Azu Etd 14496 Sip1 M
Azu Etd 14496 Sip1 M
Azu Etd 14496 Sip1 M
by
__________________________
Copyright © Nathan Chad Goldman 2016
DEPARTMENT OF ACCOUNTING
DOCTOR OF PHILOSOPHY
2016
1
THE UNIVERSITY OF ARIZONA
GRADUATE COLLEGE
As members of the Dissertation Committee, we certify that we have read the dissertation
prepared by Nathan Chad Goldman, titled The Effect of Tax Aggressiveness on
Investment Efficiency and recommend that it be accepted as fulfilling the dissertation
requirement for the Degree of Doctor of Philosophy.
Final approval and acceptance of this dissertation is contingent upon the candidate’s
submission of the final copies of the dissertation to the Graduate College. I hereby certify
that I have read this dissertation prepared under my direction and recommend that it be
accepted as fulfilling the dissertation requirement.
2
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of the requirements for
an advanced degree at the University of Arizona and is deposited in the University
Library to be made available to borrowers under rules of the Library.
Brief quotations from this dissertation are allowable without special permission,
provided that an accurate acknowledgment of the source is made. Requests for
permission for extended quotation from or reproduction of this manuscript in whole or in
part may be granted by the head of the major department or the Dean of the Graduate
College when in his or her judgment the proposed use of the material is in the interests of
scholarship. In all other instances, however, permission must be obtained from the author.
3
ACKNOWLEDGEMENTS
I am thankful for all the support and encouragement I have received from my family,
friends, fellow doctoral students, faculty, co-authors, and dissertation committee
members over the past four years The University of Arizona.
I am particularly grateful to my dissertation committee: Dan Dhaliwal (chair), Katharine
Drake, and Jayanthi Sunder for their support and guidance on my dissertation. My
dissertation has also benefited from the helpful comments and suggestions of Brad
Badertscher, Erik Beardsley, Jeff Burks, John Campbell, Lin Cheng, Robert Chirinko,
Ted Christensen, Matthew Erickson, Andrew Finley, Max Hewitt, Stephen Lusch, Frank
Murphy, Jim Stekelberg, Bridget Stomberg, Erin Towery, Robert Whited, Jeff Yu, the
University of Arizona Tax Readings Group, group participants at the Deloitte and J.
Michael Cook Foundation AAA Doctoral Consortium, and workshop participants at the
George Washington University, North Carolina State University, Southern Methodist
University, University of Arizona, University of Connecticut, University of Georgia,
University of Illinois at Chicago, University of Notre Dame, University of Southern
California, and University of Texas at Dallas.
I would also like to thank all of the University of Arizona accounting faculty and doctoral
students for everything that I have learned from you all. I owe additional recognition to
the accounting (Matthew Erickson) and finance (Charles Favreau and Austin Shelton)
Ph.D. students who entered the Ph.D. program in Fall 2012. It was a pleasure to work
through the coursework with you. Additionally, I would like to acknowledge the doctoral
students in the classes above me, in particular, James Brushwood, Dane Christensen,
Andrew Finley, Curtis Hall, Timothy Hinkel, Scott Judd, David Kenchington, Mindy
Kim, Phil Lamoreaux, Stephen Lusch, Pablo Machado, Landon Mauler, and Sarah
Shaikh, for providing strong guidance to me throughout my time in the program. Lastly, I
would like to thank all of the former Arizona Accounting Ph.D. students who helped
make this program a great place to get a doctoral degree. I would like to specifically
thank Susan Albring, John Campbell, James Chyz, Merle Erickson, Fabio Gaertner,
Jenny Gaver, Shane Heitzman, Linda Krull, Bill Moser, and Logan Steele for the direct
support and guidance that they have provided during my time in the doctoral program.
Finally, words cannot express the gratitude I owe to my family for their love and support
during my time in the Ph.D. program. None of this would have been possible if it were
not for you.
4
DEDICATION
To my wife, Lindsay, for your relentless support in pursuing my dreams, and being the
best friend I could ever ask for
To my father, Jay, for being such a strong role model and motivator in being the best
person that I can be
To my mother, Jody, for always being so positive and willingness to selflessly serve
others
5
TABLE OF CONTENTS
LIST OF TABLES………………………………………………………………………..8
ABSTRACT………………………………………………………………………………9
I. INTRODUCTION……………………………………………………………………..10
II. LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT…………………16
Literature Review………………………………………………………………...16
Tax Aggressiveness Overview……………………………………………16
Agency Costs of Free Cash Flows……………………………………….20
Investment Decisions…………………………………………………….22
Hypothesis Development…………………………………………………………23
Tax Aggressiveness and Investment Efficiency………………………….23
Tax Aggressiveness, Investment, Efficiency, and Tax Monitoring……..25
Tax Aggressiveness, Investment Efficiency, and Tax Disclosure: A Quasi-
Natural Experiment………………………………………………………27
Buy-And-Hold Abnormal Return Analysis……………………………...28
III. RESEARCH DESIGN……………………………………………………………….30
Tax Aggressiveness……………………………………………………………....30
Model……………………………………………………………………………..31
Sample Selection…………………………………………………………………34
IV. PRIMARY ANALYSIS……………………………………………………………...36
Descriptive Statistics and Correlations…………………………………………..36
Hypothesis Testing……………………………………………………………….36
V. ADDITIONAL ANALYSES………………………………………………………….42
Other Cash Flow Sources………………………………………………………...42
Managerial Characteristics……………………………………………………….42
Alternative Measures of Tax Aggressiveness…………………………………….43
Unconditional Investment Efficiency Model…………………………………….43
VI. CONCLUSION………………………………………………………………………45
6
APPENDIX A: VARIABLE DEFINITIONS……………………………………………47
APPENDIX B: TABLE LISTING……………………………………………………….50
REFERENCES……………………………………………………………………………59
7
LIST OF TABLES
TABLE 1: Sample Selection……………………………………………………………..51
TABLE 2: Descriptive Statistics…………………………………………………………52
TABLE 3: Correlation Table……………………………………………………………..53
TABLE 4: The Effect of Tax Aggressiveness on Investment Efficiency………………..54
TABLE 5: The Effect of Tax Aggressiveness on Investment Efficiency by Tax
Aggressiveness and Tax Conservativeness………………………………......55
TABLE 6: The Effect of Tax Monitoring on the Relation between Tax Aggressiveness
and Investment Efficiency………………………………..…………………..56
TABLE 7: A Quasi-Natural Experiment Examining the Relation between Tax
Aggressiveness and Investment Efficiency: Pre Versus Post FIN 48…….…57
TABLE 8: Buy-And-Hold Abnormal Returns: Tax Aggressive Versus Tax Conservative
for Firms with High and Low Access to Investable Funds…………….…….58
8
ABSTRACT
Tax aggressiveness generates significant cash savings and information asymmetry.
aggressiveness is associated with higher agency costs of free cash flows that affect
investment decisions. Using the conditional investment efficiency model, I find evidence
that tax aggressiveness is associated with more investments in firms with high access to
also provide evidence that stronger tax monitoring and a change in tax disclosures
mitigate the relation between tax aggressiveness and overinvestment. Lastly, I find that
the overinvestment is associated with lower future abnormal returns. Thus, my results
to tax aggressiveness. Additionally, I further the need for shareholders and board of
directors to exert influence to avoid compensating managers for aggressive tax strategies.
9
I. INTRODUCTION
Investment choices are important firm decisions that have implications for firm
growth and value (Tobin, 1969; Tobin and Brainard, 1977). A necessary condition for
materializing these decisions is securing funds to pay for the investments. Among the
many internal avenues used to generate cash flows, aggressive tax planning strategies, or
and generate substantial operating cash flows (Mills, Erickson, and Maydew, 1998). For
example, relative to the statutory tax rate, in my sample a tax aggressive firm saves an
average of $98 million in cash taxes paid. In perspective, this large amount of savings
dwarfs the funds that could be saved from other discretionary decisions such as
eliminating R&D expenses ($56 million) or advertising expenses ($51 million). Since
the average firm in my sample invests $206 million per year, aggressive tax activities
investment decisions. Tax aggressiveness reduces cash taxes paid relative to a non-tax
aggressive firm, thereby increasing cash flows available for investment (Mills et al.,
1998). Shareholders often incentivize tax aggressiveness since it represents a shift in cash
flows from the government to the firm (Rego and Wilson, 2012). While tax savings may
be beneficial for some firms, the increase in cash increases the agency problem of free
1
I define tax aggressiveness as activities that are uncertain and have a high likelihood of drawing IRS
scrutiny. This definition is consistent with those used in prior and concurrent research (Hanlon and
Slemrod, 2009; Wilson, 2009; Hanlon and Heitzman, 2010; Lisowsky, 2010; Rego and Wilson, 2012;
Chyz, Leung, Li, and Rui, 2013; Lisowsky, Robinson, and Schmidt, 2013; Donohoe and Knechel, 2014,
among many others). In my primary analysis, I measure tax aggressiveness as the difference between firm’s
actual cash effective tax rate and a benchmark for its expected cash effective tax rate, consistent with
Balakrishnan et al. (2012) and Armstrong et al. (2015), and the discretionary permanent book-tax
differences (Frank et al., 2009). My inferences are robust to a variety of other commonly used measures.
10
cash flows (Jensen, 1986). Thus, tax aggressive firms have more cash flows, relative to
non-tax aggressive firms, without necessarily having better investment opportunity sets.
