How does tax on earned income affect Swedish innovative power?

This report examines the relationship between income taxes and innovation in Sweden, with a particular focus on the "Tax Reform of the Century" in 1990/91. The report includes a comprehensive literature review identifying key mechanisms through which taxes influence innovation. These mechanisms include tax incentives for investments in research and development (R&D) and individuals’ efforts and reward systems.

The review shows that prior research has primarily focused on corporate taxes and R&D tax incentives, while significantly fewer studies explore how taxes on earned income affect individuals’ propensity to innovate. This highlights a knowledge gap that our study helps address.

Theoretically, taxes on earned income negatively impact innovation propensity by reducing the return on effort. However, the revenue generated by taxes can be used for societal investments that enhance productivity and innovation capacity, such as education. Finding the right balance in tax levels and designing the tax system is a political question, but developing a knowledge base to understand the various effects of taxes is crucial for informed decision-making. The impact of income tax on innovation is an important effect to include in the overall assessment of the societal cost of taxing earned income.

This report combines unique data from Swedish inventors and extensive microdata registers, enabling detailed analyses of the relationship between income tax and innovation across different societal groups. By leveraging the tax reform as a natural experiment combined with an instrumental variable (IV) strategy, the study examines how changes in tax rates on labor income causally affect innovation, measured through patent applications. Two groups closely examined are women, who are significantly underrepresented as inventors, and foreign-born individuals, who have become an increasingly larger share of the population and, as the study shows, among inventors.

By using detailed data on both individuals' incomes and patent applications, the report provides new insights into how marginal taxes on earned income influence individuals' propensity to innovate. The results show that lower marginal taxes had a positive and statistically significant effect on innovation. The terms patent application and innovation are used synonymously, although the analyses of patent applications do not account for patent quality or whether they actually lead to innovation outcomes—this requires further development of the underlying patent database, which the report discusses.

Main results

The results of the report can be summarized as follows:

• Positive effect of the tax reform on innovation. The results clearly indicate that the tax reform contributed positively to innovation development. Lower marginal taxes had a positive and statistically significant effect on innovation. Based on the average tax change and its percentage impact on patent development, we calculate an innovation elasticity of 0.9. This means that a one percent reduction in the tax rate results in a 0.9 percent increase in patenting, consistent with previous estimates from the literature using other methods and based on U.S. data. Given that taxes were reduced by an average of over 20 percent for the average wage earner, this implies that approximately 250 additional patents were filed during 1991–94 due to the tax reform. This can be compared to a total of just over 5,545 patents filed by Swedish inventors during the same period, according to the database.
• No clear innovation effects for foreign-born individuals. Estimates for foreign-born individuals are positive but not statistically significant, meaning we cannot rule out that the effects are zero.
• Female innovators were unaffected. Women, who make up less than 10 percent of inventors, showed no clear effect on their innovation rate from the tax reform.
• Strong effects for individuals with innovation-prone education. Individuals with university-level education in technology, natural sciences, and medicine—common among inventors—tended to innovate significantly more after the tax reform.
• Somewhat less clear effects for individuals with innovation-prone occupations. Analyses of individuals in occupations common among inventors before the tax reform show that they, like those with relevant education, also tended to innovate more. However, the effects were somewhat less pronounced than for those with innovation-prone education.
• Innovation effects only among high-income earners. Clear effects are found only in the highest income bracket. The analyses show that among high-income earners were individuals with the greatest potential to innovate, suggesting that the highest marginal tax rates (72 percent in 1989) had a clear negative effect on the willingness to exert effort to invent. This latent potential can thus be said to have been unleashed by the tax reduction. Our estimates suggest that an additional 250 patents were filed in this group during 1991–94.

A potential limitation of the study is that we cannot rule out that it also captures other effects of the tax reform beyond the reduction in income taxes. For example, corporate taxes were reduced simultaneously, increasing companies’ net profits, which may have led to higher wages for employees, particularly for highly productive and innovative workers, and possibly encouraged innovation in corporate forms. Sensitivity analyses, however, suggest that entrepreneurs are not driving the effect. Thus, it seems unlikely that other tax reductions had direct effects, although indirect effects cannot be ruled out, such as companies introducing bonus systems for innovators as a result of corporate tax changes. This strengthens the causal interpretation of the estimates.

The results emphasize the importance of considering tax effects on innovation when designing tax policy, particularly in knowledge-intensive economies like Sweden. The study also contributes to a deeper understanding of how different societal groups are affected by tax reforms and how this, in turn, impacts the rate of innovation.

Our literature review highlights the broader development of the research field and identifies key areas for future studies, including the need for more analyses of how tax policy changes affect individuals in different societal groups. This complements the empirical analysis and broadens our understanding of the relationship between taxes and innovation.

The study’s results align with other negative effects of high marginal taxes found in the literature, such as effects on labor supply, education, effort, and tax avoidance. Our study shows that, in addition to these relatively well-known effects, there is also a significant negative effect on innovation propensity, particularly among high-income earners.