Tax-Minded Executives and
Corporate Tax Strategies: Evidence
from the 2013 Tax Hikes
Gerardo Pérez Cavazos
Andreya M. Silva
Working Paper 16-034
Tax-Minded Executives and Corporate
Tax Strategies: Evidence from the 2013
Tax Hikes
Gerardo Pérez Cavazos
Harvard Business School
Andreya M. Silva
Working Paper 16-034
Copyright © 2014, 2015 by Gerardo Pérez Cavazos and Andreya M. Silva
Working papers are in draft form. This working paper is distributed for purposes of comment and discussion only. It may
not be reproduced without permission of the copyright holder. Copies of working papers are available from the author.
Tax-minded executives and corporate tax strategies:
Evidence from the 2013 tax hikes
Gerardo Pérez Cavazos1
Andreya M. Silva
March 25, 2015
ABSTRACT
Exploiting the increase in personal tax rates due to the American Taxpayer Relief Act and
Healthcare Act, we identify tax-minded executives who exhibit a preference for personal tax
savings. We find that 2,281 top executives strategically realized their built-in capital gains prior
to the tax hikes to save nearly $741 million in personal taxes in 2012. These executives also (1)
make payout policy choices that save their shareholders taxes and (2) make tax strategy choices
that save their firms cash taxes. Their firms altered payout policies in 2012, distributing $8
billion in special and accelerated dividends, to save shareholders nearly $700 million in taxes.
Further, each tax-minded executive reduces a firm’s cash effective tax rate by 0.28%.
Keywords: Executives, capital gains, dividends, effective tax rates, tax avoidance.
JEL classification: G35, H24, H25, K34.
1Please
address all correspondence to Gerardo Perez Cavazos, Department of Accounting, The University of
Chicago Booth School of Business, 5807 South Woodlawn Avenue, Chicago, IL 60637; Phone: 773-870-1023;
Email: gp@chicagobooth.edu. Andreya M. Silva can be reached at 305-753-0492 or
andreya_silva@outlook.com. We thank Jawad Addoum, Phil Berger, Andreas Bodmeier, Jennifer Blouin
(discussant), Vidhi Chhaochharia, George Constantinides, Christopher Cotton, Merle Erickson, Joseph Gerakos,
George Korniotis, Alok Kumar, Doug Skinner, the participants at the 2014 UNC Tax Symposium, and the
workshop participants at the University of Chicago and University of Miami for helpful comments.
1. Introduction
We implement a unique identification strategy that allows us to provide novel insights
on executives’ tax-related corporate policies. We exploit the 2013 increase in personal income
tax as established by The American Taxpayer Relief Act of 2012 (ATRA) and the Healthcare
and Education Reconciliation Act of 2010 (Healthcare Act), to identify executives that
revealed their tax preferences by actively implementing strategies to save personal taxes. We
define these executives as “tax-minded.” By identifying this tax preference we obtain an
understanding of a characteristic that is imbedded in management’s decision making and;
therefore, can empirically examine firm dividend and tax outcomes as a function of executives
and their tax-related policy choices.
Our identification strategy builds on the extensive literature that analyzes year-end
security sales (e.g., Constantinides, 1984; Seyhun and Skinner, 1994; and Poterba and
Weisbenner, 2001). In particular, Dyl (1977), Lakonishok and Smidt (1986), and Badrinath et
al. (1991) find that tax loss selling occurs, a strategy to reduce one’s tax liability at year-end
whereby losers are sold to offset prior capital gains. The setting surrounding ATRA and the
Healthcare Act is distinct from that of these papers, because the optimal tax strategy changes
when there is an anticipated increase in the future tax rate. Built-in capital gains should be
realized in the period before the increase, whereas losses should be held to offset higher taxes
in the future period. We exploit this change to identify tax-minded executives who adopted a
tax-efficient strategy by pre-emptively realizing their built-in capital gains prior to 2013.
We use the Securities and Exchange Commission (SEC) required Form 4 filings to
reconstruct each executive’s insider holdings portfolio and, most importantly, compute the
1
built-in capital gains of their holdings. We subsequently observe each executive’s selling
behavior in 2012 and identify executives who strategically lowered their future tax liability. We
find that 2,281 tax-minded executives reduced their long term built-in gains to save nearly
$741 million in capital gains tax in 2012, an average of $325 thousand per executive.2
In line with the corporate consistency literature (e.g., Malmendier and Tate, 2005;
Cronqvist et al., 2012; Davidson et al., 2013; and Hutton et al., 2013), we expect that
tax-minded executives exhibit their personal tax preference when making corporate decisions.
Thus, their dividend and tax policies should reflect an emphasis on shareholder and corporate
tax savings, respectively. Specifically, our first conjecture is that these executives manage
companies that are more considerate of the tax burden of their shareholders and,
consequently, take actions consistent with this view. To test this prediction we first identify
firms that altered their dividend policy to save their shareholders’ taxes prior to the tax rate
increases legislated by ATRA and the Healthcare Act. Firms that opted to award a special
dividend in 2012 or accelerate a dividend from the first quarter of 2013 to the fourth quarter
of 2012, allowed shareholders to lock-in the lower dividend tax rate before the tax hike. Similar
to Hanlon and Hoopes (2014) we document that there is indeed a subset of firms who alter
their dividend policy; however, our interest lies in examining the simultaneous actions of
executives and firms surrounding the event. Specifically, we examine the link between taxminded executives and the firms that implement strategies to save their shareholders taxes.
Our empirical results support our conjecture. We find that firms managed by
tax-minded executives were more likely to take actions in 2012 that reduced their shareholders’
This estimate is based on comparing the realized tax liability to the “as if” realized in 2013 tax liability.
Considering a discount rate of 5% and an investment horizon of 5 and 10 years, the total savings are reduced to
$581 and $455 million respectively.
2
2
tax liabilities. The 1,010 companies managed by tax-minded executives awarded a total of $3.7
billion in special dividends and accelerated $4.3 billion in regular dividends. These dividend
payouts collectively saved shareholders nearly $700 million in taxes.
Relatedly, both the media and prior research (e.g. Chetty and Saez, 2005 and Brown et
al., 2007) point out that executives also substantially benefit from additional dividend awards
when there is a change in personal tax rates. In particular, the media suggest that the
distributions are mostly self-serving. For example, in November of 2012 The Wall Street
Journal singled out the recently retired CEO of Costco, Jim Sinegal, because the company
borrowed $3.5 billion to finance a special dividend whereby Sinegal would receive a payout of
$14 million and consequently save over $4 million in taxes. While we do not dispute the
benefits to executives, we maintain that the additional dividend payments in 2012 were largely
due to executive tax preferences and are, therefore, an element within a series of actions aimed
at reducing the tax burden of its shareholders. As such, we expect that the dividend payouts
were in line with the executives’ preexisting corporate tax policies which emphasized tax
savings.
Formally stated, our second conjecture is that tax-minded executives manage
companies that prioritize corporate tax savings and, therefore, employ effective tax planning
strategies to lower firm tax liabilities. To test this conjecture, we conduct a panel analysis that
tracks executives through time. We examine the link between tax-minded executives and each
firm’s cash tax payments over time, as measured by their cash effective tax rate (Cash ETR).
As predicted, we find that tax-minded executives manage firms that implement incremental
cash tax savings strategies. On average, the presence of each tax-minded executive lowers a
firm’s Cash ETR by 0.28% per year, which is equivalent to cash tax savings of $2 million per
3
year for the average firm in the sample.3
The results in this paper contribute to the understanding of executives’ corporate
policies by academics and legislators. The first implication of our results is that an executive
who places importance on personal tax savings is more likely to manage a company that
undertakes actions that benefit their shareholders’ tax outcomes. This finding provides
evidence in support of the longstanding theory that managers internalize the effects of
personal shareholder taxes when making corporate decisions (e.g., Miller, 1977; DeAngelo and
Masulis, 1980; Masulis and Trueman, 1988; Green and Hollifield, 2003), as we identify a subset
of executives who display this behavior in their decision-making.
Second, our findings with respect to firm Cash ETRs indicate that an executive that
actively implements strategies to save on personal taxes will also implement strategies to lower
firm taxes. Prior evidence shows that executives impact firm effective tax rates (Dyreng et al.,
2010), but does not disentangle whether the effect is due to executives’ tax policy or other
correlated decisions that impact the outcome (Armstrong et al., 2010). Our results more
specifically demonstrate that executive tax preferences, which are imbedded in firms’ policies,
affect firms’ tax outcomes. As such, we shed light on the open question of why some firms
pay less in taxes than other firms. Moreover, our findings also provide evidence that the special
and accelerated dividend payouts in 2012 were largely in line with the executives’ pre-existing
corporate tax policies that reduced firm cash taxes.
Third, shareholders are subject to double taxation, tax on corporate earnings and tax
on dividend distributions. Thus, an executive can take two steps to save its shareholders taxes,
3
The range in Cash ETR, 0.28% to 0.65%, depends on different model specifications.
4
the first step is to save corporate taxes and the second step is to make dividend distribution
choices which reduce shareholders’ tax burden. As such, corporate tax policy and dividend
payout policy are related executive choices. Unlike prior papers that examine each of these
policies independently, we show that an executive who is tax-minded is consistent in setting
both policies. Therefore, we establish an important link between setting corporate tax policy
and setting dividend payout policy.
