Journal of Financial Economics 124 (2017) 441–463

Contents lists available at ScienceDirect

Journal of Financial Economics

journal homepage: www.elsevier.com/locate/jfec

Changes in corporate effective tax rates over the past 25

yearsR
Scott D. Dyreng a,∗, Michelle Hanlon b, Edward L. Maydew c, Jacob R. Thornock d
a
Duke University, Fuqua School of Business, 100 Fuqua Drive, Durham, NC 27708, United States
b
Massachusetts Institute of Technology, Sloan School of Management, 100 Main Street, Cambridge, MA 02142, United States
c
University of North Carolina, Kenan-Flagler College of Business, Chapel Hill, NC 27599, United States
d
Brigham Young University, Marriott School of Business, 519 Tanner Bldg., Provo, UT 84602, United States

a r t i c l e i n f o a b s t r a c t

Article history: We investigate systematic changes in corporate effective tax rates over the past 25 years
Received 5 November 2014 and find that effective tax rates have decreased significantly. Contrary to conventional wis-
Revised 6 June 2015
dom, the decline in effective tax rates is not concentrated in multinational firms; effective
Accepted 1 July 2016
tax rates have declined at approximately the same rate for both multinational and domes-
Available online 13 April 2017
tic firms. Moreover, within multinational firms, both foreign and domestic effective rates
JEL Classification: have decreased. Finally, changes in firm characteristics and declining foreign statutory tax
G38 rates explain little of the overall decrease in effective rates.
H32 © 2017 Elsevier B.V. All rights reserved.
H25
H26

Keywords:
Tax
Effective tax rates
Tax avoidance
Time trend
Foreign taxes
Domestic taxes

1. Introduction

R
We appreciate comments from Jeff Hoopes, Tom Neubig, Terry Shevlin,
Nemit Shroff, Jaron Wilde, and seminar participants at the 2013 Euro-
pean Institute for Advanced Studies in Management Workshop on Cur-
rent Research on Taxation at the University of Münster, the 2015 Mas-
sachusetts Institute of Technology Asia Conference in Accounting, Bocconi
University, Cambridge Judge Business School, Cardiff University, Cornell
University, the Duke Fuqua Summer Brownbag series, ETH Zurich KOF
Swiss Economic Institute, London School of Economics, Michigan State
University, Oxford University Centre for Business Taxation, Renmin Uni-
versity, the European Accounting Association 2015 Annual Congress, the
International Tax Policy Forum, and WHU-Otto Beisheim School of Man-
agement. We thank Brady Williams for foreign statutory tax rate data. We
thank Madeline Hartman, Tyson Mayers, Meredith Mayer-Salman, Theresa
Richards, and Tommy Saintsing for research assistance. Scott D. Dyreng
acknowledges funding from the Fuqua School of Business at Duke Univer-
sity, Michelle Hanlon from the Howard W. Johnson Chair at the MIT Sloan
School of Management, Edward L. Maydew from the Accounting Alumni
In this paper, we examine the trend in corporate ef-
fective tax rates (ETRs) over the past 25 years. While US
statutory tax rates have remained relatively constant dur-
ing this period, indications are that some firms have been
able to reduce their effective tax rates through tax plan-
ning strategies and by taking advantage of favorable provi-
sions in the tax code (Gravelle, 2013; Bloomberg, 2010; US
Senate, 2014a; Dyreng, Hanlon, and Maydew, 2008). Many
Faculty Research Fund at the University of North Carolina, and Jacob R.
Thornock from the University of Washington and Brigham Young Univer-
sity.
∗
Corresponding author.
E-mail address: scott.dyreng@duke.edu (S.D. Dyreng).

http://dx.doi.org/10.1016/j.jfineco.2017.04.001
0304-405X/© 2017 Elsevier B.V. All rights reserved.
442 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
have suggested that the US tax system is broken because
of the disparity between the high statutory rate and low
effective rates that some companies realize, resulting in
an inequitable system of taxation, perverse incentives, and
undesirable consequences.1 Despite recent attention paid
to these issues by the US government, international tax
bodies such as the Organization for Economic Co-operation
and Development (OECD), the academic literature, and the
popular press, surprisingly little evidence exists on how ef-
fective tax rates have evolved over time, even though such
data would be informative to academics and policy makers
when evaluating the current system relative to a set of al-
ternatives. A great deal of research examines tax avoidance
and effective tax rates, but almost all of this research ex-
amines the cross section of firms (e.g., Chen, Chen, Cheng,
and Shevlin, 2010; Desai and Dharmapala, 2006; Kim, Li,
and Zhang, 2011).
To investigate the existence of, and potential explana-
tions for, a trend in effective tax rates over time, we ex-
amine a sample of 54,028 US firm-years over the 25-year
period from 1988 to 2012. We primarily study the trend in
the cash effective tax rate (worldwide cash taxes paid di-
vided by pretax accounting earnings).We also examine do-
mestic and foreign current effective tax rates, which we
use to test foreign versus domestic rates for US corpo-
rations that have income from foreign sources (multina-
tional firms).2 To be clear, we do not examine taxes paid
to the Internal Revenue Service (IRS) relative to taxable in-
come reported to the IRS. Broadly speaking, taxes paid to
the IRS will be equal to taxable income reported to the
IRS multiplied by the statutory tax rate, reduced by tax
credits. Examining the ratio of these two numbers does
not capture tax avoidance strategies that reduce or de-
lay the recognition of taxable income (e.g., shifting income
to a low-tax foreign country, investment in tax favored
activities such as municipal bonds, and investments that
are subjected to accelerated depreciation for tax purposes)
because those types of strategies reduce both taxes paid
and taxable income reported to the taxing authority. In
contrast, cash effective tax rates capture all reductions in
taxes paid relative to pretax financial accounting income.
Our measures are intentionally broad, so that they capture
any form of tax reduction relative to pretax accounting in-
come, whether through tax sheltering, location decisions,
income shifting, tax preferences within the tax code, or
rule changes.
We begin by testing whether or not cash effective tax
rates have declined over the sample period. Overall, we
find a significant downward trend in effective tax rates us-
ing our broad sample of publicly traded US corporations.
On average, cash effective tax rates of our sample firms
have decreased by about 0.4 percentage points per year
over the past 25 years. The trend is economically large,
1 3
For discussions and references, see H.R. 1, The Tax Reform Act of 2014
(113th Congress), US Senate (2014b, 2015), and Citizens for Tax Justice
(2015).
2
We explain these measures more fully and we test multiple alternate
measures, such as the Generally Accepted Accounting Principles (GAAP)
effective tax rate and a cash effective tax rate based on cash flows from
operations, in the sections that follow.
representing a cumulative decline of between 5 and 10
percentage points, depending on the specification. A 10
percentage point decline translates into about $109 billion
less in taxes paid in 2012 compared with what would have
been paid had the effective rate remained constant at the
1988 level.3
Much attention has been focused on multinational
firms because they are able to shift income from high-tax
jurisdictions to low-tax jurisdictions, including tax havens
(Dharmapala and Riedel, 2013; Dyreng and Markle, 2016;
Dharmapala, 2014), especially when cross-border transac-
tions are based on intangible assets which are more diffi-
cult to value and easier to move across borders (Grubert
and Slemrod, 1998; Kleinbard, 2011; De Simone, Mills, and
Stomberg, 2014). Recent international tax reform guidance
has focused on these issues for multinationals, including
the Base Erosion and Profit Shifting (BEPS) initiative at
the OECD (OECD, 2013). If effective tax rates are declin-
ing more rapidly for multinational firms than for purely
domestic firms, then multinationals can increasingly have
a cost advantage over domestic firms. However, the litera-
ture provides little evidence about changes in effective tax
rates of multinational firms compared to purely domestic
firms over time.
We examine the time trend in effective tax rates sep-
arately for multinational and domestic firms. Surprisingly,
we find essentially the same decrease in effective tax rates
over time for purely domestic firms as for multination-
als. We probe this finding further by comparing the trend
in current effective tax rates on domestic income versus
foreign income within our subsample of US multinational
firms. As expected, we find that the decline in foreign ef-
fective tax rates of multinationals is largely explained by
the well-documented decline in statutory tax rates of for-
eign countries. However, we also find a downward trend
in the domestic effective tax rates of multinationals, which
cannot be explained by declining foreign statutory tax
rates.
These primary results are informative for tax policy.
While multinationals have access to international tax plan-
ning opportunities, our results show that domestic corpo-
rations have declining effective tax rates as well. This sug-
gests that significant and increasing opportunities exist to
reduce (via planning or from provisions in the tax laws) ef-
fective tax rates on domestic income. Thus, perhaps policy
makers should not focus exclusively on tax avoidance by
multinationals and should consider the extent of current
provisions for domestic-only firms.
We perform a number of additional tests to explain
the trend in effective tax rates. First, we examine whether
changing firm characteristics can explain the trend. Extant
research shows that the characteristics of publicly traded
firms today are different than they were 25 years ago (e.g.,
Our univariate analysis shows a decline in cash effective tax rates of
about 5 percentage points (i.e., from about 32% at the beginning of the
sample period to about 27% at the end of the sample period). Our multi-
variate analysis shows an average annual decline of 0.4 percentage points,
which amounts to about 10 percentage points over the 25-year sample
period. The $109 billion is calculated by aggregating pretax income of our
sample firms in 2012 of $1,088 billion and multiplying by 0.10.
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
Fama and French, 2001; Graham, Leary, and Roberts, 2015;
DeAngelo and Roll, 2015). If effective tax rates vary with
those characteristics, then they could help explain the time
trend. Consistent with prior cross-sectional research, we
find that a multitude of firm-level characteristics are as-
sociated with effective tax rates, such as intangible assets
and the existence of losses. However, we find that these
characteristics are unable to account for much of the de-
crease in tax rates over time.
Second, we test whether the mapping of certain firm
characteristics into tax rates (i.e., the time-varying effect
of those characteristics on effective tax rates) has changed
over time. For example, a common assertion is that global-
ization has made it easier to shift the income from intan-
gible assets into tax havens or to escape taxation entirely
by creating so-called stateless income (Kleinbard, 2011). If
so, tax rates will decrease with respect to intangible as-
sets, even if the level of intangible assets stays constant.
We find some evidence of changes in the mapping of cer-
tain firm characteristics into tax rates over time, but, again,
the changes explain little of the overall decrease in effec-
tive tax rates.
Finally, we examine whether the decline in effective
tax rates can be explained by legislative and regulatory
changes that took place during the sample period, includ-
ing the check-the-box regulations, the American Jobs Cre-
ation Act of 2004 (AJCA), and the two recent bonus de-
preciation regimes. Overall, we find little evidence that the
trends in effective tax rates coincide with these legislative
and regulatory changes.
Our paper contributes to the literature by executing a
thorough examination of time trends in corporate effec-
tive tax rates.4 Our findings that effective tax rates have
declined for purely domestic firms at essentially the same
rate as for multinationals and that multinational firms do
not on average have significantly lower cash effective tax
rates than purely domestic firms suggest that the current,
almost singular focus on multinational corporations for tax
avoidance and low tax rates could be misplaced. In ad-
dition, our study is important for researchers interested
in studying the determinants of effective tax rates. It is
common to pool observations across years when studying
effective tax rates (and tax avoidance), but the changing
nature of effective tax rates over time suggests that re-
searchers should be careful to examine whether results are
time period specific, and they should interpret their results
accordingly.
The paper proceeds as follows. Section 2 develops the
hypotheses. Section 3 discusses the sample, the variables,
and presents descriptive statistics. The main results are
presented in Section 4. Section 5 examines potential de-
terminants of declining rates. Section 6 presents additional
tests and robustness and Section 7 concludes.
4
A few studies examine changes in tax avoidance over time, but not as
the main focus (e.g., Desai, 2003; Desai and Dharmapala, 2009; Graham,
Raedy, and Shackelford, 2012; Klassen and LaPlante, 2012). Also, a few
studies that examine trends in aggregate tax collections (e.g., Auerbach
and Poterba, 1987; Auerbach, 2007).
443
2. Hypotheses development
Corporate tax avoidance has recently received substan-
tial attention in the academic literature, in the popular
press, and from policy makers. It is well known in these
circles that substantial cross-sectional variation exists in
corporate effective tax rates (e.g., Dyreng, Hanlon, and
Maydew, 2008).This variation can be partially explained
by a multitude of firm-specific factors, including own-
ership structure (Chen, Chen, Cheng, and Shevlin, 2010;
Badertscher, Katz, and Rego, 2013), compensation structure
(Rego and Wilson, 2012; Armstrong, Blouin, and Larcker
2012), corporate governance (Desai, Dyck, and Zingales,
2007), labor organization (Chyz, Leung, Li, and Rui, 2013),
subsidiary locations (Dyreng and Lindsey, 2009), business
model characteristics (Higgins, Omer, and Phillips, 2015),
and management characteristics (Chyz, 2013). Despite the
progress made in understanding the cross-sectional deter-
minants of effective tax rates, surprisingly little is under-
stood about how effective tax rates have evolved over time.
Though little is known about time trends in effective
tax rates of US corporations at the microeconomic level,
some research has examined trends in macroeconomic tax
outcomes, such as changes in corporate tax revenues. For
example, Auerbach (2007) extends Auerbach and Poterba
(1987) to examine the question of whether corporate tax
revenues have declined over time, and finds that they did
so through 1982 but leveled off thereafter. Furthermore, he
finds offsetting trends in the ratio of profits to gross do-
mestic product (which is declining) and in the average tax
on profits (which is increasing). He interprets the latter as
evidence that tax planning is not as prevalent as some sug-
gest. In contrast, Slemrod (2004) finds evidence to suggest
that misreporting of corporate income taxes has increased
over time from $61 billion in 1988 to $146 billion in 20 0 0.
Desai (2003) argues that the corporate tax receipts as per-
centage of the budget have decreased over time. Overall,
the research examining the macroeconomic trends in tax
outcomes has produced somewhat mixed results.
Several reasons support the notion that, at the microe-
conomic level, the effective tax rates of US firms are de-
creasing over time. First, many developed countries, other
than the US, have reduced their statutory tax rates. For a
multinational firm, a decline in foreign statutory tax rates
can lead to a lower effective tax rate by reducing its tax
on foreign income, at least until it repatriates the income
to the US. Moreover, a declining foreign statutory tax rate
could increase the incentives to locate more real opera-
tions in that foreign country or to source or shift income
to that country to take advantage of the lower tax rate,
or both. Second, firms are becoming more global in na-
ture. Fig. 1 shows that the fraction of sample firms that
are considered multinational has steadily increased over
time, suggesting a clear and steady upward trend in multi-
national status over the sample period. In 1988, approx-
imately 40% of the sample firms were multinationals; by
2012, nearly 70% of sample firms were multinationals.5
5
Regressing the average annual percentage of firms that are classified
as multinationals on a linear time trend variable, TIME, reveals an upward
trend that is highly statistically significant (t-statistic = 21.17).
444 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463

Fig. 1. Fraction of firms that are classified as multinational enterprises (MNE = 1) over the sample period. This figure plots the fraction of firms that are
classified as multinationals (MNE = 1) over the 25-year sample period. In this figure, a firm-year is considered multinational if the absolute value of pretax
foreign income is greater than zero or if the absolute value of foreign tax expense is greater than zero; otherwise it is considered domestic. The sample
period covers 1988–2012 and all observations are subject to the criteria described in Table 1.

