The Effect of Tax Aggressiveness on Investment Efficiency

Item Type text; Electronic Dissertation

Authors Goldman, Nathan Chad

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material

is made possible by the University Libraries, University of Arizona.
Further transmission, reproduction or presentation (such as
public display or performance) of protected items is prohibited
except with permission of the author.

Download date 10/12/2023 09:15:27

Link to Item http://hdl.handle.net/10150/612100

THE EFFECT OF TAX AGGRESSIVENESS ON INVESTMENT EFFICIENCY

by

Nathan Chad Goldman

__________________________
Copyright © Nathan Chad Goldman 2016

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF ACCOUNTING

In Partial Fulfillment of the Requirements

For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

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.

_________________________________________________________ Date: 4/22/2016

Dan S. Dhaliwal

_________________________________________________________ Date: 4/22/2016

Katharine D. Drake

_________________________________________________________ Date: 4/22/2016

Jayanthi Sunder

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.

_________________________________________________________ Date: 4/22/2016

Dissertation Director: Dan S. Dhaliwal

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.

SIGNED: NATHAN CHAD GOLDMAN

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.

Combining these two consequences of tax aggressiveness, I suggest that tax

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

investable funds, thus suggesting tax aggressiveness is associated with overinvestment. I

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

suggest that poor managerial investment decision making is an unintended consequence

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

tax aggressiveness,1 is an interesting opportunity because these actions are discretionary

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

appear to be an economically meaningful activity with the potential to significantly affect

cash flows and consequently investment decisions.

Based on the importance of tax aggressiveness, I explore how it affects

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

moderated by an indirect effect of tax aggressiveness, external information opacity

(Balakrishnan, Blouin, and Guay, 2012; Hope, Ma, and Thomas, 2013). For example, tax

aggressive firms in my sample, relative to non-tax aggressive firms, have significantly

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).

Accordingly, I question whether weaker investment decisions are an unintended

consequence of tax aggressiveness.

I examine my research question using the conditional investment efficiency

model. This model conditions on firms’ access to investable funds to examine whether a

particular factor is associated with firms investment decisions, specifically

overinvestment and underinvestment (Biddle et al., 2009; Cheng et al., 2013;

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,

Jagolinzer, and Larcker, 2015), and discretionary permanent book-tax differences, or

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

more investment, which the literature interprets as overinvestment. In addition,

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.

Naturally, a weaker information environment moderating the effect of cash flows

on investments is just one possible mechanism to explain the relation between tax

aggressiveness and investment efficiency. Furthermore, the relation may be endogenous

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

aggressiveness on investment efficiency. First, I examine cross-sectional variation in

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

to a firm without APTS. Additionally, I use a quasi-natural experiment surrounding the

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

aggressiveness on investment efficiency are significantly reduced in the post-FIN 48

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

with high and low access to investable funds.

I find evidence that conditional on access to investable funds, tax aggressiveness

is associated with investment inefficiency. Specifically, I find results consistent with my

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

correlated omitted variables and an endogenous relation, as well as to provide a

foundation for tax aggressiveness causing overinvestment. Next, I examine subsequent-

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

expenses) are not associated with inefficient investment, as well as differential

managerial ability or overconfidence. I also provide evidence that the findings are robust

to alternative tax aggressiveness measures and other investment efficiency models.

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

affects other corporate decisions, such as investment efficiency. Because overinvestment

may negatively affect profitability, I offer a potential explanation for the adverse market

reaction to tax aggressiveness.

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;

Desai and Dharmapala, 2009). However, concerns related to empirical specification

(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

aggressiveness on managerial decision making. Since tax aggressiveness is associated

with greater cash flows and more information opacity, I posit and find evidence that tax

aggressiveness is associated with investment decisions. Because these investment

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

affect the degree to which firms engage in tax sheltering activities.

My results should also be of interest to investors and board of directors. Prior

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

aggressiveness is associated with inefficient investment activities, investors and board of

directors may want to design executive pay structures to limit tax aggressiveness.

Additionally, my empirical evidence complements Crocker and Slemrod’s (2005)

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-

decreasing behavior. I extend their analysis by providing some empirical evidence

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

and board action to limit aggressive tax activities.