In sum, the key innovation of this paper is that we implement a unique empirical
strategy to identify tax-minded executives. Using this strategy we are able to analyze an
executive’s decision that is analogous in the personal and the corporate spaces: choosing to
take legal tax strategies to lower the tax liability. This is starkly different from tax studies that
focus on executives that misappropriate firm assets to evade taxes, such as Dhaliwal et al.
(2009) and Chyz (2013), because our identification is not dependent on the lack of corporate
governance at the firm level. As a result, we are able to make broader inferences about
executive tax preferences. We provide evidence to support that executives’ personal tax
preference is an important characteristic that directly maps into corporate tax and dividend
policies.
Section 2 provides background on the 2013 tax rate changes. Section 3 describes sample
formation and data. Section 4 discusses our identification strategy and develops our measures
of tax-mindedness. Section 5 explores the impact that tax-minded executives have on
shareholders’ taxes, while Section 6 explores the impact on firm-level taxes. Section 7
concludes the paper.
5
2. Summary of 2013 tax rate changes
The 2013 tax year is marked by legislation that significantly increased tax rates for
high-income taxpayers (Fig. 1). First, set to commence in 2013, the revenue provision for the
Healthcare Act established a 3.8% personal income tax on passive investment income for
taxpayers with an adjusted gross income (AGI) of $250 thousand for married joint filers
(MFJs) and $200 thousand for single filers.4 Consequently, a surtax was imposed on income
from capital gains and dividends.
Moreover, the tax provisions in the Economic Growth and Tax Relief Reconciliation
Act of 2001 (EGTRRA) and Jobs and Growth Tax Relief Reconciliation Act of 2003
(JGTRRA) were set to expire on December 31, 2012. These expirations would eliminate the
favorable ordinary income, qualified dividend, and long-term capital gain tax rates for
individuals in the top income bracket. The highest ordinary income tax bracket was scheduled
to increase by 4.6% to 39.6% and the highest capital gains rate was scheduled to increase by
5% to 20%. Qualified dividends were to lose their preferential tax treatment and increase to
ordinary tax rates, rising from 15% to 39.6%.
On July 24, 2012 the ATRA bill was proposed to address the expiration of EGTRRA
and JGTRRA. The bill called for a one-year temporary extension of both acts. The House
passed the bill on August 1, 2012; however, the Senate put it on hold due to the impending
presidential election and the dichotomous views of the candidates. A tight race existed
between the current president, Barack Obama, who pushed for raising taxes on the wealthiest
Americans and his opponent, Mitt Romney, who championed lower tax rates. On November
4
Passive investment income includes: interest, dividends, capital gains, rental and royalty income, non-qualified
annuities, income from businesses involved in trading of financial instruments or commodities, and businesses
that are passive activities to the taxpayer (within the meaning of IRC section 469).
6
6th, the uncertainty was resolved as the president was re-elected to office.
The bill was revisited and subsequently became a political bargaining chip in the
negotiations aimed at avoiding the fiscal cliff. Numerous scenarios were discussed and
considered over the last two months of 2012, but the federal tax policy remained unresolved
through the end of the year. On January 1st, 2013, the Senate passed the bill after incorporating
changes that raised tax rates. The next day, the president signed ATRA into law. The final
version of ATRA affected taxpayers with an AGI over $450 thousand for MFJs and $400
thousand for single filers by raising the ordinary income rate from 35% to 39.6%, the qualified
dividend rate from 15% to 20%, and the capital gain rate from 15% to 20%. Thus, the
Healthcare Act and ATRA jointly raised the maximum tax rate on capital gains and dividends
to 23.8%, nearly a 60% increase in tax on these types of passive investment income.
3. Sample formation and data
We use several data sources in this study. Our primary data source is Thomson
Reuters’ Insider Data. This database contains corporate insiders’ trades, as reported on SEC
required Form 4 filings of insider transactions. We extract all insider acquisitions and
dispositions of shares from 1986 through 2012. We merge the insider transaction data with
executive data from Execucomp, which covers the five highest compensated officers of each
firm in the S&P 1500. All executives who are not employed through the end of 2012, our
period of identification, are eliminated.
We match each executive’s firm with stock return data from the CRSP monthly stock
files. We also retrieve dividend data from CRSP. Consistent with DeAngelo et al. (2000), we
categorize CRSP distribution codes 1262 and 1272 as special dividends. Regularly paid and
7
taxable U.S. dividends are identified by CRSP distribution codes 1212, 1222, 1232, 1242, and
1252. Lastly, we match each firm to the accounting data on Compustat and create the
corporate tax variable Cash ETR. Table A.1. provides further detail on the definition and data
sources of the variables used in this paper.
4. Identifying tax-minded executives
4.1. Executive reactions to ATRA
The enactment of ATRA was a significant event for high-income taxpayers as it
represented the end of favorable tax rates on ordinary income, capital gains, and qualified
dividends. Prior studies by Bolster et al. (1989) and Seida and Wempe (2000) find that the
unfavorable change in tax rates following the Tax Reform Act of 1986 had a direct impact on
trading behavior. We similarly expect that the 2013 tax hike affected trading behavior;
however, our focus is on the change in behavior leading up to the event. We conjecture that
executives, on average, pre-emptively reacted to the impending 2013 increase in capital gains
tax rates by shifting stock sales to 2012. Due to the timing of the tax legislation, the effect is
likely most prominent during December of 2012. Therefore, we analyze the historical
difference in executive stock sales from December to January.
Table 1 presents the December and January differences in all insider stock sales from
2002 to 2012. We find that $1.08 billion additional stock sales occurred in December 2012
compared to January 2013. A t-test confirms that 2012 is the only year in which the difference
is statistically larger than the mean difference of $275 million. Figure 2 shows these results.
8
4.2. Framework
Having established that the impending 2013 tax rate increases spurred executives to
increasingly sell their shares, we direct our attention to identifying which individuals took
action. To do so, we present a simple tax framework.
In a world where taxes are the only friction, the profits of selling a stock at time 𝑡 are
given by the built-in capital gains after tax:
𝛱={
(𝑃𝑟𝑖𝑐𝑒𝑡 − 𝑃𝑟𝑖𝑐𝑒0 ) × (1 − 𝜏𝑡 )
(𝑃𝑟𝑖𝑐𝑒𝑡 − 𝑃𝑟𝑖𝑐𝑒0 )
if (𝑃𝑡 − 𝑃0 ) > 0
,
if (𝑃𝑡 − 𝑃0 ) ≤ 0
(1)
where 𝑃𝑡 is the price at time 𝑡, 𝑃0 is the price at the time of acquisition (i.e. tax basis), and 𝜏𝑡
is the tax rate at time 𝑡.
If an investor has positive built-in gains, i.e. 𝑃𝑡 − 𝑃0 > 0, and the tax rate will change
such that 𝜏𝑡 < 𝜏𝑡+1, there is a tax-saving strategy whereby he sells in 𝑡 and repurchases in 𝑡 +
1 if the following condition holds:
(𝑃𝑡 − 𝑃0 ) × 𝛥𝜏 > 𝑃𝑡+1 − 𝑃𝑡
(2)
If the time interval between 𝑡 and 𝑡 + 1 is sufficiently small (e.g. December 31 at 11:59
and January 1st at 12:00), the prices are likely to remain constant, and thus by construction
there would be an opportunity for tax-savings:
𝛱 𝑠𝑡𝑟𝑎𝑡𝑒𝑔𝑦 = (𝑃𝑡 − 𝑃0 ) × 𝛥𝜏
(3)
Consider a simple example: when an individual purchases 100 thousand shares at $30
per share, he has a stock basis of $3 million. If over a year later the value of the stock is $40
per share, he then has a long-term built-in gain of $10 per share, or $1 million. If he sells any
shares he is required to pay capital gains tax. On December 31st, 2012 the prevailing long-term
capital gains rate of 15% would result in a tax liability of $150 thousand. However, if the sale
9
occurred on January 1, 2013 the new rate would significantly increase the tax liability to
$205,700, i.e. $200 thousand in capital gains tax and $5,700 in net investment tax. The strategy
would result in tax savings of $55,700 or over 25% of the 2013 tax liability.
A second strategy exists whereby the investor permanently defers the realization of
capital gains. However, this alternative is unlikely for at least two reasons. First, there are
benefits to diversification that cannot be attained if an investor’s portfolio is not rebalanced.
Second, an investor’s investment horizon may be different from that of the firm.