Combining these two insights suggests that, to the extent
that US multinational firms are increasingly able to source
income into countries outside the US, their effective tax
rates could decline over time. Third, recent evidence sug-
gests that domestic firms have tax planning opportuni-
ties available to them including tax shelters (Wilson, 2009;
Lisowsky, 2010; Brown, 2011) and provisions enacted to
benefit domestic firms (e.g., domestic production activities
deduction and bonus depreciation). If domestic firms are
increasingly utilizing these strategies over time, their ef-
fective tax rates could decline over time.
In sum, changes to the global economic, competitive,
and tax landscape over the sample period suggest a de-
cline in effective tax rates. This discussion leads to our first
hypothesis (all hypotheses are stated in the alternative).
advantage of the large statutory tax rate disparities that
exist across countries. This leads to our second hypothesis.
H2: The effective tax rates of multinational firms are de-
clining more over time than those of purely domestic
firms.
To probe our second hypothesis more deeply, we con-
sider the tax rates on domestic income of multination-
als relative to the tax rates on foreign income of multi-
nationals. If the declines in effective tax rates are driven
by strategies involving income shifting across international
borders, then more stark declines should be evident in for-
eign taxes paid on foreign income than in domestic taxes
paid on domestic income. Accordingly, we offer our third
hypothesis.

H1: The effective tax rates of US corporations have de-
creased over time.
Anecdotes in the popular press, as well as evidence in
academic research, suggest that multinationals use transfer
pricing, intracompany debt, cost-sharing agreements, and
other tactics to shift income from high-tax jurisdictions to
low-tax jurisdictions (e.g., Hines and Rice, 1994; Huizinga
and Laeven, 2008; Dharmapala and Riedel, 2013). As US
firms have become more multinational, and as the dispar-
ity in statutory tax rates between the US and foreign coun-
tries has increased, the opportunities and incentives to im-
plement cross-border tax saving strategies has also likely
increased. Multinational companies likely have tax strate-
gies at their disposal that are simply unavailable to purely
domestic firms because purely domestic firms cannot take
H3: For multinationals the effective tax rate on foreign in-
come declines more over time than does the effective tax
rate on domestic income.
Beyond testing our three hypotheses, we examine sev-
eral other possible explanations for the changes in tax
rates over time, including changes in foreign tax rates,
changing firm characteristics, changing tax and accounting
rules, and other factors. We also conduct several sets of ad-
ditional analyses and robustness tests.
3. Sample and descriptive statistics
Table 1 describes our sample selection criteria. We
begin with all nonfinancial, nonutility firm-years listed in
Compustat during the period 1988–2012 that have total
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
Table 1
Sample composition.
The sample covers the period 1988–2012. All financial statement data
are acquired from the annual fundamentals database produced by Com-
pustat. Compustat data item pneumonics are reported in parentheses in
all caps.
Number
Criteria of firms Firm-years
All US incorporated nonfinancial, nonutilities 11,643 102,925
Compustat observations between 1988 and
2012 with assets greater than $10 million
Require non-missing values of control variables 11,573 101,364
Require cash tax paid (TXPD) to be 10,171 89,114
non-missing
Require pretax income (PI) to be greater than 8497 62,564
zero
Require each firm to have at least five 4643 54,028
observations
assets of at least $10 million. We start the sample in
1988 because it follows the last major overhaul of US tax
system, the Tax Reform Act of 1986, which took effect in
mid-1987. A total of 102,925 firm-years meet these initial
requirements. We require non-missing values of cash taxes
paid, pretax income, and other control variables, which
reduces the sample to 89,114 firm-year observations.6 We
also exclude firm-years with negative pretax income be-
cause effective tax rates are difficult to interpret when the
denominator is negative. This criterion further reduces the
sample to 62,564 observations.7 Finally, we require each
firm to have at least five years of data. These criteria result
in a sample of 54,028 firm-year observations from 4643
unique firms, for an average of 11.6 years of data per firm.
Our primary variable of interest is the firm’s cash ef-
fective tax rate, CASH ETR, which is computed as the ratio
of cash taxes paid to pretax accounting income.8 CASH ETR
is widely used in the literature because it captures a broad
range of tax avoidance activities, including income shifting
from high-tax to low-tax jurisdictions (e.g., strategic trans-
fer pricing arrangements, cost-sharing agreements, income
stripping using intracompany debt), investment in tax
favored assets (e.g., municipal bonds), accelerated depreci-
ation deductions, tax credits (e.g., research and experimen-
tation credits), and so on. This measure is advantageous
because it allows us to speak to changes in tax avoidance
generally, without specifying ex ante precise tax avoidance
strategies or rule changes, which could have evolved over
time. To be clear, CASH ETR uses financial accounting
income in the denominator, not taxable income. While
this has the advantage of allowing the measure to capture
tax avoidance strategies that cause reductions to taxable
income (e.g., income shifting), it has the disadvantage
6 10
Compustat variable pneumonics are included in the caption to
Table 2. Unless otherwise specified, all variables are acquired from the
annual fundamentals database maintained by Compustat.
7
Thus, our results do not apply to loss years. In Section 6, we conduct
tests to examine whether losses are driving the results that we observe,
and we find that they do not.
8
We do not adjust CASH ETR for special items but instead include spe-
cial items as a control variable in our multivariate regressions. We win-
sorize all effective tax rates at zero and one, following, among others,
Dyreng, Hanlon, and Maydew (2008).
445
of being affected by regulatory changes to financial ac-
counting rules and by managers’ incentives to report high
income to shareholders. CASH ETR also does not capture
conforming tax avoidance, in which firms engage in tax
avoidance strategies that reduce both taxable income and
financial accounting income (e.g., using external debt cap-
ital instead of equity capital to take advantage of interest
deductions).9
Table 2, Panel A, presents descriptive statistics for the
sample. CASH ETR averages 29.1% during the sample period,
with a median of 27.5%, both of which are below the statu-
tory tax rate of 35% and are consistent with rates reported
in prior research (e.g., Dyreng, Hanlon, and Maydew, 2008).
The data demonstrate considerable variation in CASH ETR.
The 25th percentile of CASH ETR is only 12.6% and the 75th
percentile is 38.6%, which is also consistent with quantiles
of CASH ETR reported in prior research.
The remainder of Panel A of Table 2 presents descrip-
tive statistics for firm characteristics that could explain
variation in effective tax rates.10 The variable MNE, which
indicates whether the firm is a multinational enterprise
(MNE) (MNE = 1) or a purely domestic firm (MNE = 0), has
a mean of 0.51. We classify a firm as a MNE in a given
year if either its pretax foreign income is greater than zero
or its absolute value of foreign tax expense is greater than
zero.11 As noted in Donohoe, McGill, and Outslay, (2012),
the classification of firms as being strictly multinational or
purely domestic is more complex than it seems at first, and
we therefore consider six alternative measures of MNE (see
Section 6.1).
LOG ASSETS is the natural log of total assets. The me-
dian LOG ASSETS is 5.803, corresponding to approximately
$331 million of total assets. The average firm-year reports
2.6 dollars of research and development (R&D) spend-
ing for every 100 dollars in sales (mean R&D EXPENSE
= 0.026). However, the median firm-year reports no re-
search and development spending. For the mean (median)
firm-year, PP&E constitute 29.4% (23.5%) of total assets. The
mean (median) firm-year has INTANGIBLE ASSETS equal to
11.8% (4%) of total assets. The mean firm-year in the sam-
ple has LEVERAGE of 21.8% of total assets, with a median
of 19.3%. The average (median) firm-year in the sample
has CAPITAL EXPENDITURES of 25.7% (21.2%) of net property,
plant, and equipment (PPE). The mean (median) firm-year
has ADVERTISING EXPENSE of 1% (0%) of total assets. Special
items, when they exist, are usually negative (i.e., income-
9
We know of no established tax burden measure that captures all
types of tax avoidance. Hence, we focus on explicit tax avoidance ac-
tivities. Our measure incorporates implicit taxes to the extent that the
implicit taxes reduce financial accounting income. However, our measure
does not fully capture these taxes. We leave an investigation of implicit
taxes to future research.
All variables are defined in the notes to Table 2.
11
Because these Compustat data items are gathered directly from the
firm’s audited financial statements filed with the Securities and Exchange
Commission (SEC), we expect them to be generally accurate. We manu-
ally checked some of the largest domestic firms in our sample, such as
CVS, Macy’s, Time Warner, and Lowes, and find that the numbers re-
ported by Compustat are generally accurate. In some cases, we find ev-
idence of small amounts of foreign activity in other parts of the financial
statements, but the amounts were relatively small (generally less than 2%
of total income).
446 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463

Table 2
Descriptive statistics and correlations.
This table reports descriptive statistics (Panel A) and correlations (Panel B) for the variables used in our study. The sample covers the period 1988–2012,
and all observations are subject to the criteria described in Table 1. CASH ETR is the ratio of cash taxes paid (TXPD) to pretax income (PI). MNE is an
indicator for multinational firm-years and is equal to one if the absolute value of pretax foreign income (PIFO) is greater than zero or if the absolute value
of foreign tax expense (TXFO) is greater than zero. R&D EXPENSE is the amount of research and development expense (XRD; if missing, it is set to zero)
scaled by the sales (SALE). LOG ASSETS is the natural log of total assets (AT). PP&E is the ratio of net property, plant, and equipment (PPENT) to total assets
(AT). INTANGIBLE ASSETS is intangible assets (INTAN) scaled by total assets (AT). LEVERAGE is total debt (DLTT + DLC) scaled by total assets (AT). CAPITAL
EXPENDITURES is the amount spent on capital assets (CAPX) scaled by net property, plant, and equipment (PPENT). ADVERTISING EXPENSE is the ratio of
advertising expense (XAD; if missing, it is set to zero) to sales (SALE). SPECIAL ITEMS is the ratio of special items (SPI; if missing, it is set to zero) to average
total assets (AT). NOL is an indicator variable equal to one if Compustat reports a tax-loss carryforward (TLCF) at the end of the previous year. NOL is the
change in net operating losses and is the difference between the current and lagged tax-loss carryforward (TLCF), scaled by lagged total assets (AT). All
financial statement data are acquired from the annual fundamentals database maintained by Compustat. Compustat data item pneumonics are reported in
parentheses in all caps.

