Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

Contents lists available at ScienceDirect

Journal of International Accounting,

Auditing and Taxation

Tax risk management and the multinational enterprise

Haroldene F. Wunder
College of Business Administration, California State University, 6000 J Street, Sacramento, CA 95819-6088, United States

a r t i c l e

i n f o

Keywords:
Risk management
Tax risk management
Multinational enterprise

a b s t r a c t
The nancial scandals in the United States and other countries ushered in nancial reporting
and corporate governance reforms that extend beyond the U.S. Sarbanes-Oxley Act of 2002
(SOX). These initiatives have increased the international nancial communitys awareness
of the importance of risk management and internal controls. Tax risk management and
related internal controls have been accorded less focus than risk management generally.
The purpose of this research is to describe the current state of tax risk management of
multinational enterprises (MNEs) by reporting survey responses from chief nancial ofcers
(CFOs) of U.S. and non-U.S. MNEs. The research shows that signicant progress has been
made by large MNEs in developing and implementing both general and tax risk management
policies. The results provide guidance in identifying the loci and impact of organizational
tax risk and indicate that respondents do not perceive alarming degrees of tax risk in their
organizations. The study reveals a remarkable degree of similarity in U.S. and foreign rm
responses and demonstrates, unexpectedly, that existing reporting structures enable CFOs
to shift a signicant degree of tax risk management to heads of tax.
2009 Elsevier Inc. All rights reserved.

1. Introduction
As a result of nancial scandals such as the Enron and Worldcom scandals in the United States, the Parmalat implosion in
Italy, the Maxwell pension debacle in the United Kingdom, and the HIH Insurance failure in Australia, a new era of nancial
reporting and corporate governance reforms was ushered in that extend beyond the U.S. Sarbanes-Oxley Act of 2002 (SOX).
For example, in early 2006, the European Union (EU) approved expansion of its 8th Council Directive on Company Law, which
pertains to the approval of auditors in EU member states. Member states have 2 years from its formal adoption (summer
2006) to incorporate its requirements into their national laws (Wyman, 2006).
The most signicant result of these laws is the increased awareness by the international nancial community of the
importance of risk management and internal controls. Risk management and other principles of corporate governance
are being addressed around the world. According to the European Corporate Governance Institute (ECGI), a total of 66
countries and geographical areas, including the United States, have published new or updated existing codes of corporate
governance to include and adequately address risk management principles. Additionally, in July 2005, the International
Corporate Governance Network (ICGN) updated its 1999 ICGN Statement on Global Corporate Governance Principles, to charge
corporate boards with ensuring that appropriate systems of control are in place, including systems for risk management.
Tax risk management and related internal controls have been accorded less focus than risk management generally. It
has been common practice to subsume tax risk management under the general subject of risk management rather than
addressing tax risk management as a discrete aspect of risk management. Tax risk management and related internal controls

Tel.: +1 916 278 7134.

E-mail address: wunderh@csus.edu.
1061-9518/$ see front matter 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.intaccaudtax.2008.12.003

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

15

have been described historically as a bit of black art, not necessarily understood even by those in the tax function. . ..
(PricewaterhouseCoopers, 2004, p. 2).
Furthermore, material weakness disclosures by management frequently identify internal control problems in complex
accounts, such as income tax expense (Ge & McVay, 2005). Quimby and Pearce (2006, p. 26) conclude that [a]rguably, no
one has felt the sting of SOX 404 more acutely than the tax departments. Elgood (2006) reports that more than 20% of
material weaknesses reported under section 404 of SOX pertained to how tax is controlled. Quimby and Pearce (2006, p. 26)
concur and conrm that for the second year in a row, tax departments held the dubious distinction of being one of the most
common areas of internal control failure (material weakness) cited in the adverse opinions led in 2005 (Year 2).
The impetus for the current research is the recent recognition by commercial organizations and revenue authorities that
the tax function has its own unique prole, which necessitates a separate inquiry into tax risk management. Since formally
classifying tax risk management as a separate element of corporate governance is a recent phenomenon, a dearth of empirical
research exists as to how individual rms rate various types of tax risk and have incorporated tax risk management into their
governance policies and procedures. The contribution of this research is that it lls the information and decision making
guidance gap by describing the state of tax risk management of some of the worlds largest multinational enterprises
(MNEs). The research is relevant to accounting, auditing, and taxation academicians and practitioners because tax risk
management is at the conuence of all three disciplines.
Specically, this research elicits information about and compares U.S. and non-U.S. MNEs to determine the current state
of tax risk management and if a global view of, and strategies pertaining to, tax risk management exists among MNEs. The
inquiry addresses the following aspects of tax risk management for U.S. and non-U.S. rms: (i) the structure of and the chain
of reporting pertaining to the tax function, (ii) the existence and nature of a documented general and tax risk management
policies, and (iii) rms ratings of seven types of tax risk and six specic tax issues. For example, the research describes the
signicance of transfer pricing rules and their disproportionate effect on the tax risk experienced by non-U.S. multinationals
rms. The study also provides demographic and organizational information as a perspective from which to draw conclusions
about the aspects of tax risk management identied above.
The study provides insights for identifying the loci and signicance of organizational tax risk and developing and
implementing tax risk management policies to adequately control risk. It is also relevant to corporate boards and senior
management. Neubig and Sangha (2004) conclude that boards and senior managements lack of understanding of the complicated and technical nature of tax may expose companies to unexpected outcomes (i.e., risk). Finally, the current research
updates the results of a 2005 KPMG study that looked at enterprises adoption of formal tax risk management policies and
demonstrates that signicant progress has been made since 2005.
In effect, the study constitutes a benchmark against which decision makers can assess organizational tax risk efforts.
The information about U.S. and foreign MNEs development and implementation of documented risk management
policies constitutes a set of best practices. The signicance of these risk management best practices is conrmed by
Ernst & Youngs (2007) series of 2006 risk management surveys, which demonstrate that 66% of surveyed companies
plan to increase their investment in risk management in order to manage risk in a more coordinated and integrated
manner.
2. Background
The current relevance and importance of tax risk management is demonstrated by the number of current initiatives
occurring worldwide. Revenue authorities and commercial organizations around the world have begun to address tax risk as
a discrete element of risk management. For example, in June 2003, the Australian Taxation Ofce (ATO) issued Large Business
and Tax Compliance to launch a campaign to elevate tax governance to the boardroom. The ATO has since augmented its
original initiative by examining companies tax risk management procedures as part of its audit and risk review process
and has published the 2006 version of Large Business and Tax Compliance, which systematically addresses tax risk concepts
(Australian Taxation Ofce, 2006).
In 2005, Ireland Revenue Authority issued The Cooperative Approach to Tax Compliance, requiring that businesses prepare
and implement annual tax risk management plans, focusing on identied risk areas. Revenues risk review will determine
the precise combination of self-audits and Revenue audits (Fennell, 2005).
From a commercial perspective, two 2005 Big-Four rm surveys produced similar results. A Deloitte survey of tax directors
of 350 U.K. companies revealed that corporate attitudes toward tax risk in the U.K. have changed signicantly and that the
proportion of tax directors that describe their approach to tax as conservative has increased from 17% to 35% from 2003 to
2005 (Crest, 2005b). Similarly, 44% of respondents to an Ernst & Young survey of 354 tax directors showed that companies
have become more risk averse in the 2 preceding years, with 75% of respondents also stating that risk management is a
criterion upon which their performance is measured (OSullivan, 2005). More recently, in connection with its 2006 global
tax risk survey, Ernst & Young (2006, p. 4) concludes that establishing a written, comprehensive and sustainable tax risk
management framework is a best practice.
In a similar vein, International Tax Review received 162 responses from tax executives in connection with its 2005 survey
of international tax services in Asia. An issue that rst surfaced in that survey is tax executives desire for their tax advisers
to understand better their appetite for tax risk. The responding tax executives want their tax advisers to understand and
appreciate the effect of tax risk on how clients structure operations and the extent to which they will expose themselves to