Thus, in the absence of non-tax costs, individuals with built-in capital gains should
realize their gains to save on taxes. However, empirically we observe that only 16% of
executives realized more than 80% of their long term built-in capital gains in 2012. This
indicates that there must be additional factors that affect each investor 𝑖 , such that their
expected profits from exercising the strategy are better described by:
𝐸𝑖 [𝛱 𝑟𝑒𝑎𝑙 𝑤𝑜𝑟𝑙𝑑 ] = [(𝑃𝑡 − 𝑃0 ) × 𝛥𝜏] × 𝑃𝑟𝑜𝑏𝑖 (𝛥𝜏) − 𝐶𝑜𝑠𝑡𝑠𝑖
𝑠. 𝑡. 𝐼𝑛𝑠𝑖𝑑𝑒𝑟 𝑡𝑟𝑎𝑑𝑖𝑛𝑔 𝑙𝑎𝑤𝑠 𝑎𝑛𝑑 𝐵𝑜𝑎𝑟𝑑 𝑟𝑒𝑠𝑡𝑟𝑖𝑐𝑡𝑖𝑜𝑛𝑠
(4)
These additional factors can include a variety of unobservable costs such as the
political or image costs of being perceived as a greedy or unpatriotic executive (Graham et al.,
2013), or market frictions such as transaction costs or the costs associated with hiring an expert
to gather tax knowledge. These costs need not be the same for all executives, as they depend
on each executive’s tax preferences. For example, managers’ political preferences affect
corporate policies (Hutton et al., 2013), family owners are more concerned with the reputation
costs of being tax aggressive (Chen et al., 2010), and CEOs with military experience tend to
avoid less tax (Law and Mills, 2013). Furthermore, executives are also subject to insider trading
laws and other restrictions set by the Board of Directors that may prevent them from selling
10
and immediately repurchasing shares. For example, executives are not allowed to trade in
opposite directions within a six-month period. This constraint becomes more costly in cases
where the executive has insider information that indicates a positive abnormal return in the
next six months.
Therefore, analyzing the variation in executives’ reactions to the 2013 tax hikes
provides an insight into their personal cost and benefit functions related to tax strategy. We
develop measures of tax-mindedness that are based on executives’ revealed tax preferences
and intend to capture the idiosyncratic costs they face. Note that we exclusively focus on
executives with built-in capital gains. Executives with zero gains or built-in losses throughout
2012, which make up less than 5% of the sample, would not have the opportunity to lower
their tax liability by selling prior to 2013. As such, their tax preferences are unobservable and
these executives must be excluded from the analysis.
4.3. Personal Tax-Minded Measures
We define tax-mindedness as a preference for tax savings that is revealed through
strategic actions to improve tax outcomes. In our setting, executive tax-mindedness can be
observed via the reduction of built-in gains before the increase in personal tax rates in 2013.
This identification strategy requires that we quantify each executive’s long-term
built-in capital gains as of December 31, 2011 and throughout 2012. To do so, we reconstruct
each executive's insider holdings portfolio starting with his first insider transaction.
Specifically, for every acquisition that an executive makes, we create a batch that records the
number of shares and their corresponding purchase price. This is referred to as tax basis. Note
that all 2012 acquisitions are excluded because a minimum holding period of one year must
11
be met for gains on stock sales to qualify for long-term capital gains treatment. When an
executive sells shares we adjust the number of shares in each batch using the First-In FirstOut (FIFO) method, in accordance with the Internal Revenue Code (IRC). 5
At each period of interest we observe the number of shares that remain in the
executive’s portfolio and record the current price per share. We then subtract the tax basis
from the current value of the shares to obtain the long-term built-in capital gains of each
portfolio. Thus, at each point in time the built-in capital gains for a particular executive are
given by:
𝑡
𝑡
𝐵𝑢𝑖𝑙𝑡-𝑖𝑛 𝑐𝑎𝑝𝑖𝑡𝑎𝑙 𝑔𝑎𝑖𝑛𝑠𝑡 = ∑𝑁
𝑖=1(𝑃 − 𝑃𝑖 ) × 𝑥𝑖
(5)
where 𝑃𝑡 is the price at time 𝑡, 𝑃𝑖 is the price at which shares in batch 𝑖 were acquired, 𝑥𝑖𝑡 is
the number of shares in batch 𝑖 at time 𝑡, and 𝑁 is the total number of batches.
We then observe which executives decreased their built-in capital gains prior to the tax
hike and create three tax-mindedness variables. Tax_Minded_2012 identifies executives who
reduced their built-in capital gains throughout 2012. It is calculated as the difference between
the initial built-in capital gain during 2012 and the built-in capital gain at January 1, 2013.6 The
following two variables capture executives who more aggressively reacted to the tax change.
Tax_Minded_Dec identifies executives whose built-in gains increased through December 1,
2012 and who subsequently realized their built-in gains during the month of December.
Tax_Minded_80 identifies executives who reduced more than 80% of their built-in gains by
For the purpose of calculating capital gains, the IRC calls for the FIFO method or exact identification of the
shares to be sold. We use the FIFO method because we cannot observe whether exact shares were identified for
every sale of stock. Nonetheless, we consider the FIFO method to be a reasonable assumption as it is the default
method used at brokerage houses and, per the IRC, to change from FIFO the client is required to identify the
exact shares at each moment they trade and no later.
6 Executives with positive built-in capital gains at the start of the year are assigned the capital-gain value at January
1, 2012. All other executives are assigned the value from their first positive month of capital gains.
5
12
the 2012 year-end.7 The definitions of these variables are summarized as follows:
Tax Variable
Definition
𝐶𝐺12 − 𝐶𝐺𝑗𝑎𝑛13 > 0
Tax_Minded_2012
Tax_Minded_Dec
𝐶𝐺12 − 𝐶𝐺𝑑𝑒𝑐12 < 0 & 𝐶𝐺12 − 𝐶𝐺𝑗𝑎𝑛13 > 0
Tax_Minded_80
𝐶𝐺𝑗𝑎𝑛13 ≤ 0.2 × 𝐶𝐺12
4.4. Descriptive Statistics
Table 2 provides summary statistics of our sample. Panel A reports our tax-minded
variables as well as other executive characteristics. The sample is comprised of 4,435 executives
that manage 1,485 firms. Of these executives, 557 are CEOs and 867 are CFOs. On average,
each executive holds a portfolio with $766 thousand in long term built-in capital gains. Slightly
more than half of the sample is defined as Tax_Minded_2012, while only 4% and 16% are
categorized as Tax_Minded_Dec and Tax_Minded_80, respectively.
Panel B provides descriptive data on the characteristics of the 1,485 firms managed by
the executives in Panel A. Overall, the firms in our sample are large and profitable. Each
company has on average 1.54, 0.13, and 0.47 executives classified as Tax_Minded_2012,
Tax_Minded_Dec, and Tax_Minded_80, respectively. Of the 1,050 firms that regularly award
dividends, 12% accelerated their payouts in 2012. On average, 2% of firms paid special
dividends in 2012.
4.5. Validation of the measures
The larger the executive’s built-in gain during 2012, the greater the incentive to sell
before January 1st, 2013. In other words, as the incentives increase executives are more likely
80% is a natural breakpoint in the distribution of built-in capital gains changes. The main results are robust to
changing that threshold in the 60% to 90% range.
7
13
to derive enough tax benefits to offset the potential costs associated with realizing their capital
gains. Therefore, to validate our measures we test whether there is an increase in the likelihood
that an executive 𝑗 is tax-minded as his built-in capital gains increase. We use a logistic
regression model with the following specification:
Pr(𝑇𝑎𝑥_𝑀𝑖𝑛𝑑𝑒𝑑𝑗 ) = 𝛼 + 𝛽 log(𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝐺𝑎𝑖𝑛𝑠𝑗 ) + 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑗
(6)
The dependent variable, Tax_Minded, takes the value of one for those executives who
reduced their capital gains as defined by Tax_Minded_2012 and Tax_Minded_Dec. The main
explanatory variable is the logarithm of each executive’s initial built-in capital gains value. This
variable captures the tax savings, or benefits, associated with exercising the optimal tax
strategy. Unlike the tax benefits, the costs are not observable. Therefore, we include several
executive and firm characteristics as proxies for possible costs.
The controls include executive characteristics such as age, tenure at the firm, and
whether the executive serves as Chief Financial Officer (CFO), Chief Executive Officer
(CEO), or dualy as CEO and Chairman of the Board. These variables can impact
tax-mindedness. For example, the Board can subject CEOs and CFOs to additional
restrictions on insider selling because of the visibility of their actions. A CFO may have more
tax knowledge or tax awareness compared to that of a vice president in an operational area of
the firm. Moreover, older or more experienced executives may have previously faced tax
changes which impact their perception of the relevance of taxes and the cost of becoming
informed. An older executive is also more likely to have a shorter investment horizon, which
reduces the foregone investment returns associated with making an early tax payment on
capital gains. We include firm controls for size and the firm’s six-month post period return.
Firm size controls for potential costs associated with the visibility of the company. Lastly, the
14
company’s stock return during the first six months of 2013 controls for the cost of realizing
capital gains when an executive has positive private information.
Note that Tax_Minded_80 is excluded from this analysis because by construction it
captures two opposing effects, costs and benefits. The greater the built-in capital gains the
greater the incentive to sell and the greater the cost of realizing 80% of those gains. Therefore,
there would be no clear interpretation for the coefficients of our main explanatory variable
when using Tax_Minded_80 as the dependent variable.
There is a positive and statistically significant relationship between the tax-mindedness
measures and the level of built-in capital gains (Table 3). These results validate that our
measures of tax-mindedness capture the economic tradeoff that exists in our theoretical
construct. Notably, the area under the curve (AUC) is relatively low across models, 56% to
62%, and the control variables do not provide large incremental explanatory power to the
regressions. This indicates that our tax-mindedness measures capture a significant element of
tax preferences that is not captured by readily observable characteristics of the executives and
the firms.
4.6. Economic significance of executive tax savings
To assess the economic magnitude of the effect, we quantify the tax savings of the
executives identified as Tax_Minded_2012. We find that these executives reduced their
long-term built-in gains in excess of $8.4 billion or 47%. The reduction of these gains in 2012
effectively reduced their tax liability by nearly $741 million, which is equivalent to an average
savings of $325 thousand per executive.