Panel A: Descriptive statistics

Standard 25th 50th 75th

Variable N Mean deviation percentile percentile percentile

CASH ETR 54,028 0.291 0.227 0.126 0.275 0.386

MNE 54,028 0.510 0.500 0.0 0 0 1.0 0 0 1.0 0 0
LOG ASSETS 54,028 5.938 1.869 4.515 5.803 7.197
R&D EXPENSE 54,028 0.026 0.050 0.0 0 0 0.0 0 0 0.026
PP&E 54,028 0.294 0.223 0.119 0.235 0.417
INTANGIBLE ASSETS 54,028 0.118 0.164 0.0 0 0 0.040 0.180
LEVERAGE 54,028 0.218 0.193 0.042 0.193 0.337
CAPITAL EXPENDITURES 54,028 0.257 0.169 0.136 0.212 0.335
ADVERTISING EXPENSE 54,028 0.010 0.025 0.0 0 0 0.0 0 0 0.009
SPECIAL ITEMS 54,028 −0.003 0.020 −0.004 0.0 0 0 0.0 0 0
LAGGED SPECIAL ITEMS 54,028 −0.009 0.036 −0.005 0.0 0 0 0.0 0 0
NOL 54,028 0.274 0.446 0.0 0 0 0.0 0 0 1.0 0 0
NOL 54,028 −0.001 0.138 0.0 0 0 0.0 0 0 0.0 0 0

Panel B: Pearson (above the diagonal) and Spearman (below the diagonal) correlations

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13

1 CASH ETR 0.0295∗ 0.0014 −0.072∗ −0.053∗ 0.0022 −0.021∗ −0.017∗ 0.03∗ −0.209∗ 0.0838∗ –0.141∗ 0.0196∗
2 MNE 0.0202∗ 0.3436∗ 0.2584∗ −0.204∗ 0.1656∗ −0.063∗ 0.0109∗ 0.0401∗ −0.075∗ –0.054∗ 0.166∗ 0.004
3 LOG ASSETS 0.0107∗ 0.3479∗ −0.04∗ 0.1524∗ 0.2525∗ 0.2093∗ −0.193∗ 0.0662∗ −0.078∗ –0.002 0.0862∗ 0.025∗
4 R&D EXPENSE −0.067∗ 0.3793∗ −0.026∗ −0.297∗ −0.012∗ −0.274∗ 0.2625∗ −0.026∗ −0.073∗ –0.09∗ 0.1103∗ –0.003
5 PP&E −0.019∗ −0.164∗ 0.1499∗ −0.289∗ −0.323∗ 0.2556∗ −0.298∗ −0.068∗ 0.0382∗ 0.0638∗ –0.106∗ 0.0171∗
6 INTANGIBLE ASSETS 0.0081 0.2246∗ 0.2822∗ 0.0382∗ −0.287∗ 0.1799∗ −0.042∗ 0.0674∗ −0.062∗ –0.05∗ 0.1577∗ 0.0104∗
7 LEVERAGE −0.05∗ −0.046∗ 0.2551∗ −0.264∗ 0.2954∗ 0.1567∗ −0.26∗ 0.0029 0.0077 –0.023∗ 0.051∗ 0.0133∗
8 CAPITAL EXPENDITURES 0.0242∗ 0.035∗ −0.156∗ 0.1555∗ −0.346∗ −0.046∗ −0.3∗ 0.0451∗ −0.044∗ 0.0219∗ 0.0019 –0.01∗
9 ADVERTISING EXPENSE 0.0378∗ 0.0181∗ 0.0612∗ 0.0124∗ −0.068∗ 0.0545∗ −0.017∗ 0.0427∗ −0.027∗ –0.008 –0.01∗ –4E-04
10 SPECIAL ITEMS −0.109∗ −0.132∗ −0.166∗ −0.092∗ 0.068∗ −0.168∗ −0.032∗ −0.015∗ −0.044∗ 0.0046 0.0043 –0.025∗
11 LAGGED SPECIAL ITEMS 0.0976∗ −0.121∗ −0.121∗ −0.1∗ 0.0734∗ −0.15∗ −0.043∗ 0.0354∗ −0.037∗ 0.1693∗ –0.067∗ 0.0126∗
12 NOL −0.187∗ 0.166∗ 0.0906∗ 0.1112∗ −0.115∗ 0.1634∗ 0.0464∗ −0.003 0.0119∗ −0.058∗ –0.102∗ –0.093∗
13 NOL 0.089∗ 0.0423∗ 0.0935∗ −0.019∗ 0.033∗ 0.0371∗ 0.0331∗ −0.022∗ −0.005 −0.048∗ –0.016∗ –0.257∗

decreasing), which is reflected in the mean SPECIAL ITEMS
of −0.3% of average total assets. The median firm-year does
not report any special items. LAGGED SPECIAL ITEMS has
an average value in the sample of −0.9%, and the median
value is zero. NOL indicates that the firm has a tax-loss
carryforward as reported by Compustat, on average, 27.4%
of the years in the sample. NOL reflects the change in
net operating losses (NOLs) from the prior year to the cur-
rent year and has a mean (median) of −0.001 (0.000) in
the sample, suggesting that tax loss carryforwards equal to
about 0.1% of assets either were utilized to offset the tax
bill or expired in the current year.12
12
These NOLs are as reported in Compustat, which are subject to data
concerns (e.g., Mills, Newberry, and Novack, 2003). We cannot determine
whether the NOLs are foreign, domestic, or state NOLs. In addition, we
cannot determine whether the losses are usable or limited under certain
tax code provisions, such as Internal Revenue Code Section 382 that limits
Table 2, Panel B, presents the Pearson (above the di-
agonal) and Spearman (below the diagonal) correlations
for these variables. Using Pearson correlations, CASH ETR is
negatively and significantly correlated with R&D EXPENSE,
PP&E, LEVERAGE, CAPITAL EXPENDITURES, SPECIAL ITEMS,
and NOL. CASH ETR is positively and significantly corre-
lated with MNE, ADVERTISING EXPENSE, LAGGED SPECIAL
ITEMS, and NOL. LOG ASSETS and INTANGIBLE ASSETS are
insignificantly correlated with CASH ETR.
4. Main results of hypotheses tests
This section presents the results of our primary tests of
changes in corporate effective tax rates over the past 25
years. In particular, we present the results of testing each
of our primary hypotheses in the subsections that follow.
the use of NOLs acquired as part of a merger or acquisition of a company
or via a technical ownership change outside of an acquisition, or both.
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
Fig. 2. Mean annual CASH ETR over the sample period. This figure plots the annual mean CASH ETR over the sample period, 1988–2012. CASH ETR is the
ratio of cash taxes paid to pretax income. All observations are subject to the criteria described in Table 1.
4.1. Changes in cash effective tax rates over time
We begin by testing our first hypothesis, which is that
CASH ETRs of US corporations have decreased over time.
In Fig. 2, we provide visual evidence on this hypothesis by
plotting the sample mean of CASH ETR for each year from
1988 to 2012. The graphic shows a clear downward trend
in CASH ETR during the past 25 years. The average CASH
ETR was just above 32% in 1988 but declined to about 27%
by 2012. The figure reveals some periods of increases and
decreases in average CASH ETR. The average CASH ETR goes
as low as 22% by 2004 before increasing to nearly 31% by
2008. Moreover, the figure shows that, in some partitions
of the sample period, the trend in cash effective tax rates
varies in steepness. For example, the trend in effective tax
rates is quite steep before 2001 and somewhat flat after
2002. However, even with the year-to-year variation in the
sample average CASH ETR, the trend over the full sample
period is clearly negative.
To test the trend more formally, we estimate the fol-
lowing regression.
CASH ET Rit = α0 + α1 T IMEt + it , (1)
where CASH ETR is as defined above and TIME is calculated
as the fiscal year for a given firm-year observation less the
number 1988, which is the first year in the data set. Thus,
the coefficient on TIME captures the linear time trend in
CASH ETR over the sample period.
We present the results from estimating Eq. (1) for the
full sample in Table 3, Panel A, Model 1.13 The slope on
13
To simplify presentation of the regression coefficients, we multiply
the dependent variable in all regressions by one hundred.
447
the time trend, TIME, is −0.414 (t-statistic = 6.99), confirm-
ing the statistical significance of the visual evidence pre-
sented in Fig. 2. The decrease in cash effective tax rates is
economically large (about 0.4 percentage points per year)
and likely resulted in positive cash flows for the average
firm.14 For example, a decline in the CASH ETR from 32%
to 27% corresponds to a 7.4% increase in average after-tax
cash flows, all else constant.15 This finding suggests that,
for the average firm, lower cash effective tax rates have be-
come an increasingly important source of cash flow in the
past 25 years.
4.2. Cash effective tax rate trends for multinational versus
domestic firms
Having shown a downward trend in effective tax rates,
we now turn to testing our second hypothesis, which is
that the effective tax rates of multinationals are declining
more rapidly than the effective tax rates of purely domes-
tic firms. In Fig. 3, we present visual evidence for our sec-
ond hypothesis by plotting the average CASH ETR over time
separately for multinationals and purely domestic firms.
The dark (light) line in Fig. 3 plots the average CASH ETR
over time for multinationals (purely domestic firms). Two
striking results are evident in Fig. 3. First, multinationals
on average have higher CASH ETRs than purely domestic
firms in almost every year during the 25-year sample pe-
14
This is consistent with Hanlon, Maydew, and Saavedra (2016), who
find a significantly negative correlation between after-tax cash flows and
a firm’s 5-year cash effective tax rate.
15
(1−0.27)/(1−0.32) = 1.074. For simplicity, we assume accruals are zero
in expectation.
448 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463

Table 3
Regressions of CASH ETR on TIME.
In this table, we report the results of estimating Eq. (1), which is an ordinary least squares regression of CASH ETR on a linear time trend (TIME). CASH
ETR is cash taxes paid divided by pretax income. TIME is calculated as the fiscal year for a given firm-year observation less the number 1988, which is the
first year in the data set. Thus, TIME takes on values of 0–24, which corresponds to sample years 1988–2012. We identify multinational firms as those with
either the absolute value of pretax foreign income greater than zero or the absolute value of foreign tax expense greater than zero. In Panel A, we use all
available observations. In Panel B, we use only observations in the largest quartile of total assets. All other variables are as defined in Table 2. In all three
models, standard errors are clustered by firm and year. The t-statistics are reported in parentheses below the coefficient estimates. ∗∗∗ , ∗∗ , and ∗ represent
statistical significance at the 1%, 5% and 10% level, respectively.

Panel A: All firms

Model 1 Model 2 Model 3 Difference

Variable (all firms) (multinationals) (domestics) (Model 2 - Model 3)

INTERCEPT 33.857∗ ∗ ∗ 35.038∗ ∗ ∗ 33.372∗ ∗ ∗ 1.667∗

(62.40) (45.30) (57.03) (1.98)
TIME −0.414∗ ∗ ∗ −0.415∗ ∗ ∗ −0.485∗ ∗ ∗ 0.070
(−6.99) (−6.11) (−7.72) (1.17)
N 54,028 27,561 26,467
Adj. R2 0.015 0.016 0.019

Panel B: Large firms (upper quartile of total assets)

Model 1 Model 2 Model 3 Difference

Variable (big firms) (big multinationals) (big domestics) (Model 2 - Model 3)

INTERCEPT 35.791∗ ∗ ∗ 36.057∗ ∗ ∗ 36.150∗ ∗ ∗ −0.093

(36.18) (34.85) (24.46) (−0.07)
TIME −0.521∗ ∗ ∗ −0.484∗ ∗ ∗ −0.735∗ ∗ ∗ 0.251∗ ∗
(−6.85) (−6.28) (−6.19) (2.26)
N 13,505 9,507 3,998
Adj. R2 0.027 0.024 0.047

Fig. 3. Mean annual CASH ETR over the sample period separately for multinational and domestic firms. This figure plots the annual mean CASH ETR over
the sample period, 1988–2012. CASH ETR is the ratio of cash taxes paid to pretax income. In this figure, a firm-year is considered multinational if the
absolute value of pretax foreign income is greater than zero or if the absolute value of foreign tax expense is greater than zero; otherwise it is considered
domestic. All observations are subject to the criteria described in Table 1.
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
riod. Thus, while multinationals have tax planning oppor-
tunities not available to purely domestic firms, the results
suggest the average multinational firm has a higher cash
effective tax rate than a purely domestic firm.16
Second, in stark contrast to conventional wisdom, both
multinationals and purely domestic firms exhibit declining
CASH ETRs over time at approximately the same rate. At
the beginning of the period in 1988, the average multina-
tional firm had a CASH ETR of approximately 34% and the
average purely domestic firm had a CASH ETR of approxi-
mately 32%. By the end of the sample period, both multi-
nationals and purely domestic firms were able to lower
their average CASH ETR to about 28% and 24%, respectively.
This result contradicts a commonly held belief that a de-
cline in CASH ETRs over time is due to the rising prevalence
of multinationals or increased tax avoidance by multina-
tionals, or both. We observe essentially the same decline
in effective tax rates in purely domestic firms as in multi-
nationals.
In Table 3, Panel A we present the results from esti-
mating Eq. (1) for only multinational firms (Model 2) and
only domestic firms (Model 3). In Model 2, the slope on
TIME is −0.415 for multinationals; in Model 3, the slope
is −0.485 for purely domestic firms. To test the difference
in the TIME coefficients across Models 2 and 3, we employ
a stacked regression approach, which reveals that the time
trends are not significantly different from one another (dif-
ference = 0.070, t-statistic = 1.17). The intercept from es-
timating Eq. (1) represents the conditional mean CASH ETR
at the beginning of the sample period. The intercept for
multinationals, presented in Model 2, is 35.038, and the in-
tercept for domestic firms, presented in Model 3, is 33.372.
The difference of 1.667 is marginally statistically signifi-
cant (t-statistic = 1.98). Thus, at the beginning of the sam-
ple period, the average multinational has a slightly higher
CASH ETR than the average purely domestic firm.
As striking as the results in Fig. 3 and Table 3 are
in suggesting that both multinational and domestic firms
have declining CASH ETRs, multinational and domestic
firms likely differ on a number of dimensions. For example,
multinationals are on average much larger than purely do-
mestic firms. To address this potential confound, in Panel
B of Table 3, we retain only firm-years in the upper quar-
tile of total assets for the sample and examine whether the
trend in effective tax rates for large multinationals differs
from that of large domestic firms. In the first column, large
firms are similar to other firms in that they exhibit a sta-
tistically significant declining time trend in their CASH ETR,
which is similar in magnitude to the trend reported for
the full sample (that in the first column of Panel A). Mod-
els 2 and 3 show that both sets of firms (large multina-
tionals and large purely domestic firms) experience a de-
crease in CASH ETR over time, with the slope on the time
trend being −0.484 for large multinationals and −0.735 for
large domestic firms (t-statistics of −6.28 and −6.19, re-
spectively). The slopes are significantly different from one
another (t-statistic = 2.26) as shown in the rightmost col-
16
This is consistent with Markle and Shackelford (2012), who show that
the CASH ETR for multinationals is greater than that for domestic firms for
the period 20 05–20 09, and with Rego (2003), over an earlier period.
449
umn, suggesting that large domestics have a faster rate
of decline than large multinationals. Overall, these results
support H1, that effective tax rates are declining over time,
but fail to support H2, as effective tax rates of domestic
firms are declining at least as much effective tax rates of
multinational firms.
4.3. ETRs on foreign income versus ETRs on domestic income
To test our third hypothesis, we focus exclusively on
multinationals and examine changes in their effective tax
rates on two different types of income: domestic versus
foreign.17 Therefore, in this subsection we use the domes-
tic or foreign portion of current tax expense in place of
cash taxes paid as the numerator in our effective tax rate
measure.18 Specifically, we examine three measures of cur-
rent effective tax rates for multinationals: 1) the world-
wide current effective tax rate, ETRww , which is the ratio
of current worldwide tax expense to total pretax income;
2) the domestic current effective tax rate, ETRdom , which is
current federal tax expense divided by pretax domestic in-
come; and 3) the foreign current effective tax rate, ETRfor ,
which is current foreign tax expense divided by pretax for-
eign income.
Fig. 4 plots the average ETRdom and ETRfor for our sam-
ple each year over the sample period. Three results imme-
diately stand out. First, the evidence suggests that multi-
nationals on average pay tax at a higher rate on their for-
eign income than their domestic income. At the beginning
of the sample period, in 1988, the average ETRdom was ap-
proximately 30% and the average ETRfor was approximately
38%. Second, multinationals experienced declines in both
their domestic and foreign current effective tax rates over
time, with both ETRdom and ETRfor declining to approxi-
mately 28% by the end of the sample period. Third, the
decline in current effective tax rates of multinationals is
greater in magnitude with respect to foreign income than
domestic income.
We report more formal empirical tests of H3 in Table 4,
Panel A. In Model 1, we show the results of a regression
of ETRww on TIME, which reveals that ETRww has a similar
downward-sloping trend to that shown in Table 3 for the
full sample of firms using CASH ETR. In Model 2 the depen-
dent variable is ETRdom ; in Model 3, ETRfor . Comparing the
coefficient on TIME between the two models, the results
show that the downward trend is steeper for ETRfor than
for ETRdom . In Panel B, we repeat this analysis but use only
large multinationals (i.e., those in the upper quartile of
17
In this subsection, we restrict the sample to the 15,339 multinational
firm-years with non-missing federal tax expense, foreign tax expense, and
positive values for both pretax domestic income and pretax foreign in-
come.
18
Although cash taxes paid are not disclosed by jurisdiction, an analy-
sis of the differential trend on foreign versus domestic income is possible
because GAAP requires firms to break down their pretax income and cur-
rent tax expense into domestic and foreign components in their financial
reports. Current tax expense can differ from cash taxes paid because of
accruals for uncertain tax positions and other current accrued taxes, but
in general current tax expense is highly correlated with cash tax paid. In
our sample, for example, the correlation between current tax expense and
cash tax paid is about 0.85.
450 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463

