University of Pennsylvania ScholarlyCommons

Wharton Public Policy Initiative Issue Briefs Wharton Public Policy Initiative

10-2013

Tax Policy and the Dividend Clientele Effect

Laura Kawano University of Pennsylvania

Follow this and additional works at: https://repository.upenn.edu/pennwhartonppi

Part of the Securities Law Commons, Taxation Commons, Taxation-Federal Commons, and the Tax Law Commons

Recommended Citation Kawano, Laura, "Tax Policy and the Dividend Clientele Effect" (2013). Wharton Public Policy Initiative Issue Briefs. 21. https://repository.upenn.edu/pennwhartonppi/21

This paper is posted at ScholarlyCommons. https://repository.upenn.edu/pennwhartonppi/21 For more information, please contact [email protected]. Tax Policy and the Dividend Clientele Effect

Summary The Jobs and Growth Tax Relief Reconciliation Act of 2003 (JGTRRA) significantly changed tax policy by cutting long-term capital gains tax rates and taxing dividend income at the same rates as long-term capital gains. Following the reduction in the tax disadvantage of dividends, investors gravitated toward dividend-paying investments—especially high-income investors who previously had faced the highest tax rates on dividends.

The behavior of investors before and after the passage of JGTRRA suggests that they divide into “clienteles” based on dividend payouts when the tax disadvantage of dividends varies across investors. Policymakers therefore need to build a proper appreciation of investor behavior, particularly among affluent households, into their thinking about any tax reform proposal affecting capital income. If dividend clientele effects are ignored, estimates of the revenue that can generated by changes in capital tax rates will be off-base.

Keywords tax policy, capital tax rates, JGTRRA, Jobs and Growth Tax Relief Reconciliation Act of 2003

Disciplines Securities Law | Taxation | Taxation-Federal | Tax Law

License

This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

This brief is available at ScholarlyCommons: https://repository.upenn.edu/pennwhartonppi/21 NOVEMBER 2013 publicpolicy.wharton.upenn.edu Volume 1, number 10

about the author TAX POLICY AND THE DIVIDEND CLIENTELE EFFECT Laura Kawano

In January of 2003, with the economy sagging and the need for some sort of stimulus becoming a pressing issue, the Laura Kawano, PhD Visiting Assistant Professor of Business Economics and Administration of President George W. Bush proposed a package Public Policy, The Wharton School of tax cuts that would reduce personal taxes, provide a $400-per- Dr. Laura Kawano is a Visiting Assistant Profes- sor in the Business Economics and Public Policy child rebate to most families, and increase the rate at which Department at the Wharton School of the University of Pennsylvania for the Fall 2013 semester. She is a certain businesses could depreciate equipment, so as to stimulate financial economist in the Business and International small business. Tax Division in the Office of Tax Analysis at the U.S. Department of the Treasury. “We cannot be satisfied until every part of as a way of bringing immediate benefit to Dr. Kawano’s research focuses on the impacts of tax policy on individual and firm choices. She our economy is healthy and vigorous,” Bush the middle class, though Democrats were has written on capital income taxes, taxes on said. “We will not rest until every business more than skeptical. New York Representa- high-income individuals, corporate income taxes, has a chance to grow and every person who tive Charles B. Rangel termed it “an assault and whether taxpayers understand the tax system. 1 wants to find work can find a job.” on the middle-class to the benefit of the She also studies the effects of unemployment and 2 The centerpiece of the Jobs and wealthy.” natural disasters on household income. Dr. Kawano received her BA in Economics Growth Tax Relief Reconciliation Act of The aggregate effect of JGTRRA on from Occidental College in 2002, and her PhD in 2003 ( JGTRRA)—passed by Congress the overall economy remains debated. But Economics from the University of Michigan in 2010. and signed into law by President Bush six the second set of Bush tax cuts—aspects She previously worked at the Federal Reserve Board months after he proposed it—was a vast of which were made permanent during the of Governors. reduction in taxes on investment income. Obama Administration—had a large effect Long-term capital gains tax rates were on individuals’ portfolio choices. High- cut, and dividend income was now to be income investors substantially increased the taxed at the same rates as long-term capital dividend yields on their equity portfolios, gains (rather than being taxed as ordinary according to a study I undertook on the income). The act, which built upon the 2001 impact of changes in dividend and capital Bush tax cuts, was hailed by Republicans gains tax rates brought about by JGTRRA. This behavior on the part of wealthy In sum, the top marginal rate for capital gains when the tax disadvantage of investors carries important implications for dividends fell from 35 percent to 15 percent, dividends varies across investors. To study tax policy—especially now, as tax reform and, for lower-income taxpayers, from 10 the impact of these tax rate changes on has become a widely-discussed issue. My percent to 5 percent. The change extended equity portfolio choices, I used data from research on the effects of JGTRRA shows across dividends from directly owned before and after JGTRRA to estimate the that investors respond to shifts in tax policy equities, as well as those owned through a relationship between the dividend yield on to reduce their tax burdens. Such behaviors mutual fund, partnership, real estate invest- a household’s equity portfolio and the gap should be taken into account by policymak- ment trust or common trust fund. This was between dividend and long-term capital ers seeking to redress the nation’s fiscal defi- a remarkable break with past tax policy. For gains rates. 3 cit by changing tax rates on capital income. Figure 1: top tax rates and dividend yields 1987-2012 POLICY CHANGES UNDER 75 4.0 Top Statutory Tax Rate JGTRRA on Ordinary Income 70 (including Dividend Income, 1981-2002) 65 3.5 In the U.S., dividends generally have been and on Qualified 60 Dividend Income (2003-2013) taxed at a higher rate than long-term capital 3.0 55 gains. Due to the progressivity of the tax Top Statutory Tax Rate 50 on Long-Term Capital system, this tax disadvantage of dividends— 2.5 Gains 45 the gap between the tax rates on dividends Dividend yield of the and those on long-term capital gains—has 40 S&P 500 2.0 increased with income. The difference 35 between the dividend tax rate and the long- 30 1.5 term capital gains tax rate has been greatest 25

