from the R&D deduction discussed previously, which provides tax savings on all R&D expenditures.) The tax credit rate ranges from 11.1 percent to 15.8 percent of qualified expenditures exceeding the benchmark amount. 

Prior academic research finds the R&D tax credit has been effective in stimulating additional R&D investment, with each $1 of tax credit generating more than $1 of additional R&D spending (e.g., Hall 1993). In tax year 2021, U.S. corporations claimed over $32 billion in R&D tax credits. By comparison, this amount is less than half the tax savings provided by the R&D deduction ($32 billion vs. $69 billion in 2021).

Despite its success in stimulating additional spending, the U.S. R&D tax credit has several notable limitations:

• Penalty for highly innovative firms. Since the credit is based on incremental R&D spending relative to past investment levels, companies with consistently high R&D investments receive relatively smaller benefits than companies increasing R&D from a lower base.
• Complexity and compliance costs. The R&D tax credit is very complex, creating challenges for both taxpayers and the IRS. Many firms struggle to substantiate qualifying expenses, while the IRS faces significant hurdles in auditing R&D credit claims. This complexity has real economic implications — recent research by co-author Cowx documents that the risk of IRS scrutiny discourages firms from claiming the credit and also dampens their R&D investment (Cowx 2025).
• Non-refundability limits access for startups and small firms. Historically, the R&D tax credit was non-refundable, meaning only firms with positive taxable income could benefit. This structure disadvantaged startups and small innovative firms, which often operate at a loss in their early years and thus do not benefit from the tax credit. To partially address this issue, recent law changes in 2015 and in 2023 permit small businesses to now use the R&D tax credit to offset payroll taxes.  This is an important change because companies pay payroll taxes regardless of profitability.

Orphan drug credit

The U.S. also offers a targeted incentive for pharmaceutical innovation: the Orphan Drug Credit, which was enacted in 1983.  This credit supports the development of treatment for rare diseases. The Orphan Drug Credit is relatively small, providing only $2 billion in total tax savings to corporations in tax year 2021.

Foreign-derived intangible income

In addition to the input-based incentives, the U.S. has one output-based innovation incentive known as Foreign-Derived Intangible Income, or FDII.  This incentive was passed in the Tax Cuts and Jobs Act of 2017 and allows for intangible-related foreign income to be taxed at 13.125 percent rather than the regular U.S. corporate rate of 21 percent. 

The goal of the incentive is to encourage U.S. companies to retain their domestically developed intellectual property (“IP”) in the U.S., rather than move the IP to offshore countries with lower tax rates, like Ireland, Switzerland, and Singapore. Thus far, there is no conclusive evidence about the effects of FDII.  Furthermore, because the incentive applies only to income earned on foreign sales, but not domestic sales, the incentive is much less generous than comparable incentives in Europe.

Other recent incentives: The CHIPS Act and the Inflation Reduction Act

Two recently enacted laws, the Chips and Science Act of 2022 (CHIPS) and the Inflation Reduction Act of 2022 (IRA), include substantial tax incentives. While both are frequently discussed in the context of innovation incentives, these policies do not directly incentivize R&D investment. Instead, they focus on scaling up the production and adoption of existing technologies within key industries, rather than stimulating new technological advancements. For example, the IRA tax credits for renewable energy production and carbon capture in practice would benefit companies investing in things like solar panels, wind turbines, and battery storage systems.  Similarly, the CHIPS Act includes a 25 percent tax credit for facilities and equipment in domestic semiconductor manufacturing.

We summarize the estimated annual tax cost to the U.S. government of all of these policies in the table below, which uses estimates from the Joint Committee on Taxation (2024).

Figure 5: Estimated Annual Tax Cost to the U.S. Government

The bars report the estimated annual tax cost to the U.S. government as reported by the U.S. Joint Committee on Taxation (JCT).  The first three bars relate to input-based R&D tax incentives (the R&D deduction estimated based on pre-2022 tax law; the R&D tax credit; and the Orphan Drug Credit); Output-based R&D incentive (Foreign Derived Intangible Income benefit); and incentives contained in the Inflation Reduction Act (various energy and manufacturing credits) and in the CHIPS Act (Advanced Manufacturing Credit) for capital investment. Amounts are obtained from the 2024 JCT estimates, except for the R&D deduction, which is estimated from 2017 JCT estimates related to the change in the R&D deduction. Specifically, the JCT estimated $24.2 billion tax revenue raised over the first nine months of the policy, or $32.3 billion annually.  This translates to $170.7 billion of total R&D spending impacted by the change ($32.3 billion of foregone tax benefits divided by 21% = $153.8B of spending.  As this is 90% of the total spending, because 10% was deducted in the first year, this is equivalent to total R&D spending of $170.7 billion ($153.7B/90%).  If instead the $170.7 billion were immediately deducted, it would result in a tax cost to the government of $36 billion ($170.7 x 21%). 