16

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

tax risk (Crest, 2005a). Tax executives rated understanding clients tax risk appetite rst among the ten top ways that tax
advice could be improved (Crest, 2005a, p. 16).
Shareholders and their agents are also increasingly concerned with how corporations manage tax and the extent to
which management exposes them to greater risks. In 2004, Henderson Global Investors Ltd. contacted the chairmen of the
350 largest companies by capitalization that have their primary listing on the London Stock Exchange (i.e., the FTSE 350),
inquiring about how their respective companies manage tax since the effect of tax management is an important aspect of its
overall strategy to deliver long-term shareholder value (KPMG, 2004). Their results indicate that companies have adopted a
conservative approach to tax and tax risk and deliberately avoid aggressive tax planning that entails excessive risk.
Campbell, Durst, and Miller (2002) describe the need for the following practices in the post-SOX environment: regular
review of document retention policies, formal guidelines about who should draft and review tax-sensitive documents, and
written protocols for communications between tax departments and auditors.
The considerations described above differ signicantly from the results reported by Arlinghaus (1998), whose survey
asked respondents to identify the principal goals of a corporations tax department. Tax risk and tax risk management were
not among those goals, which included such objectives as managing worldwide effective tax rates.
Despite the realization that tax risk management represents an increasingly important aspect of rms risk management
initiatives, a recent KPMG survey of senior tax executives reveals that only 24% of those surveyed have a formal tax risk
management policy. The survey also reveals, however, that 60% of the tax departments are increasing training programs for
their staffs (PR Newswire, 2005, p. 1).
At present, Elgood, Paroissien, and Quimby (2005) is essentially the only study to address tax risk management as it
pertains to multinational entities. They conclude that the three most important types of local tax risk are operational risk,
compliance risk, and nancial accounting risk and identify communication as the most signicant practical issue in managing
overseas tax risk (2005).
3. Research design
3.1. Denition of tax risk
Arlinghaus (1998) acknowledges that no universal denition of the term tax risk exists but offers the following denition
of the term: the likelihood that tax outcome differs from what is expected, due to a variety of reasons, for example, the
judicial process, changes in the law, changes in business assumptions, an increased intensity of audits, and uncertainty in the
interpretation of the law; and any action emanating from the tax function that subjects the company to adverse publicity.
Goodman (2004) addresses tax risk in terms of the three internal control objectives specied in the Committee of
Sponsoring Organizations (COSO) Enterprise Risk Management Integrated Framework (COSO Framework): effectiveness and
efciency of operations, reliability of nancial reporting, and compliance with laws and regulations. He identies tax-related
risks associated with each of the three COSO objectives.
The current research addresses tax risk from the perspective of Tax Risk Management (TRM), PricewaterhouseCoopers
(2004) treatise on the concept of tax risk and the elements of managing such risks, the foundation of which is the COSO
Framework. TRM identies the following seven areas of tax risk, adopting the common practice of dening business risk in
terms of uncertainties accompanying decisions, activities, and operations undertaken by an enterprise:
Specic risk areas:
Transactional risk (e.g., acquisitions, mergers).
Operational risk (e.g., new business ventures, new operating models, new operating structure).
Compliance risk (e.g., weak records and controls, data integrity issues, legislative changes).
Financial accounting risk (e.g., changes in systems and policies).
Generic risk areas:
Portfolio risk (e.g., combination of any of the risks).
Management risk (e.g., changes in personnel, new/inexperienced resources).
Reputational risk (e.g., revenue authority investigation, press comment, legal actions).
The current research addresses tax risk from the perspective of those seven risk areas.
The TRM structure was selected because all rms listed on a U.S. stock exchange must include in their annual nancial
reports an assessment of the effectiveness of the rms internal controls; the COSO framework constitutes the gold standard
for assessing internal controls (to manage risks). The essentially universal adoption of the COSO framework ensures that the
respondents have a working knowledge of the types of risk addressed in the current research and a common baseline from
which to respond.
3.2. Methodology
Information about the state of tax risk management among MNEs was obtained from a questionnaire sent to chief nancial
ofcers (CFOs) of not only U.S. Fortune 500 companies, but also non-U.S. rms that list on the New York Stock Exchange (NYSE

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

17

Euronext). The questionnaire asked for demographic and organizational information as well as specic issues concerning
the TRM seven areas of tax risk and six specic tax issues. (Copies of the questionnaire are available from the author upon
request.) Since risk management and internal control concerns are not uniquely U.S. phenomena, it was essential to obtain
data from not only domestic, but also foreign, MNEs. The questionnaires were written in English.
CFOs were selected because of the CEO and CFO duty (section 404 of SOX) to assess the effectiveness of the rms
internal control system, and the fact that heads of tax have historically reported to the CFO. CFOs represent the highest level
individual with the greatest knowledge of the rms tax function and the greatest responsibility for effective management
of tax risk. CFOs were given the option of delegating completion of the survey to encourage response by a knowledgeable
rm representative.
The research summarizes and describes responses to determine the extent to which the two MNE populations are similar/dissimilar. The value of obtaining demographic and organizational data is two-fold. First, the data are informative in and
of themselves; for example, does a global tax prole exist, or are U.S. and non-U.S. rms unique with regard to their approach
to tax matters? Second, the demographic and organizational responses provide a context for explaining and interpreting the
two MNE populations responses to inquiries about the TRM seven tax risk elements and the six specic tax issues.
Responses to inquiries about the risk appetite of rms heads of tax and the risk associated with the TRM tax risk elements
and the six specic tax issues take the form of Likert-scale responses. The responses of the two MNE populations are compared
using both a nonparametric and a parametric test. The nonparametric Wilcoxon rank-sum test is applied because of the ordinal
nature of Likert-scale responses. However, the number of total responses is more than sufcient to justify the parametric
t-test (discussed below), which increases the power of the tests. Results of both tests are presented.
4. Results and analysis
4.1. Respondents
Surveys were sent to the Fortune 500 companies and to 453 non-U.S. NYSE-listed rms. Usable responses were received
from 112 rms: 48 U.S. rms, representing 20 different industries, and 64 non-U.S. NYSE listed rms, representing 26 different
countries and 24 different industries. The 22% response rate is based on the number of surveys deemed delivered; that is, it
did not take into account surveys returned unopened because of changes in company mailing address and/or personnel, a
common phenomenon for the non-U.S. rms. The number of responses is more than sufcient for the Wilcoxon rank-sum
test and the t-test.
The number of responses to various survey items varies, as not all responding companies provided a response to all
survey inquiries. For example, a Greek company provided narrative information about its risk management policy but did
not respond to other survey inquiries. All respondents requested anonymity; specic rm responses identify the company
by reference to its home country.
4.2. Responses to demographic and organizational inquiries
The demographic and organizational information was obtained to avoid gathering information about rms perceptions
of tax risk in isolation and without context, and identify any particular characteristics of the U.S. and the non-U.S. MNEs
that would account for the nature of rms responses about tax risk. In other words, the demographics and organizational
information provide a general context from which to explain and interpret rms responses about tax risk.
4.2.1. Locations of foreign subsidiaries and alliances
The two MNE populations are similar in terms of where they conduct business. The U.S. and non-U.S. rms identify 61 and
77 countries and/or areas, respectively, as locations for foreign subsidiaries and foreign alliances. Fifty of those countries/areas
are common locations for both the U.S. and the non-U.S. rms.
4.2.2. Country-specic tax-risk proles
Firms are acknowledged to associate particular tax risk characteristics with individual countries in which they do business.
For example, in his March 23, 2006, presentation at the Lexis-Nexis Tolley Conference: Tax Risk Management, Paul Morton,
Head of Group Tax for Reed Elsevier Group plc, described his personal view about the tax characteristics of the following
countries:

U.S. Technical, adversarial.