As a result, a tax aggressive firm has a greater risk of utilizing cash flows to increase
manager utility rather than shareholder value. Furthermore, this agency problem may be
(Balakrishnan, Blouin, and Guay, 2012; Hope, Ma, and Thomas, 2013). For example, tax
higher analyst forecast errors and analyst forecast dispersions.2 As a result, if firms
choose aggressive tax strategies that generate greater cash flows and external information
asymmetry, then tax aggressiveness may adversely affect investment decisions (Biddle
and Hilary, 2006; Biddle, Hilary, and Verdi, 2009; Cheng, Dhaliwal, and Zhang, 2013).
model. This model conditions on firms’ access to investable funds to examine whether a
Balakrishnan, Core, and Verdi 2014). I proxy for tax aggressiveness using two measures:
the difference between expected and actual cash effective tax rates (Armstrong, Blouin,
DTAX (Frank, Lynch, and Rego, 2009). Conditional on high access to investable funds, I
2
This untabulated analysis is performed by splitting firms into high, medium and low tax aggressiveness
groups. Using a t-test to compare the mean analyst forecast error and dispersion for the high and low
groups, I find that analyst forecast error is 15.12% (p < 0.10) higher and analyst forecast dispersion is
17.42% (p < 0.05) higher for tax aggressive firms relative to tax conservative firms. This analysis
complements Balakrishnan et al. (2012) and Hope et al. (2013), by confirming that tax aggressiveness is
associated with higher external information asymmetry in my sample population.
11
predict that tax aggressive firms’ higher cash flows (Mills et al., 1998) and greater
information opacity (Balakrishnan et al., 2012; Hope et al., 2013) are associated with
conditional on firms’ lack of access to investable funds, tax aggressive firms may use the
cash flows from tax savings to address financial constraint (Edwards, Schwab, and
Shevlin, 2015), thereby having more subsequent year investments. Conversely, these
firms may also have lower investments due to precautionary savings (Hanlon, Maydew,
and Saavedra, 2014) and more expensive external capital (Hasan, Hoi, Wu, and Zhang,
2014; Hutchens and Rego, 2015), which extant literature interprets as underinvestment.
on investments is just one possible mechanism to explain the relation between tax
since firms could structure investments to lower cash taxes paid or increase permanent
book-tax differences. Hence, I conduct other analysis to delineate the effect of tax
firms that have strong tax monitoring using firms that engage their external auditor for
tax services. If a firm engages its auditor for tax services, thus prompting board of
director approval of those specific tax positions, then the firm has stronger monitoring of
its tax activities. Thus, I posit that firms with auditor-provided tax services (APTS)
exhibit a weaker relation between tax aggressiveness and investment efficiency, relative
onset of FIN 48, a FASB pronouncement which requires firms to disclose significantly
more information about their aggressive tax positions. Given the exogenous shock to the
12
quality and quantity of tax information disclosed, I expect the effects of tax
period, relative to the pre-FIN 48 period. Finally, I study the link between tax
aggressiveness, investment efficiency, and lower firm value by examining whether tax
aggressiveness is associated with lower buy-and-hold abnormal returns (BHAR) for firms
expectations for firms with more investable funds, tax aggressiveness is associated with
more investment. Following the prior literature that uses and develops the conditional
investment model, the positive coefficient suggests that these firms overinvest (Biddle et
al., 2009; Cheng et al., 2013). In addition, I do not find consistent evidence that tax
aggressive firms with low access to investable funds have different subsequent year
investments. This evidence suggests that tax aggressiveness may not be associated with
underinvestment. Furthermore, I yield results that the relation between tax aggressiveness
and overinvestment is significantly mitigated for firms with APTS as well as for firms in
the post-FIN 48 period. These findings are important to address concerns surrounding
year BHAR and find tax aggressive firms with strong access to investable funds have
lower future shareholder returns. This evidence suggests that for firms with access to
investable funds, the value diminishing investment activities resulting from tax
aggressiveness outweigh the cash flow benefits. Lastly, in additional analyses I provide
13
evidence that other forms of discretionary savings (i.e. cutting R&D or advertising
managerial ability or overconfidence. I also provide evidence that the findings are robust
This study makes several contributions to the literature. First, I document a non-
tax consequence of tax aggressiveness. Hanlon and Slemrod (2009) and Graham, Hanlon,
Shevlin, and Shroff (2014) identify adverse capital market consequences for firms
engaging in tax aggressiveness. However, these studies focus on the market reaction to
the perception of tax aggressiveness, rather than the outcomes associated with tax
aggressiveness. I extend the tax aggressiveness literature by examining how tax behavior
may negatively affect profitability, I offer a potential explanation for the adverse market
Also, I extend the literature examining the agency costs of tax aggressiveness.
Extant finance, law, and accounting literature claim that tax planning strategies generate
significant agency costs (Desai and Dharmapala, 2006; Desai, Dyck, and Zingales, 2007;
(Armstrong et al., 2015), and generalizability (Blaylock, 2015) limit the usefulness of
their evidence. In a departure from the literature, I examine the effect of tax
with greater cash flows and more information opacity, I posit and find evidence that tax
decisions may be interpreted as inefficient, and thus a significant non-tax cost of tax
14
aggressiveness that likely hurts firm value (Tobin, 1969; Tobin and Brainard, 1977), I
provide evidence that tax aggressiveness generates significant agency costs that may
research suggests that pre-tax versus after-tax earnings metrics may influence CEO tax
risk preferences (Crocker and Slemrod 2005; Gaertner 2014; Powers, Robinson, and
Stomberg 2015; Brown, Drake, and Martin, 2015). Given my evidence that tax
directors may want to design executive pay structures to limit tax aggressiveness.
theoretical model of the agency costs of tax aggressiveness. Specifically, their results
suggest that stricter executive contracts can act as a mechanism to limit the value-
consistent with their propositions because my results suggest that tax aggressiveness is
associated with value-decreasing decisions. As a result, I further their call for investor
literature review and hypothesis development. Section III discusses the data and research
15
II. LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT
Literature Review
Hanlon and Heitzman (2010) suggest that tax strategies exist on a continuum with
lower explicit tax savings, and perfectly legal positions are at one end, and higher explicit
tax savings and questionable legal positions are at the other end. They further state that “a
tax planning activity or a tax strategy could be anywhere along the continuum depending
on how aggressive the activity is in reducing taxes” (pg. 137). Examples of tax
aggressiveness include sheltering activities (Weisbach, 2002; Wilson, 2009; Dyreng and
Lindsey, 2009; Lisowsky, 2010; Lisowsky et al., 2013), positions with uncertain IRS
audit outcomes (Rego and Wilson, 2012), and complex financial reporting (Frank et al.,
2009; Mills, Robinson, and Sansing, 2010; Donohoe and Knechel, 2014).
Tax aggressiveness may take many forms. For example, a New York Times article
liabilities. What makes these actions unique from other non-aggressive tax planning
is that “The Double Irish” attracted regulatory scrutiny, and it was unclear whether Apple
16
would be able to keep all the funds it obtained through the tax activities. Therefore, this
complex, but not uncommon, tax planning technique may be labeled as aggressive.3
Tax aggressiveness has the potential to be a beneficial firm financial activity. For
instance, Mills et al. (1998) provide evidence that for every dollar invested in tax
planning, the firm saves an average of four dollars in tax liabilities. This result suggests
increase cash flows through lower explicit taxes. Another New York Times article on
Apple anecdotally substantiates this positive effect. It states, “Even as Apple became the
nation’s most profitable technology company, it avoided billions in taxes in the United
States and around the world” (Schwartz, 2013). Since tax expense is often one of the
largest expenses on firms' income statements, it appears reasonable that tax planning
benefits are substantial. Furthermore, Robinson, Sikes, and Weaver (2010) document that
many firms consider their tax department as a profit center. Essentially, firms
increasingly view taxes as a contributor to the bottom line rather than as a measurement
system designed to minimize costs. Additionally, Goh, Lee, Lim, and Shevlin (2016) find
that tax planning is associated with a lower cost of equity capital. Lastly, Edwards et al.
(2015) provide evidence that tax planning can be used as an internal source of financing,
as shown by tax savings allowing constrained firms to access good investment projects.
consequences of tax aggressiveness. For example, Rego and Wilson (2012) identify
3
This example is merely one of many possible tax aggressiveness activities. Large multinational firms are
likely to be engaging in countless tax planning transactions on an annual basis that can range from
aggressive to conservative. While it is outside of the scope of this study to document all tax transactions in
which firms engage in, I do provide two proxies that capture the approximate annual level of firm tax
aggressiveness.
17
several direct costs to tax aggressiveness, such as fees paid to accountants and attorneys,
employee time spent resolving IRS audits, or even the penalties paid when the IRS
“challenges” or “overturns” an aggressive tax position. These direct costs are further
substantiated by Hoopes, Mescall, and Pittman (2012) who find that tax aggressiveness is
positively associated with I.R.S. audits and corresponding penalities. Other studies
examine indirect or non-tax costs of tax aggressiveness. For instance, Scholes, Wilson,
and Wolfson (1990) provide evidence that firms consider non-tax costs within tax
and minimizing non-tax costs. Other non-tax costs of tax aggressiveness include adverse
(Balakrishnan et al., 2012; Hope et al., 2013), more expensive capital (Hasan et al., 2014;
Hutchens and Rego, 2015), higher firm risk (Frank et al., 2009; Guenther, Matsunaga,
and Williams, 2015), and higher external audit fees (Donohoe and Knechel, 2014).
example, Scholes, Wolfson, Erickson, Hanlon, Maydew, and Shevlin (2014) state:
“Aggressive tax planning and tax shelters are structured so as to obfuscate the
underlying transaction so that the Internal Revenue Service has difficulty
identifying the transaction and fully unraveling the transaction. Such complex
transaction structuring could also obfuscate management’s actions and obscure
underlying firm performance in the financial statements, thus facilitating
opportunism or even rent extraction by management. (Pg. 133)
They suggest that tax aggressiveness is both an agency cost of free cash flows problem
and an external information asymmetry problem (Slemrod, 2004; Chen and Chu, 2005;
Crocker and Slemrod, 2005). Numerous studies empirically examine the possibility that
tax strategies as a whole result in higher agency costs. For example, Desai and
Dharmapala (2006) examine the link between corporate tax planning and managerial
18
incentives. The study finds evidence of a complementary relation between rent-extraction
and tax sheltering, as primarily demonstrated by firms with weak corporate governance.