Fig. 4. Estimated mean current ETR of multinational firms over time, separated into domestic and foreign components. This figure plots the annual mean
current ETR of multinational firms, split into domestic and foreign components, over the sample period, 1988–2012. ETRdom is the ratio of domestic current
tax expense to domestic pretax income. ETRfor is the ratio of foreign current tax expense to foreign pretax income. All observations are subject to the
criteria described in Table 1 and are further constrained by the availability of expense and income measures in the ratios.

Table 4 total assets for the sample).19 We observe a similar pattern

Regressions of ETRww , ETRdom , and ETRfor on TIME. in the time trends in Panel B as we do for the full sample
In this table, we report the results of estimating an ordinary least of multinationals in Panel A.
squares regression of ETRww , ETRdom , and ETRfor on a linear time trend
(TIME). TIME is calculated as the fiscal year for a given firm-year obser-
Overall, the results in Fig. 4 and Table 4 support H3,
vation less the number 1988, which is the first year in the data set. Thus, suggesting that the slope in current effective tax rates is
TIME takes on values of 0–24, which corresponds to sample years 1988– negative with respect to both foreign and domestic income
2012. ETRww is current worldwide tax expense (TXFED + TXFO) divided by over time and is decreasing at a faster rate with regard
pretax income (PI). ETRdom is current domestic tax expense (TXFED) di-
to foreign income. Thus, when we compare multination-
vided by pretax domestic income (PIDOM). ETRfor is current foreign tax
expense (TXFO) divided by pretax foreign income (PIFO). In Panel A, we als with domestic firms (in Table 3), the evidence suggests
use all available observations. In Panel B, we use only observations in the that domestic firms have a similar or a slightly steeper de-
largest quartile of total assets. In all three models, standard errors are cline over time in cash effective tax rates than do multi-
clustered by firm and year. The t-statistics are reported in parentheses nationals. However, within multinationals (in Table 4), the
below the coefficient estimates. All financial statement data are acquired
from the annual fundamentals database maintained by Compustat. Com-
evidence suggests that the current effective tax rate on
pustat data item pneumonics are reported in parentheses in all caps. ∗∗∗ , foreign-sourced income has declined more over the last 25
∗∗
, and ∗ represent statistical significance at the 1%, 5% and 10% level, re- years than has the current effective tax rate on domestic-
spectively. sourced income.20
Model 1 Model 2 Model 3
Variable (ETRww ) (ETRdom ) (ETRfor )
19
Panel A: All firms Because the quartiles are based on the full sample, the number of
INTERCEPT 32.159∗ ∗ ∗ 31.395∗ ∗ ∗ 39.159∗ ∗ ∗ observations might not be equal in particular tables.
(68.84) (49.63) (53.85) 20
Income shifting by multinational firms would not generally change
TIME −0.371∗ ∗ ∗ −0.298∗ ∗ ∗ −0.451∗ ∗ ∗ the effective rate on domestic income. Domestic income would be lower
(−9.46) (−5.12) (−10.66) or higher (if income shifted out of or into the US), and overall US taxes
N 15,336 15,336 15,336 would be lower or higher, but the tax per dollar of domestic income re-
Adj. R2 0.030 0.010 0.019 ported would not necessarily be lower or higher. In addition, the shift-
ing of income to low-tax foreign jurisdictions yields lower cash taxes and
Panel B: Large firms (upper quartile of total assets)
lower current tax expense, but if the firm later repatriates those foreign
INTERCEPT 32.824∗ ∗ ∗ 31.885∗ ∗ ∗ 40.179∗ ∗ ∗ earnings, US taxes (less a foreign tax credit) would likely need to be paid.
(50.86) (38.16) (39.42) Thus, there is a potential future US tax liability that is unrecorded in our
TIME −0.424∗ ∗ ∗ −0.309∗ ∗ ∗ −0.574∗ ∗ ∗ data. See Foley, Hartzell, Titman, and Twite (2007) and Graham, Hanlon,
(−8.30) (−4.24) (−9.35) and Shevlin (2011) for further explanation. Our objective is to examine
N 6,184 6,184 6,184 trends in taxes paid (as well as we can measure it). Thus, this measure-
Adj. R2 0.040 0.010 0.034 ment issue related to future potential repatriation taxes is not directly
relevant to our study.
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
Fig. 5. Gross domestic product-weighted average statutory tax rate of Organisation for Economic Co-operation and Development countries over the sample
period. This figure plots the weighted (by GDP) average statutory tax rate over the sample period, 1988–2012. The data are from the OECD Tax Database
and KPMG, as in Williams (2015).
Some important caveats should be kept in mind regard-
ing the data on foreign effective tax rates. First, the effec-
tive tax rate on foreign income can obscure significant het-
erogeneity in taxes across jurisdictions. For example, a firm
can face a 10% tax rate on income earned in one country,
while facing a 40% tax rate in another, but these are av-
eraged away in computing ETRfor . Thus, even when a firm
exhibits an effective tax rate on its foreign income that is
greater than its effective tax rate on domestic income, the
firm may still have incentives to shift income (from the
US or high-tax-rate foreign countries) into low-tax foreign
jurisdictions. Second, the effective tax rate on foreign in-
come, ETRfor , may obscure variation in profitability across
countries. For example, a firm may report losses in one
country and income in another, but will typically not be
able to use the loss in the first country to offset the in-
come in the second.21 The result of this friction can result
in an increased ETR on the firm’s total income.22
5. Examination of possible determinants of declining
rates
In the following subsections, we examine the effect on
declining cash ETRs of declining foreign statutory tax rates,
changes in firm characteristics, changes in the US tax sys-
21
While the incomes cannot be directly offset, see De Simone, Klassen,
and Seidman (2016) for a discussion of how firms could shift income via
transfer pricing to utilize losses cross-country.
22
Country-by-country data could offer better data for use in similar
tests but such data are not available now.
451
tem, and changes in US GAAP, as well as the effect of dif-
ferent types of book-tax differences.
5.1. The effect of declining foreign statutory tax rates
One explanation for declining CASH ETRs is the decline
in foreign statutory tax rates. To examine the overall effect
of foreign statutory tax rates, we first plot the weighted-
average statutory tax rates of OECD countries (excluding
the US) over the sample period in Fig. 5.23 Countries are
weighted each year by their GDP. The figure shows a
clear downward trend in weighted-average foreign statu-
tory tax rates over time among other developed coun-
tries. At the beginning of the sample period, the weighted-
average statutory rate in the OECD was about 45%, declin-
ing steadily to about 30% by 2012.24
We test this proposition directly by constructing a firm-
year-specific foreign statutory tax rate, FORSTAT, as the
simple average of the foreign statutory tax rates for the
countries in which the firm has significant subsidiaries.
The subsidiary data are gathered from Exhibit 21 of Form
10-K following Dyreng and Lindsey (2009) and Dyreng,
Lindsey, and Thornock, (2013) and include subsidiaries
23
Statutory tax rate and GDP data come from the OECD Tax
Database (http://www.oecd.org/tax/tax-policy/tax-database.htm#C_
CorporateCaptial, accessed October 18, 2014) as well as from KPMG, as in
Williams (2015).
24
The trend is not likely due to the weighting schema. Re-examining
the trend using an unweighted average results in a trend in foreign statu-
tory tax rates that begins at 44% in 1988 and declines to 25% in 2012.
452 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463

Table 5 tax rates for multinationals. In contrast, in Model 4, when

Regressions of CASH ETR, ETRww , ETRdom , and ETRfor on TIME and FORSTAT.
ETRfor is the dependent variable, the coefficient on FORSTAT
In this table, we report the results of estimating an ordinary least
squares regression of CASH ETR, ETRww , ETRdom , and ETRfor on a linear time is positive and highly significant, with a value of 58.448 (t-
trend (TIME) for the subsample of multinational firms with available data. statistic = 7.85).
TIME is calculated as the fiscal year for a given firm-year observation less However, the TIME coefficient is uniformly negative
the number 1988, which is the first year in the data set. Thus, TIME takes across all models in Panel A of Table 5, indicating that
on values of 0–24, which corresponds to sample years 1988–2012. CASH
changes in statutory foreign tax rates are only a partial
ETR is cash taxes paid divided by pretax income. ETRww is worldwide tax
expense divided by pretax income. ETRdom is domestic tax expense di- explanation for the decrease in current effective tax rates
vided by pretax domestic income. ETRfor is foreign tax expense divided over time. The TIME coefficient is smaller in magnitude
by pretax foreign income. In Panel A, we use all available observations. in the ETRfor regression. Nevertheless, it is still negative
FORSTAT is the average of the foreign statutory tax rates for the countries
and significant (coefficient = −0.144, t-statistic = −2.08). In
in which the firm has subsidiaries listed in Exhibit 21 of Form 10-K. In
Panel B, we use only observations in the largest quartile of total assets.
Panel B, we repeat the analysis for large firms and reach
In all three models, standard errors are clustered by firm and year. The t- similar conclusions.
statistics are reported in parentheses below the coefficient estimates. ∗∗∗ , Together, the results in Fig. 5 and Table 5 suggest that
∗∗
, and ∗ represent statistical significance at the 1%, 5% and 10% level, re- the decline in effective tax rates is associated with the de-
spectively.
cline in foreign statutory tax rates. However, the decline
Model 1 Model 2 Model 3 Model 4 in foreign statutory tax rates is at best a partial explana-
Variable (CASH ETR) (ETRww ) (ETRdom ) (ETRfor ) tion for the declining effective rates observed in our data.
Panel A: All firms It does not explain the declining rates in purely domes-
INTERCEPT 26.790∗ ∗ ∗ 27.872∗ ∗ ∗ 36.286∗ ∗ ∗ 16.929∗ ∗ ∗ tic US firms and even among US multinationals, evidence
(11.96) (16.43) (13.08) (5.48)
still exists of declining effective rates after controlling for
TIME −0.356∗ ∗ ∗ −0.332∗ ∗ ∗ −0.408∗ ∗ ∗ −0.144∗ ∗
(−5.28) (−7.50) (−6.40) (−2.08)
variation in foreign statutory rates, both with respect to US
FORSTAT 15.397∗ ∗ ∗ 12.492∗ ∗ ∗ −10.323 58.448∗ ∗ ∗ multinational firms’ domestic effective tax rates and their
(2.93) (2.86) (−1.43) (7.85) foreign effective tax rates.
N 11,172 11,172 11,172 11,172
Adj. R2 0.030 0.034 0.012 0.037
5.2. The effect of changes in firm characteristics
Panel B: Large firms (upper quartile of total assets)
INTERCEPT 27.993∗ ∗ ∗ 30.268∗ ∗ ∗ 36.096∗ ∗ ∗ 20.462∗ ∗ ∗
We next examine the possibility that firms themselves
(9.00) (10.74) (8.64) (4.58)
TIME −0.401∗ ∗ ∗ −0.385∗ ∗ ∗ −0.373∗ ∗ ∗ −0.250∗ ∗ have changed over time (e.g., Graham, Leary, and Roberts,
(−4.87) (−6.01) (−4.15) (−2.63) 2015), altering their mix of characteristics and abilities to-
FORSTAT 13.640∗ 6.602 −11.155 50.006∗ ∗ ∗ wards those that yield lower effective tax rates.26 As one
(1.75) (0.92) (−1.04) (4.55) example, US firms receive a tax credit for certain expen-
N 4,840 4,840 4,840 4,840
ditures related to research and experimentation. If firms
Adj. R2 0.037 0.036 0.009 0.041
have shifted investment away from fixed assets and into
research and experimentation, then more can firms qual-
in developed countries and all countries, including tax ify for and receive tax credits, causing effective tax rates to
havens.25 drop on average. Another example could arise if rules or
In Table 5, Panel A, we present the results of a re- abilities of firms change. For example, if firms are required
gression of effective tax rates on TIME and FORSTAT. In to depreciate capital assets over a specified period of time,
Model 1, the dependent variable is CASH ETR, which in- and that period of time is reduced by the government,
cludes both domestic and foreign taxes paid relative to the then the effect of capital expenditures on cash effective
firm’s worldwide pretax income. The results indicate that tax rates will also change over time. Furthermore, improve-
the coefficient on FORSTAT is positive and statistically sig- ments in tax avoidance technology could have evolved over
nificant, with a value of 15.397 (t-statistic = 2.93). This time. For example, if firms are better able to shift income
finding, together with the inference from Fig. 5 of a de- related to intangible assets to low-tax jurisdictions than
crease in foreign statutory tax rate, supports the explana- they were in the past, then, even holding the intangibil-
tion that the decrease in foreign statutory tax rates is at ity of firms constant, a decrease in effective tax rates could
least partially responsible for the decrease in effective tax be evident over time. Thus, changes in firm characteristics
rates over time. In Model 2, we repeat the exercise using could account for a decline in effective tax rates over time,
ETRww as the dependent variable and reach similar con- simply because firms today look and act differently than
clusions. In Models 3 and 4, we present the results sep- firms did 25 years ago. Moreover, even holding the effect
arately for multinationals’ foreign and domestic current ef- of firm characteristics constant, effective tax rates could
fective tax rates, ETRfor and ETRdom , respectively. In Model change if the mapping of firm characteristics into taxes has
3, we find that when ETRdom is the dependent variable, changed over time.
the coefficient on FORSTAT is insignificant, suggesting that
foreign statutory tax rates do not affect domestic effective
26
These variables are, to some extent, a result of choices by the firm.
As an example, as discussed above, an increase in the opportunity for tax
25
Exhibit 21 is required by Item 601 of SEC Regulation S-K to accom- avoidance behavior because of an exogenous change, such as a decline in
pany the form 10-K filed with the SEC. The sample is restricted to those foreign statutory tax rates, could prompt the firm to take tax avoidance
that have foreign subsidiary data from the Exhibit 21, which reduces the actions, such as shifting operations overseas, that in turn affect the firm’s
sample size to 11,172 firm-year observations. characteristics.
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463 453