Netherlands Consensual.
U.K. Pragmatic.
Russia Confrontational.
France Legalistic.
Japan Patriotic.

18

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

Table 1
Countries with conicting tax-risk proles.
Country

Population(s) rating as least tax risk

Population(s) rating as most tax risk

Australia
Brazil
Canada
Cayman Islands
Chile
China
Ireland
Korea
Mexico
Netherlands
Spain
U.K.
U.S.

Some non-U.S. rms

Some non-U.S. rms
Some U.S. rms
Both populations
Some U.S. rms
Some non-U.S. rms
Both populations
Some U.S. rms
Some non-U.S. rms
Both populations
Some U.S. rms
Both populations
Some non-U.S. rms

Some non-U.S. rms

Both populations
Some U.S. rms
Some non-U.S. rms
Some non-U.S. rms
Some non-U.S. rms
Some non-U.S. rms
Some non-U.S. rms
Some U.S. rms
Both populations
Some non-U.S. rms
Both populations
Both populations

To determine if country-specic tax-risk proles exist, the current research asked U.S. and non-U.S. rms to identify
the country associated with the least and the greatest degree of tax risk. Although a U.K. mining rm stated that tax risks
may occur in all countries and that no single country is inherently the greatest or least risky, other respondents identied
the following ten countries as characterized by the least tax risk (with the number of countries indicating least risky in
parentheses): Ireland (5), Bermuda (4), Germany (4), Cayman Islands (3), Luxembourg (3), Netherlands (3), Singapore (3),
Switzerland (3), U.K. (3) and France (2). The following six countries were identied as associated with the greatest degree of
tax risk: U.S. (11), China (7), India (6), U.K. (6), Brazil (5), and Bolivia (2).
Table 1 identies the 13 countries that have the dubious distinction of being classied as both the least and the most risky
with regard to tax.
The Netherlands and the U.K. have the further dubious distinction of lacking a clear tax-risk prole since both countries
were rated as both least and most risky by rms in both MNE populations. No doubt the precise nature of the companies
operations in those countries account for their perceptions about the tax risk associated with each country.
Only seven countries appear to have an unambiguous tax-risk prole for the two populations. Bermuda, France, Luxembourg, Singapore, and Switzerland appear only in the least-tax-risk category, whereas, Bolivia and India appear only in the
most-tax-risk list. No systematic difference between the U.S. and foreign rms emerges with regard to their classication of
countries as most- or least-risky.
4.2.3. Organization of tax function
Firms were asked to describe their tax functions as centrally or decentrally organized to gauge the trend toward centralization that other research has previously discerned. Elgood et al. (2005, p. 24) describe a trend toward centralizing control
of global tax matters because, in their opinion, CFOs expect their heads of tax to be responsible for all taxes and all tax risk,
whatever tax that may be and wherever it arises. Lambert and Lucas (2006,p. 34) also nd a trend toward centralization
of the tax function, notwithstanding their assertion that historically many multinational groups have taken a decentralized
approach. . . delegating the management of tax compliance obligations to local in-country nance teams.
Table 2 summarizes the responses of rms when asked to classify the companys tax function as centralized or decentralized and demonstrates the dominance of centralized tax functions.
A clear image of the structure of the tax function of U.S. rms emerges: the tax function is almost universally centralized.
Only three U.S. rms report a purely decentralized tax function. Thirty-nine U.S. rms (81%) report an exclusively centralized
tax function, and six U.S. rms (13%) respond that the tax function is centralized for some local businesses and decentralized
for others.
In contrast, the organization of the tax function of non-U.S. rms is characterized by more variation, although 32 responding rms (52%) use a centralized tax function exclusively. Fourteen non-U.S. rms (23%) report using a centralized tax function
for some local businesses and a decentralized tax function for others. Some of the variation in non-U.S. rm responses is, of
course, explained by multi-country inuences.
Table 2
Structure of tax function.
Population

Structure of tax function

Centralized

Decentralized

Centralized for some local businesses, decentralized for others

U.S. rms
Non-U.S. rms

39
32

3
16

6
14

Total

71

19

20

Number of U.S. respondents = 48. Number of foreign respondents = 62.

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

19

Table 3
Title of head of tax.
Title

Number

U.S. rms
Vice President
Director
Chief Accounting Policy Ofcer
General Manager Tax
General Tax Counsel
General Tax Ofcer

37
6
1
1
1
1

Foreign rms
Vice President
Head of Tax or Head of Group Tax
Director
Tax Manager
CFO
Group General Manager
Corporate Ofcer
Corporate Controller
Executive Manager
Deputy Ofcer
Global Tax Director
Graduate (undened)

20
11
10
9
3
3
2
1
1
1
1
1

Number of U.S. respondents = 47. Number of foreign respondents = 63.

When the two populations of MNEs are combined, 71 rms (65%) use a centralized tax function, 19 rms (17%) use a
decentralized tax function, and the remaining 20 rms (18%) use a combination of centralized and decentralized tax structure.
Centralized tax functions dominate for the MNEs responding to the research questionnaire since 83% of responding MNEs
use a centralized tax function at least partially.
4.2.4. Head of tax prole and reporting
Of central signicance to this research is the rms tax reporting prole, which encompasses the characteristics of the
head of tax and the lines of reporting. The recent focus on tax risk management suggests that reporting structures may need
to be realigned to encompass not only the compliance aspect of taxation, but also the management of risk appetite associated
with tax matters.
Firms were asked to provide information about the following:

Title of global head of tax.

To whom does the head of tax report?
How risk-averse is the head of tax?
Does the head of tax sign off on totality of global organization tax position?

Table 3 shows that the most common designation for both U.S. and non-U.S. rms head of tax is vice president, including
qualiers such as assistant, executive, senior or taxation. Seventy-three of the 110 responding MNEs (66%) indicate that their
head of tax is at the level of vice president or director.
More variation exists for the foreign rms, no doubt due to some extent to country-specic organizational characteristics
and cultural inuences. In any case, for both populations of MNEs, the head of tax is a high-level executive, most often at the
vice-president level and in some cases at the director level, consistent with the spirit of effective governance owing from
SOX and global governance initiatives.
Researchers have begun to re-evaluate the most appropriate reporting structure for the head of tax, given the recent
shift in the focus and responsibilities of the tax function. Cummings (2005, p. 625) notes that as recently as the latter 1990s, the primary focus of the tax function was management of the corporations worldwide effective tax rate
(i.e., rms manage what they can measure). He concludes that with such a focus, it was most common for the head
of tax to report to the CFO, with a few tax ofcers reporting to the controller, the treasurer and the CEO. With the
increased focus on risk management, Cummings questions whether reporting to the CFO is still the best natural t for
the tax function. He offers, as viable alternatives, reporting to the corporate legal function, the internal auditor, or audit
committees.
The current research demonstrates that the vast majority of U.S. and foreign MNE heads of tax still report to the CFO,
which, notwithstanding Cummings (2005) inquiry, seems particularly appropriate since the CFO and the CEO must assess
the effectiveness of internal controls, including those pertaining to managing tax risk. Table 4 shows that 71% of U.S. heads
of tax and 66% of foreign heads of tax report to the companys CFO; importantly, none of the respondents indicated that a
change in reporting structure was imminent or even contemplated.