for people in the highest tax bracket. 20 1.0

Previous studies have shown that an 15 investor’s optimal portfolio is a function of 10 .5 the difference between dividend and capital 5 gains tax rates; for a given level of expected 0 0 returns, portfolio dividend yields increase ‘87 ‘88 ‘89 ‘90 ‘91 ‘92 ‘93 ‘94 ‘95 ‘96 ‘97 ‘98 ‘99 ‘00 ‘01 ‘02 ‘03 ‘04 ‘05 ‘06 ‘07 ‘08 ‘09 ‘10 ‘11 ‘12 as the relative tax disadvantage of divi- dends falls. Before the passage of JGTRRA, Tax Rate % Dividend Yield of the S&P 500 therefore, high-income individuals had a particularly strong incentive to select stocks decades prior to 2003, the long-term capital The data come from the Surveys of based on dividend payouts because of their gains tax rate had been much lower than Consumer Finances (SCF) from 2001, 2004 tax implications, avoiding those with high the ordinary income rate (except for a brief and 2007. The SCF is a triennial survey dividend yields, and investing instead in period after the Tax Reform Act of 1986, conducted by the Federal Reserve Board those that would deliver returns in the form when dividends and capital gains were both of Governors; each survey samples about of long-term capital gains. taxed at 28 percent). 4,500 households. An advantage of using JGTRRA changed this calculation. the SCF for my analysis is that it provides With the new legislation, the top marginal DIVIDEND CLIENTELES detailed information on household invest- tax rate on long-term capital gains fell ments and allows for accurate marginal tax from 20 percent to 15 percent, while the This change in dividend tax rates provides rate calculations. Also included is informa- 10 percent rate for lower-income taxpayers a rare opportunity to test the “dividend tion that allows me to control for other, fell to 5 percent (and then to zero percent clientele hypothesis,” the idea that inves- non-tax-related factors that can influence in 2008). Also, new qualified dividends now tors sort into “clienteles” based on dividend portfolio choices, such as age, marital status, were taxed in the same way as capital gains payouts. Some have preferences for stocks household size, educational attainment, risk (rather than at the traditional ordinary- that pay dividends while others prefer stocks preferences, and optimism about the future income marginal tax rates). whose expected returns come in the form of of the economy.