The remainder of this paper is organized as follows. Section II provides a

literature review and hypothesis development. Section III discusses the data and research

design. Section IV presents the results. Section V provides a summary of additional

analysis performed. Section VI concludes.

15
 II. LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT

Literature Review

Tax Aggressiveness Overview

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

documents Apple Inc.’s “Double Irish” tax strategy. It states:

“This strategy [“The Double Irish”] … involves setting up a shell subsidiary in an

offshore tax haven — a.k.a. Ireland — and transferring most of Apple’s
intellectual property rights to the dummy subsidiary. The subsidiary, in turn,
charges “royalties” that allows it to capture billions of dollars in what otherwise
would be taxable profits in the United States. In Ireland, according to Apple, it
pays an astonishing 2 percent in taxes, thanks to a deal it has with the
government.” (Nocera, 2013)
This tax strategy reflects a complex set of actions designed to reduce Apple’s tax

liabilities. What makes these actions unique from other non-aggressive tax planning

activities (i.e. investment in municipal bonds or deductions for accelerated depreciation)

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

that tax aggressiveness is a value-enhancing activity due to the activity’s ability to

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.

Thus, tax aggressiveness may be a value increasing firm activity.

Despite the benefits of tax aggressiveness, numerous studies document adverse

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

strategy decision making, suggesting an equilibrium between maximizing tax planning

and minimizing non-tax costs. Other non-tax costs of tax aggressiveness include adverse

market reactions (Hanlon and Slemrod, 2009), weaker information environments

(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).

Additional research investigates agency concerns of tax aggressiveness. For

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

However, other studies question this evidence due to a lack of generalizability or

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

Dharmapala (2009).5 As a result, whether tax planning activities generate significant

agency costs remains an empirical question.

To overcome the generalizability and empirical specification concerns, as well as

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

potentially unintended negative consequences, for managerial decision making such as

investment decisions.6

Agency Costs of Free Cash Flows

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

preferences generates agency costs of free cash flows.

Numerous studies document the adverse consequences of excess cash flows.7 For

example, Richardson (2006) finds evidence that overinvestment commonly occurs in

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

mechanism for managers to choose self-serving projects.

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

holdings trapped overseas in a value decreasing manner. Additionally, Yost (2015)

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

flows, which may affect investment decisions.

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

agency concerns such as moral hazard (Jensen, 1986; Blachard, Lopez-de-Silanes,

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

value projects. Either occurrence of these actions is often referred to as inefficient

22
investment. Because investments comprise a large firm expenditure each year, investment

inefficiency has the potential to be a significant non-tax cost, and unintended

consequence, of tax aggressiveness.

Hypothesis Development

Tax Aggressiveness and Investment Efficiency

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

aggressiveness is also associated with higher external information asymmetry because

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

unintended consequences for managerial actions, specifically investment decisions.8

To assess the effects of tax aggressiveness on investment decisions, I follow

Biddle et al. (2009), Cheng et al. (2013), and Balakrishnan et al. (2014) by focusing on

how a specific factor effects subsequent investment choices conditional on access to

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.

Additionally, tax aggressiveness exacerbates these concerns due to information

asymmetry between managers and monitors.10

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

me to suggest that, conditional on high access to investable funds, tax aggressiveness is

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

hypothesis (stated in the alternate form):

H1a: Conditional on high access to investable funds, tax aggressiveness is

associated with a higher subsequent year level of investments, or

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

result in an insignificant relation between tax aggressiveness and underinvestment.12 As

a result, I make the following hypothesis (stated in the null form):

H1b: Conditional on firms inability to access to investable funds, tax

aggressiveness is not associated with the subsequent year level of investments,

or underinvestment.

Tax Aggressiveness, Investment Efficiency, and Tax Monitoring

Alternative explanations may exist for the relation between tax aggressiveness

and investment efficiency. To help improve identification of my proposed mechanism, I

examine cross-sectional variation in firms tax-related information asymmetry as

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:

“A registered public accounting firm may engage in any non-audit service,

including tax services, that is not expressly prohibited, after audit committee pre-
approval. Accordingly, accountants will be able to continue to provide tax
compliance, tax planning and tax advice to audit clients, subject to audit
committee pre-approval requirements.”