We examine the influence of firm characteristics on the Table 6

trend in cash effective tax rates by estimating a regression
of CASH ETR on TIME while allowing for variation in firm
characteristics. The approach follows that of prior time-
trend studies (e.g., Collins, Maydew, and Weiss, 1997; Fran-
cis and Schipper, 1999; Collins, Li, and Xie, 2009), which
examine the continued significance of TIME with the inclu-
sion of additional explanatory variables. Thus, the model
is
CASH ET Rit = βindustry + β1 T IMEt + β2 LOG ASSET Sit
+ β3 R&D E X P E NSEit + β4 P P &Eit
Regressions of CASH ETR on TIME and firm characteristics.
In this table, we report the results of estimating Eq. (2), which is an or-
dinary least squares regression of CASH ETR on a linear time trend (TIME)
and firm characteristics. CASH ETR is cash taxes paid divided by pretax
income. TIME is calculated as the fiscal year for a given firm-year obser-
vation less the number 1988, which is the first year in the data set. Thus,
TIME takes on values of 0–24, which corresponds to sample years 1988–
2012. All other variables are as defined in Table 2. The estimation includes
industry fixed effects based on the Barth, Beaver, Hand, and Landsman
(2005) classification schema. In all three models, standard errors are clus-
tered by firm and year. The t-statistics are reported in parentheses below
the coefficient estimates. ∗∗∗ , ∗∗ , and ∗ represent statistical significance at
the 1%, 5% and 10% level, respectively.

+ β5 INT ANGIBLE ASSE T Sit + β6 LEV E RAGEit Model 1 Model 2 Model 3

Variable (all firms) (multinationals) (domestics)
+ β7CAP IT AL EX P ENDIT URE Sit
TIME −0.439∗ ∗ ∗ −0.443∗ ∗ ∗ −0.432∗ ∗ ∗
+ β8 ADV ERT ISING EX P ENSEit (−8.07) (−6.77) (−7.86)
3.022∗ ∗ ∗
+ β9 SP ECIAL IT EMSit MNE
(7.21)
+ β10 LAGGE D SP ECIAL IT EMSit−1 LOG ASSETS 0.212 −0.181 0.458∗ ∗ ∗
(1.62) (−1.29) (2.78)
+ β11 N OLit + β12 N OLit + it . (2) R&D EXPENSE −30.668∗ ∗ ∗ −25.268∗ ∗ ∗ −34.137∗ ∗ ∗
(−7.68) (−5.83) (−5.07)
When we estimate Eq. (2) on the full sample, we in-
PP&E −9.328∗ ∗ ∗ −9.374∗ ∗ ∗ −8.824∗ ∗ ∗
clude MNE as an explanatory variable; it is not included (−7.34) (−5.35) (−6.38)
when we split the sample into multinationals and purely INTANGIBLE ASSETS 1.582 3.079∗ 0.282
domestic firms. We employ a set of controls that prior (1.07) (1.93) (0.13)
research has identified to be important drivers of tax LEVERAGE −4.440∗ ∗ ∗ −0.791 −6.644∗ ∗ ∗
(−4.28) (−0.58) (−5.24)
avoidance and tax sheltering (e.g., Graham and Tucker, CAPITAL EXPENDITURES −4.266∗ ∗ ∗ −7.040∗ ∗ ∗ −1.595
2006; Chen, Chen, Cheng, and Shevlin, 2010; Kim, Li, and (−4.18) (−5.01) (−1.49)
Zhang, 2011; Hasan, Hoi, Wu, and Zhang, 2014; Desai and ADVERTISING EXPENSE −0.238 1.877 3.819
Dharmapala, 2009). The model includes industry fixed ef- (−0.05) (0.27) (0.50)
SPECIAL ITEMS −241.444∗ ∗ ∗ −258.616∗ ∗ ∗ −216.576∗ ∗ ∗
fects based on the classifications in Barth, Beaver, Hand,
(−14.18) (−14.86) (−11.10)
and Landsman, (2005) to account for any changes in the LAGGED SPECIAL ITEMS 45.016∗ ∗ ∗ 34.161∗ ∗ ∗ 58.858∗ ∗ ∗
composition of firms across industries over time. We clus- (4.44) (2.97) (7.03)
ter standard errors by firm and year. All variables are de- NOL −5.506∗ ∗ ∗ −2.572∗ ∗ ∗ −9.559∗ ∗ ∗
fined in Table 2. (−10.41) (−6.26) (−12.97)
NOL 1.020 3.035∗ −0.131
Table 6 presents the results of the pooled ordinary least (1.30) (1.84) (−0.17)
squares estimation of Eq. (2). Model 1 presents the results
Industry fixed effects Yes Yes Yes
for the full sample of firms, and Models 2 and 3 present N 54,028 27,561 26,467
the results for multinationals and purely domestic firms, Adj. R2 0.105 0.102 0.129
respectively. Similar to the models we estimated without
firm-level controls (i.e., in reduced form), we find that
the coefficient on the time trend is negative and signifi-
cant, equal to −0.439 in Model 1, −0.443 in Model 2, and velopment are associated with reduced taxes paid on in-
−0.432 in Model 3, with t-statistics ranging from −6.77 to come. The coefficient on PP&E is negative and significant,
−8.07. This suggests, that even after controlling for firm as are the coefficients, generally, on LEVERAGE and CAPI-
characteristics, we continue to see a significant decrease in TAL EXPENDITURES. The coefficient on INTANGIBLE ASSETS is
effective tax rates over time, both in the sample as a whole marginally significant in Model 2 and insignificant in the
and in the subsamples of multinationals and purely do- other models. The coefficient on SPECIAL ITEMS is nega-
mestic firms. Moreover, the decrease in effective tax rates tive and significant, and the coefficient on LAGGED SPECIAL
is economically significant. For example, a coefficient of ITEMS is positive and significant, both of which are consis-
−0.439 on the time trend corresponds to an approximately tent with special items being deductible for tax purposes
10% decrease in the CASH ETR of a firm over the 25-year later than they are recognized for financial reporting pur-
sample period, holding other variables constant. poses. As expected, the coefficient on NOL is negative and
The coefficients on the control variables provide some significant, while NOL is generally not significant.
important insights. In Model 1, the coefficient on MNE is Two additional points are worth mentioning. First,
positive and significant, suggesting that multinationals dur- while firm characteristics do help explain variation in
ing our sample period exhibited higher cash effective tax cash effective tax rates across firms (e.g., the adjusted R2
rates than did domestic-only firms. LOG ASSETS is posi- increases to 10.5% in Model 1 of Table 6, compared with
tive in Model 3, suggesting that larger domestic firms have 1.5% in Model 1 of Table 3), they apparently do little to
higher cash effective tax rates. In all three models, the co- explain the decrease in cash effective tax rates over time.
efficient on R&D EXPENSE is negative, consistent with our The estimated decline in CASH ETR is the same magnitude
expectation that greater expenditures on research and de- in Table 6, after controlling for changes in firm character-
454 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463

Fig. 6. Mean annual CASH ETR over the sample period conditional on firm characteristics. This figure plots the annual conditional mean CASH ETR over
the sample period, 1988–2012. CASH ETR is the ratio of cash taxes paid to pretax income. The data points in the figure represent the intercepts from the
following annual regressions:
CASH ET Rit = βt0 + βt2 LOG ASSET Sit + βt3 R&D EXPENSEit + βt4 P P &Eit + βt4 INT ANGIBLE ASSE T Sit + βt6 LEV E RAGEit + βt7CAPIT AL E XPE NDIT URE Sit
+βt8 ADV ERT ISING EXP ENSEit + βt9 SP ECIAL IT EMSit + βt10 LAGGED SPECIAL IT EMSit + βt11 N OLit + βt12 N OLit + ∈it .
All continuous variables are mean centered so that the intercept captures a hypothetical firm with all indicator variables equal to zero and the mean value
of each of the continuous variables.

istics, as it was in Table 3, which reported the results of
regressions without controls. Second, the results in Table 6
reinforce the earlier finding that the decrease in effective
tax rates over time is, at best, only partially explained by
the rise in multinationals over time. Specifically, Table 6
continues to show that effective tax rates decreased over
time for both US multinationals (reported in Model 2)
and domestic-only US firms (reported in Model 3) and at
roughly the same rate.
Fig. 6 plots the changes in the conditional mean of
CASH ETR over time. The figure is generated by estimat-
ing Eq. (2) each year (without industry fixed effects) and
plotting the intercept.27 Fig. 6 shows a downward trend
in CASH ETR over time, conditional on firm characteristics,
consistent with earlier results.
We also investigate whether the effect of firm charac-
teristics on tax rates has changed over time by adopting a
methodology used by Rajgopal and Venkatachalam (2011),
in which we interact firm characteristics with the time
trend variable, TIME, as follows:
CASH ET Rit = βindustry + β1 T IME

+ (βcCit + γcCit ∗ T IMEt ) + it . (3)
c wards inversions (e.g., Desai and Hines, 2002; Seida and Wempe, 2004).
27
All continuous variables are mean centered so that the intercept cap-
tures a hypothetical firm with all indicator variables equal to zero and the
mean value of each of the continuous variables.
where C represents the same vector of firm characteristics
as used in Eq. (2): R&D EXPENSE, INTANGIBLE ASSETS, LOG
ASSETS, PP&E, LEVERAGE, CAPITAL EXPENDITURES, ADVERTIS-
ING EXPENSE, SPECIAL ITEMS, LAGGED SPECIAL ITEMS, NOL,
and NOL. The important difference between Eqs. (2) and
(3) is that we interact each of the firm characteristics with
TIME.
In untabulated tests, we estimate Eq. (3) across the
entire sample (Model 1), multinationals (Model 2), and
purely domestic firms (Model 3). Across all three models,
the coefficient on the main time trend effect, TIME, is neg-
ative and significant and maintains a consistent level rang-
ing from −0.484 to −0.570. On the interactions, little con-
sistent evidence exists of changes in the relation between
firm characteristics and effective tax rates over time. The
coefficients on the majority of the interaction terms are in-
significant, with a few exceptions. Overall, the results sug-
gest a decline in effective tax rates over time that is unex-
plained by changes in the mapping of firm characteristics
into taxes.28
28
We also examine if the decline is explained by the recent trend to-
Some inverted firms in the sample likely experienced a substantial de-
cline in effective tax rates post-inversion. To the extent that the inversions
are a recent phenomenon and decrease effective tax rates, they could af-
fect the over-time trend in effective tax rates, as discussed in Sullivan
(2010). In untabulated analyses, we examine the trend after excluding a
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
Fig. 7. CASH ETR over time with different regulatory regimes. This figure plots mean CASH ETR over the sample period. Each panel highlights periods when
different regulatory regimes were effective that could have had an effect on CASH ETR. All observations are subject to the criteria described in Table 1.
AJCA = American Jobs Creation Act of 2004.
5.3. The effect of changes in the US tax system over time
We chose the beginning of our sample period, 1988, to
coincide with the last major overhaul of the US corporate
tax system.29 Since that time, the US statutory tax rate has
remained effectively flat. In Fig. 7 and Table 7, we exam-
ine three important changes to the US tax system that are
most likely affect the trend in effective tax rates. These are
the advent of the “check-the-box” rules in 1997, the repa-
triation tax holiday in 20 04–20 05, and the bonus deprecia-
tion rules in place during the periods 20 01–20 04 and after
2007.
Fig. 7 reproduces the annual average CASH ETRs for our
sample previously presented in Fig. 1, but with overlays for
the periods of change in US tax laws. A visual inspection
does not appear to support any of these tax or regulatory
changes as producing a major shift in cash effective tax
rates. However, any shifts could be more subtle than can
list of recent inverting firms. The trend for the remaining firms is nearly
identical to that reported in Fig. 1 and conclude that inversions are not
likely driving the downward trend in effective tax rates that we observe.
Moreover, because our sample is composed solely of US domestic firms
and US multinationals, most inverted firms were already excluded from
our sample, as Compustat classified them as foreign following the inver-
sion. In addition, the actual number of completed inversions during our
sample period was relatively modest.
29
The Tax Reform Act of 1986 took effect on July 1, 1987.
455
be perceived visually. Accordingly, we augment Eq. (2) to
include variables for each of the tax changes, and present
the results in Table 7. For expositional ease, we present
only the coefficients on the newly added variables and the
time trend (although all control variables are included in
the regression as estimated). The variable CHECKBOX takes
on a value of one beginning in 1997 and a value of zero
prior to 1997, marking the implementation of the check-
the-box regulations on January 1, 1997. The variable AJCA
takes on a value of one in the years 2004 and 2005 and
zero otherwise. It reflects the tax holiday that allowed for a
reduced tax rate on certain repatriations taking place in ei-
ther 2004 or 2005, as specified by the American Jobs Cre-
ation Act of 2004 (Blouin and Krull, 2009).30 We create
indicator variables for the years when Congress allowed
for more rapid (i.e., bonus) depreciation of certain new as-
sets to stimulate investment. BONUSDEPN1 takes on a value
of one in the period 20 01–20 04 and zero otherwise, and
BONUSDEPN2 takes on a value of one in the period 2007–
2012 and zero otherwise.
30
Section 199 of the AJCA also provides for a deduction of part of the
income related to Domestic Production Activities. This provision was ef-
fective starting in 2005 (and is still in effect as of the end of 2016) so the
time trend test surrounding the Sarbanes-Oxley Act of 2002 subsumes
testing for the effects of Section 199.
456 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463

