How Effective are State R&D Tax Credits?

Over the past thirty years, state research and development (R&D) tax credits have become one of the most common state tools to boost their local technology economy. Few studies, however, have attempted to provide a comprehensive evaluation of their effectiveness. In this article, SSTI provides an overview of the literature on state R&D credits, examining why states introduce these incentives and whether or not tax credits are an effective policy tool to promote high-tech businesses and jobs.

The first state R&D tax credits were introduced in the early 1980s. One year after the introduction of the federal Research and Experimentation Tax Credit in 1981, Minnesota passed its own credit, followed by six other states over the next five years. Earlier this year, a review by the Texas Legislative Budget Board reported that 43 states now offer R&D tax incentives.

As the number of states offering R&D credits has increased, the generosity of the incentives has grown as well. A study by Dan Wilson of the Federal Reserve Bank of Boston found that between 1981 and 2005, the average effective credit rate in states offering R&D tax credits increased four-fold. Wilson, in a separate 2005 paper, noted that these credits tend to have a negative impact on private R&D spending in neighboring states, leading to inter-state competition over which state is able to offer the most generous incentives. A 2010 paper by Jungbu Kim concluded that states tend to offer larger incentives based on declining state revenues rather than the influence of organized private interests within the state, a conclusion that is compatible with Wilson's suggestions that these credits can influence firm location decisions.

Of course, the nature of these incentives varies from state-to-state. Many states use the federal R&D credit as a model. The federal credit allows corporations to take a credit against their tax liability equal to 20 percent of their current year's qualified R&D expenditures in excess of a base amount on their federal corporate R&D tax return. Qualified research includes research undertaken for the purpose of discovering information that is technological in nature and could be used in the development of new or improved processes or products. This includes wages paid to researchers and 65 percent of contract research expenses. Credits of 20 percent also are available for payments made to perform basic research and amounts paid to an energy research consortium. In 2009, the last year for which information is available, 12,359 companies claimed $7.8 billion in federal credits for increasing research activities.

State business tax-based R&D incentives typically follow this model. Companies pay corporate income or franchise taxes to states based on their federal taxable income and the income generated by transactions in a particular state. As with the federal credit, states often provide tax credits that may be taken against their state income or franchise tax equal to a percentage of their qualified R&D expenses over a certain base amount. The percentage varies from state-to-state, as does the formula for determining the base amount. Some states also adopt a definition of qualified R&D expenses that differs from the federal definition. According to the Texas Legislative Budget Board report, 40 states currently offer corporate tax credits for qualified research expenses. Twenty-seven states offer a separate sales tax incentive for research activities, including three that do not also offer a corporate tax credit.

States that introduce R&D tax credits tend to be motivated by a desire to leverage their research activity to create new jobs. In a 2010 article, authors Chad R. Miller and Brian Richard tracked the spread of R&D tax credits across the U.S. and the motivations behind their passage. The authors found that states generally adopt R&D tax credits out of a desire to promote economic development, particularly when the state's economy is suffering from high unemployment rates. Increasing a state's unemployment rate by 1 percent increases its likelihood of introducing an R&D tax credit by about 66 percent. States with greater research and manufacturing activity are also more likely to introduce R&D tax credits than states without a significant research base. Miller and Richard conclude that these kinds of tax incentives are used by manufacturing-intensive states to create new jobs by fostering research activity in states that already have significant R&D activity. It was unclear to the authors if states with little research activity often had introduced R&D credits with the intention of creating a new high-tech or manufacturing business community.

While new jobs appear to be the ultimate goal in most cases, state metrics for these tax credits programs also have included other criteria to evaluate their success. One of the more notable academics to take on the subject of state R&D tax credits is Yonghong Wu, who authored papers in 2005 and 2008 on the correlation between state R&D tax incentives, private R&D expenditures and high-tech business establishments. In 2008, Wu reported that the Washington legislature had established seven metrics for the success of their R&D credit: total job creation, new jobs for Washington residents, company growth, diversification of the state economy, growth in R&D investment, introduction of new products and movement of firms into the state. The diversity of these metrics is indicative of how wide-ranging the expectations are for research credits and how difficult it can be to capture the full impact of technology-based economic development incentives, particularly in the short term.