These findings are further delineated by Desai et al. (2007), who find that rent extraction
through tax strategies and corporate governance are inversely related. Lastly, Desai and
Dharmapala (2009) provide evidence that tax planning strategies are associated with a
higher firm value, but only for firms with strong corporate governance.4
empirical specification. Specifically, Blaylock (2015) suggests that some of the prior
findings may not apply to firms in the United States since Desai et al. (2007) focus on
firms in Russia, a country with weak investor protection. Furthermore, Armstrong et al.
(2015) identify numerous limitations of the prior studies (i.e. adequately identifying
governance mechanisms, non-linear relation, etc…), thus casting doubt on the theoretical
framework of Desai and Dharmapala (2006), Desai et al. (2007), and Desai and
help, answer this empirical question, I use tax aggressiveness. Because tax
aggressiveness is a subset of tax strategies that particularly affects both cash flows and
information opacity, it may be a more specific construct to examine the effect of tax
strategies on managerial decision making since it is a setting where I would expect the
4
Numerous other studies also investigate the relation between agency costs and tax planning strategies (i.e.
Wilson, 2009; Chen, Chen, Cheng, and Shevlin, 2010; Kim, Li, and Zhang, 2011; Donohoe and McGill,
2011). Each of these studies use the framework developed by Desai and Dharmapala (2006), Desai et al.
(2007), and Desai and Dharmapala (2009).
5
Gallemore and Labro (2015) find evidence of a positive relation between internal information quality and
tax avoidance. This also questions established theory. Because Gallemore and Labro’s (2015) findings are
primarily related to internal (rather than external) information asymmetry, they may not translate to my
research setting. However, it is important to acknowledge their findings.
19
relation to be especially strong. Furthermore, I concentrate on how tax aggressiveness
affects the agency costs of free cash flows, an area which has implications, and
investment decisions.6
Jensen (1986) highlights and analyzes the agency costs of free cash flows. These
costs increase when cash flows exceed profitable investment opportunities. While
shareholders prefer payouts to reduce resources under managers’ control, managers have
incentives to grow firms. This growth increases the resources under the manager’s
control, thus yielding more compensation and executive power over the firm (Jensen and
Meckling, 1976). As a result, Jensen (1986) suggests that managers and shareholders may
possess different preferences over what to do with excess cash flows. This difference in
Numerous studies document the adverse consequences of excess cash flows.7 For
firms with excess free cash flows. This finding is consistent with Jensen’s (1986)
discussion of oil and gas firms investing in value decreasing diversification after a cash
flow surprise. Also, Lang, Stulz, and Walking (1991) provide evidence that successful
6
I acknowledge that tax aggressiveness is just one of numerous ways for firms to increase internal
liquidity. Even narrowing down to discretionary managerial decisions, tax aggressiveness is similar to
managers cutting spending, such as R&D or advertising expenses. Different from other accounts, tax
aggressiveness generates information opacity (Balakrishnan et al., 2012; Hope et al., 2013), in addition to
the prevailing cash flows. I empirically test the differences in additional analysis.
7
It is also worth noting that several studies document significant benefits to excess cash flows, such as
more effective leverage adjustments (Faulkender, Flannery, Hankins, and Smith, 2012), greater asset
tangibility to increase investment ability (Almeida and Campello, 2007), and stronger ability to invest in
positive net present value projects among constrained firms (Dennis and Sibilkov, 2010). Because I expect
the information opacity from tax aggressiveness to adversely affect the cash flows, I focus my review on
the negative consequences.
20
tender offers for firms with high cash flows and low investment opportunities incur
negative abnormal returns because concerns exist over what managers may do with these
cash flows. Additionally, Leuz, Triantis, and Wang (2008) find that excess free cash
flows are a common reason firms deregister from the SEC and that deregistering is a
Other studies use the relation between investments and cash flows to examine
whether firms are investing efficiently (Fazarri, Hubbard, Peterson, 1988; Kaplan and
Zingales, 1997; Biddle and Hilary, 2006). The general presumption in these models is
that firms follow q-theory, which states that firms should formulate investment decisions
based upon investment opportunities, rather than cash flows (Tobin, 1969; Tobin and
Brainard, 1977). Thus, a significant positive relation between investments and cash flows
may suggest inefficient investment. Lastly, excess cash flows lead to a higher likelihood
of holding more cash, which can have significant costs related to the agency costs of free
cash flows (e.g. Harford, 1999; Faulkender and Wang, 2006; Acharya, Davydenko, and
Strebulaev, 2012).
An area that has received some attention is how tax planning strategies influence
the agency costs of free cash flows. Different from changes to revenues that affect cash
flows (i.e. Lamont, 1997) tax planning strategies are unique in that they are voluntary
activities that generate cash flows, which can be used to affect firms. For example,
Edwards et al. (2015) document that financially constrained firms choose tax planning
strategies to increase internally-generated funds. Law and Mills (2015) find similar
results while examining the relation between financial constraints and the financial
reporting consequences of tax planning. However, tax related activities may also generate
21
agency costs of free cash flows. For instance, Hanlon, Lester, and Verdi (2015) and
Edwards, Kravet, and Wilson (2015) provide evidence that firms use foreign cash
examines the market reaction to acquisitions made by tax avoiding firms and finds that
they are, on average, negative relative to non-tax avoiding firms, suggesting that
investors are innately concerned with how managers use tax savings. This evidence
highlights that cash savings from tax strategies might generate agency costs of free cash
Investment Decisions
The literature suggests that firms should invest until the marginal benefits of
capital investment equal the marginal costs (Yoshikawa, 1980; Hayashi, 1982; Abel,
1983). Extant research documents that firms depart from this investment theory due to
Shleifer, Vishny, 1994; Hope and Thomas, 2008) and adverse selection (Myers and
Majluf, 1984; Baker, Stein, Wurgler, 2003). Recent literature examines these agency
costs and investigates how firm-specific attributes affect capital investment decisions
such as financial reporting quality (Biddle and Hilary, 2006; Biddle et al., 2009;
Balakrishnan et al., 2014) earnings management (McNichols and Stubben, 2008), and
internal control weaknesses (Cheng et al., 2013). Notably, Biddle et al. (2009) and Cheng
et al. (2013) use the conditional investment efficiency model to provide evidence on
whether a firm-specific attribute (i.e. accounting quality) affects firm decisions to invest
sub-optimally in negative net present value projects or not to invest in positive net present
22
investment. Because investments comprise a large firm expenditure each year, investment
Hypothesis Development
Firms that choose aggressive tax strategies have more cash flows, relative to tax
conservative firms (Mills et al., 1998). Shareholders often incentivize managers to choose
aggressive tax strategies since it represents a shift in cash flows from the government to
the firm (Rego and Wilson, 2012). In addition to the higher cash flows, tax
managers must conceal aggressive actions from the regulatory agencies, thereby also
concealing actions from shareholders (Desai and Dharmapala, 2006; Desai et al., 2007;
Desai and Dharmapala, 2009; Balakrishnan et al., 2012; Hope et al., 2013). I posit the
combination of increased cash flows and increased external information asymmetry has
Biddle et al. (2009), Cheng et al. (2013), and Balakrishnan et al. (2014) by focusing on
investable funds. I posit that for firms with access to investable funds,9 the additional free
cash flows from tax aggressiveness represent an agency concern because management is
8
A recent anecdote from The Texas Tribune cites that the transition of Oncor from a public company to a
real estate investment trust, a common tax planning strategy, is expected to increase investments by way of
lower cash taxes paid. While utility firms are not included in my sample due to their membership in a
regulated industry, this anecdote is just one example of firms using tax planning to affect investment
decisions (Malewitz, 2016).
9
I define firms having access to investable funds as those firms with high amounts of cash holdings and
low amounts of leverage. This is consistent with prior studies that use the conditional investment efficiency
model (Biddle et al., 2009; Cheng et al., 2013; Balakrishnan et al., 2014).
23
incentivized to spend these funds on growing the size of the firm, even if such growth is
suboptimal (Jensen, 1986; Hope and Thomas, 2008), or other self-serving behavior.
Ceteris paribus, I posit, managers choose aggressive tax strategies that generate
excess cash flows. However, due to the inherent complexity of these strategies and
incentives to hide such activities from monitors and tax authorities, these positions are
difficult to understand and increase external information asymmetry. This difficulty leads
associated with more subsequent year investment. Extant literature using this model
suggests that this result is evidence of overinvestment. As a result, I make the following
overinvestment.
While there is a clear direction for the H1a theoretical prediction, the relation
between tax aggressiveness and subsequent year investments when there is low access to
investable funds is less clear. For example, for firms with low access to investable funds,
tax aggressiveness may be associated with higher subsequent year investment because
they are currently less able to access external markets. As a result, the additional cash
flows may provide these firms the opportunity to fund positive net present value projects
10
A recent Wall Street Journal article anecdotally substantiates the presence of opacity (Mann, 2015). The
article suggests that General Electric has long used GE Capital to reduce its effective tax rates by way of
convoluted and complex overseas subsidiaries. However, most expert analysts are unable to identify the
mechanisms for these strategies. As a result, the effects of tax aggressiveness on taxes saved and
information opacity appear consistent with practice.
24
that the firm may have otherwise not been able to finance (Edwards et al., 2015).
Conversely, low access to investable funds may result in lower subsequent year
investment due to the incentives to exercise precaution with the additional cash flows in
case they need to be used towards future tax claims on prior and current uncertain tax
positions (Hanlon et al., 2014).11 The cash savings motive, combined with the more
expensive capital associated with tax aggressiveness (Hasan et al., 2014; Hutchens and
Rego, 2015) may intensify firm capital constraints. Together, these two explanations may
or underinvestment.