Firms have numerous outlets for implementing a tax strategy, including their

external audit firm, other accounting or law firms, or self-implementation. Although

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

aggressive than firms with APTS.

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

moderates the relation between tax aggressiveness and investment inefficiency. As a

result, I make the following hypothesis (stated in the alternate form):

H2: Firms with APTS have a significantly weaker relation between tax

aggressiveness and investment inefficiency, relative to firms without APTS.

Tax Aggressiveness, Investment Efficiency, and Tax Disclosure: A Quasi-Natural

Experiment

While I posit that tax aggressiveness affects investment decisions, it is possible

that firms structure investment decisions to enhance tax aggressiveness opportunities.

Even though I lessen the potential for an endogenous relation by examining how tax

aggressiveness affects subsequent year investments, I further alleviate these concerns

through a quasi-natural experiment surrounding the onset of FIN 48.15

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,

increase in the quality and quantity of tax-related information disclosed, specifically

related to uncertain and potentially aggressive tax positions (Blouin, Gleason, Mills, and

Sikes, 2007; Erickson, Goldman, and Stekelberg, 2015).16

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

reserve in 2007, tax aggressiveness is significantly less associated with inefficient

investment, relative to the pre-period. As a result, I make the following hypothesis

(stated in the alternate form):

H3: Firms with a FIN 48 disclosure in 2007 have a significantly weaker

relation between tax aggressiveness and investment inefficiency, relative to

2006.

Buy-And-Hold Abnormal Return Analysis

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

tax aggressiveness is associated with poorer managerial decisions, such as investment

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

following hypothesis (stated in the null form):

H4: Tax aggressiveness is not associated with subsequent BHAR

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;

Armstrong et al., 2015), and discretionary permanent book-tax-differences, or DTAX

(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

residual from a regression of permanent book-tax-differences on nondiscretionary book-

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

discretionary-permanent book-tax-differences, thus suggesting tax aggressiveness

30
(conservativeness). See the Appendix for a more detailed explanation of the DiffETR and

DTAX calculations. 17

Model

To test my hypotheses, I follow the conditional investment efficiency 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

investment conditional on firms access to investable funds. Using this methodology, I

estimate the following model:

Investmenti,t+1 = α + β1 DiffETR / DTAXi,t + β2 OverFirmi,t + β3 DiffETR /

DTAXi,t *OverFirmi,t + Controls + Year F.E. + Industry F.E.+ ɛi,t (1)

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

observations where overinvestment or underinvestment activity may be more likely.18

In particular, OverFirm reasonably captures the firm’s financial state relative to

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

(Biddle et al. 2009; Cheng et al. 2013).

I control for numerous variables known to affect investments and tax

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)

to be negatively associated with subsequent year investments. Conversely, I include

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

income (ForIncome). I expect each of these variables to be positively associated with

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).

In addition to these variables, I also consider two managerial characteristics. First,

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

negatively associated with subsequent year investments. Lastly, it is possible that

investments are sticky from year to year. Therefore, I control for current year investments

(Investment) and expect it to be positively associated with subsequent 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

coefficient on β1 measures the effect of DiffETR / DTAX on Investmentt+1 when firms

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

less efficiently. Meanwhile, a negative and significant β1 + β3, or a positive and

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

APTS,20 and firms in the pre versus post-FIN 48 era, respectively.

Lastly, H4 examines buy-and-hold abnormal returns for tax aggressive firms,

relative to non-tax aggressive firms. Following the literature, I calculate BHAR as the

difference in returns between the observation and the return on a buy-and-hold

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

decile (RankDiffETR = 0, RankDTAX = 0) to firms in the highest decile (RankDiffETR =

9, RankDTAX = 9) across values of OverFirm (ranging from 0 to 1 in 0.1 increments). I

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

the cash flows.

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

sample size of 12,876 firm-year observations.