Table 7 or goodwill accounting.31 We conduct several tests in an

Regressions of CASH ETR on TIME, regimes, and the interaction of TIME
attempt to assuage concerns that our results are driven by
and regimes.
In this table, we report the results of estimating an augmented ver- accounting rule changes.
sion of Eq. (2), which is an ordinary least squares regression of CASH ETR We first eliminate the effect of accounting changes by
on a linear time trend (TIME) and regulatory/tax regimes. CASH ETR is developing a cash flow based ETR measure that uses oper-
cash taxes paid divided by pretax income. TIME is calculated as the fis-
ating cash flows in the denominator instead of pretax ac-
cal year for a given firm-year observation less the number 1988, which is
the first year in the data set. Thus TIME takes on values of 0–24, which
counting income (CASH FLOW ETR). This measure is unaf-
corresponds to sample years 1988–2012. CHECKBOX is equal to one for all fected by GAAP changes, because it uses cash flow from
years after 1996, zero otherwise. AJCA is equal to one for all years after operations as the benchmark, not pretax income based on
2003, zero otherwise. BONUSDEPN1 is equal to one for the years 2001– financial accounting rules. In Fig. 8, we plot the CASH FLOW
2004, zero otherwise. BONUSDEPN2 is equal to 1 for the years 2008–2012,
ETR over the sample period, separately for multinationals
zero otherwise. All other variables are as defined in Table 2. The estima-
tion includes industry fixed effects based on the Barth, Beaver, Hand, and and domestic firms. The plot shows a declining trend for
Landsman (2005) classification schema. In all three models, standard er- CASH FLOW ETR over the sample period, and the trend is
rors are clustered by firm and year. The t-statistics are reported in paren- very similar for both multinationals and domestic firms.
theses below the coefficient estimates. ∗∗∗ , ∗∗ , and ∗ represent statistical The fact that we observe this declining pattern for the
significance at the 1%, 5% and 10% level, respectively.
CASH FLOW ETR suggests that changes in financial account-
Variable Model 1 Model 2 Model 3 ing rules are unlikely to be driving the results.32
TIME −0.633∗ ∗ ∗ −0.673∗ ∗ ∗ −0.425 We also examine the effect of several accounting or fi-
(−4.40) (−6.45) (−6.39) nancial reporting changes on our main results. Specifically,
CHECKBOX −3.883∗ ∗ we examine the effects of changes in the accounting treat-
(−2.14)
ment of goodwill (i.e., was amortized over 40 years until
CHECKBOX∗ TIME 0.353∗ ∗
(2.05) 2002, when the rules changed to goodwill not being amor-
AJCA −12.672∗ tized but instead tested for impairment) and the expens-
(−1.78) ing of stock options (i.e., compensation related to employee
AJCA∗ TIME 0.823∗ ∗ stock options was not a required expense until 2005).33
(2.29)
BONUSDEPN1 21.045∗ ∗ ∗
The results are included in the Online Appendix. We sum-
(4.39) marize the results briefly here.
BONUSDEPN1∗ TIME −1.735∗ ∗ ∗ To capture firms with stock options, we follow an ap-
(−5.28) proach similar to that in Graham, Lang, and Shackelford
BONUSDEPN2 19.965
(2004) by comparing firms listed on the NASDAQ to other
(1.66)
BONUSDEPN2∗ TIME −0.855 sample firms. The results show that the pattern in effec-
(−1.50) tive tax rates is very similar between NASDAQ firms and
Control variables Yes Yes Yes other sample firms, suggesting that changes in stock op-
Industry fixed effects Yes Yes Yes tion accounting are not driving our results. To account for
N 54,028 54,028 54,028 changes in accounting for goodwill, we examine the differ-
Adj. R2 0.106 0.108 0.112
ential trend in CASH ETR for firms with intangible assets
relative to those without. The results show that the down-
ward trend in effective tax rates is very similar between
The results in Table 7 show some evidence of changes intangible-laden firms versus other sample firms, suggest-
in the time trend for some of the regimes. For the ing that goodwill accounting is not driving our results.34
check-the-box period (Model 1), the results show a neg-
ative coefficient on TIME, with a positive coefficient on
CHECKBOX∗ TIME. The results for the AJCA time periods also 31
However, the direction of the effect of changes in GAAP on the trend
show a negative coefficient on TIME, with a positive co- in CASH ETR is unclear because some changes in GAAP, such as stock op-
tion expensing would increase CASH ETR (by reducing accounting earn-
efficient on the interaction. In Model 3, the slope on the
ings), while others such as fair value accounting might potentially de-
time trend is significantly negative even after accounting crease CASH ETR (depending on market value changes).
for both bonus depreciation regimes, with a negative coef- 32
We also examine other tax rates using alternative denominators in
ficient on the first interaction. Overall, the evidence sug- the Online Appendix. See Hanlon and Heitzman (2010) for a detailed dis-
gests that the trend in tax rates over time is negative, cussion of the pros and cons of common empirical measures of tax avoid-
ance.
notwithstanding some differences in the time trend during 33
We also examine the effect of the Sarbanes Oxley Act (SOX) in the
several regulatory regimes. Online Appendix. Although SOX was primarily a response to financial
reporting scandals of that period, it could conceivably affect tax avoid-
5.4. The effect of changes in financial accounting rules over ance by affecting firms’ overall attitudes toward regulatory enforcement
or conservatism, as well as other indirect effects such as interactions with
time
auditors (Maydew and Shackelford, 2007). To examine the effect of SOX,
we conduct tests similar to those for tax law changes above. We find no
As noted in Dyreng, Hanlon, and Maydew, (2008), the evidence to suggest that SOX is responsible for the downward trends in
numerator in CASH ETR, cash taxes paid, is a cash-based effective tax rates.
34
amount of total taxes paid across all jurisdictions. However, We note that long-run changes in the denominator, financial account-
ing income, would be hard-pressed to explain the full declining trend in
the denominator in CASH ETR, pretax financial income, is CASH ETR. For example, for financial accounting income to explain the 5
affected by regulatory changes in financial accounting rules percentage point decline in CASH ETR (i.e., from 32% to 27% in Fig. 2), fi-
(i.e., US GAAP), such as changes in stock option expensing nancial accounting income would have had to increase by 19 percentage
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
Fig. 8. Mean annual CASH FLOW ETR separately for multinational and domestic firms. This figure plots the mean CASH FLOW ETR over the sample period,
1988–2012. CASH FLOW ETR is the ratio of cash taxes paid to pretax cash flow from operations. All observations are subject to the criteria described in
Table 1.
5.5. Examination of types of book-tax differences
In this subsection, we investigate what types of book-
tax differences could be contributing to the decline in
CASH ETRs. By doing so, we provide a more detailed ex-
amination of the types of accounts and transactions that
could lead to that trend. For example, the trend could be
different for temporary versus permanent differences for
MNEs, because income-shifting strategies are often classi-
fied as creating permanent differences. Therefore, we ex-
amine whether permanent or temporary book-tax differ-
ences, or both, differentially contribute to the declining
trend in tax rates.
We follow Edwards, Schwab, and Shevlin (2016) in pars-
ing book-tax differences into those that are temporary in
nature (i.e., those that will reverse in the future, such as
depreciation) and those that are permanent in nature (i.e.,
differences that never reverse, such as municipal bond in-
terest). PERMDIFF equals negative one multiplied by the
difference between the statutory tax rate and the ratio of
total tax expense to pretax income. TEMPDIFF equals neg-
ative one multiplied by the ratio of deferred tax expense
to pretax income. By multiplying each measure by neg-
ative one, lower values of PERMDIFF and TEMPDIFF indi-
cate increased tax avoidance (just as lower values of CASH
points, holding constant the numerator [(32/100)−(32/119) = 5%]. In other
words, GAAP rule changes would have had to increase financial account-
ing income by about 200 billion dollars in 2012 alone [= 1,088 billion
aggregate pretax income times 19%] to fully explain that year’s decrease
relative to aggregate income in 1988.
457
ETR indicate increased tax avoidance). It is important to
note where the effect of unpaid (deferred) US tax on for-
eign earnings of a US multinational will be reflected. Be-
cause the US has a worldwide tax system with deferral,
the US tax on foreign earnings is not paid until earnings
are repatriated to the US. When earnings are not repatri-
ated, the deferral of US taxation on those earnings reduces
cash taxes paid and reduces current tax expense relative
to the US statutory tax rate of 35% (assuming foreign rates
are lower than the US statutory tax rate). Companies are
required to accrue the US tax on unremitted foreign earn-
ings as a deferred tax expense unless the company asserts
that the unremitted foreign earnings are indefinitely rein-
vested abroad. If the company accrues the US taxation on
some or all of its unremitted foreign earnings, then that
portion of foreign earnings will be reflected as a temporary
difference between book and taxable incomes (because the
foreign earnings are included in accounting income). If the
company asserts some or all of its unremitted foreign earn-
ings are indefinitely reinvested and does not accrue the
residual US tax, then that portion of unremitted foreign
earnings will be reflected as a permanent difference be-
tween book and taxable incomes. If a permanent differ-
ence, the GAAP effective tax rate as well as the cash and
current effective tax rate will be reduced. If a temporary
difference, the GAAP effective tax rate is not affected (cur-
rent tax expense is lower but the deferred tax expense is
higher).
In the interest of parsimony, we discuss the results here
and present the related figures in the Online Appendix. We
find that, in the early years of the sample, the median firm
458 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
had permanent differences that increased the cash effec-
tive tax rate by about 3 percentage points. By the end of
the sample period, the median firm in our sample had per-
manent differences that decreased the cash effective tax
rate by about 2 percentage points. This pattern suggests
that the effect of permanent book-tax differences (e.g., un-
remitted foreign earnings in which the US tax is not ac-
crued, tax credits, and domestic production incentives) has
increased over time, driving effective tax rates lower.
Temporary differences had almost no effect on cash ef-
fective tax rates for the median firm at the beginning of
the sample period, but the effect changes over time. We
find that, during some periods, firms were able to take
substantial temporary tax benefits, such as accelerated de-
preciation, possibly a result of special tax provisions or in-
creased capital investment in some years. Both temporary
and permanent differences have linear time trends that
suggest they are driving effective tax rates down over the
sample period, but the effect of permanent differences is
greater than the effect of temporary differences.
We also split the sample into multinational firms and
purely domestic firms to examine whether the results vary
across these two types of firms. For multinational firms,
the largest effect is driven by permanent differences, with
the median firm showing a positive effect of 4 percentage
points at the beginning of the sample period and about
negative 4 percentage points by the end. The data suggest
that temporary differences have relatively little effect on
the cash effective tax rates of multinational firms over
time. In contrast, the effect of temporary items is large
for purely domestic firms, but the effect of permanent
differences is relatively small. Taken together, these re-
sults suggest that multinational firms are able to take
advantage of permanent differences (e.g., earnings in
countries with low tax rates that are then designated as
indefinitely reinvested) and that purely domestic firms
appear to take advantage of items that are temporary
differences (e.g., accelerated depreciation, including bonus
depreciation).
To further probe the causes of the declining time trend
in tax rates for multinational firms versus purely domes-
tic firms, we conduct a more detailed analysis on a small
subset of firms. We hand-collect detailed book-tax differ-
ences from the Form 10-Ks for 50 firms at five different
times from income tax rate reconciliation contained in the
income tax footnote. The rate reconciliation requires firms
to reconcile the firm’s reported GAAP effective tax rate to
the US statutory tax rate by disclosing the material items
that cause the difference in rates. With this subsample, we
examine which types of accounts appear to be more asso-
ciated with declining effective tax rates.
Substantial variation emerges in the types of items that
firms disclose. However, in our subsample of firms, we find
some recurring themes, which have changed over time.
For example, it has become more common for multina-
tional firms to report foreign tax rate differences and for
domestic firms to report domestic production tax incen-
tives. These findings are consistent with different types of
tax planning leading to the decline in tax rates for multi-
nationals versus domestic firms. More detailed figures are
included in the Online Appendix.
6. Additional tests and robustness
In the following subsections, we conduct additional
tests and robustness tests regarding alternative measures
of MNE firms, as well as sensitivity to losses and NOLs. We
also briefly examine foreign incorporated firms relative to
US firms.
6.1. Alternative measures of multinationality
Much of the recent debate about tax avoidance has cen-
tered on cross-border tax planning by large multination-
als. The results above suggest that such planning is likely
only a partial explanation for the decreasing trend in effec-
tive tax rates. However, from an empirical perspective, it
is challenging to measure multinational activity precisely
(e.g., Donohoe, McGill, and Outslay, 2012). In this subsec-
tion, we employ several variants of MNE to ensure that our
conclusions are not sensitive to a measurement issue re-
garding firms’ multinationality.
We first repeat our interacted model regression spec-
ified in Eq. (3) using six different measures of multina-
tional activity. We include, one by one, the following alter-
native MNE variables: (1) an indicator variable for whether
or not the firm records foreign sales in its geographic seg-
ment data, (2) an indicator variable for whether or not the
firm discloses at least one subsidiary in a foreign country
in Exhibit 21 of Form 10-K, (3) the ratio of foreign sales
disclosed in the geographic segment data to total sales, (4)
the ratio of pretax foreign income to total pretax income,
(5) the ratio of the absolute value of pretax foreign income
to sales, and (6) the log of the total number of foreign
countries in which a firm discloses a significant subsidiary
in Exhibit 21 of its 10-K. The results of this additional anal-
ysis are reported in Table 8. Across all specifications, the
coefficient on TIME is negative and significant, with val-
ues ranging from −0.526 to −0.582, and all the t-statistics
have values greater in magnitude than 8.56. The interac-
tion coefficient on X∗ TIME is significant in only two of the
specifications (both using variation in the same underlying
data from Exhibit 21 of Form 10-K). Thus, using different
measures of multinational activity, we find that increases
in multinational activity explain only a portion of the de-
cline in effective tax rates.
6.2. Sensitivity of results to financial accounting losses and
net operating loss carryforwards
Because CASH ETRs are well defined only when pretax
income is positive, in our main tests we eliminate obser-
vations in which the denominator of the CASH ETR, pretax
accounting income, is not positive. However, perhaps elim-
inating loss firm-years from the sample yields results that
are somehow not representative of the population of firms.
Another possible issue related to tax losses is the ability to
utilize current period tax losses against taxable profits in
other periods (i.e., tax loss carryforwards or carrybacks).35
35
See Erickson, Heitzman, and Zhang (2013) for a discussion and tests
of tax loss carryback incentives and Erickson and Heitzman (2010) for an
analysis of how companies use and value NOLs.
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463 459