Alternative explanations may exist for the relation between tax aggressiveness
identified as firms that use their external auditor for tax services (APTS).13
11
While the precautionary motive theory exists for all firms, I expect it to more significantly affect firms
with low access to investable funds because they are already more likely to spend less. Furthermore, to the
extent this precautionary motive affects firms more prone to overinvest, it would likely bias against finding
significant results for H1a.
12
Similar to H1a, I use the conditional investment efficiency model (Biddle et al., 2009). Different from
H1a, I examine firms with low amounts of cash holdings and high amounts of leverage.
13
Prior literature uses a cross-section on firm governance, proxied as the G-Score, to examine whether
managerial oversight mitigates the agency costs of tax planning activities (e.g. Desai and Dharmapala,
2006; 2009). Because general firm governance can significantly affect both tax planning activities and
investment decisions, I use APTS so that I may more directly examine the consequences to the decision to
choose a tax strategy, rather than all firm decisions.
25
Title II of the Sarbanes-Oxley Act of 2002 focuses on enhancing auditor
independence. In this chapter, Section 201 specifically focuses on services that external
auditors may no longer provide (e.g. bookkeeping, information systems design and
implementation, actuarial services, etc…). One gray area is tax services. Section 201
states:
Firms have numerous outlets for implementing a tax strategy, including their
many differences exist between each of these avenues, the most discernable difference is
that APTS requires additional approval, and thus greater information transparency, by the
firms’ board of directors. The information that the external auditor must provide to the
audit committee is not trivial as substantiated by Klassen, Lisowsky, and Mescall (2015),
which finds that firms without APTS take advantage of manager to board of directors
information asymmetry.14 Their result is that firms that do not use APTS are more tax
I posit that the additional disclosure of this information to the audit committee
may have two outcomes. First, the approval stage gives the audit committee the
opportunity to decline the service if it is too aggressive. As a result, for firms with APTS,
an approved tax position may be less aggressive relative to a position not provided by its
auditor. More importantly, the APTS approval provides a greater external insight into the
14
See PCAOB Rule 3524 for a full description of the information required to be provided to the audit
committee.
26
firms tax strategies, thus potentially mitigating the adverse effects of tax aggressiveness
on managerial decisions. These two possible actions lead me to suggest that APTS
H2: Firms with APTS have a significantly weaker relation between tax
Experiment
Even though I lessen the potential for an endogenous relation by examining how tax
FIN 48 was enacted beginning in 2007 and required firms to separately disclose
their reserve for uncertain tax positions. Previous disclosure requirements under SFAS 5
required firms to include this reserve in their contingent liability reserves. The prior
disclosure was not disaggregated. Thus, FIN 48 marked a notable, and exogenous,
related to uncertain and potentially aggressive tax positions (Blouin, Gleason, Mills, and
15
Roberts and Whited (2012) suggest that a natural experiment can be used to help mitigate endogeneity
concerns as well as generate causal inferences in empirical corporate finance research. Furthermore, Gow,
Larcker, and Reiss state “papers using these methods [quasi-experimental methods] are considered stronger
research contributions (pg. 11).”
16
FIN 48 marks a plausible quasi-natural experiment because the passage of this pronouncement was not
certain. In the years leading up to its passage, a significant number of objections were raised, thus clouding
the likelihood of approval (Erickson et al., 2015).These actions run parallel to Michels (2015), who exploits
27
I posit that the additional disclosure of the tax financial statement information
informs external parties regarding the aggressive nature of firm tax strategies. Similar to
APTS, this may result in firms not choosing tax positions that are as aggressive.
However, for firms that make these disclosures in 2007 and are therefore continuing to
choose aggressive tax positions, I suggest that the disclosure raises financial statement
user awareness for the cash flows generated from tax aggressiveness. The result may be
that managers are less able to use the additional cash flows towards negative net present
value projects. This theory leads me to suggest that, for firms that disclose a FIN 48
2006.
Under q-theory, perfectly efficient firms invest in all positive net present value
projects and do not invest in all negative net present value projects (Tobin, 1969; Tobin
and Brainard, 1977). As a result, firms investing efficiently are expected to have higher
firm value, relative to firms investing inefficiently. I apply this theory to my setting. If
inefficiency, then I anticipate that a tax aggressive firm has a lower market adjusted
return relative to a tax conservative firm. Consistent with the primary analysis, I expect
a difference in disclosure requirements. Michels (2015) is a study highlighted by Gow et al. (2016) for
having adequately executed a natural experiment research design to infer causality.
28
this effect is most pronounced among firms with high and low access to investable funds
because the extant literature suggests that these firms are more prone to inefficient
investment (Biddle et al., 2009; Cheng et al., 2013). However, it is also possible that the
benefits from tax aggressiveness mitigate these costs. For example, if a firm is tax
aggressive and uses the additional cash flows to purchase a negative net present value
asset, then it is not clear that the firm would have lower returns (i.e. tax aggressiveness
may insignificantly affect firms value) since the firm essentially transfers funds from the
I.R.S. to the seller of the asset. In fact, firms may still benefit from tax aggressiveness if
some of the cash flows are efficiently allocated to shareholders. As a result, I make the
29
III. RESEARCH DESIGN
Tax Aggressiveness
I proxy for tax aggressiveness two ways: (1) the difference between a firm’s
expected and actual cash effective tax rate, or DiffETR (Balakrishnan et al., 2012;
(Frank et al. 2009). These two definitions follow the literature that defines tax
aggressiveness as “a subset of tax positions have weak support (Lisowsky et al., 2013, pg.
589)” or positions that are “pushing the envelope of tax law (Hanlon and Heitzman, 2010,
pg. 137).”
For DiffETR, I define expected cash effective tax as the three-year average cash
effective tax rate for each industry, year, size-decile grouping of firms. I consider this
average to be the expected rate for all firms in that particular group. I define actual cash
effective tax rate as the three-year cash effective tax rate (Dyreng, Hanlon, and Maydew
2008). For each observation, I subtract the actual cash effective tax rate from the
expected cash effective tax rate to generate tax aggressiveness, where a positive
(negative) DiffETR is considered aggressive (conservative) since it suggests that the firm
has a lower (higher) cash effective tax rate than its industry-year-size decile would
suggest. Lastly, I scale the difference by the expected cash effective tax rate to generate
the percentage difference. I define DTAX consistent with Frank et al. (2009), which is the
tax-differences unrelated to aggressive tax planning. The regression is run by year and
industry and generates a residual which is positive (negative) if the firm has more
30
(conservativeness). See the Appendix for a more detailed explanation of the DiffETR and
DTAX calculations. 17
Model
(Biddle et al., 2009; Cheng et al., 2013; Balakrishnan et al., 2014). This model examines
the relation between tax aggressiveness and the subsequent year level of capital
Investmentt+1 measures subsequent year capital and non-capital investment. DiffETR and
DTAX measures the firm-specific tax aggressiveness. Lastly, Overfirm proxies for access
to investable funds. This measure uses cash holdings and leverage to distinguish between
its available investment opportunities. For example, if a firm has high cash and low
leverage (i.e., high OverFirm) then it has cash that can be readily spent and may be
under-levered (Graham, 1996; Blouin, Core, Guay, 2010). These conditions suggest that
the firm can invest in positive net present value projects if they exist, and the presence of
17
To further validate these measures, I examine the sample and find that well-known tax aggressive firms
within my sample (e.g., Starbucks, Apple, Amazon, etc…) have positive DiffETR and DTAX.
18
Operationalized, I separately rank firm cash holdings and debt from one to 1,000. To keep the directions
consistent, debt is multiplied by negative one so that a high value of OverFirm represents a firm with high
cash and low leverage, thus having high access the investable funds. I then sum each firms score and rank
the scores from 1 to 11. Finally, I adjust the groups to take a value from 0 to 1 in increments of 0.1. This
formulation is similar to that used in Cheng et al. (2013). In addition, it allows the effect for firms with the
lowest value of OverFirm to be captured by β1, and firms with the highest value of OverFirm to be
captured by β1 + β3 (Burks, Randolph, and Seida, 2015). See additional discussion of the composition of
OverFirm in the Appendix.
31
such attributes means that the firm is more likely to have exhausted the possibilities to
invest in projects that are positive net present value. On the contrary, if a firm has low
cash and high leverage (i.e., low OverFirm), then the firm does not have cash that can be
spent and might have limited additional access to the debt market. These conditions
suggest that the firm is more likely to have quality investment projects available, but may
not have the ability to finance the projects. In summary, a high OverFirm may indicate
that a firm has limited positive net present value projects available, and a low OverFirm
may indicate that a firm has more positive net present value projects available, thus firms
at the extremes are more at risk for tax aggressiveness affecting investment efficiency
aggressiveness. First, I control for firm maturity. Specifically, I posit that large mature
firms that are more stable are significantly less likely to purchase assets in the future. As
a result, I expect firm size (Size), likelihood of bankruptcy (ZScore), and firm age (Age)
controls for firm growth opportunities, as proxied by return on assets (ROA), cash-flows
from operations (CFO), Tobin’s Q (TobinQ),19 sales growth (SalesGrowth) and foreign
subsequent year investments. Additional variables that may positively affect subsequent
year investments include firm tangibility (PPE) and financial constraint (Rating).
Meanwhile, other variables that may negatively affect subsequent year investments
19
Hayashi (1982) and Chirinko (1993) argue that the inclusion of Tobin’s Q as a control variable may not
be appropriate because it is a forward-looking variable that may reflect all firm characteristics, including
those related to tax aggressiveness. In untabulated analysis, I remove TobinQ as a control variable, and note
that my inferences remain unchanged.
32
include sales volatility (StdDevSales), institutional holdings (TotalIO), and dividend
issuance (Dividend). Each of these control variables and predictions follow prior
literature (see Biddle and Hilary, 2006; Biddle et al. 2009; Cheng et al., 2013).