20

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

Table 4
To whom does the head of tax report?
Head of tax reports to

Number

U.S. rms
CFO
Vice President or Senior Vice
President, Finance
Treasurer
Executive Vice President & Controller
Chief Accounting Ofcer

34
8
4
1
1

Foreign rms
CFO
Finance Director/Manager
Controller
Vice President
Treasurer
CEOa
Head of Finance
Head of Legal & General Affairs and
Members of Executive Committee
Chief Accounting Ofcer
Head of Administration

41
6
4
3
3
1
1
1
1
1

Number of U.S. respondents = 48. Number of foreign respondents = 62.

a
The Chilean rms head of tax is the CFO, who reports to the CEO.

Although a Swiss pharmaceutical company indicates that its head of tax reports to the Head of Legal & General Affairs
and Members of the Executive Committee, the other three Swiss companies indicate that their heads of tax report to the
CFO. Perhaps the characteristics of the pharmaceutical industry (e.g., extensive legal issues such as patents and cross-border
distribution restrictions, as well as extensive research and development periods) explain its seemingly unique reporting
structure. The reporting structures of both populations of MNEs are consistent with the increased level of responsibility
accorded the CFO under SOX.
Firms were also asked to rate their heads of tax with regard to risk aversion, using a ve-point Likert scale, where one is
very risk averse and ve is risk taker. The mean and standard deviations of responses appear in Table 5.
Both populations heads of tax are described as having moderate risk appetites, with the U.S. rms being somewhat less
risk averse. The standard deviations reveal somewhat more variation in non-U.S. rm responses, which is not surprising
given that 26 countries are represented in the foreign population.
The degree of risk aversion reported is similar to the results of the Ernst & Young (2006) Global Tax Risk Survey, which
identies a growing trend toward risk aversion in the area of taxes. Fifty-four percent of the senior tax decision makers who
responded to the Ernst & Young survey indicate that they have become more risk averse during the last 2 years (Ernst & Young,
2006, p. 4). Ernst & Young attributes the heightened aversion to tax risk to legislative and regulatory developments, increased
enforcement activity, heightened press scrutiny, more involvement of stakeholders, greater focus on nancial statements
and internal controls, and increased transparency and disclosure (2006 Foreword). It also notes that audit committees are
requiring more information about tax matters, resulting in more time spent by corporate tax departments on tax nancial
reporting matters. Additionally, Ernst & Young concludes that for SEC registrant companies, the percentage of time allocated
to nancial reporting will probably increase because of the adoption of FASB Interpretation No. 48 Accounting for Uncertainty
in Income Taxes (2006).
Two statistical tests were applied to determine if the two populations responses about head-of-tax risk aversion are
statistically different. The research hypothesis is that since a total of 27 countries are represented in this research (26 foreign
countries and the United States), country-specic business and cultural inuences result in statistical differences between
the risk aversion of U.S. and foreign heads of tax. The null form of the hypothesis is:
H1.

U.S. and foreign heads of tax do not differ statistically in their degrees of risk aversion.

Table 6 reports the results of applying the nonparametric Wilcoxon rank-sum test and the parametric t-test to the
data to test the null hypothesis. Both tests are used because of the large number of ordinal responses. The nonparametric Wilcoxon rank-sum test is appropriately applied because it is designed to test whether two independent groups
Table 5
How risk averse is the head of tax? Means and standard deviations of responses.
Population

Mean

Standard deviation

U.S. rms
Foreign rms

2.75000
2.51695

0.689202
0.835454

1 = very risk averse and 5 = risk-taker.

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

21

Table 6
How risk averse is the head of tax? Results of Wilcoxon rank-sum test and t-test.
Panel Aa
Wilcoxon rank-sum test
Computed W-statistic (Wx )
-Value

2694.00000
0000.09695

Panel Bb
t-Test results
Computed t-statistic
-Value

0001.52900
0000.12940

a
b

The null hypothesis is rejected at the .10 signicance level.

The null hypothesis is not rejected; however, the -value of .1294 is near the boundary of the .10 signicance level.

Table 7
Does the head of tax sign off on totality of global organization tax position?
Population

Yes

No

Other response

U.S. rms
Foreign rms

47
46

1
14

0
Only to the extent reported in published nancials and if/as required by tax laws.

Total

93

15

Number of U.S. respondents = 48. Number of foreign respondents = 61.

have been drawn from the same population when responses are ordinal measures (e.g., the Likert-scale responses in this
research).
Since nonparametric tests lack the explanatory power of parametric tests, the t-test was also applied to the data. When
the size of either sample exceeds ten, it has been shown that the sampling distribution of Wx (the Wilcoxon statistic) rapidly
approaches that of the normal distribution (Siegel & Castellan, 1988). Forty-six U.S. and 59 foreign rms provided responses
to the risk aversion question, resulting in a total sample of 105. Therefore, parametric t-test is also appropriate. The two
tests are undertaken as complements. When the two tests yield the same result, the t-test adds power to the results of the
nonparametric Wilcoxon rank-sum test. When the tests yield different results, the specic nature of the results would need
to be addressed.
Both statistical tests conrm the similarity in the degree of risk aversion evinced by the U.S. and foreign MNE heads of
tax. The Wilcoxon test permits rejection of the null hypothesis of no difference at only the .10 signicance level. While the
t-test does not permit rejection of the null at a .10 or greater signicance level, the computed t-statistic is at the boundary of,
but not actually in, the rejection region of the .10 level ( = .1294). Viewed together, the results constitute a tepid rejection of
the null hypothesis.
The nal inquiry pertaining to the head of tax prole and reporting structure pertains to whether the rms head of tax
has the responsibility to sign off on the totality of the global organizations tax position. Table 7 shows that for both MNE
populations, it is the responsibility of the rms head of tax to sign off on the totality of the global organizations tax position.
Of the 109 rms that provided a response to the inquiry, 93 rms (85%) responded that the head of tax is responsible for
signing off on the totality of the global organization tax position without regard to, for example, the number of geographical
areas encompassed by the global organization. Forty-seven of the 48 responding U.S. rms responded in the afrmative,
although one of the 47 indicated that the head of taxs global sign-off responsibility is only to income taxes. Responses of
foreign rms are similar, with 75% of respondents indicating that their rms head of tax is ultimately responsible for the
global organizations tax position.
Since the vast majority of U.S. and foreign heads of tax report to their rms CFO (Table 4), the data show that CFOs are
able ultimately to shift a signicant amount of tax risk to heads of tax because of their responsibility with regard to the global
organizations tax position. That is certainly the case in the United States, as demonstrated by 47 of 48 responding U.S. rms.
In summary, the demographic and organizational responses reveal that U.S. and foreign MNEs are remarkably similar
along the following dimensions: (i) where they do business, (ii) organization of the tax function, (iii) head of tax proles
and reporting structure, and (iv) head of tax sign-off on the totality of the organizations global tax position. They differ
somewhat in the degree of risk aversion evinced by the head-of-tax, with foreign heads of tax characterized as slightly more
risk averse than their U.S. counterparts. The foregoing demographic and organizational information constitutes the context
for the following discussion of tax risk management procedures and risk assessment of U.S. and foreign MNEs.
5. Responses to tax risk management procedures and risk assessment
The survey questions for this category fall into two broad categories with two specic categories under each. Firms were
asked for information about documented risk management policies (general and tax) and assessments of tax risk (the seven
TRM tax risk elements and six specic tax issues).