While U.S. tax policies aim to stimulate private-sector R&D investment, recent changes and the structural limitations discussed above have weakened their effectiveness and threaten U.S. competitiveness.

Comparison of U.S. policy to global policies

While the U.S. offers R&D tax deductions and credits, many other countries have implemented more generous policies that provide larger and more immediate tax benefits to innovative companies. Two policies that are gaining traction internationally are “innovation box regimes" and “super deductions.”

• Innovation box regimes. Innovation box regimes (sometimes called patent boxes) lower the tax rate on profits earned from innovation-related intellectual property, such as patents and proprietary technology. Similar to the U.S. FDII, these are output-based incentives because they reward the income generated from innovation, rather than the investment that led to it.  Countries such as the Netherlands and the United Kingdom have adopted this approach. The U.K.’s patent box, for example, reduces the corporate tax rate on qualifying innovation income from 25 percent to just 10 percent — a cut of 60 percent. 
  
  The goals of these regimes are threefold: to increase innovation; to increase economic activity through the co-location of investment and employment; and to retain income that companies would otherwise shift to lower-taxed jurisdictions. On the first goal, research generally shows that these incentives are associated with increased innovation, although questions remain about whether companies simply move patents across borders to claim the benefit. For the second goal, finds that innovation box regimes are associated with increased capital investment and higher compensation paid per employee (Chen et al. 2023).  Whether the incentives achieve the third goal remains an open question.
• Super deductions. While the U.S. tax code currently requires R&D costs to be deducted over five years, some countries offer super deductions that allow firms to deduct more than the amount they spend on R&D. These incentives are designed to significantly lower the after-tax cost of R&D investment. For example, China and Brazil allow companies to deduct up to 200 percent of their qualifying R&D expenditures; in China, in particular, the super deduction has dramatically expanded in scope and benefit over the last ten years. Currently, a company that spends $1 million on R&D in China receives a $2 million tax deduction in the year of the investment. In stark comparison, under current U.S. rules, that same $1 million of R&D spending results in a deduction of only $100,000 in the year of investment, with the deduction for the remaining $900,000 spread over five years at a lower present value. Super deductions create powerful upfront tax savings, reducing the cost of innovation and encouraging companies to invest more.

Strengthen U.S. innovation incentives to compete globally

The U.S. faces two pressing challenges to the competitiveness of its innovation tax incentives. 

First, domestic tax policy must be modernized, with a particular focus on the R&D capitalization and amortization requirement introduced in 2022. Delaying R&D deductions raises the cost of innovation, discourages private-sector investment, and weakens the incentive structure that has supported U.S. technological leadership for decades. 

Second, even if immediate R&D expensing is restored, U.S. innovation incentives remain far less generous than those provided by other OECD countries and China. Competitor nations offer super deductions, refundable credits, and generous innovation box regimes — tools that more effectively lower the after-tax cost of R&D and help retain innovation profits domestically. Without reform, the U.S. risks falling further behind in the global race for innovation-driven growth.

Adding to this challenge, the OECD’s Pillar Two Global Minimum Tax framework introduces new tax pressures on U.S. multinationals. Under Pillar Two, nonrefundable tax credits — like the U.S. R&D credit — are treated less favorably than refundable credits. This could trigger additional tax liabilities for U.S. firms operating globally, further diminishing the competitiveness of U.S. R&D tax policy.

The U.S. must act swiftly to restore and enhance its R&D tax incentives — both to support domestic innovation and to ensure that U.S. companies remain competitive. Failure to act risks ceding America’s innovation leadership to countries that prioritize R&D investment.

Footnote

[1] As defined under §174 of the Internal Revenue Code.

References

• Chen, S., L. De Simone, M. Hanlon, and R. Lester. 2023. "The effect of innovation box regimes on investment and employment activity." The Accounting Review, 98(5), pp. 187-214.
• Cowx, M. 2025. Tax enforcement and R&D credits. Journal of Accounting and Economics, 101784.
• Cowx, M., R. Lester, and M. Nessa. 2025. The Consequences of Limiting the Tax Deductibility of R&D. Working paper.
• .
• Hall, B.H. 1993. "R&D tax policy during the 1980s: success or failure?" Tax policy and the economy, 7, pp. 1-35.
• .
• Lucking, B., N. Bloom, and J. Van Reenen. 2019. "Have R&D spillovers declined in the 21st century?" Fiscal Studies, 40(4), pp. 561-590.
• .
• Romer, P. M. (1990). Endogenous Technological Change. Journal of Political Economy, 98(5), pp. S71-S102.
• Solow, R. M. (1957). Technical Change and the Aggregate Production Function. The Review of Economics and Statistics, 39(3), pp. 312-320.