35
 IV. PRIMARY ANALYSIS

Descriptive Statistics and Correlations

Table 2 presents the descriptive statistics for the sample population. The mean

value of Investment is 13.906 suggesting that, on average, capital and non-capital

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

from 0 appears reasonable. Conversely, OverFirm is calculated on the testing sample to

ensure equal distribution, and accordingly takes an average value of 0.500. In examining

the remaining variables, I do not note any unusual descriptive statistics.

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

aggressiveness and access to investable funds are appropriate determinants of subsequent

year investments. Furthermore, current year Investmentt is highly correlated with

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

examining the remaining variables, I do not note any unusual associations.

Hypothesis Testing

In Table 4, I present the results of estimating Model (1). First, I find that the

combination of the coefficients on DiffETR and DiffETR*OverFirm is positive and

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

with access to investable funds. Additionally, the coefficient on DiffETR / DTAX is

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

governance (Armstrong et al., 2015). Furthermore, assuming a linear relation may be

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;

Desai and Dharmapala, 2009).

To overcome concerns related to non-linearity, I split the same at 0 for both

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

without access to investable funds.

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

tax aggressiveness and investment decisions.

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

the estimation when the observations do not have APTS.

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

positively associated with subsequent year investment, or overinvestment. Meanwhile,

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

between tax aggressiveness and investment efficiency.

For H3, I examine a quasi-natural experiment surrounding the implementation of

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

the post period.

Consistent with expectations, the results significantly persist in the pre-period. For

firms with high access to investable funds, tax aggressiveness (DiffETR +

DiffETR*OverFirm, p < 0.05; DTAX + DTAX*OverFirm, p < 0.05) is associated with

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

specification being a quasi-experiment, the combined findings from Table 7 provide

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,

tax aggressiveness leads to overinvestment.

For H4, I cluster all firms into one of 110 groups based on the firm’s

RankDiffETR / RankDTAX decile ranking and 0 to 1 OverFirm ranking. For presentation

purposes, I only display the cells when RankDiffETR or RankDTAX is 0 (most

conservative) or 9 (most aggressive). I then provide the difference between RankDiffETR

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 <

0.05 for both specifications when OverFirm equals 1).

This finding has numerous important implications. First, the positive differences

when OverFirm equals 1 yields confirmatory evidence to the conditional investment

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

significantly outweighed by poor managerial decision making. While inefficient

investment is one-way managers may make value-destroying decisions, my findings

imply that firms that have access to investable funds subsume the cash flow benefits due

to direct and indirect costs of a more opaque information environment.

41
 V. ADDITIONAL ANALYSES

Other Cash Flow Sources

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

decreases as an indicator variable taking the value of 1, and 0 otherwise. Finally, I

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

An alternative explanation for my primary results is that managers make

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

overconfident managers.22 As a result, while managerial ability may affect tax

aggressiveness or investment efficiency, it does not appear to affect the relation between

the two constructs.

Alternative Measures of Tax Aggressiveness

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

untabulated analysis, my inferences remain unchanged when using these alternative

proxies for tax aggressiveness.

Unconditional Investment Efficiency Model

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.

The unconditional investment model assesses how a firm-specific attribute contributes to

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

subject to numerous criticisms. To assess the robustness of my analysis, I also examine

my research question using the unconditional investment model. In untabulated analysis,

I find my inferences remain unchanged. As a result, my evidence from the unconditional

model appears to complement the primary results.

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,

tax aggressiveness is associated with higher levels of subsequent year investments

relative to firms that are not tax aggressive. Prior research interprets this relation as

evidence of overinvestment. Thus, I provide evidence that tax aggressiveness is

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

position disclosure requirements, I find that an increase in tax aggressiveness disclosure

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

to investable funds are significantly outweighed by some of the unintended

consequences, such as inefficient investment.

This study provides evidence consistent with investment inefficiency as a

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

explanation for the undersheltering puzzle (Weisbach, 2002). If tax aggressiveness is

45
required to maximize sheltering, then managers making suboptimal investments may

incentivize shareholders to shift managerial incentives away from choosing aggressive

tax strategies.