I control for percentage of compensation derived from equity earnings since managers
with high equity compensation (EqComp) are more likely to grow the size of the firm
(Murphy, 1999; Coles, Daniel, and Naveen, 2006). As a result, I expect EqComp to be
positively associated with subsequent year investments. Additionally, I control for the
CEOs age (CEOAge) because older executives have a shorter incentive horizon (Dechow
and Sloan, 1991; Yim, 2013) and may invest less. Thus, I expect CEOAge to be
investments are sticky from year to year. Therefore, I control for current year investments
See the Appendix for a more detailed definition of each the control variables. I also
include industry and year fixed effects and cluster standard errors at the firm level.
Beginning with H1a, the sum of β1 and β3 measures the effect of DiffETR / DTAX
on Investmentt+1 when firms have high access to investable funds (i.e. when OverFirm
takes the value of 1). As a result, I expect a positive and significant β1 + β3. For H1b, the
have low access to investable funds (i.e. OverFirm takes the value of 0). In corollary with
the null hypothesis for H1b, I do not make a signed prediction for β1. While the direct
interpretations of the model correspond to the affect tax aggressiveness has on subsequent
year investment, prior literature argues that the results can speak to firm investment
33
efficiency (Biddle et al., 2009; Cheng et al., 2013). Following these studies, a positive
and significant β1 + β3, or a negative and significant β1 suggests the firm may be investing
significant β1 suggest the firm may be investing more efficiently. Additionally, H2 and
H3 follow a similar research design while splitting the sample on firms with or without
relative to non-tax aggressive firms. Following the literature, I calculate BHAR as the
investment in a portfolio of similar firms in the same period (Barber and Lyon, 1997). I
use BHAR from year t+2 so that the investment decisions can be realized.21 For this
analysis, I rank firms into groups based on decile of TAX and compare firms in the lowest
do not make signed predictions, but a t-test showing high DiffETR / DTAX is greater than
low DiffETR / DTAX when OverFirm equals 1 (0) would be consistent with firms having
high (low) access to investable funds making value decreasing investment decisions with
Sample Selection
Table 1 summarizes the sample selection procedure. The initial sample consists of
all Compustat and Execucomp firms with fiscal year ends between 1992 and 2014. The
20
I define firms as having APTS if their tax fees are in excess of 1% of their total audit fees. Requiring
firms to have more than a trivial amount of tax fees mitigates the concern that the APTS are not significant
services that do not trigger the attention of the audit committee.
21
In untabulated analysis, I use BHAR from year t + 3 and the aggregate BHAR from both t+2 and t+3, and
the inferences remain unchanged.
34
sample period begins in 1992 subsequent to the implementation of SFAS 109. The
sample period ends in 2014 since that is the most current data available I also exclude
observations in regulated industries (Fama-French 48 #’s 31,32, 44, 45, 46, 47, and 48).
Additionally, I exclude observations with negative pre-tax income over a rolling three-
year window from t-2 to t as well as negative pre-tax income in year t. Lastly, I exclude
observations that do not have enough information to calculate DiffETR, DTAX, or any of
the other variables used in this study. See the Appendix for a full list of these variables.
After removing observations that do not meet all of the criteria, I am left with a total
35
IV. PRIMARY ANALYSIS
Table 2 presents the descriptive statistics for the sample population. The mean
expenditures equal about 14% of firm assets each year. The average of the two primary
dependent variables is -0.005 and 0.045 for DiffETR and DTAX, respectively. By
construction, these two variables should approximate to 0. However, they are calculated
on a full sample of firms in the COMPUSTAT database with available data to calculate
each variable, and before other sample cuts in the study. As a result, the minor deviation
ensure equal distribution, and accordingly takes an average value of 0.500. In examining
Table 3 reports the Pearson and Spearman correlations for my sample population.
Consistent with expectations, DiffETR, DTAX and OverFirm are positively correlated
with Investment (p < 0.01). This result provides preliminary evidence that tax
subsequent year Investmentt+1 (p < 0.01). This correlation suggests that investments are
sticky from year to year and controlling for the prior year activity is appropriate. In
Hypothesis Testing
In Table 4, I present the results of estimating Model (1). First, I find that the
36
significant (p < 0.01), and insignificant when examining DTAX. This result yields some
evidence for H1a that tax aggressiveness is associated with more investment for firms
insignificant in both specifications. As a result, I fail to reject the null hypothesis for H1b.
Overall, Table 4 provides limited evidence that tax aggressiveness is associated with
overinvestment.
While the findings in Table 4 are not completely in line with the hypothesized
results, I posit that the lack of significant results may be due to a non-linear relation
between investment decisions and tax aggressiveness. Numerous studies examine the
effects of tax aggressiveness and find that it has a non-linear relation to firm
characteristics like the cost of equity capital (Cook, Moser, and Omer, 2015), or even
how tax aggressiveness affects the relation between managerial incentives and firm
problematic because it would suggest that tax aggressive firms invest less efficiently,
while tax conservative firms invest more efficiently, a finding that runs counter to prior
research documenting the positive effects of reducing cash taxes paid (Mills et al., 1998;
DiffETR and DTAX. I then estimate Model (1) separately for each of the four groups
(DiffETR > 0, DiffETR < 0, DTAX > 0, DTAX < 0), and present my findings in Table 5.
Column (1) and (2) present the findings when DiffETR and DTAX are greater than 0 (tax
aggressive subset), while Columns (3) and (4) present the findings when DiffETR and
DTAX are less than 0 (tax conservative subset). When examining only tax aggressive
37
firms, I find evidence consistent with H1a. Specifically, for this subset of firms, I find
that conditional on high access to investable funds, tax aggressiveness (DiffETR, p <
0.01; DTAX, p < 0.05) is associated with higher subsequent year investments. Because
these firms are more likely to have already exhausted positive net present value projects,
the prior literature suggests that these firms are investing inefficiently by way of
overinvestment (Biddle et al., 2009; Cheng et al., 2013). Additionally, I find limited
evidence to reject the null hypothesis for H1b that firms with low access to investable
funds that are tax aggressive have lower subsequent year investments (DiffETR, p <
0.05). In addition to statistical significance, the results are also economically significant.
For example, a one standard deviation increase in DiffETR (DTAX) is associated with
36.0% (33.5%) increase in subsequent year investments for firms with high access to
investable funds, and a 14.4% (12.5%) decrease in subsequent year investments for firms
Meanwhile, for firms that are tax conservative, the primary inferences do not
persist. Interestingly for DTAX, I find results suggesting an opposite effect in that tax
aggressiveness is associated with less overinvestment for firms with access to investable
funds (p < 0.05). This significant result further reflects the non-linear relation between
For H2, I examine the effect of strong tax monitoring on the relation between tax
aggressiveness and investment efficiency. Operationalized, I split the sample into firms
with or without APTS. This test does involve a reduced sample because the Audit
Analytics database does not begin tracking this information until 2000. I conservatively
make an additional cut to only firms on or after 2003 since the Sarbanes-Oxley Act of
38
2002 significantly changed the types of APTS allowed. These cuts result in a sample size
of 8,975. Using this reduced sample, I re-estimate Model (1) separately for firms with
and without APTS. See Table 6. Columns (1) and (2) present the findings of observations
that have APTS (DiffETR and DTAX, respectively), while Columns (3) and (4) present
Consistent with expectations, the results strongly persist when observations do not
have APTS. That is when APTS is not present, and firms have access to investable funds,
DiffETR + DiffETR*OverFirm (p < 0.01) and DTAX + DTAX*Overfirm (p < 0.01) are
the relation fails to persist when APTS is present. The aggregate of findings in Table 6
provides evidence consistent with H2, in that APTS significantly moderates the relation
FIN 48. To better capture the effect of the exogenous shock, I make numerous sample
cuts. First, I only examine observations in 2006 (pre-period) and 2007 (post-period). This
sample cut helps to mitigate other events confounding the inferences. I also require the
firm to exist in both the pre-period and post-period. Lastly, I remove any observations
that did not disclose an uncertain tax position reserve in 2007. These restrictions result in
1,280 firms for a total of 2,560 observations. Consistent with the H2 analysis, I split the
sample into the pre and post periods and separately analyze the relation between tax
aggressiveness and investment efficiency for each sub-group. See Table 7. Columns (1)
and (2) present the findings from observations in the pre-period for DiffETR and DTAX,
39
respectively, while Columns (3) and (4) present the same findings, but for observations in
Consistent with expectations, the results significantly persist in the pre-period. For
more subsequent year investment, or overinvestment. Meanwhile, the same firms in the
post-period no longer have the same positive relation. Due to the nature of the empirical
some evidence that tax aggressiveness generates agency costs of free cash flows that
cause more investment for firms with high access to investable funds; Said another way,
For H4, I cluster all firms into one of 110 groups based on the firm’s
and RankDTAX at 0 and 9 across each value of OverFirm, along with corresponding t-
statistic. For both definitions of tax aggressiveness, I reject the null hypothesis that there
are no differences in BHAR, but only for firms with high access to investable funds (p <
This finding has numerous important implications. First, the positive differences
efficiency model that a positive coefficient for β1 + β3, suggests overinvestment. While
prior literature posits this statistical effect to be associated with inefficient investment, I
40
am the first to connect the result directly to an effect on firm value, thus providing
credence to the model. Additionally, this finding provides evidence that, among firms
with high access to investable funds, the cash flow benefits of tax aggressiveness are
imply that firms that have access to investable funds subsume the cash flow benefits due
41
V. ADDITIONAL ANALYSES
While my study relates to the cash obtained from tax aggressiveness, it is possible
that this cash flow source is not any different from other discretionary managerial
decisions. For example, instead of being tax aggressive managers can cut advertising
expense or R&D expense, and incur similar cash flow benefits (Bublitz and Ettredge,
1989). To mitigate the concern that tax aggressiveness is not any different from other
discretionary cash flow sources, I examine firms with significant decreases in advertising
and R&D expenses. To operationalize this, I calculate the change in each expense from
year t-1 to t. I then decile rank the changes and identify firms with the most significant
separately replace tax aggressiveness with each indicator variable and re-estimate model
(1). In untabulated analysis, I find that decreasing advertising expense or R&D expense
do not significantly affect subsequent year investments for both firms with low and high
access to investable funds. Because tax aggressiveness differs from the other two
discretionary expenditure accounts due to the action’s opacity, this analysis provides
some evidence that the combination of cash flows and opacity is important to my
findings.