Given thirty years of history, however, it may be appropriate to ask how successful these credits have been in boosting state economies. Compared to the numerous studies that have ranked the effectiveness and generosity of national R&D tax credits around the world, relatively few have taken on a comprehensive evaluation of state credits. In his 2005 study on R&D credits and private R&D spending, Wu concluded that state R&D credit programs, in addition to other state initiatives focused on higher education and high-tech research, were effective in stimulating industrial R&D expenditure. In a 2008 article on R&D credits and high-tech business establishments, Wu focused on business creation within high-tech industries in particular, concluding that the introduction of a state R&D credit has a significant and positive effect on the number of state high-tech establishments. His analysis found that the average estimated effect of a state R&D credit was about 1.35-1.47 percent of total high-tech establishments in the state. Though this figure appears small, Wu uses the example of Washington's tax credit, initiated in 1996, to estimate that in its first year the credit induced the creation of 2,437 new high-tech jobs at a cost of about $9,000 in potential state revenues per job. Retaining one of these jobs for 20 years would cost the state an estimated $112,000 in current dollars.

In a 2006 dissertation, Yujeung Ho also defended the effectiveness of R&D credits in inducing research spending and job creation. Ho found that state R&D tax credits are effective in increasing R&D investment across all industry sectors and in both large and small companies. Positive effects on employment, however, were limited to high-tech companies, particularly larger high-tech companies. Ho thus concludes that state R&D credits can serve as a valuable component of economic strategy, depending on the local mix of industries and the size of local high-tech businesses. This mix would determine whether or not the growth induced by the credits were worth the loss in tax revenue.

Jennifer Weiner, in a 2009 New England Public Policy Center discussion paper, examined the effectiveness of many varieties of state business tax incentives, reviewing many reports produced and commissioned by states to evaluate their own incentives. Weiner finds that the available evidence suggests that state R&D credits are effective in inducing new in-state research and related employment. The cost effectiveness of R&D credits compared to other types of tax incentives, however, was less clear. Also, the reports she reviewed seem to suggest that state credits do little for the national economy. States appear to compete for corporate research activity using these incentives in a zero-sum game that does little to generate new research.

Since Weiner's report, a number of other states have undertaken reviews of their R&D tax incentives, often demonstrating that implementation matters a great deal for the effectiveness of tax credits. A December 2011 report from the Iowa Department of Revenue examined the effectiveness of the state's Research Activity Tax Credit and indicates some of the difficulty of judging a credit's success. Iowa's credit ran close to the middle of the pack in a ranking of all fifty states, except in the case of large, multi-state firms. For these firms, Iowa ranked closer to the top and surpasses all of its neighboring states. Because of this approach, more than 60 percent of all claimed research expenses in Iowa are reported by large, multi-state businesses. Three-fourths of businesses claiming the credit conducted all of their research activities in Iowa. Iowa's credit appeared to be effective in attracting these companies, but had not enticed smaller companies with the credit.

Last year, Timothy Bartik and Kevin Hollenbeck of the W.E. Upjohn Institute authored a working paper on Washington's High Technology Business and Occupation Tax Credit, finding that the tax credit had created jobs, but at the rather high cost of $40,000-$50,000 per job-year. Bartik and Hollenbeck cite several factors that appear to limit the effectiveness of the credit. The credit appears to only have a moderate effect in lowering business costs, while state business activity as a whole appears only modestly responsive to cost-lowering subsidies. More importantly, the credit is nonrefundable, as it is in most other states with R&D credits, which effectively caps its cost-lowering potential.

In light of the literature available on state R&D tax credits, it would seem that these credits can be an effective tool in a state's economic development strategy, but only when designed with a particular state's economy in mind. R&D incentives are most effective in states that already have a significant level of research activity, and a substantial high-tech business community. As seen in Iowa and Washington, the design of these credits should be tailored to fit the state's needs, whether that means attracting larger or smaller firms.

The evidence indicates, however, that state R&D tax credits are more properly categorized as a tool of conventional economic development and business attraction than technology-based economic development. States appear to compete in a zero-sum game to attract businesses with these incentives, instead of commercializing technologies through home-grown startups. This approach may complement, but not replace, state and regional innovation strategies.