bypass breadcrumb links
  InfoBrief SRS Logo  
  NSF 05-316 | July 2005   PDF format PDF format | See Related Reports  

The U.S. Research and Experimentation Tax Credit in the 1990s

by Francisco Moris Send an e-mail message to the Francisco Moris

Industrial R&D and the policy tools that stimulate it are increasingly important, owing to the role of R&D in economic growth (Smith and Barfield 1996). In recent decades, industry has funded and performed a growing share of R&D in the United States (NSB 2004). The new knowledge that results from R&D and other innovation activities, however, is often intangible and may benefit its users without fully compensating its producers. In such cases, private incentives for R&D are diminished, resulting in levels of R&D that may not maximize potential economy-wide or social benefits.

Various policy tools are needed in fostering R&D because of the diversity of R&D practices across industries, technologies, and innovation environments (Martin and Scott 2000, Tassey 1996). Fiscal policy tools used by the U.S. federal government include direct funding, such as grants or contracts, and indirect incentives, such as tax credits or allowances.[1] U.S. R&D tax credits, and those in other advanced economies, vary in how they are structured or targeted, their effect on public budgets, and their effectiveness in stimulating innovation (Bloom et al. 2002, OECD 2003).

One of the better-known indirect federal incentives is the research and experimentation (R&E) tax credit for corporations, examined here. U.S. corporate claims for the R&E tax credit totaled an estimated $6.4 billion in 2001, the latest year for which data are available. From 1990 to 2001 the annual dollar amount of R&E tax credit claims grew twice as fast as company and other nonfederally funded R&D expenditures (industry-funded R&D), after adjusting for inflation. In contrast, direct federal funding for industrial R&D declined throughout the 1990s, both in absolute terms and relative to industry-funded R&D.

This InfoBrief examines R&E tax credit data from the Statistics of Income (SOI) program of the U.S. Internal Revenue Service (IRS) and selected data on R&D funding from the National Science Foundation (NSF). For tax purposes, R&D expenses are restricted to research and experimental expenditures (see "Data Notes" for more detail). Federal funding for industrial R&D performers can be accounted for by either federal outlays from agency sources or by R&D expenditures data from industrial recipients. The latter set of funding statistics is used here.

R&E Tax Credit History and Structure

The R&E tax credit, part of the U.S. Internal Revenue Code, was established by the Economic Recovery Tax Act of 1981.[2] The credit was one of several policy tools put in place in the 1980s to address perceived problems in the competitive position of U.S. companies (Guenther 2005).[3] It is subject to periodic extensions given its temporary status (it was allowed to expire between July 1995 and June 1996). The credit was renewed most recently by the Working Families Tax Relief Act of 2004 (Public Law 108-311) through 31 December 2005. The Bush administration and several congressional bills propose making the R&E tax credit permanent (Guenther 2005, OMB 2005).

The credit is designed to stimulate company R&D over time by reducing after-tax costs. Specifically, companies that qualify for the credit can deduct or subtract from corporate income taxes an amount equal to 20 percent of qualified research expenses above a base amount.[4] For established companies, the base amount depends on historical expenses over a statutory base period relative to gross receipts, whereas start-up companies follow other provisions (Guenther 2005).

An alternative R&E tax credit has been available since 1996 (Small Business Protection Act, Public Law 104-188). This credit has a lower base amount and a maximum statutory rate of 3.75 percent. The alternative credit benefits established companies that have smaller annual increases relative to their base period (Hall 2001). Companies may select only one of these two credits on a permanent basis, unless the IRS authorizes a change. Both credits include provisions for basic research payments paid to qualified universities or other research organizations.

Trends in the R&E Tax Credit and Industrial R&D Funding

According to SOI/IRS, R&E tax credit claims reached an estimated $6.4 billion in 2001 ($6.2 billion in 2000 constant or inflation-adjusted dollars), compared with a high of $7.1 billion in 2000 (table 1).

Table 1. R&E tax credit claims and U.S. corporate tax returns claiming the credit: 1990–2001.
  Table 1 Source Data: Excel file

From 1990 to 2001 the average annual growth rate for claims was 11 percent, compared with 5 percent for company-funded R&D expenditures (table 2), after adjusting for inflation. The number of corporate tax returns claiming the credit grew at a slower rate than claims, fluctuating between 8,000 and 10,000 over most of the 1990s. From 1990 to 1996 companies claimed between $1.5 billion and $2.4 billion in R&E tax credits annually; since then annual claims have exceeded $4 billion (table 1). Even so, R&E tax credit claims accounted for less than 4 percent of industry-funded R&D expenditures as of 2001 (figure 1).