22

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

Table 8
Documented risk management policy.
Population

Single policy for the global

organization

Multiple, country-specic
policies

No documented risk
management policy

Other response

U.S. rms
Foreign rms

27
37

3
12

17a
13c

1b
2d , e

Total

64

15

30

Number of U.S. respondents = 48. Number of foreign respondents = 64.

a
One U.S. rm indicates that it is currently documenting its risk management policy.
b
One U.S. rm states that specic policies exist for specic types of risk.
c
One Indian rm states that development of a single risk management policy is in process.
d
One Swiss rm indicates a single policy for the global organization, except where required locally.
e
One Swiss rm indicates a single policy for the global organization and multiple, country-specic policies.

5.1. Risk management policies

Ernst & Young (2007) concludes that developing an integrated and systematic approach to risk management is still an
important challenge facing management. The current research inquires about the existence of risk management policies to
ascertain the extent to which U.S. and non-U.S. MNEs approach risk systematically, rather than on an informal or ad hoc
basis. The questionnaire also elicits information about whether rms operate under a single global risk management policy
or country-specic policies.
The data in Table 8 demonstrate that signicant progress had been made toward systematizing risk management concerns.
Seventy-nine (71%) of the 112 responding rms have a documented general risk management policy either in place or in
process, 30 (63%) of 48 U.S. rms and 49 (77%) the 64 foreign rms. The majority of those policies (81%) are single policies that
apply to the global organization. The applicability of a single risk management policy to the entire organization constitutes
strong support for the notion of a global, non-country-specic view of managing risk.
The following responses are notable because of the specicity with which the rms address risk management. One U.S.
rm has specic policies for specic types of risk, and one Swiss rm operates under a single policy for the global organization
and multiple, country-specic policies. At a different point along the U.S. continuum, a respondent states that although it
lacks a documented risk management strategy, the corporation has a Code of Conduct or credo that applies to all aspects
of the business.
The responses pertaining to documented tax risk management policies (as shown in Table 9) are less impressive. They
are, however, more positive than the results of the 2005 KPMG survey, which reported that only 24% of senior tax executives
surveyed indicated that their rms have a formal tax risk management policy (PR Newswire, 2005).
Forty-three percent of respondents indicate that their rm has a documented tax risk management policy in place or in
process, 14 (29%) of the 48 U.S. rms and 34 (53%) of the 64 foreign rms. Seventy-eight percent of the documented policies
are single policies that are applicable to the global organization, notwithstanding the multitude of different taxation systems
that impinge on MNE operations.
One Canadian rms tax risk management policy is included in its documented global policy, a model consistent with
Ernst & Youngs view that [t]ax risk management is inseparable from overall business risk management (Ernst & Young,
2004, p. 20). Also notable is the response from a Swiss rm that has a single tax risk management policy for the global
organization and multiple, country-specic tax risk management policies, the same model that the company follows for its
general risk management.
Based on survey responses, MNEs have made signicant progress in documenting formal general risk management
policies and somewhat less progress in formalizing their tax risk management policies. In any case, the general and
tax risk management policies that have been and continue to be developed by the largest MNEs, such as the responTable 9
Documented tax risk management policy.
Population

Single policy for the global

organization

U.S. rms
Foreign rms

11
24

Total

35

Multiple, country-specic
policies

No documented risk
management policy

Other response

2
8

33a
31c

2b
1d

10

64

Number of U.S. respondents = 48. Number of foreign respondents = 64.

a
One U.S. rm is currently in the process of documenting a tax risk management policy.
b
One U.S. rm has no documented tax risk management policy, but aspires to it.
c
One Canadian rms tax risk management policy is included in its documented global policy.
d
One Swiss rm has a single tax risk management policy for the global organization and multiple, country-specic tax risk management policies (the
same model that the company follows for its general risk management policies).

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

23

dents in this research, constitute best practices to which other rms can refer as guidance in their own risk management
initiatives.
The non-U.S. rms report more progress than the U.S. rms, perhaps because they are experiencing greater urgency. All
of the foreign rms surveyed are NYSE-listed; the stakes for retaining their U.S. listing is characterized by a dimension not
present for U.S. rms, that is, satisfying host-country reporting standards. While U.S. Fortune 500 rms may be able to satisfy
the SOX Section 404 attest requirement with regard to internal controls (and by implication risk management) without
formal, documented risk management policies, non-U.S. rms may be reticent to assume their similar success.
A telephone interview with the head of tax of one of the ten largest U.S. Fortune 500 companies conrms that presumption. The tax executive acknowledged that his rm lacks a formal, documented tax risk management policy, although he
is condent that he can demonstrate that his rm systematically manages tax risk based on his rms documented general
risk management policy. In a similar vein, another U.S. respondent that lacks a documented tax risk management policy
attributes its ability to ensure a satisfactory level of risk mitigation to its headquarters tax function being staffed by highly
qualied attorneys and accountants who work with some of the best outside advisors in the world.
NYSE-listed foreign rms lack a home country advantage and may be more condent in their ability to satisfy U.S.
authorities of their internal control and management of tax (and other) risk with a documented policy in place, hence their
greater emphasis on developing formal risk management policies. The fact that SOX is U.S. legislation no doubt also increases
the urgency that U.S.-listed foreign rms infer with regard to the need for formal, documented policies.
5.2. Assessments of tax risk
Firms were asked to assess tax risk on two different dimensions: the seven TRM tax risk elements and six specic tax
issues. First, each rm was asked to rate the seven TRM tax risk elements, using a ve-point Likert scale, where one is not
likely or has insignicant impact and ve is very likely to occur or have a signicant impact. Each type of risk was dened in
the questionnaire by example, as follows:

Transactional risk (examples: acquisitions, mergers).

Operational risk (examples: new business ventures, new operating models, new operating structure).
Compliance risk (examples: weak records and controls, legislative changes).
Financial accounting risk (examples: changes in systems and policies).
Portfolio risk (example: combination of any of the risks).
Management risk (examples: changes in personnel, new/inexperienced resources).
Reputational risk (example: revenue authority investigation).

Second, rms were asked to rate the following six specic tax issues on a ve-point Likert scale, where one is an
insignicant risk factor and ve is a major risk factor:

Transfer pricing.
Corporate income tax rates.
Value-added taxes.
Foreign tax credits.
Payroll taxes.
Cultural norms effect on tax compliance in foreign countries.

The research hypotheses are based on the presumption that U.S. and non-U.S. rms are so culturally and operationally
different that their assessments of tax-related risk management also differ. For example, although Enrons effect was global,
it was a U.S. corporate experience that put in motion events that led to the enactment of SOX and the immediate, intense
focus on risk management principles in the United States. Although other countries are in the process of formalizing and
promulgating risk management principles, none has progressed to the U.S. level of sophistication.
In null form, the hypotheses are:
H2. U.S. and non-U.S. MNEs do not differ statistically as to their rating of the likelihood and signicance of the seven TRM
tax risk elements.
H3.