Managerial Characteristics
aggressive tax decisions and inefficient investments primarily because they have a low
managerial ability or because they or overconfident with their ability to be tax aggressive
and choose good investments. To mitigate this concern, I repeat the analysis using a
42
proxy for managerial ability (Demerjian, Lev, and McVay, 2012; Demerjian, Lev, Lewis,
and McVay, 2012; Koester, Shevlin, and Wangerin, 2016) and managerial
overconfidence (Ahmed and Duellman, 2013; Hrirbar and Yang, 2015). In untabulated
analysis, I find no significant difference in the relation between tax aggressiveness and
subsequent year investments among high and low ability managers, or high and low
aggressiveness or investment efficiency, it does not appear to affect the relation between
While DiffETR and DTAX are appropriate measures of tax aggressiveness that
align with the definitions provided by Hanlon and Heitzman (2010) and Lisowsky et al.
(2013), the literature also uses other measures. To ensure my results are robust to other
specifications, I also examine my results when tax aggressiveness takes the form of a
three-year cash effective tax rate (Dyreng et al., 2008), low three-year cash effective tax
rate (Donohoe and Knechel, 2014), and tax haven usage (Dyreng and Lindsay, 2009. In
Biddle et al. (2009) and Cheng et al. (2013) both employ the conditional
investment model, but Biddle et al. (2009) also use an unconditional investment model.
22
I caveat my results such that the calculation of these two managerial characteristics involves significant
data cuts that reduce the power of my variables of interest. Therefore, in a broader sample, it may be
possible that ability or overconfidence do moderate the relation between tax aggressiveness and investment
decisions. However, I fail to find any statistical differences.
43
deviation from the expected level of investment. While valid, this model suffers from the
assumption that researchers can document the expected level of investment, and thus is
44
VI. CONCLUSION
This study finds evidence that tax aggressiveness is significantly associated with
investment decisions. Specifically, I find that for firms with access to investable funds,
relative to firms that are not tax aggressive. Prior research interprets this relation as
associated with inefficient investment. To further explore these results, I find that the
relation weakens for firms with stronger monitoring of tax accounts, thus offering
credence to the theory that tax aggressiveness is associated with higher agency costs of
free cash flows. Furthermore, using a natural experiment around the uncertain tax
is associated with a weaker relation between tax aggressiveness and overinvestment, thus
providing some evidence that tax aggressiveness may cause investment inefficiency by
way of high agency costs of free cash flows. Lastly, I examine subsequent year buy-and-
hold abnormal returns and find that the overinvestment firms have a significantly lower
return, thus suggesting that the benefits of tax aggressiveness for firms with high access
significant non-tax cost of tax aggressiveness. The results suggest that the negative
market reaction associated with the disclosure of firm tax aggressiveness may be more
than reputational concerns (Hanlon and Slemrod, 2009). Also, I offer a potential
45
required to maximize sheltering, then managers making suboptimal investments may
tax strategies.
46
APPENDIX A
Variable Definitions
Where:
47
OverFirm The average of rankings for firm cash holdings (CHE/AT)
and leverage (DLTT/AT). Firms are ranked from one to
1,000 for each attribute. Leverage is multiplied by negative
1 so that both variables are increasing in the propensity to
overinvest. The two rankings are combined into a single
score. That score is then ranked from 1 to 11. I rescale the
ranking to take a value between 0 (low cash & high
leverage) and 1 (high cash & low leverage) in increments of
0.1.
Control Variables
Age The natural log of the number of years the firm has
appeared in the CRSP database.
SalesGrowth Current year sales (REVT) less prior year sales (REVT)
scaled by prior year sales (REVT).
48
PPE Net property, plant, and equipment (PPENT) scaled by total
assets (AT).
StdDevSales The standard deviation of the sales (REVT) from years t-2,
t-1, and t.
49
APPENDIX B
TABLE LISTING
50
TABLE 1
Sample Selection
Criteria:
1992 - 2014 Intersection of Compustat and Execucomp 38,704
Less: Observations in regulated industries -8,818
Less: Observations with negative 3 year pre-tax income -12,479
Less: Observations without enough data to calculate DiffETR and DTAX -2,487
Less: Observations without enough data to compute testing variables -2,044
Total Sample Size 12,876
51
TABLE 2
Descriptive Statistics
Notes: this table presents the descriptive statistics for the primary testing sample compiled using the
criteria outlined in Table 1. All variables are as defined in the Appendix. All continuous variables
are winsorized at the 1% and 99% levels.
52
TABLE 3
Correlation Table
Variable (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20)
(1) Investmentt+1 1 0.09 0.06 0.14 -0.13 0.18 -0.15 0.19 0.22 0.23 0.16 0.07 0.08 -0.10 -0.02 0.02 -0.13 0.02 -0.07 0.27
(2) DiffETR 0.09 1 0.02 0.01 -0.03 0.03 -0.08 -0.04 0.00 0.05 0.10 0.05 -0.01 0.00 -0.01 0.03 -0.14 0.07 -0.06 0.08
(3) DTAX 0.04 0.09 1 0.02 -0.01 0.02 -0.04 0.01 0.01 0.02 0.02 0.03 -0.04 -0.01 0.01 0.04 -0.04 -0.01 -0.03 0.02
(4) OverFirm 0.18 0.01 0.10 1 -0.32 0.55 -0.17 0.36 0.06 0.40 0.05 0.16 -0.37 -0.45 0.01 0.06 -0.22 0.00 -0.08 -0.04
(5) Size -0.11 -0.03 0.00 -0.34 1 -0.31 0.43 -0.18 0.01 -0.14 -0.07 0.22 0.18 0.66 -0.14 -0.03 0.30 0.36 0.11 -0.09
(6) ZScore 0.25 -0.05 0.05 0.67 -0.40 1 -0.22 0.56 0.28 0.73 0.19 0.05 -0.19 -0.35 0.05 0.01 -0.15 -0.05 -0.07 0.08
(7) Age -0.15 -0.09 -0.01 -0.18 0.42 -0.20 1 -0.17 -0.09 -0.20 -0.23 0.11 0.07 0.34 -0.16 -0.04 0.44 0.11 0.17 -0.19
(8) ROA 0.26 -0.08 0.02 0.35 -0.17 0.63 -0.14 1 0.72 0.66 0.30 0.17 -0.08 -0.21 0.10 0.03 -0.05 -0.03 -0.05 0.19
(9) CFO 0.29 -0.01 0.01 0.06 0.00 0.29 -0.08 0.71 1 0.41 0.31 0.11 0.34 -0.05 0.03 -0.02 0.03 -0.01 -0.03 0.34
(10) TobinQ 0.30 0.01 0.08 0.42 -0.14 0.70 -0.17 0.67 0.41 1 0.24 0.18 -0.17 -0.20 0.01 -0.01 -0.11 0.06 -0.10 0.16
(11) SalesGrowth 0.18 0.07 0.04 0.09 -0.08 0.18 -0.22 0.31 0.28 0.25 1 0.06 -0.05 -0.09 0.13 0.05 -0.19 0.02 -0.06 0.36
(12) ForIncome 0.07 0.05 0.15 0.10 0.28 -0.01 0.19 0.06 -0.01 0.12 -0.01 1 -0.14 0.08 -0.11 0.07 0.04 0.16 -0.02 0.06
(13) PPE 0.14 -0.02 -0.11 -0.35 0.15 -0.18 0.11 -0.03 0.32 -0.16 -0.09 -0.17 1 0.18 -0.12 -0.04 0.17 -0.08 0.07 0.09
(14) Rating -0.10 0.00 -0.02 -0.44 0.68 -0.43 0.35 -0.20 -0.05 -0.20 -0.11 0.15 0.19 1 -0.06 -0.01 0.26 0.21 0.09 -0.09
(15) StdDevSales -0.04 -0.02 -0.04 0.04 -0.20 0.11 -0.15 0.08 -0.01 -0.01 0.09 -0.15 -0.10 -0.10 1 -0.03 -0.10 -0.08 -0.06 0.07
(16) TotalIO 0.04 0.04 0.04 0.07 0.01 0.01 -0.07 0.02 -0.04 0.00 0.10 0.07 -0.09 -0.01 -0.02 1 -0.08 0.08 -0.04 0.00
(17) Dividend -0.14 -0.15 -0.04 -0.22 0.30 -0.13 0.46 -0.02 0.04 -0.08 -0.21 0.09 0.21 0.26 -0.11 -0.12 1 -0.03 0.18 -0.16
(18) EQComp 0.02 0.08 0.06 0.03 0.39 -0.05 0.10 0.00 0.01 0.11 0.04 0.21 -0.13 0.21 -0.09 0.12 -0.04 1 -0.11 0.02
(19) CEOAge -0.06 -0.06 -0.03 -0.08 0.10 -0.06 0.17 -0.03 -0.01 -0.09 -0.06 0.01 0.09 0.10 -0.05 -0.05 0.17 -0.10 1 -0.07
(20) Investment 0.48 0.09 0.03 0.06 -0.10 0.14 -0.19 0.25 0.36 0.24 0.32 0.05 0.15 -0.12 0.00 0.00 -0.19 0.04 -0.07 1
Notes: this table presents Pearson (above identity) and Spearman (below identity) correlations among our variables of interest and control variables. All variables are as defined
in the Appendix. All continuous variables are winsorized at the 1% and 99% levels. Correlations in bold are statistically significant at the 5% level or better.