15
5. Tax-minded executives and shareholders’ taxes
The main thesis of the paper is that tax-minded executives have a preference for tax
savings which affects their corporate decisions. Specifically, our first conjecture is that these
executives manage companies that are more considerate of the tax burden of their
shareholders and, consequently, take actions consistent with this view. To test this hypothesis
we examine firm dividend payout policy in 2012.
The period before the 2013 tax change is an optimal setting to test this hypothesis
because the impending increase in dividend tax rates created an opportunity for companies to
lessen their shareholders’ dividend tax burden. Any additional dividend awards in 2012 would
allow shareholders to lock-in the lower dividend tax rate. Specifically, firms could choose to
award a special dividend in 2012 or accelerate a dividend from the first quarter of 2013 to the
fourth quarter of 2012.
Figure 3 shows evidence consistent with firms taking both of these actions. The
historical total value of both ordinary and special dividends awarded increased substantially in
2012, the effect being most pronounced during the fourth quarter. Additionally, the number
of special dividend awards reached a historical high of 383 in 2012, surpassing the 342
observed in 2010 when JGTRAA and EGTRRA were initially set to expire. We therefore
examine whether tax-minded executives were more likely to manage the companies that
distributed special dividends or accelerated the payment of their regular dividends in
anticipation of their shareholders’ tax increases.
16
5.1. Special Dividend Awards in 2012
We test whether there is an increase in the probability that company 𝑖 awards a special
dividend when there are more tax-minded executives in the firm. We use the following logistic
model:
𝑃𝑟(𝑆𝑝𝑒𝑐𝑖𝑎𝑙𝑖 ) = 𝛼 + 𝛽(𝑁𝑜. 𝑇𝑎𝑥 𝑀𝑖𝑛𝑑𝑒𝑑 𝐸𝑥𝑒𝑐𝑠𝑖 ) + 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐹𝐸
(7)
The dependent variable takes the value of one for those firms that awarded a special
dividend, and zero otherwise. The main explanatory variable, Tax_Minded_Execs, is the
number of tax-minded executives present in each company. We control for several firm
characteristics such as firm size, market-to-book-ratio (MTB), cash on hand, financial leverage,
return on assets, cash from operations, and institutional stock holdings. We also include
industry fixed effects.
We expect that companies with more cash holdings (Jensen, 1986) and stronger past
performance (Fama and Babiak, 1968) are more likely to make special payouts. Moreover, we
predict that companies with low leverage will use dividends as a means to rebalance their
capital structure. We include institutional investors as a control, because they have different
tax implications than that of individual investors (Allen et al., 2000; Grinstein and Michaely,
2005). MTB controls for growth opportunities. Lastly, industry fixed effects control for
possible industry-wide shocks that could have triggered the distribution of special dividends
in 2012.
We find strong evidence that the presence of each tax-minded executive significantly
increases the likelihood of a company awarding a special dividend throughout 2012 (Table 4).
The odds ratio is between 1.06 and 2.36 across the different models. The coefficients on the
17
control variables are not always significant; however, we find that their signs are consistent
with our economic predictions. For instance, the coefficient on the institutional holdings
variable is negative, which is consistent with the dividend catering theory (Baker and Wurgler,
2004a and 2004b). In this setting institutional investors do not demand an increase in special
dividends because they do not obtain a significant tax benefit. Lastly, the AUC is between 78%
and 82% for the models that include controls and industry fixed effects. This indicates that
our models are useful to predict which companies are more likely to award a special dividend.
5.2. Acceleration of regular dividends to the fourth quarter of 2012
We use a logistic regression model to examine whether the probability of a company
accelerating the payment of its regular dividends increases with the presence of tax-minded
executives. To define a company as a dividend accelerator, all three of the following conditions
must be met: (1) the company paid a dividend during the first quarter of 2011 and 2012, (2)
the company decreased or did not pay a dividend in the first quarter of 2013, and (3) the
company increased a payment in the last quarter of 2012. We include the same set of controls
that appear in the special dividends regression; however, we do not expect that any of the nontax variables will play a critical role in the decision to accelerate a dividend by less than one
quarter. Therefore, in contrast to the analysis of special dividends, we expect the coefficients
of the control variables to be statistically insignificant across models. The general specification
is as follows:
𝑃𝑟(𝐴𝑐𝑐𝑒𝑙𝑒𝑟𝑎𝑡𝑖𝑜𝑛𝑖 )
= 𝛼 + 𝛽(𝑁𝑜. 𝑇𝑎𝑥 𝑀𝑖𝑛𝑑𝑒𝑑 𝐸𝑥𝑒𝑐𝑠𝑖 ) + 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐹𝐸
(8)
There is a positive and significant relation between the number of tax-minded
executives in a firm and the likelihood that a company will accelerate dividends (Table 5). The
18
odds ratio that a regular dividend paying company will accelerate the payment of dividends
increases by 1.2 to 1.5 with the presence of each tax-minded executive. As predicted, the
coefficients associated with the control variables are insignificant. This is explained by the low
cost of accelerating a dividend, which is equivalent to the cost of borrowing funds for a period
shorter than 3 months.
5.3. Economic significance of tax savings from special dividends and dividend
accelerations
The tax-minded executives in our sample manage 1,010 companies. These companies
awarded a total of $3.7 billion in special dividends and accelerated $4.3 billion in regular
dividends in 2012. Due to the lower dividend tax rate, the dividend payouts collectively saved
shareholders nearly $700 million in taxes.8
We perform additional analyses on the 194 Tax_Minded_Dec executives to quantify
their impact. We find that these executives saved roughly $78 million in taxes by realizing
capital gains during December 2012. They manage 144 companies. Of these companies, 100
award dividends regularly. Of the regular dividend issuers, 28 awarded special dividends or
accelerated dividends in 2012. Of the 44 non-regular dividend issuers, five granted a special
dividend.
8
This estimate is based on the assumption that all shareholders were individuals subject to the 23.8% tax on
dividends. Also note that these awards were granted when the dividend rate was scheduled to increase to a
maximum of 43.4%; therefore, there may have been an expectation of much larger tax savings for shareholders.
19
5.4. Historical analysis of specials and accelerations
Finally, to verify that our results with respect to special and accelerated dividends are
driven by a tax choice and not due to other unique characteristics of tax-minded executives,
we perform a falsification test. We examine whether the presence of tax-minded executives
explains the distribution of special dividends or acceleration of regular dividends in prior years.
While we expect that our tax-mindedness variables are able to explain payouts in 2012 and
potentially in 2010, we do not expect to find a relation between these variables in other years.9
Table 6 reports the results of a logit regression where the dependent variable takes the
value of one if the company paid a special or accelerated a dividend during the year. The choice
of controls is the same as in Tables 4 and 5. Evidence from Table 6 is consistent with our
prediction. There is a positive and significant relation between the number of tax-minded
executives and the likelihood of a company distributing a special dividend or accelerated
dividend in 2012 and 2010, but not in other years.
5.5. Robustness checks
In this section we perform additional tests to analyze the robustness of our results.
5.5.1. Tax-mindedness or price decline?
A potential concern is that our identification strategy can possibly misidentify executives
as tax-minded if the stock price of their company had a sharp decline during 2012, thus
reducing their built-in capital gains. Even though this reduction in built-in capital gains is
consistent with an optimal tax strategy, it can be argued that these executives need not take
In 2010 there was a strong belief that the favorable tax rates were going to expire; however an extension was
signed into legislation on December 16th 2010.
9
20
any action and therefore should not be categorized as tax-minded. To address this concern,
we repeat the analyses from Sections 6.1. and 6.2., excluding all companies whose stock price
decreased during 2012.
An issue with this type of test is that it reduces the sample size; however, only 28% of
the firms in our sample are excluded since the S&P 500 rose over 11% in 2012. Models (1)
and (3) in Table 8 show the results for specials and accelerations respectively. In both models
the coefficient associated with the number of executives is positive, although it is only
significant for accelerations.
5.5.2. Tax-mindedness or restricted stock holdings?
A second potential concern is that an executive’s portfolio is likely to have restricted
stock. This could impact our identification if the restricted stock holdings hinder some
executives from strategically selling in 2012. To mitigate this concern, we conduct additional
analyses that do not include the value of restricted stock in the calculation of built-in capital
gains.
A couple of caveats are in order. First, Execucomp only shows the restricted stock
holdings on an annual basis; therefore, only Tax_Minded_2012 and Tax_Minded_80 can be
adjusted. Second, we assume that the tax basis of the restricted stock is equal to zero, as it is
common that restricted stock is awarded at no cost or very little cost. Table 8 shows that our
results are robust to the exclusion of restricted stock.
21
6. Tax-minded executives and corporate taxes
The spike in dividend awards in 2012 provokes the question of whether executives
were primarily acting in their own interest when granting payouts prior to the tax change. For
example, an article by Forbes highlights that Oracle’s CEO, Lawrence Ellison, would largely
benefit from its three-quarter dividend acceleration as he would receive $198.9 million from
the payout, resulting in a $17.5 million tax break. A similar article by the New York Times
reported that Steve Wynn, founder and CEO of Wynn Resorts, received nearly a $20 million
tax break in 2012 due to the special dividend award by his company.