U.S. and non-U.S. MNEs do not differ statistically as to the risk associated with six specic tax provisions.

The hypotheses were tested using both the Wilcoxon rank-sum test and the t-test. The data in Tables 10 and 11 demonstrate
virtually identical results for both tests.
As shown in Table 10, for the seven TRM tax risk elements, the Wilcoxon rank-sum test results indicate rejection of the
null with regard to operational risk at the .05 signicance level (-value of .03973) and rejection of the null pertaining to
reputational risk at the .10 signicance level (-value of .08924). As noted in Table 11, when a t-test is applied to the data,
the same hypotheses are rejected and at the same signicance levels, -values of .04848 and .077942, respectively.

24

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

Table 10
Results of Wilcoxon rank-sum test.
Type of risk

Computed W-statistic

-Value

Panel A
TRM tax risk management elements
Transactional risk
Operational risk
Compliance risk
Financial accounting risk
Portfolio risk
Management risk
Reputational risk

2440.5
2342.0
2519.5
2714.0
2668.5
2591.5
2402.0

0.1789
0.03973a
0.3173
0.8747
0.9760
0.5677
0.08924b

Tax provision

Computed W-statistic

-Value

Panel B
Risk associated with specic tax issues
Transfer pricing
Corporate income tax rates
Value-added taxes
Foreign tax credits
Payroll taxes
Cultural norms effect on tax compliance in foreign countries

2098.0
2556.5
2429.0
2734.5
2834.0
2259.0

0.002914c
0.519
0.1575
0.562
0.3041
0.02022d

a
b
c
d

The null hypothesis was rejected for operational risk at the .05 signicance level.
The null hypothesis was rejected for reputational risk at the .10 signicance level.
The null hypothesis was rejected for transfer pricing at the .01 signicance level.
The null hypothesis was rejected for cultural norms effect on tax compliance at the .025 signicance level.

The conclusion is that the two populations of MNEs differ statistically only with regard to their assessments of operational
risk and reputational risk. In other words, the results suggest that a global view exists among MNEs about ve of the seven
TRM tax risk elements: transactional risk, compliance risk, nancial accounting risk, portfolio risk, and management risk.
It stands to reason that U.S. and foreign rms would have different perspectives about the tax risk associated with their
operations. Virtually by denition, operational risk (tax or otherwise) is not only country but rm specic. Firms craft their
operational structures to achieve their missions and satisfy stakeholders by leveraging relevant tax authorities strictures.
The more truly multinational a rm is, the more signicant the impact of operating within multiple jurisdictions and under
multiple tax laws. The responses to the 2005 International Tax Review survey conrm this conclusion. Respondents to that
survey indicated specically that tax executives want their tax advisers to understand and appreciate the effect of tax risk
on how the client structures operations (Crest, 2005a).

Table 11
Results of t-tests.
Type of risk

Computed t-statistic

-Value

Panel A
TRM tax risk management elements
Transactional risk
Operational risk
Compliance risk
Financial accounting risk
Portfolio risk
Management risk
Reputational risk

1.413
1.996
1.142
0.263
0.003566
0.6933
1.78

0.1605
0.04848a
0.2558
0.793
0.9972
0.4896
0.077942b

Tax provision

Computed t-statistic

-Value

Panel B
Risk associated with specic tax issues
Transfer pricing
Corporate income tax rates
Value-added taxes
Foreign tax credits
Payroll taxes
Cultural norms effect on tax compliance in foreign countries

3.051
0.8296
1.207
0.8203
0.7714
2.662

0.002881c
0.4086
0.2300
0.4139
0.4422
0.008988d

a
b
c
d

The null hypothesis was rejected for operational risk at the .05 signicance level.
The null hypothesis was rejected for reputational risk at the .10 signicance level.
The null hypothesis was rejected for transfer pricing at the .01 signicance level.
The null hypothesis was rejected for cultural norms effect on tax compliance at the .025 signicance level.

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

25

As to the difference between U.S. and foreign rms, it is recognized that individual countries and their tax authorities
and laws are associated with particular characteristics. As noted previously, Morton (2006) views the U.S. as technical and
adversarial and the Netherlands and Japan as consensual and patriotic, respectively. With the U.S. as their home country, U.S.
MNEs obviously possess an edge in structuring their operations to exploit the U.S. tax law to their advantage in that they
are inured, by history and culture, to the technical, adversarial nature of the U.S. tax law and tax authorities. In contrast, the
Netherlands and Japanese MNEs, accustomed to more consensual and patriotic tax regimes, respectively, must adapt to the
more technical and adversarial U.S. tax mosaic.
Firms were provided the opportunity to elaborate on any aspect of the survey. The following comment of a German rm
conrms the above analysis when it states that its responses to questions about tax risk may be inuenced by the degree
to what our company is engaged in [in] the specic countries (e.g., large degree of involvement in the U.S.).
The results pertaining to reputational risk are also not surprising. The Wilcoxon and the t-tests revealed that U.S. and
foreign heads of tax differ (at around a .10 signicance level) with regard to their risk appetites, with foreign heads of tax
being somewhat more risk averse (mean of 2.51695 compared to mean of 2.75000 for U.S. heads of tax). More risk averse
decision makers are likely to perceive more risk in rms activities than do those with a greater appetite for risk. Compounding
this phenomenon is the fact that foreign rms are expected to feel more vulnerable to, for example, the prospect of negative
publicity since they are concerned about the publicitys affect on their continued listing in the United States.
When the Wilcoxon and t-tests are applied to the rms assessments of the six specic tax issues, results of the two tests
are again virtually identical. The Wilcoxon test results in Table 10 reveal that the two MNE populations differ statistically with
regard to their assessments of risk associated with transfer pricing (.01 level) and cultural norms effect on tax compliance
in foreign countries (.025 level). The t-test results in Table 11 are almost identical. The null hypotheses for transfer pricing
and cultural norms effect are rejected at the .01 level.
The results pertaining to transfer pricing and the effect of cultural norms are straightforward and obvious. With regard
to transfer pricing, Lowell, Martin, and Donohue (2006, p. 1) state: An obvious concern for all multinationals after SOX is
whether there are effective controls in place to deal with transfer pricing exposure (i.e., risk). The combination of Internal
Revenue Code (IRC) section 482, which Avi-Yonah (2007, p. 103) describes as the most sweeping grant of power to the IRS
of any section in the Internal Revenue Code, and the Treasury Department 1996 transfer pricing penalty regulations creates
formidable tax risk.
The SOX element that is most signicant pertains to the requirement of transfer pricing documentation relating to
the inuence of legislation, ethical standards, and associated matters that do not specically target transfer pricing but
nonetheless impose a range of incremental requirements on transfer pricing compliance, documentation opinions, and
implementation of controls to make sure that the intended transfer pricing results are actually achieved (Lowell et al., 2006,
p. 1).
Although transfer pricing is a signicant potential exposure for all MNEs, the exposure differs for U.S. and foreign MNEs.
The U.S. transfer pricing documentation requirements, which are highly evolved and rigorous, precede SOX. The Treasury
Departments 1996 transfer pricing penalty nal regulations (i.e., Regulation Section 1.6662-6) created a urry of activity,
including a veritable cottage industry of external consultants offering their own version of documentation compliance (Lowell
et al., 2006, p. 1) (emphasis provided).
Even prior to SOX and its focus on documentation of internal control, the U.S. tax regulations enacted in 1996 identied
ten documents required to describe the taxpayers transfer pricing analysis and avoid imposition of penalties. The regulation
also requires that the primary documents ordinarily be supported by additional background documents, and any other
documents that are necessary to establish that the taxpayers method was selected and applied to ensure the most reliable
measure of an arms length result. The foregoing demonstrates the massive scope and consequent burden of U.S. transfer
pricing documentation requirements.
In contrast, other countries are scurrying in response to SOX (and the U.S. regulations) to develop their own transfer pricing
documentation regimes. Several foreign respondents elaborated on various aspects of transfer pricing. For example, a German
respondent clearly and unequivocally describes the foreign-rm dilemma when it states that [t]ax risk in transfer pricing
is high, because there is often not enough expertise in tax authorities, i.e., to [sic] little understanding of relevant factors.
The rm does not have a documented tax risk management policy. Two Indian rms refer to the compliance cost associated
with managing the risk associated with transfer pricing (e.g., cost of outside experts review of transfer pricing, exacerbated
by the lack of an advance pricing agreements (APAs) mechanism). One of the Indian respondents does not have a tax risk
management policy, but such a policy is in the process of development. On a more positive note, an Australian rm with a
tax risk management policy concludes, Transfer pricing is high risk but controllable through tax policies.
U.S. MNEs are steeped in experience pertaining to the U.S. transfer pricing rules and regulations concerning documentation
requirements and are expected to assess less risk from transfer pricing. The mean transfer pricing responses conrm that
assertion with a mean of 2.51064 for the U.S. rms versus a mean of 3.25806 for the foreign rms. None of the U.S. rms
provided commentary about the tax risk associated with transfer pricing.
Differences between the two MNE populations with regard to their assessments of the tax risk associated with transfer
pricing are likely to become more pronounced in the future. On August 1, 2006, the U.S. Internal Revenue Service (IRS) and
Treasury Department issued proposed regulations pertaining to the transfer of services and intellectual property between
multinational afliates. The result is that [m]ultinational companies operating in the U.S. will likely face stricter taxcompliance rules and bigger tax bites as a result of [the] proposed changes (Matthews, 2006, p. A.2). The importance