53
TABLE 4
The Effect of Tax Aggressiveness on Investment Efficiency
(1) (2)
Dependent Variable = Investmentt+1 Prediction Coefficient Coefficient
(t-stat) (t-stat)
Notes: This table presents results of estimating OLS regression model 1. Investmentt+1 is measured as total investments
(capital expenditures, research and development, and acquisitions) less the sale of property, plant, and equipment in
year t + 1. Column (1) is the regression when tax aggressiveness is proxied as DiffETR, the difference between
expected and actual three-year cash effective tax rates scaled by the expected three-year cash effective tax rate.
Column (2) is the regression when tax aggressiveness is proxied as DTAX, the discretionary permanent book-tax
differences (Frank et al., 2009). OverFirm is the ability to invest measured as the average ranking of cash and leverage
rescaled to take a value between 0 and 1. All dependent, independent, and control variables are defined in the
Appendix. All continuous variables are winsorized at the 1% and 99% levels. T-statistics are reported in italics below
coefficient estimates. The symbols *, ** and *** denote statistical significance at the 10%, 5%, and 1% levels,
respectively.
54
TABLE 5
The Effect of Tax Aggressiveness on Investment Efficiency by Tax
Aggressiveness and Tax Conservativeness
(1) (2) (3) (4)
Dependent Variable = Investmentt+1 Prediction Coefficient Coefficient Coefficient Coefficient
(t-stat) (t-stat) (t-stat) (t-stat)
55
TABLE 6
The Effect of Tax Monitoring on the Relation between Tax Aggressiveness and
Investment Efficiency
(1) (2) (3) (4)
Dependent Variable = Investmentt+1 Prediction Coefficient Coefficient Coefficient Coefficient
(t-stat) (t-stat) (t-stat) (t-stat)
Notes: This table presents results of estimating OLS regression model 1. Investmentt+1 is measured as total investments (capital
expenditures, research and development, and acquisitions) less the sale of property, plant, and equipment in year t + 1. Column (1) and (3) is
the regression when tax aggressiveness is proxied as DiffETR, the difference between expected and actual three-year cash effective tax rates
scaled by the expected three-year cash effective tax rate. Column (2) and (4) is the regression when tax aggressiveness is proxied as DTAX,
the discretionary permanent book-tax differences (Frank et al., 2009). The columns are additionally separated by whether the firm does not
have APTS (Columns (1) and (2)) or has APTS (Columns (3) and (4)). OverFirm is the ability to invest measured as the average ranking of
cash and leverage rescaled to take a value between 0 and 1. All dependent, independent, and control variables are defined in the Appendix.
All continuous variables are winsorized at the 1% and 99% levels. T-statistics are reported in italics below coefficient estimates. The
symbols *, ** and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively.
56
TABLE 7
A Quasi-Natural Experiment Examining the Relation between Tax
Aggressiveness and Investment Efficiency: Pre Versus Post FIN 48
(1) (2) (3) (4)
Dependent Variable = Investmentt+1 Prediction Coefficient Coefficient Coefficient Coefficient
(t-stat) (t-stat) (t-stat) (t-stat)
Notes: This table presents results of estimating OLS regression model 1. Investmentt+1 is measured as total investments (capital
expenditures, research and development, and acquisitions) less the sale of property, plant, and equipment in year t + 1. Column (1)
and (3) is the regression when tax aggressiveness is proxied as DiffETR, the difference between expected and actual three-year cash
effective tax rates scaled by the expected three-year cash effective tax rate. Column (2) and (4) is the regression when tax
aggressiveness is proxied as DTAX, the discretionary permanent book-tax differences (Frank et al., 2009). The columns are
additionally separated by whether the observation is in 2006 (Columns (1) and (2)) or in 2007 (Columns (3) and (4)). OverFirm is the
ability to invest measured as the average ranking of cash and leverage rescaled to take a value between 0 and 1. All dependent,
independent, and control variables are defined in the Appendix. All continuous variables are winsorized at the 1% and 99% levels. T-
statistics are reported in italics below coefficient estimates. The symbols *, ** and *** denote statistical significance at the 10%, 5%,
and 1% levels, respectively.
57
TABLE 8
Buy-And-Hold Abnormal Returns: Tax Aggressive Versus Tax Conservative for Firms with High and Low Access to
Investable Funds
Rank OverFirm
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0 0.1611 0.1645 0.1824 0.1094 0.1099 0.2681 0.1354 0.1831 0.2157 0.2302 0.1976
DiffETR
9 0.0086 0.0419 0.2115 0.2464 -0.0101 0.3257 0.0484 0.1436 0.1246 0.0942 0.0030
Difference 0.1525 0.1226 -0.0291 -0.1370 0.1200 -0.0576 0.0870 0.0395 0.0911 0.1360 0.1946**
T-Stat 1.52 1.40 -0.28 -1.33 1.21 -0.19 1.18 0.35 0.97 1.53 2.01
0 0.0718 -0.00315 0.2342 -0.1137 -0.1121 0.0132 0.2509 0.126 0.0958 0.1603 0.1659
DTAX
9 0.2848 0.1306 0.0106 0.047 0.146 -0.0377 -0.0149 -0.0145 0.0146 -0.0659 -0.1534
Difference -0.2130 -0.1338 0.2236 -0.1607 -0.2581 0.0509 0.2658 0.1405 0.0812 0.2262* 0.3193**
T-Stat -0.56 -0.88 1.24 -1.22 -1.61 0.48 1.40 1.06 0.90 1.90 2.08
Notes: This table presents the results of a difference in means for firms that are tax aggressive versus tax conservative across different levels of access to investable
funds (OverFirm). Each firm is independently grouped by OverFirm and decile rank of tax aggressiveness (DiffETR or DTAX). For presentation purposes, I present the
most conservative (DiffETR,DTAX = 0) and aggressive (DiffETR,DTAX = 9) groups and display the difference in means. Additionally, I calculate a t-statistic for
whether that difference is significantly different from 0. The symbols *, ** and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively.
58
REFERENCES
59
Chen, S., Chen, X., Cheng, Q., & Shevlin, T. (2010). Are family firms more tax
aggressive than non-family firms? Journal of Financial Economics, 95(1), 41–61.
Chen, K., & Chu, C. (2005). Internal Control versus External Manipulation: A Model of
Corporate Income Tax Evasion. The RAND Journal of Economics. 36(1): 151 –
164.
Cheng, M., Dhaliwal, D., & Zhang, Y. (2013). Does investment efficiency improve after
the disclosure of material weaknesses in internal control over financial reporting?
Journal of Accounting and Economics, 56(1), 1–18.
Chirinko, R. (1993). Business Fixed Investment Spending: Modeling Strategies,
Empirical Results, and Policy Implications. Journal of Economic Literature, 31(4),
1875–1911.
Chyz, J., Leung, W., Li, O., & Rui, O. (2013). Labor unions and tax aggressiveness.
Journal of Financial Economics, 108(3), 675–698.
Coles, J., Daniel, N., & Naveen, L. (2006). Managerial incentives and risk-taking.
Journal of Financial Economics, 79(2), 431–468.
Cook, K., Moser, W., & Omer, T. (2014). Towards an Optimal Level of Tax Avoidance.
Miami University of Ohio Working Paper.
Crocker, K., & Slemrod, J. (2005). Corporate tax evasion with agency costs. Journal of
Public Economics, 89(9-10), 1593–1610.
Dechow, P., & Sloan, R. (1991). Executive incentives and the horizon problem. Journal
of Accounting and Economics, 14, 51–89.
Demerjian, P., Lev, B., Lewis, M., & McVay, S. (2013). Managerial Ability and Earnings
Quality. The Accounting Review, 88(2), 463–498.
Demerjian, P., Lev, B., & McVay, S. (2012). Validity Tests Quantifying Managerial
Ability : A New Measure and Validity Tests. Management Science, 58(7), 1229–
1248.
Denis, D., & Sibilkov, V. (2010). Financial constraints, investment, and the value of cash
holdings. Review of Financial Studies, 23(1), 247–269.
Desai, M., & Dharmapala, D. (2006). Corporate tax avoidance and high-powered
incentives. Journal of Financial Economics, 79(1), 145–179.
Desai, M., & Dharmapala, D. (2009). Corporate tax avoidance and firm value. The
Review of Economics and Statistics, 91(3), 537–546.
Desai, M., Dyck, A., & Zingales, L. (2007). Theft and taxes. Journal of Financial
Economics, 84(3), 591–623.
Donohoe, M., & Knechel, R. (2014). Does Corporate Tax Aggressiveness Influence
Audit Pricing? Contemporary Accounting Research, 31(1), 284–308.
Donohoe, M., & McGill, G. (2011). The Effects of Increased Book-Tax Difference Tax
Return Disclosures on Firm Valuation and Behavior. The Journal of the American
Taxation Association, 33(2), 35–65.
Dyreng, S., Hanlon, M., & Maydew, E. (2008). Long Run Corporate Tax Avoidance. The
Accounting Review, 83(1), 61–82.
Dyreng, S., & Lindsey, B. (2009). Using Financial Accounting Data to Examine the
Effect of Foreign Operations Located in Tax Havens and Other Countries on U.S.
60
Multinational Firms’ Tax Rates. Journal of Accounting Research, 47(5), 1283–
1316.
Edwards, A., Kravet, T., & Wilson, R. (2016). Trapped Cash and the Profitability of
Foreign Acquisitions. Contemporary Accounting Research, 33(1), 44–77.
Edwards, A., Schwab, C., & Shevlin, T. (2015). Financial Constraints and Cash Tax
Savings. Forthcoming at The Accounting Review.