46
 APPENDIX A

Variable Definitions

Dependent Variables of Interest

Investmentt+1 Capital expenditures (CAPX) plus research and

development expenses (XRD), plus acquisitions (AQC) less
the sale of property, plant, and equipment (SPPE). All
scaled by lagged total assets (AT). The product is
multiplied by 100.

Independent Variables of Interest

DiffETR The difference between expected and actual three-year cash

effective tax rate scaled by expected cash effective tax rate.
Expected three-year cash effective tax rate is calculated by
placing all firms into industry-year-size decile buckets. The
average of each group's three-year cash effective tax rate is
considered expectation. The actual three-year cash effective
tax rate is defined as the sum of cash taxes paid (TXPD) for
years t-2 through year t. It is them scaled by the sum of
pre-tax income (PI) for years t-2 through year t less special
items (SPI) for year t-2 through year t.

DTAX Discretionary permanent book-tax differences, as created by

Frank et al. (2009). This is calculated by regressing total
permanent book-tax differences on nondiscretionary items
that are known to cause permanent differences, but are
unrelated to tax reporting aggressiveness. The resulting
residual is averaged from year t-2 to year t to generate
DTAX. Specifically, I calculate it as follows:

PERMDIFF = Intercept + Intang + Uncon + MI + Cste +

ChangeNOL + LagPERM + Residual

Where:

PERMDIFF = (PI - [(TXT + TXO) / 0.35] - (DTA / 0.35)) /

AT
Intang = INTAN / AT
Uncon = ESUB / AT
MI = MII / AT
Cste = TXS / AT
ChangeNOL = (TLCF - LagTLCF) / AT.

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

Size The natural log of total assets (AT).

ZScore Mathematical formula to help predict bankruptcy calculated

as the following:

1.2*A + 1.4*B + 3.3*C + 0.6*D + 0.999*E

A = Working Capital (ACT-LCT) / Total Assets (AT)

B = Retained Earnings (RE) / Total Assets (AT)
C = Earnings Before Interest and Taxes (EBIT) / Total
Assets (AT)
D = Market Value of Equity (PRCC_F*CSHO) / Total
Liabilities (LT)
E = Sales (REVT) / Total Assets (AT).

Age The natural log of the number of years the firm has
appeared in the CRSP database.

ROA Pre-tax income scaled by prior year assets.

CFO Cash flows from operations (IB + DPC + TXBD) scaled by

prior year total assets.

TobinQ The market value of equity (PRCC_F*CSHO) plus book

value of assets (AT) less book value of equity (CEQ) less
deferred taxes (TXDB), all scaled by book value of assets
(AT).

SalesGrowth Current year sales (REVT) less prior year sales (REVT)
scaled by prior year sales (REVT).

ForIncome Pre-tax foreign income (PIFO) scaled by lagged total assets.

48
PPE Net property, plant, and equipment (PPENT) scaled by total
assets (AT).

Rating Indicator variable equal to 1 if the firm has a credit rating,

and 0 otherwise.

StdDevSales The standard deviation of the sales (REVT) from years t-2,
t-1, and t.

Total_IO The percentage of ownership held by institutional investors.

Dividend Indicator variable equal to 1 if dividends (DVC) or cash

dividends (DV) is greater than 0, and 0 otherwise.

EqComp Percentage of CEO salary derived from equity-based

compensation (TDC1 - Salary - Bonus - OthComp) scaled
by total compensation (TDC1).

CEOAge Natural log of the CEOs age.

Investment Capital expenditures (CAPEX) scaled by lagged net

property, plant, and equipment (PPENT), all in year t.

Other Variables of Interest

Indicator variable equal to 1 if a firm's auditor-provided tax

services (TAX_FEES, audit analytics) are more than 1% of
APTS
total audit fees (TOTAL_FEES, audit analytics), and 0
otherwise.

Indicator variable equal to 1 if the observation is in 2007,

Post and 0 if the observation is in 2006.

The buy-and-hold abnormal return year t+2 adjusted for

BHAR
market returns.

