Gradient power of tax responsiveness

One of the most fundamental areas of economic life is human interaction with the tax system. A vast empirical literature studies people’s response to tax policy, with a recent interest in how people understand, reason and learn the tax code (Saez et al. 2012, Bernheim and Taubinsky 2018, Stantcheva 2021). Several papers emphasize the role of acquired skills, such as education and the transmission of information among peers, but little is known about innate traits and the role of non-acquired skills.

In a new study (Bastani and Waldenstrom 2021), we analyzed whether individuals with different cognitive abilities reacted differently to a larger and more major isolation in the Swedish marginal tax rate schedule (marginal tax increase of 20 percentage points). Our data comes from administrative population registration on taxable income and cognitive ability from military enlistment at age 18.

People with higher powers respond more to taxes

Figure 1 captures an interesting stylized pattern. Given the taxpayers’ ranking relative to their taxable income, we statistically notice a significant increase in the distribution of income (left panel) and the average cognitive capacity (right panel) of contemporary cognitive ability. In other words, intelligent people are more likely to be a bunch.

Figure 1 Bunch and average cognitive ability at income tax kink point

Note:: Taxable income in 2012-2016 is thousands of Swedish kronor (EUR / SEK = 10).

To examine whether there is a systematic difference between tax responses across power levels, we divide the population into ten equal sizes of cognitive power decile (minimum decile 1 to maximum decile 10) and then estimate the bunch in each power decile.

Figure 2 shows that people with higher cognitive skills tend to be more responsive to income tax. The kink point of the Ability Decile 10 has an extra mass which is twice as much as the full population and almost three times as much as the minimum power decile. The growth is almost monotonous. We refer to this pattern as the tax gradient of tax responsiveness and this is the main result of our study.

Figure 2 Gradient power of tax responsiveness

Note:: Excess mass and 95% confidence interval (+/- 1.96 times standard error, bootstrapped with 500 replicas) estimated separately for each decile of male cognitive ability distribution (for all males in our main sample population) and for labor 2012-2016 Earnings in (pooled data). The dashed line is the approximate (average) cluster of the full male population

Transfer of income between high-powered self-employed

How can this power of tax responsiveness be explained to the gradient? We begin by dividing income earners into wage earners and self-employed, distinguishing between incorporated and unorganized business owners. Figure 3 reveals a slight tendency for a power gradient among wage earners, which we see in the paper is partly driven by the labor supply response. More significantly, however, we find a clear gradient among the included business owners but no power gradient among the disorganized self-employed (although they are most clustered, as shown by the high extra mass estimates).

What accounts for the sharp power gradient among corporate business owners? A key difference between these groups is that corporation owners can more humbly reclassify a portion of their high-taxed labor income as taxable capital income. Furthermore, due to the sharp rise in marginal income tax rates, the incentive to reclassify income has been steadily increasing (capital income is taxed at a proportional rate). Our analysis shows that high-capacity unincorporated business owners who are pushed by a group of high-capacity business owners are much more likely to have large capital gains, suggesting that the transfer of income is a major explanation of the tax responsiveness gradient within itself. -Appointed.

Figure 3 Power gradient between wage earners and self-employed

Men exhibit stronger gradients than women

We also study whether there are gender differences in the power gradient. Since we do not observe women in military examinations, we use high school GPAs as a proxy for cognitive abilities. Figure 4 shows that there is no power gradient for women, while the power gradient for men is still visible when using this alternative measure of cognitive ability. A closer look reveals that this result is not driven by gender differences in self-employment.

Figure 4 Power gradient between men and women

Conclusions and policy implications

Our study is the first in a series of documents that show that people with higher cognitive abilities respond more strongly to tax incentives than those with lesser abilities, suggesting a power gradient of tax responsiveness. Our empirical analysis combines the cognitive ability measured at 18 years of age with income records after 20-50 years, emphasizing the importance of skills acquired before entering the labor market for taxpayer behavior and the relevance of these skills throughout careers.

The results highlight a fundamental contradiction in tax design. The government wants to pay higher taxes to high-powered people and lower taxes to high-resilience people. However, if high-power individuals are also high-resilience individuals, the effectiveness of using progressive labor income tax to achieve capacity-based redistribution may be questioned.

An important lesson for policy makers is that taxes on labor and capital labor must be carefully graded to achieve the desired distributional objectives, especially in the context of the dual income tax system. Our research paper highlights that Achilles heels of complex tax systems with ambitious distribution objectives is that they invite high-powered individuals to avoid the advancement of the tax code through careful tax planning. More broadly, opportunities for liberal tax avoidance could undermine the long-term legitimacy of the tax system and affect tax compliance and social norms (e.g., Slemrod et al. 2019, Sarin and Summers 2020, De Neve et al. 2021).


Bachas, P, M Fisher-Post, A Jensen and G Zucman (2022), “Globalization and Capital vs. Labor Tax”,, 6 April.

Bastani, S and D Waldenstrom (2021), “The Gradient of Tax Responsiveness”, Journal of Public Economics Plus 2, 100007.

Burnheim, BD and D. Toubinsky (2018), “Behavioral Public Economics”, BD Burnheim, S. Delavigna and D & Libson (Edis), Behavioral Economics-Foundation and Handbook of Applications 1, North-Holland.

De Neve, JE, C Imbert, J Spinnewijn, T Tsankova and M Luts (2021), “How to Improve Tax Compliance: Evidence of Population-Based Examination in Belgium”,, 20 May.

Saez, E, J Slemrod and SH Giertz (2012). “Elasticity of taxable income with marginal tax rates: a critical review”, Journal of Economic Literature 50 (1): 3-50.

Sarin, N. and LH Summers (2020), “Increasing Tax Compliance in the United States”,, 24 April.

Slemrod, J, O Ur Rehman amd M Waseem (2019), “Financial and Non-Financial Motivation for Tax Compliance: Evidence from Pakistan”,, 15 May.

Stancheva, S. (2021), “Understanding Tax Policy: How Do People Reason?”, Quarterly Journal of Economics qjab033.

Leave a Reply

Your email address will not be published.