Erickson, M., Goldman, N., & Stekelberg, J. (2015). The Cost of Compliance : FIN 48
and Audit Fees. Forthcoming at The Journal of the American Taxation Association.
Faulkender, M., Flannery, M., Hankins, K., & Smith, J. (2012). Cash flows and leverage
adjustments. Journal of Financial Economics, 103(3), 632–646.
Faulkender, M., & Wang, R. (2006). Corporate Financial Policy and the Value of Cash.
The Journal of Finance, 61(4), 1957–1990.
Fazzari, S., Hubbard, R., & Petersen, B. (1988). Financing Corporate Constraints
Investment. Brookings Papers on Economic Activity, 1, 141–206.
Frank, M., Lynch, L., & Rego, S. (2009). Tax Reporting Aggressiveness and Its Relation
to Aggressive Financial Reporting. The Accounting Review, 84(2), 467–496.
Gaertner, F. (2014). CEO After-Tax Compensation Incentives and Corporate Tax
Avoidance. Contemporary Accounting Research, 31(4), 1077–1102.
Gallemore, J., & Labro, E. (2015). The Importance of the Internal Information
Environment for Tax Avoidance. Journal of Accounting and Economics, 60(1),
149–167.
Goh, B., Lee, J., Lim, C., & Shevlin, T. (2016). The Effect of Corporate Tax Avoidance
on the Cost of Equity. Forthcoming at The Accounting Review.
Gow, I., Larcker, D., & Reiss, P. (2016). Causal Inference in Accounting Research.
Forthcoming at The Journal of Accounting Research.
Guenther, D., Matsunaga, S., & Williams, B. (2015). Corporate Tax Aggressiveness and
Firm Risk. Forthcoming at The Accounting Review.
Graham, J. (1996). Proxies for the corporate marginal tax rate. Journal of Financial
Economics, 42(2), 187–221.
Graham, J., Hanlon, M., Shevlin, T., & Shroff, N. (2014). Incentives for Tax Planning
and Avoidance: Evidence from the Field. The Accounting Review, 89(3), 991-1023.
Hanlon, M., & Heitzman, S. (2010). A review of tax research. Journal of Accounting and
Economics, 50(2-3), 127–178.
Hanlon, M., Lester, R., & Verdi, R. (2015). The Effect of Repatriation Tax Costs on U .S.
Multinational Investment. Journal of Financial Economics, 116(1), 179–196.
Hanlon, M., Maydew, E., & Saavedra, D. (2014). Tax Risk and Corporate Cash
Holdings. Massachusetts Institute of Technology Working Paper.
Hanlon, M., & Slemrod, J. (2009). What does tax aggressiveness signal? Evidence from
stock price reactions to news about tax shelter involvement. Journal of Public
Economics, 93(1-2), 126–141.
Harford, J. (1999). Corporate Cash Reserves and Acquisitions. The Journal of Finance,
LIV(6), 1–29.
61
Hasan, I., Hoi, C., Wu, Q., & Zhang, H. (2014). Beauty is in the eye of the beholder: The
effect of corporate tax avoidance on the cost of bank loans. Journal of Financial
Economics, 113(1), 109–130.
Hayashi, F. (1982). Tobin’s Marginal q and Average q: A Neoclassical Interpretation.
Econometrica, 50(1) 213 – 224.
Hoopes, J., Mescall, D., & Pittman, J. (2012). Do IRS Audits Deter Corporate Tax
Avoidance? The Accounting Review, 87(5), 1603–1639.
Hope, O., & Thomas, W. (2008). Managerial Empire Building and Firm Disclosure.
Journal of Accounting Research, 46(3), 591–626.
Hope, O., Ma, M., & Thomas, W. (2013). Tax avoidance and geographic earnings
disclosure. Journal of Accounting and Economics, 56(2-3), 170–189.
Hribar, P., & Yang, H. (2016). CEO Overconfidence and Management Forecasting.
Contemporary Accounting Research, 33(1), 204–227.
Hutchens, M., & Rego, S. (2015). Does Greater Tax Risk Lead to Increased Firm Risk.
Indiana University Working Paper.
Jensen, M. (1986). Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers.
American Economic Review, 76(2), 323–329.
Jensen, M., & Meckling, W. (1976). Theory of a Firm: Managerial Behavior, Agency
Costs and Ownership Structure. Journal of Financial Economics, 3, 305–360.
Kaplan, S., & Zingales, L. (1997). Do investment-cash flow sensitivities provide useful
measures of financing constraints? The Quarterly Journal of Economics,
(February), 169 – 215.
Kim, J. B., Li, Y., & Zhang, L. (2011). Corporate tax avoidance and stock price crash
risk: Firm-level analysis. Journal of Financial Economics, 100(3), 639–662.
Klassen, K., Lisowsky, P., & Mescall, D. (2015). The Role of Auditors, Non-Auditors,
and Internal Tax Departments in Corporate Tax Aggressiveness. Forthcoming at
The Accounting Review.
Koester, A., Shevlin, T., & Wangerin, D. (2016). Does operational efficiency spill over
onto the tax return? Forthcoming at Management Science.
Lamont, O. (1997). Cash Flow and Investment : Evidence from Internal Capital. The
Journal of Finance, 52(1), 83–109.
Lang, L., Stulz, R., & Walkling, R. (1991). A test of the free cash flow hypothesis - The
case of bidder returns. Journal of Financial Economics, 29(2), 315–335.
Law, K., & Mills, L. (2015). Taxes and Financial Constraints : Evidence from Linguistic
Cues. Journal of Accounting Research, 53(4), 777 – 819.
Leuz, C., Triantis, A., & Wang, T. (2008). Why do firms go dark? Causes and economic
consequences of voluntary SEC deregistations. Journal of Accounting and
Economics, 45(2-3), 181–208.
Lisowsky, P. (2010). Seeking Shelter: Empirically Modeling Tax Shelters Using
Financial Statement Information. The Accounting Review, 85(5), 1693–1720.
Lisowsky, P., Robinson, L., & Schmidt, A. (2013). Do Publicly Disclosed Tax Reserves
Tell Us About Privately Disclosed Tax Shelter Activity? Journal of Accounting
Research, 51(3), 583–629.
62
Malewitz J. (2016). Texas Regulators Approve Oncor Deal, But Uncertainty Persist. The
Texas Tribune. 3/24/2016.
Mann, T. (2015). General Electric Earnings: What to Watch. The Wall Street Journal.
4/14/2015.
McNichols, M., & Stubben, S. (2008). Does Earnings Management Affect Firms’
Investment Decisions? The Accounting Review, 83(6), 1571–1603.
Michels, J. (2016). Disclosure Versus Recognition : Inferences From Subsequent Events.
University of Pennsylvania Working Paper.
Mills, L., Erickson, M., & Maydew, E. (1998). Investments in Tax Planning. Journal of
American Taxation Association, 20(1), 1–20.
Mills, L., Robinson, L., & Sansing, R. (2010). FIN 48 and Tax Compliance. The
Accounting Review, 85(5), 1721–1742.
Murphy, K. (1999). Executive Compensation. Handbook of Labor Economics, 3, 2485 –
2563.
Myers, S., & Majluf, N. (1984). Corporate Financing and Investment Decisions When
Firms Have Information that Investors do not have. Journal of Financial
Economics, 13(2), 187–221.
Nocera, J. (2013). Here Comes the Sun. New York Times. 5/24/2013.
Powers, K., Stomberg, B., & Robinson, J., (2015). You Get What You Pay For: How
Earnings and Cash Flow Performance Metrics Differentially Affect CEO Behavior.
Forthcoming at Review of Accounting Studies.
Rego, S., & Wilson, R. (2012). Equity Risk Incentives and Corporate Tax
Aggressiveness. Journal of Accounting Research, 50(3), 775–810.
Richardson, S. (2006). Over-investment of free cash flow. Review of Accounting Studies,
11(2-3), 159–189.
Roberts, M., & Whited, T. (2013). Endogeneity in Empirical Corporate Finance.
Handbook of the Economics of Finance, (PA), 493–572.
Robinson, J., Sikes, S., & Weaver, C. (2010). Performance Measurement of Corporate
Tax Departments. The Accounting Review, 85(3): 1035–1064.
Scholes, M., Wilson, P., Wolfson, M. (1990). Tax Planning, Regulatory Capital Planning,
and Financial Reporting Strategy for Commercial Banks. Review of Financial
Studies, 3(4): 625 – 650.
Scholes, M., Wolfson, M., Erickson, M., Hanlon, M., Maydew, E., Shevlin, T. Taxes and
Business Strategy: A Planning Approach. New York, US: Pearson Publishing,
2014.
Schwartz, N. (2013). Apple’s Web of Tax Shelters Saved it Billions, Panel Finds. New
York Times. 5/20/2013.
Slemrod, J. (2004). The Economics of Corporate Tax Selfishness. National Tax Journal,
57(4): 877-899.
Tobin, J. (1969). A General Equilibrium Approach to Monetary Theory. Journal of
Money, Credit and Banking, 1(1), 15–29.
Tobin, J., & Brainard, W. (1977). Asset Market and the Cost of Capital. Economic
Progress, Private Values and Public Policy: Essays in Honor of William Fellner.
235-262.
63
Weisbach, D. (2002). Ten truths about tax shelters. 55 Tax Law Review 215.
Wilson, R. (2009). An Examination of Corporate Tax Shelter Participants. The
Accounting Review, 84(3), 969–999.
Yim, S. (2013). The acquisitiveness of youth: CEO age and acquisition behavior. Journal
of Financial Economics, 108(1), 250–273.
Yoshikawa, H. (1980). On the “ q ” Theory of Investment. The American Economic
Review, 70(4), 739–743.
Yost, B. (2015). Tax Avoidance and Investment. Massachusetts Institute of Technology
Working Paper.
64