While the magnitude of these tax benefits undeniably provides executives with
incentives to act, it is important to note that the fundamental driver behind these decisions is
the executives’ preference for tax savings. It stands to reason that this preference is also
present when making other managerial choices. Therefore, we maintain that the additional
dividend payments in 2012 were not an opportunistic one-shot action, but rather an element
within a series of actions aimed at reducing the tax burden of its shareholders. As such, we
expect that the dividend payouts were in line with the executives’ preexisting corporate tax
policies which emphasize tax savings.
Formally stated, our second conjecture is that tax-minded executives manage
companies that prioritize tax savings and, therefore, employ incremental tax planning
strategies to lower firm tax liabilities. We test this prediction using a panel regression that
tracks executives through time. The panel includes all firm years that have at least two
executives whose tax preferences have been identified from their actions in 2012. This
restriction mitigates the issue of attributing a tax outcome in a previous year to a single
22
identified executive.10 For the same reason we use a one-year measure of cash taxes paid, as
opposed to a longer measure, as the outcome variable.
Table 8 provides the summary statistics for the panel. It consists of 6,744 firm years.
On average, each firm year has 1.6 Tax_Minded_2012, 0.14 Tax_Minded_Dec, and 0.46
Tax_Minded_80 executives. The mean Cash ETR is 24%, which is equivalent to $168.5 million
cash taxes paid for the average firm in the sample.
Using a linear model, we analyze the impact of tax-minded executives on the cash taxes
paid by their firm in each year 𝑡:
𝐶𝑎𝑠ℎ 𝐸𝑇𝑅𝑖,𝑡 = 𝛼 + 𝛽(𝑁𝑜. 𝑇𝑎𝑥 𝑀𝑖𝑛𝑑𝑒𝑑 𝐸𝑥𝑒𝑐𝑠𝑖,𝑡 ) + 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖,𝑡 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐹𝐸
(9)
+ 𝑌𝑒𝑎𝑟 𝐹𝐸
Cash ETR measures the cash taxes paid by the firm. The main explanatory variable is
the number of tax-minded executives who are present at each company. We control for firm
characteristics that are associated with tax avoidance (e.g., Gupta and Newberry, 1997; Mills
et al., 1998). These controls include firm size, leverage, multinational operations, R&D
intensity, intangibles, and advertising expense. We also include industry and year fixed effects.
Consistent with our conjecture, we find that there is a negative and statistically
significant relation between Cash ETR and Tax_Minded_2012 and Tax_Minded_80. The
presence of each tax-minded executive reduces the cash effective tax rate by 0.28% to 0.65%
per year, which is equivalent to cash tax savings of $2 million to $4.5 million for the average
firm in the sample. This result suggests that executives’ tax preferences are stable over time
10
The other executives likely contribute to the tax outcome; however, due to their inexistence in our 2012 sample
we are unable to determine their tax preference.
23
and is consistent with tax-minded executives implementing long-term corporate tax policies
that are aimed at reducing cash taxes for the firm. This finding also demonstrates that
executives were not simply acting in self-interest by awarding additional dividends in 2012. On
the contrary, these tax-minded executives maintain an overall corporate policy of tax
minimization, saving taxes for both their firms and its shareholders.
7. Conclusion
This paper examines executive tax preferences, a previously unexplored characteristic
of executives that maps into firm dividend and tax policies. We implement a unique empirical
approach to identify tax-minded executives who demonstrate a preference for tax savings.
Specifically, we exploit the setting in 2012, the year prior to large tax rate increases legislated
by ATRA and the Healthcare Act, to identify executives who lowered their tax liability by
realizing their built-in capital gains prior to year-end. We then examine the link between taxminded executives and the firms who awarded additional dividends in 2012 with the intention
of reducing the tax burden of their shareholders. Lastly, we analyze the link between taxminded executives and the firms who pay less cash taxes than their industry peers.
We find that tax-minded executives saved nearly $741 million in personal taxes by
reducing their built-in capital gains prior to 2013. They are also between 1.19 and 1.34 times
more likely to manage firms that take incremental steps to lessen shareholder taxes. This is
evidenced by their firms awarding two types of dividend payouts, $3.7 billion in special
dividends and $4.3 billion in accelerated dividends, prior to the 2013 tax hikes. These payouts
collectively saved shareholders nearly $700 million in taxes. Additionally, tax-minded
24
executives manage firms that actively strategize to reduce their tax liability. The presence of
each tax-minded executive reduces a firm’s Cash ETR by 0.28% to 0.65%. This is equivalent
to a $2 million to $4.5 million cash tax savings per tax-minded executive per year.
In sum, we present evidence that tax-minded executives save on personal taxes and
make strategic decisions at the firm level that save their shareholders taxes and firm cash taxes.
As such, we identify a contributing factor that sheds light on the open question of why some
firms pay less in taxes than other firms. Our findings also support the longstanding assumption
that managers internalize the effects of personal shareholder taxes when making corporate
decisions. In future work, it would be useful to examine other aspects of firms managed by
tax-minded executives, such as capital structure and performance. We leave these subjects for
future research.
25
Appendix
Table A.1.
Definition and data source of variables
Variable
Acceleration
Description
One if the company accelerated the payment of
a regular dividend from the first quarter of year
t+1 to the last quarter of year t, and zero
otherwise
Source
The Center for Research in
Securities Prices (CRSP)
Advertising
Advertisement expense as a percentage of sales
Compustat
Age
Executive age
Execucomp
Capital gains
Logarithm of each executive’s initial built-in
capital gains value
Created from Thomson
Reuters Insiders Data
Cash
Total cash divided by total assets
Compustat
Cash_ETR
Cash taxes paid divided by pre-tax book income
less special items
Compustat
CEO
One if the executive is a CEO, and zero
otherwise
Thomson Reuters Insiders
Data
CFO
One if the executive is a CFO, and zero
otherwise
Thomson Reuters Insiders
Data
Duality
One if the executive is both the CEO and the
chairman of the board, and zero otherwise
Thomson Reuters Insiders
Data
Institutional Investors
Percentage of institutional ownership in the
firm
Thomson Reuters
Institutional Holdings Data
Intangibles
Intangible assets divided by total assets
Compustat
Leverage
Long term debt divided by total assets
Compustat
MTB
Market value divided by shareholders’ equity
Compustat
Multinational
One if the company has positive pre-tax
income generated outside the US, and zero
otherwise
Compustat
Operating Cash Flow
Cash flow from operations divided by total
assets
Compustat
Post Period Return
Stock return from January 2013 to June 30th
2013
The Center for Research in
Securities Prices (CRSP)
R&D
Research and development expense divided by
sales
Compustat
26
ROA
Net income divided by average total assets
Compustat
Size
Logarithm of the company’s average assets
Compustat
Special
One if the company paid a special dividend
during the year, and zero otherwise
CRSP
Tax_Minded_80
One if the executive reduced his built-in capital
gains by more than 80% throughout 2012, and
zero otherwise
Created from Thomson
Reuters Insiders Data
Tax_Minded_2012
One if the executive reduced his built-in capital
gains throughout 2012, and zero otherwise
Created from Thomson
Reuters Insiders Data
Tax_Minded_Dec
One if the executive continued to increase his
built-in capital gains through December 1, 2012
and subsequently reduced his built-in gains in
the month of December, and zero otherwise
Created from Thomson
Reuters Insiders Data
TM_80
Number of Tax_Minded _80 executives present
in the company
Created from Tax_Minded_80
TM_2012
Number of Tax_Minded _2012 executives
present in the company
Created from
Tax_Minded_2012
TM_Dec
Number of Tax_Minded _Dec executives present
in the company
Created from
Tax_Minded_Dec
Tenure
Number of years of an executive’s employment
at the firm
Execucomp and Thomson
Reuters Insiders Data
27
References
Allen, F., Bernardo, A. E., & Welch, I., 2000. A theory of dividends based on tax clienteles.
The Journal of Finance 55, 2499-2536.
Armstrong, C. S., Blouin, J. L., & Larcker, D. F., 2012. The incentives for tax planning. Journal
of Accounting and Economics 53, 391-411.
Badrinath, S. G., & Lewellen, W. G., 1991. Evidence on tax‐motivated securities trading
behavior. The Journal of Finance 46, 369-382.
Baker, M., & Wurgler, J., 2004a. A catering theory of dividends. The Journal of Finance 59,
1125-1165.
Baker, M., & Wurgler, J., 2004b. Appearing and disappearing dividends: The link to catering
incentives. Journal of Financial Economics 73, 271-288.
Bolster, P. J., Lindsey, L. B., & Mitrusi, A., 1989. Tax‐induced trading: the effect of the 1986
Tax Reform Act on stock market activity. The Journal of Finance 44, 327-344.
Brown, J. R., Liang, N., & Weisbenner, S., 2007. Executive financial incentives and payout
policy: firm responses to the 2003 dividend tax cut. The Journal of Finance 62, 19351965.
Chen, S., Chen, X., Cheng, Q., & Shevlin, T., 2010. Are family firms more tax aggressive than
non-family firms? Journal of Financial Economics 95, 41-61.
Chetty, R., & Saez, E., 2005. Dividend taxes and corporate behavior: evidence from the 2003
dividend tax cut. The Quarterly Journal of Economics 120, 791-833.
Chyz, J. A., 2013. Personally tax aggressive executives and corporate tax sheltering. Journal of
Accounting and Economics 56, 311-328.