1 New York Times, January 8, 2003, http://www.ny- Act,” American Economic Journal: Economic Policy, times.com/2003/01/08/us/politics-economy-over- February 2014, http://papers.ssrn.com/sol3/papers. view-bush-unveils-plan-cut-tax-rates-spur-economy. cfm?abstract_id=1668158. html?pagewanted=all&src=pm. 4 Raj Chetty and Emmanuel Saez, “Dividend Taxes and 2 Ibid. Corporate Behavior: Evidence from the 2003 Dividend 3 This brief is based on Laura Kawano, “The Dividend Tax Cut,” Quarterly Journal of Economics 120 (3): 791- Clientele Hypothesis: Evidence from the 2003 Tax 833. Timing was important to my analysis, higher than the increase experienced by changes to dividend policies for stocks that especially since people’s expectations influ- those households one bracket below. Longer were already held. Regardless, because indi- ence their decisions. Capital income tax term, the increase for households in the top viduals were able to respond to these firm cuts were not part of Bush’s 2000 campaign tax bracket was even larger—a 35 percent changes in payout policies, the tax effects platform. In fact, reductions in dividend increase in dividend yields, or almost 18 that I estimated can still be interpreted as tax rates were not seriously discussed until percent more than those one bracket below. reflecting investor choices. the end of 2002, just before Bush unveiled These responses provide strong evidence for Nevertheless, it is an important question plans for his second tax cut in an address at the dividend clientele hypothesis. That is, to consider how much of the increase could the Economic Club of Chicago. The data the differential tax treatment of dividends be attributed to active portfolio shifts. To from the 2001 SCF survey, derived from and capital gains caused a significant degree help sort out this question, I constructed equity holdings in 2000, therefore are not of investment sorting. hypothetical portfolios from stock market at all affected by JGTRRA, or even by any At the same time, firms also responded data for households in each tax bracket anticipation that legislation like JGTRRA to the tax changes of 2003, increasing their that matched observed portfolio dividend was on the horizon. By 2003, though, it was dividend payments in response to JGTRRA. yields in the 2001 SCF. The increase in clear that dividends probably would be taxed According to research by Raj Chetty of yields for these proxy portfolios for high- at a lower rate. The 2004 and 2007 SCF Harvard University and Emmanuel Saez income households was only a portion of the surveys include dividend receipts from 2003 estimated effect of the tax changes. This sug- and 2006, respectively—both of which were gests that active decision-making accounted “If policymakers ignore affected by the 2003 act. A comparison of for a significant share of the tax effect. the data from these years can help shed light dividend clientele effects, their Was investor response to the 2003 act on the impact of the new tax rates. estimates of the revenue that temporary or permanent? Data measuring JGTRRA—the provisions of which differences between 2001 and 2004, and were made retroactive to January of 2003— will be generated by changes then 2001 and 2007, are similar, suggesting represented a major policy shift, and the in capital tax rates will be a longer-term shift toward higher dividend investor response was dramatic. Following off-base.” yields. Under the Obama administration, the the reduction in the tax disadvantage of divi- new tax rates were extended through 2010, dends, investors did in fact gravitate toward and then again through the end of 2012. dividend-paying investments. By closing More recently, the provisions of JGTRRA the gap between tax rates on dividends and of the University of California, Berkeley, eliminating the difference in the way capital long-term capital gains, dividend income JGTRRA “indeed raised dividend payments gains and dividend income are taxed were became more attractive for all investors—but significantly, and in particular induced many made permanent. (For individuals in the especially for high-income investors who firms to initiate dividend payments.” Chetty highest tax bracket, the tax rate went back previously had faced the highest tax rates on and Saez find that, following a continuous up to 20 percent.) These policy changes dividends. decline in dividend payments during two permanently reduced incentives to sort When dividends and capital gains decades, total regular dividends since 2003 into dividend clienteles on the basis of tax 4 became taxed similarly, the tax-based incen- have grown by nearly 20 percent. considerations. tives for selecting stocks on the basis of divi- There is also some evidence that dend yields were dampened. In the same way dividends were initiated at firms at which POLICY IMPLICATIONS that consumers contemplating a big purchase executive compensation was tied to stocks, often choose to drive to another, sales-tax- so executives who could gain from the intro- The investor responses to JGTRRA provide free state to do their shopping, so did high- duction of the new tax rates were the ones telling lessons as the nation considers differ- income investors seek and find alternatives who ushered in the increases in dividends. ent proposals to reduce its long-term deficit. that were in their best financial interest. Perhaps most notably, Microsoft, after long Moving forward, policymakers will have to I estimated that because of the resisting calls to pay dividends, began a divi- seriously consider options that either reduce JGTRRA tax rate changes, households in dend payout for the first time—at rates that expenditures, increase revenue from taxes, or, the top bracket increased their portfolio have grown since the initial dividend payout. as is most likely, a mixture of the two. Both dividend yields by 23 percent between As a result, part of the portfolio adjust- Democrats and Republicans have signaled a 2001 and 2004. This increase is 13 percent ments that I estimated could have reflected desire to consider fundamental tax reform,