Table 8
Sensitivity of results to different measures of multinationality.
In this table, we report the results of estimating Eq. (3), which is an ordinary least squares regression of CASH ETR on a linear time trend (TIME) and
firm characteristics, and the interaction of TIME with the firm characteristics. CASH ETR is cash taxes paid divided by pretax income. TIME is calculated as
the fiscal year for a given firm-year observation less the number 1988, which is the first year in the data set. Thus, TIME takes on values of 0–24, which
corresponds to sample years 1988–2012. Each column uses a different proxy for the firm’s multinationality. SALEFO is foreign sales from the geographic
segment disclosure. N COUNTRIES counts the number of foreign countries listed on Exhibit 21 of Form 10-K filed with the Securities and Exchange Com-
mission (SEC). We backfill N COUNTRIES for observations prior to 1995 using the oldest available SEC filing. SALE is worldwide sales, and PIFO is pretax
income from foreign operations. PI is worldwide pretax income. All other variables are as defined in Table 2. The estimation includes industry fixed effects
based on the Barth, Beaver, Hand, and Landsman (2005) classification schema. The t-statistics are reported in parentheses below the coefficient estimates.
In all models, standard errors are clustered by firm and year. ∗∗∗ , ∗∗ , and ∗ represent statistical significance at the 1%, 5% and 10% level, respectively.

X=

Variable SALEFO > 0 N COUNTRIES > 0 SALEFO/SALE PIFO/PI ABS(PIFO)/SALE log(N COUNTRIES + 1)

TIME −0.567∗ ∗ ∗ −0.582∗ ∗ ∗ −0.546∗ ∗ ∗ −0.532∗ ∗ ∗ −0.532∗ ∗ ∗ −0.526∗ ∗ ∗

(−9.02) (−8.83) (−8.90) (−8.71) (−8.69) (−8.56)
X 0.956 0.354 1.986 0.140∗ ∗ −99.832 0.274
(1.35) (0.69) (1.01) (2.56) (−0.23) (1.13)
X∗ TIME 0.066 0.100∗ ∗ 0.097 −0.003 7.915 0.056∗ ∗ ∗
(1.24) (2.54) (0.79) (−0.59) (0.34) (3.20)
Controls Yes Yes Yes Yes Yes Yes
Industry fixed effects Yes Yes Yes Yes Yes Yes
N 54,028 54,028 54,028 54,028 54,028 54,028
Adj. R2 0.108 0.108 0.108 0.109 0.107 0.109

Table 9
Sensitivity of results to losses and net operating loss (NOL) carryforwards.
In this table, we report the results of estimating Eq. (3), except we impose different sample selection criteria to evaluate the sensitivity of the results
to financial accounting losses and net operating loss carryforwards. In Model 1, we require the firm to have non-missing data, including positive pretax
income, for each year of the 25-year sample period. In Model 2, we require the firm to always have positive pretax income when data are available, but
data only need to be available for any five years during the sample period. Model 3 requires that the firm never report a net operating loss carryforward
during the sample period. The t-statistics are reported in parentheses below the coefficient estimates. In all models, standard errors are clustered by firm
and year. ∗∗∗ , ∗∗ , and ∗ represent statistical significance at the 1%, 5% and 10% level, respectively.

Model 1 (constant sample of Model 2 (firms that never Model 3 (firms that never
Variable firms that never report a loss) report a loss) report an NOL)

TIME −0.252∗ ∗ ∗ −0.431∗ ∗ ∗ −0.566∗ ∗ ∗

(−3.33) (−7.53) (−9.59)
MNE 0.274 1.136∗ ∗ 0.710
(0.28) (2.13) (1.15)
LOG ASSETS −0.520∗ −0.170 −0.174
(−1.82) (−1.00) (−0.90)
R&D EXPENSE −60.035∗ ∗ ∗ −52.338∗ ∗ ∗ −64.832∗ ∗ ∗
(−5.26) (−8.75) (−11.58)
PP&E −5.209∗ −10.228∗ ∗ ∗ −11.154∗ ∗ ∗
(−1.91) (−6.27) (−6.22)
INTANGIBLE ASSETS −8.275∗ ∗ −2.968 1.854
(−2.21) (−1.54) (0.83)
LEVERAGE −2.808 −4.894∗ ∗ ∗ −6.428∗ ∗ ∗
(−0.86) (−3.16) (−3.85)
CAPITAL EXPENDITURES −3.734 −7.926∗ ∗ ∗ −5.604∗ ∗ ∗
(−1.13) (−5.14) (−3.42)
ADVERTISING EXPENSE −4.421 3.051 9.520
(−0.36) (0.35) (1.00)
SPECIAL ITEMS −306.577∗ ∗ ∗ −258.224∗ ∗ ∗ −261.656∗ ∗ ∗
(−11.80) (−17.70) (−16.28)
LAGGED SPECIAL ITEMS 79.910∗ ∗ ∗ 76.763∗ ∗ ∗ 53.090∗ ∗ ∗
(3.87) (6.50) (4.02)
NOL −1.723 −2.418∗ ∗ ∗
(−1.63) (−3.07)
NOL 14.585∗ 9.257∗ ∗ ∗
(1.85) (4.72)
N 4,200 16,035 16,347
Adj. R2 0.150 0.124 0.108

This action can reduce taxes paid in those other periods,
resulting in a temporal shift in tax rates that could influ-
ence trend in effective tax rates that we show.
To examine the effect of accounting and tax losses on
our results, we conduct several tests (see Table 9). First,
we estimate Eq. (3) on a constant subsample of firms
that have positive pretax income in each of the 25 years
we study. The resulting sample is 42 hundred observa-
tions corresponding to 168 firms. Results in Model 1 of
Table 9 show that even for consistently profitable firms,
CASH ETR exhibits a reliable downward trend. Second, we
loosen the sample restriction so that firms are required
460 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
Fig. 9. Mean annual CASH ETR over the sample period for the 25 largest multinational and 25 largest domestic firms in the sample. This figure plots the
mean annual CASH ETR over the sample period, 1988–2012, for two subsets of firms: the largest 25 multinational firms and the 25 largest domestic firms
measured as of 2012. CASH ETR is the ratio of cash taxes paid to pretax income. All observations are subject to the criteria described in Table 1.
to always report a pretax profit but are not required to
exist in the database for the entire 25 years of the sample
period. The resulting sample is 16,035 observations corre-
sponding to 1,281 unique firms that never report a loss.
The results in Model 2 of Table 9 once again demonstrate
a reliable negative trend in CASH ETR. Third, we drop all
firms that ever report a net operating loss carryforward in
the database. The resulting sample is 16,347 observations
corresponding to 1,498 unique firms. Once again, the
results again show a negative trend in CASH ETR. Overall,
the results in Table 9 suggest that firms with losses in
some years that offset tax payments in other years are
unlikely to drive the downward time trend in CASH ETR.36
6.3. Examination of the largest multinational and domestic
firms
Certain firms, including GE, Google, and Apple, are often
mentioned in the popular press for their tax strategies.37
36
In addition, we run a separate test in which we retain all obser-
vations, including firms with pretax losses. Because traditional ratio-
based effective tax rate measures are difficult to interpret when the
denominator is negative, we instead estimate regressions of the form:

T XP Dit = α0 + α1 P Iit + α2 P Iit ∗ T IMEt + αkCONT ROLkit + it , where TXPD
k
is cash taxes paid and PI is pretax income. We find that the coefficient
α 2 , which captures the time trend on the estimated effective tax rate,
remains negative and significant. We also estimate the model allowing
the coefficients to vary for observations with a pretax loss. In those ob-
servations, we find that the effective tax rate is declining over time for
profitable observations, and that the effective tax rate is near zero and
unchanging over time for loss observations.
37
See, for example New York Times (2011, 2012) and Bloomberg (2010).
In this subsection, we examine whether these highly visi-
ble firms have a different over-time pattern in effective tax
rates. Specifically, we track the effective tax rates for the 25
largest multinationals and 25 largest domestic firms over
our sample period. The 25 largest firms are measured by
total assets as of the end of 2012. For multinationals, the
25 largest firms include such firms such as Apple, Berk-
shire Hathaway, GE, Google, and Microsoft; for domestic-
only firms, Kroger, CVS, Macy’s, Union Pacific, and Lowe’s.
Fig. 9 shows the trend in effective tax rates for these
two groups of firms. The figure shows that the trend in
effective tax rates is declining over time for these firms.
Moreover, the trend is similar for the multinational group
as it is for the domestic group. From this analysis, we con-
clude that the most visible, newsworthy firms appear to
have experienced a similar declining trend in effective tax
rates as the average firm in the sample.38
6.4. Comparison of US multinational firms with foreign firms
The primary focus of this study is on the effective tax
rates of US multinational firms compared with those of
purely domestic US firms. However, politicians, business
executives, and the business press often assert that US cor-
porations pay more tax than competing firms domiciled
outside the US.39 In this subsection, we briefly compare
38
In the Online Appendix, we also examine the trend in effective tax
rates by industry, as well as the trend in value-weighted effective tax
rates.
39
For example, PricewaterhouseCoopers (2011) conducted some re-
search on behalf of the Business Roundtable and reports that, based on a
sample of the largest two thousand companies in the world, those com-
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463 461

Fig. 10. Comparison of mean Generally Accepted Accounting Principles ETRs of US multinational firms with foreign parented firms. This figure plots the
median annual GAAP ETR over the sample period, 1988–2012 for firms from the Brazil, France, Germany, the UK, and the US. The data are from Compustat
Global. Observations with non-positive pretax income are deleted. Each firm is required to exist in the sample for five years. ETR is computed as total tax
expense (TXT) divided by pretax income (PI), as recorded by Compustat Global. ETR values greater than one (less than zero) are set to one (zero).