Constantinides, G. M., 1984. Optimal stock trading with personal taxes: implications for prices
and the abnormal January returns. Journal of Financial Economics 13, 65-89.
Cronqvist, H., Makhija, A. K., & Yonker, S. E., 2012. Behavioral consistency in corporate
finance: CEO personal and corporate leverage. Journal of Financial Economics 103, 20-40.
Davidson, R., Dey, A., & Smith, A., 2013. Executives'“off-the-job” behavior, corporate
culture, and financial reporting risk. Forthcoming in Journal of Financial Economics.
DeAngelo, H., DeAngelo, L., & Skinner, D. J., 2000. Special dividends and the evolution of
dividend signaling. Journal of Financial Economics 57, 309-354.
28
DeAngelo, H., DeAngelo, L., & Stulz, R. M., 2006. Dividend policy and the
earned/contributed capital mix: a test of the life-cycle theory. Journal of Financial
Economics 81, 227-254.
DeAngelo, H., & Masulis, R. W., 1980. Optimal capital structure under corporate and personal
taxation. Journal of Financial Economics 8, 3-29.
Dhaliwal, D., Erickson, M., & Heitzman, S., 2009. Taxes and the backdating of stock option
exercise dates. Journal of Accounting and Economics 47, 27-49.
Dyl, E. A., 1977. Capital gains taxation and year‐end stock market behavior. The Journal of
Finance 32, 165-175.
Dyreng, S. D., Hanlon, M., & Maydew, E. L., 2010. The effects of executives on corporate tax
avoidance. The Accounting Review 85, 1163-1189.
Fama, E. F., & Babiak, H., 1968. Dividend policy: an empirical analysis. Journal of the
American Statistical Association 63, 1132-1161.
Fama, E. F., & French, K. R., 2001. Disappearing dividends: changing firm characteristics or
lower propensity to pay? Journal of Financial Economics 60, 3-43.
Graham, J., Hanlon, M., Shevlin, T., & Shroff, N., 2013. Incentives for tax planning and
avoidance: evidence from the field. The Accounting Review 89, 991-1023.
Green, R. C., & Hollifield, B., 2003. The personal-tax advantages of equity. Journal of Financial
Economics 67, 175-216.
Grinstein, Y., & Michaely, R., 2005. Institutional holdings and payout policy. The Journal of
Finance 60, 1389-1426.
Gupta, S., & Newberry, K., 1997. Determinants of the variability in corporate effective tax
rates: evidence from longitudinal data. Journal of Accounting and Public Policy 16, 1-34.
Hanlon, M., & Heitzman, S., 2010. A review of tax research. Journal of Accounting and
Economics 50, 127-178.
Hanlon, M., & Hoopes, J. 2013. What do firms do when dividend tax rates change? An
examination of alternative payout responses. Journal of Accounting and Economics, 114,
105-124.
Hutton, I., Jiang, D., & Kumar, A., 2013. Corporate policies of Republican managers.
Forthcoming in Journal of Financial and Quantitative Analysis.
Internal Revenue Bulletin: 2010-47. Basis reporting by securities brokers and basis
determination for stock. November, 2010. Washington, D.C.: Internal Revenue Service.
29
Jacob, M., Michaely, R., & Alstadsæter, A., 2014. Taxation and dividend policy: The muting
effect of diverse ownership structure. Working paper. WHU – Otto Beishem School of
Management.
Jensen, M. C., 1986. Agency costs of free cash flow, corporate finance, and takeovers. The
American Economic Review 76, 323-329.
Lakonishok, J., & Smidt, S., 1988. Are seasonal anomalies real? A ninety-year perspective.
Review of Financial Studies 1, 403-425.
Law, K., & Mills, L. F., 2013. Following the rules? Corporate tax reporting by CEOs with
military experience. Working paper. Tilburg University.
Malmendier, U., & Tate, G., 2005. CEO overconfidence and corporate investment. The
Journal of Finance 60, 2661-2700.
Masulis, R. W., & Trueman, B., 1988. Corporate investment and dividend decisions under
differential personal taxation. Journal of Financial and Quantitative Analysis 23, 369-385.
Miller, M. H., 1977. Debt and taxes. The Journal of Finance 32, 261-275.
Mills, L., Erickson, M., & Maydew, E., 1998. Investments in tax planning. Journal of the
American Taxation Association 20, 1-20.
Popper, N., & Schwartz N., 2012. Investors rush to beat threat of higher taxes. The New York
Times, 18 November.
Poterba, J. M., & Weisbenner, S. J., 2001. Capital gains tax rules, tax‐loss trading, and turn‐of‐
the‐year returns. The Journal of Finance 56, 353-368.
Review & Outlook, 2012. Costco’s dividend tax epiphany. The Wall Street Journal, 30
November.
Schaefer, S., 2012. Larry Ellison, Oracle investors latest to skirt higher taxes with quicker
dividend. Forbes Magazine, 3 December.
Scholes, M., Wolfson, M., Erickson, M., Maydew, E., Shevlin, T., 2005. Taxes and business
strategy. Prentice-Hall, Englewood Cliffs, NJ.
Seida, J. A., & Wempe, W. F., 2000. Do capital gain tax rate increases affect individual
investors’ trading decisions? Journal of Accounting and Economics 30, 33-57.
Seyhun, H. N., & Skinner, D. J., 1994. How do taxes affect investors' stock market realizations?
Evidence from tax-return panel data. Journal of Business, 231-262.
30
U.S. Government, 2013. Bill Summary and Status: American Taxpayer Relief Act of 2012. The
Library of Congress. Web site, <http://thomas.loc.gov/cgi-bin/bdquery/z?d112:HR
00008:@@@S>
Wilson, R. J., 2009. An examination of corporate tax shelter participants. The Accounting
Review 84, 969-999.
31
Fig. 1. Timeline of relevant changes in tax legislation from 2001 to 2013.
This figure presents the dates of the introduction and enactment of legislation that impacted ordinary income, qualified dividend, and long-term capital gain tax rates for
high-income tax payers from 2001 to 2013. The lower half shows the tax implications of each act and the boxes span the time period when they were in effect. EGTRRA
stands for the Economic Growth and Tax Relief Reconciliation Act of 2001. JGTRRA stands for the Jobs and Growth Tax Relief Reconciliation Act of 2003. The short
title of the Healthcare Act is the Healthcare and Education Reconciliation Act of 2010. The short title of the Tax Relief Act is the Tax Relief, Unemployment Insurance
Reauthorization, and Job Creation Act of 2010. ATRA stands for the American Taxpayer Relief Act of 2012.
2012
Introduction of ATRA
July 24: introduced
Aug 1: passed in the House
Presidential election results
Healthcare Act enactment
2013
ATRA enactment
2001
2002
2003
2004
2005
2006
Jan 2: signed into law
Tax Relief Act enactment
JGTRRA enactment
EGTRRA enactment
2007
2008
2009
2010
2011
2012
Jan 1: passed in Senate with changes
2013
Extends EGTRRA and JGTRRA
Schedule to gradually reduce taxes
Additional 3.8% tax on net investment income
Accelerated the EGTRRA reduction in ordinary income tax
Highest ordinary income rate increased by 4.6%
Lowered the qualified dividend tax rate to 15%
Highest long-term capital gains rate increased by 5%
Qualified dividend rate increased by 5%
Lowered the long-term capital gain tax rate to 15%
32
Fig. 2. Executive stock sales, December and January differences, 2002 - 2012.
This figure presents the difference of December and January executive stock sales, in millions of dollars, for the
2002-2012 period. The dotted line represents the average difference for the 2002-2012 period.
Value of executive sales (millions)
$1,200
$1,000
$800
$600
$400
$200
$0
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
-$200
-$400
December - January
33
Average
2012
Fig. 3. Value of ordinary and special dividend payments, January 2001 - March 2013.
This figure shows the dollar amount, in billions of dollars, of regular and special dividends paid quarterly from
the first quarter of 2000 to the first quarter of 2013.
180
Dividend payments (in billions)
160
140
120
100
80
60
40
20
0
Ordinary
34
Special
Table 1. Executive stock sales, December and January differences, 2002 - 2012.
This table presents the historical level of executive stock sales in December and January, in millions of dollars, for the 2002-2012 period. The table shows the difference
of December and January values and reports a test of the statistic being different from the 2002-2012 average.
Year
December t
January t+1
Diff. Dec - Jan
T-stat (different from the mean)
2002
223.7
193.6
30.0
2003
828.5
673.3
155.2
2004
1436.0
914.3
521.7
2005
1280.8
1581.7
(300.9)
2006
793.5
797.0
(3.5)
2007
1264.7
860.0
404.7
2008
276.7
186.0
90.6
2009
822.7
282.0
540.7
2010
778.8
469.8
309.0
2011
799.8
600.0
199.8
2012
1744.9
667.2
1077.7
(0.67)
(0.33)
0.68
(1.59)
(0.77)
0.36
(0.51)
0.73
0.09
(0.21)
2.21
35
2002 - 2012
Mean St. Dev.
931.8
464.9
656.8
399.7
275.0
362.9
Table 2. Summary Statistics, 2012.
This table presents the summary statistics for the sample. Panel A presents executive characteristics. Panel B
presents firm-level characteristics. The sample consists of 4,435 executives and the 1,485 unique firms they
manage. Variable definitions are provided in Appendix Table A.1.
Panel A. Executive characteristics, 2012
Variable
Mean
Std. Dev.