26

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

Table 12
Mean responses to tax risk assessments.
Tax risk classication

Mean of U.S. respondents

Mean of foreign respondents

Transaction
Operational
Compliance
Financial accounting
Portfolio
Management
Reputational

2.87500
2.77083
2.39583
2.54167
2.64583
2.39583
2.13542

3.22581
3.17460
2.63492
2.49206
2.64516
2.52381
2.50794

Transfer pricing
Corporate income tax
Value-added tax
Foreign tax credits
Payroll taxes
Cultural norms

2.51064
2.37500
2.02083
2.16667
1.85417
1.91667

3.25806
2.56452
2.25806
2.00000
1.74194
2.37705

of tax risk policies and practices will increase signicantly because the scope of U.S. transfer pricing law will expand to
encompass the transfer of not only tangible property, but also services and intellectual property.
The results of this research conrm recent research pertaining to the effect of culture on business practices. Based on their
study of culture, leadership, and organizations in 62 countries, House, Hanges, Javidan, Dorfman, and Gupta (2004, p. 54)
conclude that there is no evidence of a single model of management practices or of cultural values toward which all nations
are converging and acknowledge that there still remains a great deal of stability with respect to the more fundamental
aspects of both cultural practices and psychological commonalities within cultural entities.
Much of the preceding discussion is relevant in explaining why the two populations differ statistically in their ratings of
the risk associated with cultural norms effect on tax compliance in foreign countries. Of the two populations, the foreign
rms attributed greater risk to cultural norms effect on tax compliance in foreign countries. The foreign rms and U.S. rms
mean responses were 2.37705 and 1.91667, respectively.
It is not surprising that U.S. rms would assess less impact from other countries cultural norms because in matters
of commerce the United States is accustomed to affecting, not being affected by, other countries and their cultures.
Additionally, although the United States may be referred to as a melting pot of ethnicities, the views of U.S. MNE executives
reect predominantly the values and business culture of the companys home country.
In contrast, executives of rms in 26 foreign countries responded to survey inquiries. Even without the SOX impact, the
technical, adversarial U.S. tax culture (Morton, 2006) creates the need for more formal conduct with regard to tax matters
than, for example, the consensual and patriotic tax cultures of the Netherlands and Japan. U.S. MNEs are inured by history
and culture to such strictures, whereas adapting to the U.S. tax regime and administration requires signicant education
and adaptation for foreign MNEs. A case in point is the emerging central European countries that did not even have market
economies and formal infrastructures such as taxing authorities until the early 1990s.
6. Conclusions
The implications of this research are ve-fold. First, the research demonstrates that signicant progress has been made
by large MNEs in recognizing the importance of developing and implementing not only general risk management policies,
but also separate tax risk management policies. Development and implementation of those policies contributes to the
development of tax risk management best practices.
Second, the results provide guidance in identifying the loci and impact of organizational tax risk. The mean responses in
Table 5 demonstrate that with regard to the seven types of tax risk, both U.S. and foreign respondents conclude that the most
signicant amount of risk is from transactional risk (e.g., acquisitions, mergers). Both populations of MNEs assess transfer
pricing as the specic tax provision with the greatest degree of risk.
Third, of considerable importance are respondents overall assessments of tax risk. The data in Table 12 demonstrate that
the responding CFOs (or designees) do not perceive alarming degrees of tax risk in their organizations.
Responses for the risk associated with the seven TRM classes of tax risk range from 2.13542 (U.S. mean for reputational
risk) to 3.22581 (foreign mean for transaction risk). The majority of means were mid-two to mid-three, a moderate amount
of perceived risk and impact. With regard to the risk associated with the six tax issues, responses range from 1.74194 (foreign
mean for payroll tax) to 3.25806 (foreign mean for transfer pricing).
With a rating of ve denoting a major risk factor, none of the mean responses approaches such a rating. For example, both
populations of MNEs rated transfer pricing risk highest among the six tax issues; however, only ve (10%) of 48 responding
U.S. rms and eight (13%) of 64 responding foreign rms rated it as a major risk factor. The respondents either do not perceive
the existence of alarming degrees of tax risk or conclude that tax risk is under control in their organizations. Such optimistic
ratings are probably due at least to some degree to the progress reported in systematically managing tax risk by formalizing
and implementing tax risk management policies (e.g., the Australian respondent who noted that tax risk, while high, is
controllable through tax policies).

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

27

Fourth, the research results are signicant because of the similarity in responses of U.S. and foreign MNEs. Using both
parametric and nonparametric tests, the two populations responses differed statistically with regard to only four of 13
classications of tax risk: operational risk, reputational risk, transfer pricing, and cultural norms effect on tax compliance.
In all four cases, the mean rating of foreign respondents exceeded the mean of U.S. respondents, suggesting that non-U.S.
MNEs perceive more risk in those tax elements than do U.S. rms. The similarity in assessments of various types of tax risk
and the fact that the vast majority of respondents apply a single risk policy to their global organizations constitute strong
support for the conclusion that a global view of tax risk management exists among the MNE community.
Fifth, the research results demonstrate, unexpectedly, that existing reporting structures enable CFOs to shift a signicant
degree of tax risk management to heads of tax. This shifting of risk is the result of two factors: (1) the vast majority of U.S.
and foreign heads of tax report to the CFO, and (2) 85% of respondents indicate that the head of tax is responsible for signing
off on the totality of the global organization tax position. In the United States, for example, although the CFO and the CEO
are ultimately responsible for internal control (i.e., risk management) assessment under section 404 of SOX, the CFO has the
head-of-tax sign off on the rms global tax position as a basis upon which to rely.
7. Limitations and future research
7.1. Limitations
The research questionnaire provides rms the opportunity to elaborate and/or identify relevant factors not addressed
specically in the instrument. Respondents identied the following factors as relevant to tax risk:

Accounting for deferred taxes.

