While President Trump has proposed lowering taxes on capital gains, former Vice President Joe Biden has proposed increasing them. Under Biden’s plan, all taxpayers, regardless of their annual income, would see capital gains earnings above $1 million taxed at the standard income tax rate of 39.6 percent.

Key to the debate is how changes in the tax rate will affect tax revenue (i.e., “elasticities”) and what rate is revenue-maximizing. That is, as the tax rate increases, how does it affect investor behavior and, ultimately, revenues? While some researchers have argued that investors are extremely sensitive to increases in capital gains taxes and therefore capital gains tax cuts pay for themselves by spurring investment, other researchers argue that reducing the rate would have major fiscal costs and reduce federal revenues. 

Importantly, however, the federal tax and fiscal environment has evolved significantly over recent decades, meaning findings from some past research may be less relevant today. New studies are required to understand modern investor behavior and how to maximize revenues.

In a new paper that takes these changes into account (PDF), Princeton University’s Owen Zidar and Ole Agersnap conclude that investors aren’t as sensitive to changes in the capital gains tax rate as many have argued. The authors find that a reduction in the federal capital gains tax rate would have substantial fiscal costs and that the revenue-maximizing federal capital gains tax rate is around 40%. One implication of their findings is that proposals to raise capital gains tax rates may raise substantially more tax revenue than official scorekeepers expect.

Revenue-maximizing tax rates and revenue projections

The authors’ analysis suggests that a capital gains tax rate of around 40% would maximize federal capital gains tax revenues over a ten-year budget window. (The authors’ analysis suggests a revenue-maximizing rate of 38% to 47%, whereas a recent estimate from official scorekeepers is 27%.) For reference, the average federal tax rate on capital gains was about 18% in 2018, the most recent year for which the authors have data.

Using 2018 values, the authors estimate that increasing the rate by even 5 percentage points would provide a 10% to 20% increase in annual revenue. In 2018, federal capital gains tax revenue was $158.4 billion. A 5 percentage point increase (in the current regime with unlimited deferral and step-up-basis at death) would provide an additional $18 to $30 billion. That amount would be even higher if investors fail to substantially change their behavior. Absent any behavioral response, the authors estimate tax revenues from this 5 percentage point tax increase would increase revenues by $44.5 billion. 

These revenue estimates are roughly twice as big as recent estimates from the Joint Committee on Taxation (JCT) and other tax analysts. The difference results from varying estimates of elasticity–or how sensitive investors are to a change in the tax rate. Larger elasticities mean that investors are more responsive to changes in the tax rate. An elasticity of less than -1 (1 in absolute magnitude) indicates that, if the capital gains tax rate were cut, investors would increase their capital gains realizations by so much that tax revenues would actually increase and tax cuts would pay for themselves. While JCT uses an elasticity of -0.70 to score results, Agersnap and Zidar estimate a ten-year elasticity of -0.3 to -0.5. 

Methodology

The authors estimate revenue-maximizing federal capital gains tax rates by studying the effects of 584 different capital gains tax rate changes at the state level. While many of these changes are small, 128 state tax changes exceeded 1 percentage point in absolute value. 

These changes occurred between 1980 to 2016, allowing the authors to study effects of capital gains tax changes in a more recent period than most other literature. This perspective is important because capital gains depend on inflation, which has been lower in recent decades. Further, recent years have seen a bigger role of pass-through firms, which have accounted for nearly half of capital gains realizations, and a more widespread use of diversified investment vehicles. All these changes mean that results from previous decades may be less relevant to today.

Finally, using state-level panel data provides more reforms and closer comparison groups than time-series analysis at the federal level. At the federal level, there are not only fewer reforms but also many confounding factors.

To learn more, download the full paper (PDF).