49
 APPENDIX B

TABLE LISTING

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

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

Variable N Mean Std Dev P25 Median P75

Investmentt+1 12,876 13.905 12.939 4.678 9.112 16.151

DiffETR 12,876 -0.045 0.533 -0.312 -0.024 0.261
DTAX 12,876 0.004 0.178 -0.016 0.001 0.023
OverFirm 12,876 0.500 0.318 0.200 0.500 0.800
Size 12,876 7.302 1.455 6.231 7.124 8.217
ZScore 12,876 3.823 5.153 2.945 3.300 6.614
Age 12,876 2.910 0.825 2.398 2.944 3.526
ROA 12,876 0.138 0.097 0.071 0.116 0.181
CFO 12,876 0.176 0.087 0.115 0.160 0.218
TobinQ 12,876 2.156 1.313 1.309 1.743 2.529
SalesGrowth 12,876 0.119 0.190 0.009 0.083 0.181
ForIncome 12,876 0.026 0.040 0.000 0.005 0.040
PPE 12,876 0.272 0.214 0.107 0.210 0.379
Rating 12,876 0.438 0.496 0.000 0.000 1.000
StdDevSales 12,876 0.150 0.131 0.066 0.110 0.186
TotalIO 12,876 0.553 0.346 0.283 0.668 0.829
Dividend 12,876 0.562 0.496 0.000 1.000 1.000
EQComp 12,876 0.554 0.294 0.368 0.634 0.798
CEOAge 12,876 4.015 0.131 3.932 4.025 4.111

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)

Intercept ? 20.2766*** 20.8438***

(5.90) (6.06)
DiffETR ? (H1b) 0.3656
(0.99)
DTAX ? (H1b) -0.3764
(-0.29)
OverFirm + 4.7367*** 4.7313***
(9.68) (9.68)
DiffETR*OverFirm + 1.3613**
(2.22)
DTAX*OverFirm + 0.9715
(0.52)
Size - -0.8042*** -0.8256***
(-7.09) (-7.27)
ZScore - -0.0817** -0.0753**
(-2.33) (-2.14)
Age - -0.6404*** -0.6465***
(-4.07) (-4.11)
ROA + 1.9136 0.0235
(0.82) (0.01)
CFO + 4.5959* 5.6847**
(1.89) (2.34)
TobinQ + 1.3206*** 1.3633***
(9.28) (9.58)
SalesGrowth + 0.9169 1.2356*
(1.39) (1.88)
ForIncome + 1.5297 2.7915
(0.51) (0.93)
PPE + 9.4983*** 9.6573***
(11.18) (11.37)
Rating + 0.6968** 0.7499**
(2.31) (2.48)
StdDevSales - -3.7300*** -3.6961***
(-4.38) (-4.34)
TotalIO - -0.8955** -0.8718**
(-2.22) (-2.16)
Dividend - -1.1738*** -1.3321***
(-4.61) (-5.26)
EQComp + 2.7565*** 2.9066***
(6.50) (6.86)
CEOAge - -2.1134** -2.2712***
(-2.58) (-2.77)
Investment + 0.1463*** 0.1473***
(15.74) (15.84)

Test: DiffETR + DiffETR*OverFirm = 0 + (H1a) 1.7269***

(4.10)
Test: DTAX + DTAX*OverFirm = 0 + (H1a) 0.5952
(0.45)
Fixed Effects Year & Industry Year & Industry
Clustered Standard Errors Firm Firm
N 12,876 12,876
Adjusted R-Square 18.5% 18.3%

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)