Q1
Median
Q3
Tax_minded_2012
0.51
0.50
0.00
1.00
1.00
Tax_minded_Dec
0.04
0.20
0.00
0.00
0.00
Tax_minded_80
0.16
0.36
0.00
0.00
0.00
(0.27)
2.17
(1.39)
(0.04)
1.12
Tenure
8.25
4.81
5.00
7.00
11.00
Age
48.84
6.64
44.00
49.00
53.00
CEO
0.13
0.33
0.00
0.00
0.00
CFO
0.20
0.40
0.00
0.00
0.00
Duality
0.02
0.15
0.00
0.00
0.00
Capital Gains
Panel B. Firm characteristics, 2012
Variable
Mean
Std. Dev.
Q1
Median
Q3
Special
0.02
0.14
0.00
0.00
0.00
Acceleration
0.12
0.32
0.00
0.00
0.00
No. TM_2012
1.54
1.47
0.00
1.00
3.00
No. TM_Dec
0.13
0.46
0.00
0.00
0.00
No. TM_80
0.47
0.83
0.00
0.00
1.00
Institutional Investors
0.77
0.16
0.68
0.80
0.89
Cash
0.14
0.15
0.03
0.09
0.20
Leverage
0.20
0.19
0.04
0.17
0.31
ROA
0.05
0.10
0.01
0.04
0.08
Size
8.07
1.73
6.83
7.98
9.14
MTB
2.81
3.66
1.25
1.91
3.12
Operating Cash Flow
0.09
0.09
0.05
0.09
0.14
Post Period Return
0.02
0.24
(0.10)
0.00
0.12
36
Table 3. Likelihood of an executive being tax-minded, logistic regression 2012.
This table reports the estimation results from logit regressions of the following form:
Pr(𝑇𝑎𝑥_𝑀𝑖𝑛𝑑𝑒𝑑𝑗 ) = 𝛼 + 𝛽 log(𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝐺𝑎𝑖𝑛𝑠𝑗 ) + 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑗 .
In this model the dependent variable is an indicator for executive tax-mindedness. The dependent variable in
columns 1 and 2 is Tax_Minded_2012, while in columns 3 and 4 is Tax_Minded_Dec. Variable definitions are
provided in Appendix Table A.1. Standard errors are adjusted for heteroskedasticity and autocorrelation. P-values
are reported below the coefficient estimates. *, **, *** indicate significance at the two-tailed 10%, 5% and 1%
levels, respectively. The Odds-Ratio factor for an increase of 1% in the initial built-in capital gains is presented.
It is calculated as the exponential of the coefficient and is provided for interpretation purposes.
Dependent Variable:
Model:
Intercept
Capital Gains
Tax_Aware_2012
Tax_Aware_Dec
(1)
0.081***
(2)
0.707**
(3)
-3.077***
(4)
-3.242***
0.0075
0.025
<.0001
<.0001
0.09***
0.075***
0.085**
0.049
<.0001
<.0001
0.0348
0.3267
Tenure
Age
0.014
0.052***
0.1429
0.0066
-0.005
0.008
0.3491
0.5243
CEO
0.120
-0.290
0.3028
0.3358
CFO
-0.185**
-0.067
0.0447
0.7791
-0.108
0.942**
Duality
Size
Post Period Return
0.6503
0.0269
-0.062***
-0.076
0.0036
0.166
-0.029
-0.144
0.8393
0.6106
AUC
N. Obs.
0.56
4,435
0.56
3,276
0.56
4,435
0.59
3,276
Odds Ratio: Capital Gains
1.09**
1.07**
1.08**
1.04*
37
Table 4. Likelihood of a company awarding special dividends, logistic regression 2012.
This table reports the estimation results from logit regressions of the following form:
𝑃𝑟(𝑆𝑝𝑒𝑐𝑖𝑎𝑙𝑖 ) = 𝛼 + 𝛽(𝑁𝑜. 𝑇𝑎𝑥 𝑀𝑖𝑛𝑑𝑒𝑑 𝐸𝑥𝑒𝑐𝑠𝑖 ) + 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐹𝐸.
In this model the dependent variable is an indicator for special dividends paid during 2012. The independent
variable is the number of tax-minded executives present at each firm. The independent variable in columns 1 and
2 is TM_2012, in columns 3 and 4 is TM_Dec, while in columns 5 and 6 is TM_80. Variable definitions are
provided in Appendix Table A.1. Standard errors are adjusted for heteroskedasticity and autocorrelation. P-values
are reported below the coefficients. *, **, *** indicate significance at the two-tailed 10%, 5% and 1% levels,
respectively. The Odds-Ratio factor for an increase of 1 tax-minded executive is presented. It is calculated as the
exponential of the coefficient and is provided for interpretation purposes.
Dependent variable: Special
Tax-mindedness variable:
Model:
Intercept
No. Tax-Minded Execs.
TM_2012
TM_Dec
TM_80
(1)
-3.931***
(2)
-0.277
(3)
-4.051***
(4)
-0.359
(5)
-4.062***
(6)
-0.306
<.0001
0.7881
<.0001
0.7382
<.0001
0.7707
0.069
0.183
0.837***
0.864***
0.384***
0.359**
0.5442
0.1388
<.0001
0.0005
0.0097
-0.302***
0.008
0.0155
0.012
MTB
-0.001
-0.001
-0.001
0.56
0.5594
0.5842
Cash
0.109
0.051
0.066
0.9286
0.9685
0.9571
0.028
0.035
-0.018
Leverage
-0.301**
0.0215
Size
-0.306**
0.9818
0.9782
0.9881
ROA
7.663***
7.514***
7.145***
0.0004
0.0006
0.0011
Operating Cash Flow
-4.995**
-5.008**
-4.525**
0.0127
0.013
0.0237
Institutional Investors
-1.873*
-1.805*
-1.687*
0.0567
0.0848
0.0953
Industry FE
AUC
N. Obs.
No
0.53
1,485
Yes
0.78
1,435
No
0.57
1,485
Yes
0.82
1,435
No
0.59
1,485
Yes
0.78
1,435
Odds - Ratio: No. Tax-Minded Execs.
1.06
1.2
2.29***
2.36***
1.46***
1.42**
38
Table 5. Likelihood of a company accelerating a dividend, logistic regression 2012.
This table reports the estimation results from logit regressions of the following form:
𝑃𝑟(𝐴𝑐𝑐𝑒𝑙𝑒𝑟𝑎𝑡𝑖𝑜𝑛𝑖 ) = 𝛼 + 𝛽(𝑁𝑜. 𝑇𝑎𝑥 𝑀𝑖𝑛𝑑𝑒𝑑 𝐸𝑥𝑒𝑐𝑠𝑖 ) + 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐹𝐸.
In this model the dependent variable is an indicator for accelerated dividend payment from the first quarter of
2013 to the last quarter of 2012. The independent variable is the number of tax-minded executives present at
each firm. The independent variable in columns 1 and 2 is TM_2012, in columns 3 and 4 is TM_Dec, while in
columns 5 and 6 is TM_80. Variable definitions are provided in Appendix Table A.1. Standard errors are adjusted
for heteroskedasticity and autocorrelation. P-values are reported below the coefficients. *, **, *** indicate
significance at the two-tailed 10%, 5% and 1% levels, respectively. The Odds-Ratio factor for an increase of 1
tax-minded executive is presented. It is calculated as the exponential of the coefficient and is provided for
interpretation purposes.
Dependent variable: Acceleration
Tax mindedness variable:
TM_2012
(3)
-2.116***
(4)
-0.913
TM_80
Model:
Intercept
(1)
-2.362***
<.0001
0.1738
<.0001
0.2149
<.0001
0.2043
No. Tax-Minded Execs.
0.189***
0.177**
0.388***
0.33**
0.307***
0.245**
0.0141
0.006
0.0293
0.0007
0.0037
(2)
-0.991
TM_Dec
(5)
-2.22***
0.01
Size
-0.044
0.4723
0.6542
0.513
MTB
-0.003
-0.003
-0.003
0.1572
0.1355
0.1466
0.493
0.343
0.278
0.6275
0.7445
0.7878
-1.147
-1.294
-1.264
0.2118
0.1576
0.1676
ROA
2.962
2.495
2.565
0.1142
0.1981
0.166
Operating Cash Flow
-1.812
-1.127
-1.191
0.3866
0.5893
0.5631
Institutional Investors
-0.742
-0.745
-0.668
0.2219
0.2188
0.2652
Cash
Leverage
-0.026
(6)
-0.925
-0.040
Industry FE
AUC
N. Obs.
No
0.58
1,050
Yes
0.65
998
No
0.54
1,050
Yes
0.64
998
No
0.58
1,050
Yes
0.65
998
Odds - Ratio: No. Tax-Minded Execs.
1.2***
1.19**
1.46***
1.39**
1.35***
1.27**
39
Table 6. Specials and accelerated dividend awards over time, 2008-2012.
This table reports the estimation results from logit regressions of the following form:
𝑃𝑟(𝑆𝑝𝑒𝑐𝑖𝑎𝑙 or 𝐴𝑐𝑐𝑒𝑙𝑒𝑟𝑎𝑡𝑖𝑜𝑛𝑖 ) = 𝛼 + 𝛽(𝑁𝑜. 𝑇𝑎𝑥 𝑀𝑖𝑛𝑑𝑒𝑑 𝐸𝑥𝑒𝑐𝑠𝑖 ) + 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐹𝐸.