the time trend in GAAP effective tax rates of US multi-
national firms to that for firms from five large foreign
economies: Brazil, France, Germany, Japan and the UK. We
use accounting effective tax rates because most foreign
firms do not disclose cash taxes paid, and they do not com-
monly disclose the current and deferred components of tax
expense separately.
We present results from the analysis in Fig. 10. The
analysis illustrates that the effective tax rates of US firms
have fallen over time but that the effective tax rates
of firms located in major foreign economies have also
dropped, sometimes more dramatically than those of US
firms. The most dramatic drop is observed in German
firms. The average German firm reported an accounting ef-
fective tax rate of over 50% in 1989 but less than 30% in
2012. Thus, consistent with Markle and Shackelford (2012),
the evidence suggests that currently the accounting effec-
tive tax rates of US multinationals are higher than those of
companies in many other countries. This evidence is also
consistent with the recent trend in which US multinational
firms undertake inversions so that the new parent entity
is in a foreign country, usually one with territorial taxation
and lower rates than the US.
We note several caveats to consider when interpreting
the results in Fig. 10. First, most foreign countries have
substantially reduced their statutory corporate tax rates
over the sample period, whereas the US statutory tax rate
has remained essentially constant. Second, the US has re-
panies located in the US had average GAAP effective tax rates that were
significantly higher than the accounting effective tax rates from compa-
nies located outside the US.
tained a worldwide tax system with deferral, along with
an accounting exception that allows firms to not record
deferred tax expense on the residual tax owed to the US
on foreign earnings of subsidiaries if the earnings are in-
definitely reinvested. Many other countries have moved to
a territorial taxation regime, which makes the rates com-
puted less comparable. For example, Germany, Japan, and
the UK have all moved towards territorial taxation. One
way to view the decline in effective tax rates for US firms
is as a necessary adaptation in the face of the declines in
effective tax rates in the rest of the world, coupled with
the adoption of more favorable territorial taxation by many
of the largest economies.
7. Conclusion
We examine systematic changes in corporate effective
tax rates during the past 25 years. We test widespread be-
liefs that firms, particularly large multinational firms, are
increasingly able to reduce their effective tax rates. The
data confirm one important element of the conventional
wisdom but contradict many of the other commonly held
beliefs. We find a clear decrease in effective tax rates over
time across the broad sample of US firms. On average, cash
effective tax rates of US corporations have decreased by
about 0.4% per year over the past 25 years, representing a
cumulative decline of approximately 10 percentage points.
However, the data do not support the belief that the
decrease in effective tax rates is concentrated in multi-
nationals. We find essentially the same decrease in effec-
tive tax rates over time among purely domestic firms as
among multinationals. Moreover, during most years purely
462 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
domestic firms avoid taxes at a rate equal to or greater
than multinationals. The implications of these findings are
of broad importance. While multinational firms could have
access to tax avoidance opportunities that purely domes-
tic firms do not, the end result is that the trend in their
effective tax rates is quite similar. The results suggest that
purely domestic firms do not appear to be disadvantaged
relative to multinational firms in terms of tax avoidance
and that both types of firms are benefiting from decreased
effective tax rates over time.
In addition to furthering the literature on corporate tax-
ation, the findings have important policy implications. For
one, they suggest that a singular focus on tax avoidance by
multinational firms is perhaps unwarranted. While multi-
national tax avoidance certainly does deserve attention, the
data show that it is only part of the story. Perhaps more
important, however, the results point to an increasing dis-
connect between statutory tax rates and effective tax rates
over time among US firms. While the US statutory tax
rate remained remarkably stable over the past 25 years, US
firms’ effective tax rates have continued to decline. This
finding suggests that an overhaul of the US corporate tax
system that couples base broadening with lower tax rates
could be possible with little or no tax revenue loss, al-
though there would be winners and losers and how the
base would be broadened is not clear. Periodic calls have
been made for such reform, particularly in the face of sus-
tained cuts in statutory tax rates in other countries, which
has left the US with the highest corporate statutory rate
among the largest economies of the developed world.
References
Armstrong, C., Blouin, J., Larcker, D., 2012. The incentives for tax planning.
Journal of Accounting and Economics 53, 391–411.
Auerbach, A., 2007. Why have corporate tax revenues declined? Another
look. CESifo Economic Studies 53, 153–171.
Auerbach, A., Poterba, J., 1987. Why have corporate tax revenues de-
clined?.. In: Summers, L. (Ed.), Tax Policy and the Economy, volume
1. MIT Press, Cambridge, MA, pp. 1–28.
Badertscher, B., Katz, S., Rego, S., 2013. The separation of ownership and
control and corporate tax avoidance. Journal of Accounting and Eco-
nomics 56, 228–250.
Barth, M., Beaver, W., Hand, J., Landsman, W., 2005. Accruals, account-
ing-based valuation models, and the prediction of equity values. Jour-
nal of Accounting, Auditing, and Finance 20, 311–345.
Bloomberg, 2010. Google 2.4% rate shows how $60 billion lost to tax loop-
holes. J. Drucker, October 21.
Blouin, J., Krull, L., 2009. Bringing it home: a study of the incentives sur-
rounding the repatriation of foreign earnings under the American Jobs
Creation Act of 2004. Journal of Accounting Research 47, 1027–1059.
Brown, J., 2011. The spread of aggressive corporate tax reporting: a de-
tailed examination of the corporate-owned life insurance shelter. The
Accounting Review 86, 23–57.
Chen, S., Chen, X., Cheng, Q., Shevlin, T., 2010. Are family firms more tax
aggressive than non-family firms? Journal of Financial Economics 95,
41–61.
Chyz, J., 2013. Personally tax aggressive executives and corporate tax shel-
tering. Journal of Accounting and Economics 56, 311–328.
Chyz, J., Leung, W., Li, O., Rui, O., 2013. Labor unions and tax aggressive-
ness. Journal of Financial Economics 108, 675–698.
Citizens for Tax Justice, 2015. Memo to senate permanent subcommittee
on investigations: US corporations already pay a low tax rate. July 30.
Citizens for Tax Justice. Washington, DC.
Collins, D., Li, O., Xie, H., 2009. What drives the increased informativeness
of earnings announcements over time? Review of Accounting Studies
14, 1–30.
Collins, D., Maydew, E., Weiss, I., 1997. Changes in the value-relevance of
earnings and book values over the past 40 years. Journal of Account-
ing and Economics 24, 39–67.
DeAngelo, H., Roll, R., 2015. How stable are corporate capital structures?
Journal of Finance 70, 373–418.
Desai, M.A., 2003. The divergence between book income and tax income.
In: Poterba, J.M. (Ed.), Tax Policy and the Economy, Volume 17. MIT
Press, Cambridge, MA, pp. 169–208.
Desai, M.A., Dharmapala, D., 2006. Corporate tax avoidance and high-
-powered incentives. Journal of Financial Economics 79, 145–179.
Desai, M.A., Dharmapala, D., 2009. Earnings management, corporate tax
shelters, and book-tax alignment. National Tax Journal 62, 169–186.
Desai, M.A., Dyck, A., Zingales, L., 2007. Theft and taxes. Journal of Finan-
cial Economics 84, 591–623.
Desai, M.A., Hines, J.R., 2002. Expectations and expatriations: tracing the
causes and consequences of corporate inversions. National Tax Journal
55, 409–440.
De Simone, L., Klassen, K., Seidman, J.K., 2016. Unprofitable affiliates and
income shifting behavior. The Accounting Review forthcoming.
De Simone, L., Mills, L., Stomberg, B., 2014. Measuring Income Mobility.
Stanford University, University of Texas, and University of Georgia,
Stanford, CA, Austin, TX, and Athens, GA Unpublished working paper.
Dharmapala, D., 2014. What do we know about base erosion and profit
shifting? A review of the empirical literature. Fiscal Studies 35,
421–448.
Dharmapala, D., Riedel, N., 2013. Earnings shocks and tax-motivated in-
come shifting: evidence from European multinationals. Journal of
Public Economics 97, 95–107.
Donohoe, M., McGill, G., Outslay, E., 2012. Through a glass darkly: what
can we learn about a US multinational corporation’s international op-
erations from its financial statement disclosures? National Tax Journal
65, 961–984.
Dyreng, S.D., Hanlon, M., Maydew, E.L., 2008. Long-run corporate tax
avoidance. The Accounting Review 83, 61–82.
Dyreng, S., Lindsey, B., 2009. Using financial accounting data to examine
the effect of foreign operations located in tax havens and other coun-
tries on US multinational firms’ tax rates. Journal of Accounting Re-
search 47, 1283–1316.
Dyreng, S., Lindsey, B., Thornock, J., 2013. Exploring the role Delaware
plays as a domestic tax haven. Journal of Financial Economics 108,
751–772.
Dyreng, S., Markle, K., 2016. The effect of financial constraints on income
shifting by US multinationals. The Accounting Review 91, 1601–1627.
Edwards, A., Schwab, C., Shevlin, T., 2016. Financial constraints and cash
tax savings. The Accounting Review 91, 859–881.
Erickson, M., Heitzman, S., 2010. NOL poison pills: Selectra v. Versata et
al. Tax Notes, June 23.
Erickson, M., Heitzman, S., Zhang, X., 2013. Tax-motivated loss shifting.
The Accounting Review 88, 1657–1682.
Fama, E., French, K., 2001. Disappearing dividends: changing firm char-
acteristics or lower propensity to pay? Journal of Applied Corporate
Finance 14, 67–79.
Foley, C., Hartzell, J., Titman, S., Twite, G., 2007. Why do firms hold so
much cash? A tax-based explanation. Journal of Financial Economics
86, 579–607.
Francis, J., Schipper, K., 1999. Have financial statements lost their rele-
vance? Journal of Accounting Research 37, 319–352.
Graham, J., Hanlon, M., Shevlin, T., 2011. Real effects of accounting rules:
evidence from multinational firms’ investment location and repatria-
tion decisions. Journal of Accounting Research 49, 137–188.
Graham, J., Lang, M., Shackelford, D., 2004. Employee stock options, cor-
porate taxes, and debt policy. The Journal of Finance 59, 1585–1618.
Graham, J., Leary, M., Roberts, M., 2015. A century of capital structure:
the leveraging of corporate America. Journal of Financial Economics
118, 658–683.
Graham, J., Raedy, J., Shackelford, D., 2012. Research in accounting for in-
come taxes. Journal of Accounting and Economics 53, 412–434.
Graham, J., Tucker, A., 2006. Tax shelters and corporate debt policy. Jour-
nal of Financial Economics 81, 563–594.
Gravelle, J., 2013. Tax havens: international tax avoidance and eva-
sion. Congressional Research Service. US Government Printing Office,
Washington, DC January 23.
Grubert, H., Slemrod, J., 1998. The effect of taxes on investment and in-
come shifting to Puerto Rico. Review of Economics and Statistics 80,
365–373.
Hanlon, M., Heitzman, S., 2010. A review of tax research. Journal of Ac-
counting and Economics 50, 127–178.
Hanlon, M., Maydew, E., Saavedra, D., 2016. The taxman cometh: does
tax uncertainty affect corporate cash holdings?. Massachusetts Insti-
tute of Technology, University of North Carolina, University of Califor-
nia, Cambridge, MA, Chapel Hill, NC and Los Angeles, CA Unpublished
working paper.
 S.D. Dyreng et al. / Journal of Financial Economics 124 (2017) 441–463
Hasan, I., Hoi, C.K.S., 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, 109–130.
Higgins, D., Omer, T., Phillips, J., 2015. The influence of a firm’s business
strategy on its tax aggressiveness. Contemporary Accounting Research
32, 674–702.
Hines, J., Rice, E., 1994. Fiscal paradise: foreign tax havens and american
business. The Quarterly Journal of Economics 109, 149–182.
Huizinga, H., Laeven, L., 2008. International profit shifting within multina-
tionals: a multi-country perspective. Journal of Public Economics 92,
1164–1182.
Kim, J.B., Li, Y., Zhang, L., 2011. Corporate tax avoidance and stock price
crash risk: firm-level analysis. Journal of Financial Economics 100,
639–662.
Klassen, K., Laplante, S., 2012. Are US multinational corporations becom-
ing more aggressive income shifters? Journal of Accounting Research
50, 1245–1285.
Kleinbard, E., 2011. Stateless income. Florida Law Review 11, 699–774.
Lisowsky, P., 2010. Seeking shelter: empirically modeling tax shelters
using financial statement information. The Accounting Review 85,
1693–1720.
Markle, K.S., Shackelford, D.A., 2012. Cross-country comparisons of corpo-
rate income taxes. National Tax Journal 65, 493–528.
Maydew, E., Shackelford, D., 2007. The changing role of auditors in cor-
porate tax planning. In: Auerbach, A., Hines, J., Slemrod, J. (Eds.), Tax-
ing Corporate Income in the 21st Century. Cambridge University Press,
New York, NY, pp. 307–337.
Mills, L., Newberry, K., Novack, G., 2003. How well do Compustat NOL
data identify firms with US tax return loss carryovers? The Journal of
the American Taxation Association 25, 1–17.
New York Times, 2011. GE’s strategies let it avoid taxes altogether. David
Kocieniewski March 24.
New York Times, 2012. How Apple sidesteps billions in taxes. Charles
Duhigg, David Kocieniewski April 28.
463
OECD, 2013. Action plan on base erosion and profit shifting. OECD Pub-
lishing https//dx.doi.org/10.1789/9789264202719-en.
PricewaterhouseCoopers, 2011. Report prepared for Business Roundtable.
PricewaterhouseCooopers, London, UK April 14.
Rajgopal, S., Venkatachalam, M., 2011. Financial reporting quality and id-
iosyncratic return volatility. Journal of Accounting and Economics 51,
1–20.
Rego, S., 2003. Tax avoidance activities of US multinational corporations.
Contemporary Accounting Research 20, 805–833.
Rego, S., Wilson, R., 2012. Equity risk incentives and corporate tax aggres-
siveness. Journal of Accounting Research 50 (3), 775–810.
Seida, J.A., Wempe, W.F., 2004. Effective tax rate changes and earn-
ings stripping following corporate inversion. National Tax Journal 57,
805–828.
Slemrod, J., 2004. The economics of corporate tax selfishness. National
Bureau of Economic Research, Cambridge, MA Unpublished working
paper no. w10858.
Sullivan, M., 2010. Oil drillers gain billions from immoral tax break. Tax
Notes is a weekly periodical. Tax Analysts, pp. 1183–1187 (June 14).
US Senate, 2014a. Caterpillar’s offshore tax strategy. Majority staff report
of the committee on Homeland Security and Governmental Affairs
Permanent Subcommittee on Investigations. US Government Printing
Office, Washington, DC April 1.
US Senate, 2014b. Comprehensive tax reform for 2015 and beyond. Report
prepared by Republican staff, Committee on Finance. US Government
Printing Office, Washington, DC December.
US Senate, 2015. Impact of US tax code on the market for corporate con-
trol and jobs. Majority report of the committee on Homeland Security
and Governmental Affairs Permanent Subcommittee on Investigations.
US Government Printing Office, Washington, DC July 30.
Williams, B., 2015. Income-shifting incentives and offshored US jobs. Uni-
versity of Texas, Austin, TX Unpublished working paper.
Wilson, R.J., 2009. An examination of corporate tax shelter participants.
The Accounting Review 84, 969–999.