Intercept ? 19.4799*** 19.5569*** 22.3394*** 22.1003***

(3.67) (4.15) (4.96) (4.38)
DiffETR ? (H1b) -2.0131* 0.7901
(-1.91) (1.20)
DTAX ? (H1b) -1.6820 3.2684
(-0.87) (1.58)
OverFirm + 2.0763** 4.8739*** 4.1057*** 3.2891***
(2.17) (6.98) (5.21) (4.45)
DiffETR*OverFirm + 7.0093*** -0.7116
(4.19) (-0.65)
DTAX*OverFirm + 6.3407** -8.7276***
(2.24) (-2.90)
Size - -0.9160*** -0.7896*** -0.7134*** -0.8271***
(-5.38) (-5.20) (-4.68) (-4.82)
ZScore - -0.0689 -0.1323*** -0.0848* -0.0211
(-1.36) (-2.78) (-1.71) (-0.40)
Age - -0.6922*** -0.6729*** -0.5371** -0.6242***
(-2.92) (-3.08) (-2.54) (-2.74)
ROA + 1.5617 -2.4674 5.2243 2.0167
(0.45) (-0.78) (1.63) (0.59)
CFO + 6.1426* 4.3074 1.8755 6.1983*
(1.75) (1.27) (0.55) (1.76)
TobinQ + 1.4736*** 1.4003*** 1.1018*** 1.3757***
(7.13) (7.33) (5.55) (6.42)
SalesGrowth + 0.3369 1.9203** 1.0326 0.3574
(0.36) (2.10) (1.11) (0.38)
ForIncome + 6.8161 6.8997* -3.7544 -4.7660
(1.57) (1.81) (-0.89) (-0.97)
PPE + 8.8201*** 9.7026*** 10.5221*** 9.9501***
(6.80) (8.14) (9.25) (8.19)
Rating + 0.7130 0.6898* 0.6404 0.7783*
(1.54) (1.65) (1.61) (1.78)
StdDevSales - -3.0408** -3.8353*** -4.6166*** -3.7043***
(-2.34) (-3.18) (-4.07) (-3.06)
TotalIO - -0.4093 -0.6965 -1.5248*** -0.8830
(-0.66) (-1.24) (-2.87) (-1.53)
Dividend - -0.7887** -1.4844*** -1.5371*** -1.0361***
(-2.03) (-4.24) (-4.56) (-2.82)
EQComp + 1.5054** 2.7223*** 3.9266*** 3.1085***
(2.36) (4.75) (6.91) (4.95)
CEOAge - -1.8286 -1.6693 -2.4461** -2.7666**
(-1.45) (-1.49) (-2.27) (-2.30)
Investment + 0.1590*** 0.1640*** 0.1273*** 0.1284***
(11.51) (12.73) (10.07) (9.53)

Test: DiffETR + DiffETR*OverFirm = 0 + (H1a) 4.9962*** 0.0784

(4.58) (0.10)
Test: DTAX + DTAX*OverFirm = 0 + (H1a) 4.6587** -5.4592**
(2.27) (-2.57)
Fixed Effects Year & Industry Year & Industry Year & Industry Year & Industry
Clustered Standard Errors Firm Firm Firm Firm
N 6,123 6,753 6,762 6,114
Adjusted R-Square 20.0% 19.4% 16.3% 17.7%
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 is tax aggressive (DiffETR and DTAX > 0 for
Columns (1) and (2), respectively) or tax conservative (DiffETR and DTAX > 0 for Columns (3) and (4), respectively). 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.

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)

Intercept ? 14.0627*** 15.2066*** 30.1847*** 30.5485***

(2.89) (3.12) (3.72) (3.75)
DiffETR ? (H2) 0.0546 1.5833*
(0.11) (1.77)
DTAX ? (H2) -2.8030 -0.0191
(-1.59) (-0.01)
OverFirm + 4.2275*** 4.2566*** 5.8045*** 5.7868***
(6.18) (6.23) (5.04) (5.01)
DiffETR*OverFirm + 2.1715** -0.3801
(2.57) (-0.26)
DTAX*OverFirm + 6.3753* -0.8498
(1.91) (-0.21)

Test: DiffETR + DiffETR*OverFirm = 0 + (H2) 2.2261*** 1.2032

(3.98) (1.19)
Test: DTAX + DTAX*OverFirm = 0 + (H2) 3.5723* -0.8689
(1.85) (-0.35)
Controls Yes Yes Yes Yes
Fixed Effects Year & Industry Year & Industry Year & Industry Year & Industry
Clustered Standard Errors Firm Firm Firm Firm
N 4,648 4,648 4,327 4,327
Adjusted R-Square 16.3% 16.0% 19.1% 18.8%

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)