In this model the dependent variable is an indicator for special dividend paid in 2012 or accelerated dividend payment from the first quarter of 2013 to the last quarter
of 2012. The independent variable is the number of tax-minded executives present at each firm. The independent variable in columns 1, 3, 5, 7, and 9 is TM_2012, while
in columns 2, 4, 6, 8, and 10 is TM_80. Variable definitions are provided in Appendix Table A.1. Standard errors are adjusted for heteroskedasticity and autocorrelation.
P-values are reported below the coefficients. *, **, *** indicate significance at the two-tailed 10%, 5% and 1% levels, respectively.
Dependent variable: Special or acceleration
Year:
2012
Tax-mindedness variable:
TM_2012
TM_80
Model:
Intercept
No. Tax-Minded Execs.
2011
2010
2009
2008
TM_2012
TM_80
TM_2012
TM_80
TM_2012
TM_80
TM_2012
TM_80
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
-1.804**
-1.761**
-2.992
-2.908
-1.643
-1.711
-6.274***
-6.124**
-0.803
-0.701
0.012
0.0152
0.1724
0.227
0.2208
0.2212
0.0026
0.0148
0.516
0.5822
0.178***
0.297***
-0.010
-1.150
-0.013
0.377*
-0.047
-0.808
0.042
-0.217
0.0097
0.0039
0.9655
0.2566
0.9447
0.0966
0.7971
0.1199
0.7948
0.6293
Size
0.061
0.065
0.129
0.159
-0.147
-0.166
0.471***
0.487***
-0.202
-0.202
0.3625
0.3332
0.5615
0.478
0.3698
0.2903
0.0043
0.004
0.1496
0.151
MTB
-0.002
-0.002
0.045
0.048*
0.000
0.000
-0.002
-0.002
-0.009
-0.009
0.103
0.1173
0.134
0.0765
0.876
0.9772
0.3255
0.2571
0.1792
0.2313
Cash
-0.090
-0.162
1.917
1.583
2.755**
2.738**
2.572
2.371
-1.272
-1.278
0.917
0.8531
0.2725
0.4308
0.0367
0.0431
0.1209
0.1467
0.6011
0.6096
Leverage
-1.024
-1.105
-4.230
-5.150
-0.907
-0.688
-2.363
-2.735
-0.295
-0.323
0.2853
0.2501
0.23
0.2092
0.4185
0.5551
0.4169
0.3416
0.8615
0.8494
1.664
1.561
-2.120
-3.141**
8.979**
9.28**
-6.063**
-6.274**
0.841
0.939
0.1126
0.1356
0.1437
0.0284
0.0126
0.0184
0.0256
0.02
0.6918
0.6565
-1.788
ROA
Operating Cash Flow
Institutional Investors
Industry FE
AUC
N. Obs.
0.122
0.427
-0.223
-0.779
-3.161
-3.491
8.271**
8.295**
-1.895
0.9376
0.7851
0.9582
0.8816
0.3213
0.3038
0.0363
0.0314
0.4738
0.4964
-1.218*
-1.148*
-3.593**
-3.298**
-1.411
-1.503
-2.755*
-2.723*
-0.697
-0.676
0.05
0.063
0.042
0.0286
0.2218
0.1987
0.0889
0.0974
0.3945
0.3994
Yes
0.66
1,132
Yes
0.68
1,132
Yes
0.90
1,121
Yes
0.91
1,121
Yes
0.82
1,092
Yes
0.84
1,092
Yes
0.90
1,054
Yes
0.92
1,054
Yes
0.71
1,018
Yes
0.71
1,018
40
Table 7. Robutstness tests.
This table presents results from two sets of robustness tests: (i) excluding companies with negative stock returns
in 2012 and (ii) adjusting built-in capital gains for restricted stock holdings. The dependent variable in columns
1 and 2 is an indicator for special dividends paid during 2012, while in columns 3 and 4 is an indicator for
accelerated dividends accelerated from the first quarter of 2013 to the fourth quarter of 2012. The Odds-Ratio
factor for an increase of 1 tax-minded executive is presented. It is calculated as the exponential of the coefficient
and is provided for interpretation purposes.
Dependent variable:
Robustness:
Model:
Intercept
No. Tax-Minded Execs.
Special
Price Declines
Restricted Stock
(1)
-0.876
(2)
-0.108
Acceleration
Price Declines
Restricted Stock
(3)
-1.176
(4)
-0.795
0.5082
0.9201
0.208
0.2827
0.271
0.483***
0.336**
0.277**
0.4346
0.0031
0.0196
0.0159
Size
-0.333**
-0.317**
-0.003
-0.043
0.0414
0.0131
0.9713
0.4691
MTB
-0.001
-0.001
-0.002
-0.003
0.6823
0.6014
0.2714
0.1616
Cash
-0.533
0.098
-1.240
0.252
0.771
0.9373
0.4097
0.806
0.615
-0.111
-2.141**
-1.322
0.1503
Leverage
0.6703
0.9296
0.0245
5.831***
7.295***
4.934*
2.598
0.0099
0.0007
0.0612
0.1616
-4.189*
-4.693**
-2.313
-1.165
0.0714
0.0132
0.3795
0.5651
-2.011
-1.831*
-0.485
-0.734
0.1868
0.0759
0.5144
0.2238
Industry FE
AUC
N. Obs.
Yes
1.03
1,032
Yes
1.52
999
Yes
0.08
978
Yes
0.19
728
Odds - Ratio: No. Tax-Minded Execs.
1.31
1.62***
1.39**
1.31**
ROA
Operating Cash Flow
Institutional Investors
41
Table 8. Summary Statistics, panel data 2000-2012.
This table presents the summary statistics for the panel analysis. Panel A presents firm-level characteristics.
Variable definitions are provided in Appendix Table A.1. Panel B presents the fraction of the sample that pertains
to each year from 2000 to 2012.
Panel A. Descriptive Statistics
Variable
Mean
Std. Dev.
Q1
Median
Q3
Cash ETR
TM_2012
TM_Dec
TM_80
Size
Institutional Investors
MTB
Leverage
Multinational
Intangibles
R&D
Advertising
0.24
1.60
0.14
0.46
7.56
0.75
2.98
0.20
0.51
0.18
0.04
0.01
0.18
1.35
0.45
0.79
1.58
0.20
24.09
0.17
0.50
0.19
0.17
0.03
0.11
0.00
0.00
0.00
6.42
0.64
1.45
0.05
0.00
0.02
0.00
0.00
0.23
2.00
0.00
0.00
7.46
0.79
2.19
0.18
1.00
0.12
0.00
0.00
0.33
2.00
0.00
1.00
8.57
0.90
3.47
0.29
1.00
0.29
0.03
0.01
Panel B. Time Composition of the Sample
Year
Fraction of the sample
2000
2.3%
2001
2.9%
2002
3.7%
2003
4.5%
2004
5.0%
2005
5.2%
2006
6.9%
2007
9.2%
2008
10.5%
2009
11.4%
2010
12.3%
2011
12.7%
2012
12.7%
42
Table 9. Cash tax avoidance, panel regressions 2000-2012.
This table presents the estimation results from linear regressions that take the following form:
𝐶𝑎𝑠ℎ 𝐸𝑇𝑅𝑖,𝑡 = 𝛼 + 𝛽(𝑁𝑜. 𝑇𝑎𝑥 𝑀𝑖𝑛𝑑𝑒𝑑 𝐸𝑥𝑒𝑐𝑠𝑖,𝑡 ) + 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖,𝑡 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐹𝐸 + 𝑌𝑒𝑎𝑟 𝐹𝐸.
The dependent variable in all models is Cash ETR. The independent variable is the number of tax-minded
executives present at each firm. The independent variable in column 1 is TM_2012, in column 2 is TM_Dec, and
in column 3 is TM_80. Variable definitions are provided in Appendix Table A.1. Standard errors are adjusted for
heteroskedasticity and autocorrelation. P-values are reported below the coefficients. *, **, *** indicate
significance at the two-tailed 10%, 5% and 1% levels, respectively.
Dependent variable: Cash ETR
Tax mindedness variable:
TM_2012
TM_Dec
TM_80
Model:
Intercept
(1)
0.2983***
(2)
0.2941***
(3)
0.2967***
<.0001
<.0001
<.0001
-0.0028*
-0.0016
-0.0065**
0.0896
0.6995
0.0172
0.0005
0.0004
0.0006
0.7635
0.7797
0.7071
-0.0298**
-0.0306**
-0.031**
0.0278
0.0237
0.022
0.0001*
0.0001*
0.0001*
0.098
0.0865
0.0912
-0.126***
-0.1245***
-0.1258***
<.0001
<.0001
<.0001
0.0019
0.0019
0.0017
0.7302
0.729
0.7496
0.032**
0.0325**
0.0322**
0.0127
0.0114
0.012
-0.2351***
-0.2387***
-0.232***
No. Tax-Minded Execs.
Size
Institutional Investors
MTB
Leverage
Multinational
Intangibles
R&D
Advertising
Industry FE
Year FE
N. Obs.
R2
<.0001
<.0001
<.0001
0.0767
0.0766
0.0881
0.3362
0.3357
0.273
Yes
Yes
6,744
7.62%
Yes
Yes
6,744
7.58%
Yes
Yes
6,744
7.65%
43