Table 2.  U.S. industrial R&D expenditures, by source of funds: 1990–2001.
  Table 2 Source Data: Excel file


Figure 1. Research and experimentation tax-credit claims as a percentage of industry-funded R&D expenditures: 1990–2001.
  Figure 1 Source Data: Excel file

In contrast to trends in corporate R&E tax credits, direct federal funding for industrial R&D declined throughout the 1990s, both in absolute terms (table 2) and relative to industry-funded R&D (figure 2).

Figure 2.  Federal funds for industry R&D as percentage of industry-funded R&D expenditures: 1990–2001.
  Figure 2 Source Data: Excel file

Estimates at the industry level for R&E tax credit claims and R&D expenditures and are not strictly comparable because of methodological differences (see "Data Notes"). However, industry-level trends in these indicators can be discussed separately. Table 3 shows corporate claims for the R&E tax credit from 1998 to 2001 by industry classification, using the North American Industry Classification System (NAICS). Since 1998, corporate tax returns classified in five industries have accounted for 80 percent or more of R&E tax credit claims, and in 2001 these industries accounted for $5.1 billion of the total $6.4 billion in claims:

  • computer and electronic products (26 percent)
  • information, including software (16 percent)
  • chemicals, including pharmaceuticals and medicines (16 percent)
  • transportation equipment, including motor vehicles and aerospace (12 percent)
  • professional, scientific, and technical services, including computer services and R&D services (10 percent)
Table 3.  R&E tax credit claims and U.S. corporate tax returns claiming the credit, by selected NAICS industry: 1998–2001.
  Table 3 Source Data: Excel file

The same five industries accounted for two-thirds of company-funded R&D expenditures from the NSF Survey of Industrial R&D in 2001. Federal funding for industrial R&D is even more concentrated: three industries—computer and electronic products; professional, scientific, and technical services; and transportation equipment—accounted for 94 percent of federal funding for industrial R&D in 2001.

Companies in the professional, scientific, and technical services industry filed more corporate tax returns claiming the R&E tax credit than did any other industry in 2001 (figure 3). That industry represented about 28 percent of all returns claiming the credit, followed by computer and electronic products and information, each with about 15 percent.

Figure 3.  Industries with largest R&E tax-credit claims and number of corporate tax returns claiming the credit: 2001.
  Figure 3 Source Data: Excel file

Across all industries, the average R&E tax credit claim per corporate tax return in 2001 was largest for returns classified in motor vehicles, trailers, and parts ($7.7 million per return), followed by pharmaceuticals and medicines ($2.7 million per return) and aerospace products and parts ($2.5 million per return). Among nonmanufacturers, returns classified in management of companies and enterprises; broadcasting and telecommunications; and finance, insurance, and real estate had the largest average claim per return.

Data Notes

Data reported here are from the NSF Survey of Industrial R&D and from SOI/IRS. The NSF Survey of Industrial R&D expenditures ( is based on a nationally representative sample of all for-profit companies in the United States with five or more employees. Data are collected on a calendar year basis. Industry classification has been based on NAICS since the 1999 survey. Estimates are subject to sampling and non-sampling errors. For a description of the survey methodology see

Estimates on the R&E tax credit were obtained from special tabulations of SOI/IRS. Companies requesting the credit must complete IRS Form 6765 ( Qualified costs include company-funded expenses for wages paid, supplies used in the conduct of qualified research, and certain contract expenses. For tax purposes, R&D expenses are restricted to the somewhat narrower concept of research and experimental expenditures. Such expenditures are limited to experimental or laboratory costs aimed at the development or improvement of a product (defined to include any pilot model, process, formula, or technique) in connection with the taxpayer's business (26 U.S. Code of Federal Regulations [CFR] 1.174).[5] Further, tax-credit regulations define qualified research expenses as a subset of research and experimental expenses that satisfy additional tests involving the experimental and technological nature of the activities (26 CFR 1.41-2; Oliver 2003). In February 2004 the IRS published final regulations on the definition of qualified research for purposes of the tax credit, reflecting changes to the tax code made by the Tax Reform Act of 1986 (IRS 2004).[6]

Data reported here exclude IRS forms 1120S (S corporations), 1120-REIT (Real Estate Investment Trusts), and 1120-RIC (Regulated Investment Companies. The R&E tax credit covers activities performed in the United States by domestic and foreign-owned firms but excludes those conducted abroad by U.S. companies. The Tax Relief Extension Act of 1999 extended the R&E tax credit to include qualified expenses in Puerto Rico and the U.S. possessions.