Intercept ? 5.3084 5.8031* 4.1424* 4.1414*

(1.62) (1.77) (1.65) (1.65)
DiffETR ? (H3) 0.4440 3.7438***
(0.34) (3.42)
DTAX ? (H3) -6.4266** 2.5618
(-2.79) (1.54)
OverFirm + 1.2607 0.5641 1.2792 0.8020
(0.71) (0.32) (0.90) (0.56)
DiffETR*OverFirm + 3.2889* -3.6149**
(1.68) (-2.11)
DTAX*OverFirm + 9.3905** -0.2807
(1.98) (-0.12)

Test: DiffETR + DiffETR*OverFirm = 0 + (H3) 3.7329** 0.1289

(1.97) (0.12)
Test: DTAX + DTAX*OverFirm = 0 + (H3) 2.9639* 2.2811
(1.67) (1.37)
Controls Yes Yes Yes Yes
Year & Year & Year & Year &
Fixed Effects Industry Industry Industry Industry
Clustered Standard Errors Firm Firm Firm Firm
N 1,280 1,280 1,280 1,280
Adjusted R-Square 16.6% 16.8% 17.6% 17.1%

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

Abel, A. (1983). Optimal Investment Under Uncertainty. The American Economic

Review, 73(1), 228 – 233.
Acharya, V., Davydenko, S., & Strebulaev, I. (2012). Cash holdings and credit risk.
Review of Financial Studies, 25(12), 3572–3609.
Ahmed, A., & Duellman, S. (2013). Managerial Overconfidence and Accounting
Conservatism. Journal of Accounting Research, 51(1), 1–30.
Almeida, H., & Campello, M. (2007). Financial Constraints, Asset Tangibility, and
Corporate Investment. Review of Financial Studies, 20(5), 1429–1460.
Armstrong, C., Blouin, J., Jagolinzer, A., & Larcker, D. (2015). Corporate governance,
incentives, and tax avoidance. Journal of Accounting and Economics, 60(1), 1–17.
Balakrishnan, K., Blouin, J., & Guay, W. (2012). Does Tax Aggressiveness Reduce
Corporate Transparency? University of Pennsylvania Working Paper.
Balakrishnan, K., Core, J., & Verdi, R. (2014). The relation between reporting quality
and financing and investment: Evidence from changes in financing capacity.
Journal of Accounting Research, 52(1), 1–36.
Baker, M., Stein, J., & Wurgler, J. (2003). When Does the Market Matter? Stock Prices
and the Investment of Equity-Dependent Firms. The Quarterly Journal of
Economics, 118(3), 969–1005.
Barber, B., & Lyon, J. (1997). Detecting long-run abnormal stock returns: The empirical
power and specification of test statistics. Journal of Financial Economics, 43(3),
341–372.
Biddle, G., & Hilary, G. (2006). Accounting Quality and Firm-Level Capital Investment.
The Accounting Review, 81(5), 963–982.
Biddle, G., Hilary, G., & Verdi, R. (2009). How does financial reporting quality relate to
investment efficiency? Journal of Accounting and Economics, 48(2-3), 112–131.
Blanchard, O., Lopez-de-Silanes, F., & Shleifer, A. (1994). What do firms do with cash
windfalls? Journal of Financial Economics, 36(3), 337–360.
Blaylock, B. (2015). Is Tax Avoidance Associated with Economically Significant Rent
Extraction among U.S. Firms? Forthcoming Contemporary Accounting Research.
Blouin, J., Core, J., & Guay, W. (2010). Have the tax benefits of debt been
overestimated? Journal of Financial Economics, 98(2), 195–213.
Blouin, J., Gleason, C., Mills, L., & Sikes, S. (2007). What Can We Learn about
Uncertain Tax Benefits from FIN 48? National Tax Journal, 60(3), 521 – 535.
Bublitz, B., & Ettredge, M. (1989). The Information in Discretionary Outlays:
Advertising, Research, and Development. Accounting Review, 64(1), 108.
Burks, J., Randolph, D., & Seida, J. (2015). Modeling and Interpreting Interactions in
Tradeoff Research. University of Notre Dame Working Paper.
Brown, J., Drake, K., & Martin, M. (2015). Compensation in the Post- FIN 48 Period:
The Case of Contracting on Tax Performance and Uncertainty. Contemporary
Accounting Research 33 (1), 121-151.

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