Application of tax accounting under various accounting standards (e.g., U.S. GAAP, IFRS, local GAAP).
Complexity of tax legislation in each country.
Corruption.
State and local taxes.
Uncertainty inherent in taxing authority actions.
Withholding tax.

Since respondents devoted time to identify these factors, they are obviously important in respondents decision-making.
A Finnish rm and a French rm identied accounting for deferred taxes as characterized by the most tax risk and among the
high risk areas, respectively. This study might have beneted by including deferred taxes (a highly technical topic) among
the specic tax-related issues. However, deferred taxes are a nancial reporting issue and not a tax provision (i.e., it is not a
tax code item). Consequently, the risk associated with deferred taxes should have been captured in responses pertaining to
nancial accounting and/or compliance risk.
State and local taxes and withholding taxes were identied by only one respondent each as relevant to their tax risk
exposure. Neither of these taxes is prominent in the international tax literature.
The remaining four factors (i.e., application of multiple accounting standards, complexity of tax legislation in multiple
jurisdictions, corruption, and uncertainty inherent in taxing authority actions) constitute aspects of the environment in
which every MNE operates. However, their impact is captured in assessments of various types of risk that were elements of
the research questionnaire (e.g., nancial accounting risk, compliance risk). For example, rms that are subject to multiple
accounting standards and taxing jurisdictions would be expected to assign a relatively high rating to nancial accounting
and/or compliance risk.
7.2. Future research
Seventy-eight percent of the documented tax risk management policies identied in this research are single policies that
are applicable to the global organization, notwithstanding the multitude of different taxation systems that MNEs encounter.
It is possible that rms using a single tax risk management policy do so only as a rst step in systematically addressing tax
risk management; they may develop more country-specic or area-specic policies in the future. Discerning whether single
tax risk management policies are the nal product or merely a step toward multiple tax risk management policies is an area
to be developed by future research.
References
Arlinghaus, B. P. (1998). Goal setting and performance measures by tax professionals in Fortune 500 companies. The Tax Executive, 50(6), 434442.
Australian Taxation Ofce (2006). 2006 Large Business and Tax Compliance, at http://www.ato.gov.au/content/downloads/77898 N8675-08-2006 w.pdf.
Avi-Yonah, R. (2007). International tax as international law. New York: Cambridge University Press.
Campbell, G. C., Durst, M. C., & Miller, J. D. (2002). Beyond auditor independence. . .what is really important for tax executives in Sarbanes-Oxley? The Tax
Executive, 54(6), 522527.
Committee of Sponsoring Organizations. (2004). Enterprise risk management integrated framework. Jersey City, NJ: COSO.
Crest, S. (2005a). Asia tax executives want to get intimate. International Tax Review, 16(8), 1318.

28

H.F. Wunder / Journal of International Accounting, Auditing and Taxation 18 (2009) 1428

Crest, S. (2005b). Tax-risk tolerance declines sharply in the U.K. International Tax Review, 16(10), 4.
Cummings, J. L., Jr. (2005). Chief tax ofcer reporting structures. The Tax Adviser, 36(10), 620625.
Elgood, T. (2006, July 17). Tax-risk policy. International Tax Review Special Supplement: Tax Management in Companies, 1417.
Elgood, T., Paroissien, I., & Quimby, L. (2005). Managing global risk for multinationals. Journal of International Taxation, 16(5), 2227.
Ernst & Young (2004). Tax Risk Management: The evolving role of tax directors, at http://www.ey.com/global/download.nsf/International/EY - Tax Tax Risk Managemenet/$le/EY Tax Risk Survey Report.pdf.
Ernst & Young (2006). Tax Risk: External Change, Internal Challenge Global Tax Risk Survey 2006, at http://www.ey.com/GLOBAL/content/nsf/International/
TARAS - Global Tax Risk Survey 2006.pdf.
Ernst & Young (2007). Risk management programs: What does the audit committee need to know? BoardMatters Quarterly, at https://
eyaprimo.ey.com/natlmktgaprimoey/attachments/BMQ14 Oct%202007 .pdf.
European Corporate Governance Institute (2005). Index of codes, at http://www.ecgi.org/codes/all codes.php.
Fennell, D. (2005). Revenues large cases division co-operative compliance framework. Accountancy Ireland, 37(6), 5759.
FASB Interpretation No. 48 (2006). Accounting for uncertainty in income taxes an interpretation of FASB Statement No. 109. Norwalk, CT: Financial Accounting
Standards Board.
Ge, W., & McVay, S. (2005). The disclosure of material weaknesses in internal control after the Sarbanes-Oxley Act. Accounting Horizons, 19(3), 137158.
Goodman, G. R. (2004). Internal controls for the tax department. Tax Notes, 103(May 3), 579588.
House, R. J., Hanges, P. J., Javidan, M., Dorfman, P. W., & Gupta, V. (2004). Culture, leadership, and organizations: The GLOBE study of 62 societies. Thousand
Oaks, CA: Sage Publications, Inc.
International Corporate Governance Network (2005). ICGN Statement on Global Corporate Governance Principles. London: ICGN Secretariat.
Ireland Revenue Authority (2005). The Cooperative Approach to Tax Compliance, at http://www.revenue.ie/index.htm?/services/bus large business.htm.
KPMG (2004). Tax in the Boardroom, at http://www.kpmg.com.
Lambert, C., & Lucas, J. (2006, July 17). Managing global tax compliance. International Tax Review Special Supplement: Tax Management in Companies, 3438.
Lowell, C. H., Martin, M. R., & Donohue, M. J. (2006). Managing transfer pricing. Part I. Journal of International Taxation, 17(7), 4459.
Matthews, R. G. (2006, August 2). IRS, treasury propose tightening tax rules for multinational rms. Wall Street Journal, A.2.
Morton, Paul. (2006). Tax risk management. In Presentation made at LexisNexis Tolley conference London, England,
Neubig, T., & Sangha, B. (2004). Tax risk and strong corporate governance. The Tax Executive, 56(2), 114119.
OSullivan, K. (2005, January 18). Tiptoe through the new regs. CFO, 1.
PR Newswire (2005, November 16). Corporate tax executives grow in prominence with Sarbanes-Oxley compliance work, KPMG study reveals. 1.
PricewaterhouseCoopers. (2004). Tax risk management. London: PWC.
Quimby, L., & Pearce, J. (2006, July 17). Lessons learned from SOX 404. International Tax Review Special Supplement: Tax Management in Companies, 2629.
Siegel, S., & Castellan, N. J., Jr. (1988). Nonparametric statistics for the behavioral sciences. New York: McGraw-Hill, Inc.
Wyman, P. (2006). New 8th company law directive, not the EUs Sarbanes-Oxley. Journal of International Taxation, 17(6), 6163.