and changes to how capital income is taxed Behavioral responses like the clientele could be in the mix. effect carry important consequences for the brief in brief It is important to understand investor impacts of tax policy on the distribution of • The Jobs and Growth Tax Relief Reconciliation behavior as policymakers consider future tax burdens, especially as equity holdings Act of 2003 (JGTRRA) significantly changed tax policy by cutting long-term capital gains tax reform options. If investment income have historically been concentrated in the tax rates and taxing dividend income at the tax rates increase, knowing how investors upper tail of income distribution. Moreover, same rates as long-term capital gains. are likely to react is key to forecasting the if policymakers ignore dividend clientele • Following the reduction in the tax disadvan- tax revenue implications of such policies. effects, their estimates of the revenue that tage of dividends, investors gravitated toward dividend-paying investments—especially All other things being equal, taxing capital will be generated by changes in capital tax high-income investors who previously had income more heavily will increase the tax rates will be off-base. Policymakers will need faced the highest tax rates on dividends. burden of high-income taxpayers. However, to build a proper appreciation of investor • The behavior of investors before and after the my research, and the research conducted behavior, particularly among affluent house- passage of JGTRRA suggests that they divide into “clienteles” based on dividend payouts by others, indicates that investors have holds, into their thinking about any tax when the tax disadvantage of dividends varies shown both a sophisticated awareness of tax reform proposal affecting capital income. across investors. policy and a willingness to adjust portfolio • Policymakers therefore need to build a proper positions to reduce their tax burdens. The The views expressed in this paper are those of appreciation of investor behavior, particularly among affluent households, into their thinking evidence strongly suggests a dividend clien- the author and do not necessarily reflect the about any tax reform proposal affecting tele hypothesis at work: investors respond policy of the U.S. Department of the Treasury. capital income. If dividend clientele effects are to changes in tax policy, rationally seek the ignored, estimates of the revenue that can generated by changes in capital tax rates will highest after-tax return, and actively steer be off-base. their portfolios accordingly.

Founded in 1881 as the first collegiate business school, the Wharton School of the University of Pennsylvania is recognized globally for intellectual leadership and ongoing innovation across every major discipline of business education. With a broad global community and one of the most published business school faculties, Wharton creates economic and social value around the world.

About the penn Wharton Public Policy Initiative About Penn Wharton PPI Issue Briefs

The Penn Wharton Public Policy Initiative (PPI) is a hub for Penn Wharton PPI publishes issue briefs at least once a month, research and education, engaging faculty and students across Uni- tackling issues that are varied but share one common thread: they versity of Pennsylvania and reaching government decision-makers are central to the economic health of the nation and the Ameri- through independent, practical, timely, and nonpartisan policy can people. These Issue Briefs are nonpartisan, knowledge-driven briefs. With offices both at Penn and in Washington, DC, the documents written by Wharton and Penn faculty in their specific initiative provides comprehensive research, coverage, and analysis, areas of expertise. anticipating key policy issues on the horizon. Penn Wharton PPI For additional copies, please visit the Penn Wharton PPI is led by Faculty Director Mark Duggan, the Rowan Family Foun- website at publicpolicy.wharton.upenn.edu. dation Professor and Chair of Business Economics and Public Policy, and Professor of Health Care Management at Wharton. Follow us on Twitter: @PennWharton­­PPI

Penn Wharton Public Policy Initiative The Wharton School University of Pennsylvania