Estimates on corporate tax statistics are based on a stratified probability sample of unaudited returns selected from a population of active returns. Active corporate returns include returns having current income or deductions. IRS data are for tax years, which cover accounting periods ended any month between July of the calendar year of reference through June of the following calendar year. In tax year 2001 more than 80 percent of all corporate tax returns (not just the returns claiming the R&E tax credit) were calendar-year returns, or returns with accounting periods ending in December 2001. Industry classification has been based on NAICS since tax year 1998.

Estimates are subject to sampling and non-sampling errors. For a full description of the IRS statistical methodology see

Although both IRS and NSF statistics use NAICS as the underlying industry classification system, comparisons of R&D-related estimates at the industry level are problematic due to differences in methodology. For example, the assignment of industry codes for tax purposes is based on gross receipts, whereas the classification in the NSF survey is based on dollar payrolls.


Billings, B.A., S.N. Glazunov, and M. Houston. 2001. The role of taxes in corporate research and development spending, R&D Management, 31(4):465–477.

Bloom, N., R. Griffith, and J. Van Reenen. 2002. Do R&D tax credits work? Evidence from a panel of countries 1979–1997. Journal of Public Economics 85:1–31.

Guenther, G. 2005. Research and Experimentation Tax Credit: Current Status Legislative Proposals, and Policy Issues, Washington, DC: U.S. Congressional Research Service.

Hall, B. 2001. Tax Incentives for Innovation in the United States. Available online at

Internal Revenue Service (IRS). 2004. Treasury decision 9104, Credit for Increasing Research Activities, Internal Revenue Bulletin: 2004–6. Available online at

Martin, S., and J.T. Scott. 2000. The nature of innovation market failures and the design of public support for private innovation, Research Policy 29:437–447.

National Science Board (NSB). 2004. U.S. and International Research and Development: Funds and Technology Linkages. In Science and Engineering Indicators 2004, Volume 1, NSB 04-1. Arlington, VA: National Science Foundation.

Office of Management and Budget (OMB). 2005. Analytical Perspectives: Budget of the United States Government, Fiscal Year 2006. Washington, DC: U.S. Government Printing Office. Available online at

Oliver, J.R. 2003. Accounting and tax treatment of R&D: An update, CPA Journal (July). Available online at

Organisation for Economic Co-operation and Development (OECD). 2003. Tax Incentives for Research and Development: Trends and Issues, Paris.

Smith, B.L.R., and C.E. Barfield (eds.). 1996. Technology, R&D, and the Economy. Washington, DC: The Brookings Institution and American Enterprise Institute.

Tassey, G. 1996. Choosing R&D policies: Tax incentives vs. direct funding, Review of Industrial Organization, 11(5):579–600.

For more information, contact

Francisco Moris
Research and Development Statistics Program
Division of Science Resources Statistics
National Science Foundation
4201 Wilson Boulevard, Suite 965
Arlington, VA 22230


[1] Both tax incentives and direct federal funding represent federal expenses. In terms of the budget, tax incentives generate tax expenditures—government revenue losses due to tax exclusions or deductions. For estimates of tax expenditures arising from the R&E tax credit see OMB (2005).

[2] Section 41 of the Internal Revenue Code (U.S. Code of Federal Regulations, Title 26).

[3] An older indirect incentive for which there are no data available is the expensing allowance for qualified R&E (Internal Revenue Code, Section 174), in place since 1954 (Guenther 2005). The R&E tax credit covers a more restricted subset of qualified expenses compared to the R&E expensing allowance.

[4] The effective rate is considered to be lower than this statutory rate, in part because the credit reduces the expenses available for deductions under the R&E expensing allowance (Billings et al. 2001, Guenther 2005).

[5] In contrast, R&D expenditures for purposes of NSF surveys include basic research, applied research, and development. However, similarities also exist. In particular, both the NSF Survey of Industrial R&D and IRS definitions exclude routine testing, adaptation of existing components, marketing, and research in the social sciences, arts, or humanities.

[6] These regulations do not include final rules on computer software for internal use. For guidelines on software development expenses see Revenue Procedure 2000-50, IRB 2000-52, December 26, 2000, as modified by Revenue Procedure 2004-11, IRB 2004-3, January 20, 2004.