pott - 2009 - review of empirical research on rotation and non-audit service - auditor independence in fact vs. appearance

DOI 10.1007/s11301-009-0043-0
STATE-OF-THE-ART-ARTIKEL
Review of empirical research on rotation
and non-audit services:
auditor independence in fact vs. appearance
Christiane Pott · Theodore J. Mock · Christoph Watrin
Received: 14 May 2008 / Accepted: 16 December 2008 / Published online: 29 January 2009
© Wirtschaftsuniversit
¨
at Wien, Austria 2009
Abstract Confidence in the processes of corporate reporting and auditing has rapidly
decreased recently due to front-page accounting scandals in both the United States
and Europe. The goal of audit regulations, such as the Sarbanes Oxley Act in the
United States (US) and the 8th Directive in the European Union (EU), is to restore
public trust in the auditing process. Along with other regulatory aspects, requirements
related to audit partner rotation and bans on providing concurrent non-audit services
were implemented to maintain auditor independence, both in fact and in appearance.
However, the implementation of audit regulation implies that increased requirements
are able to enhance the failed audit function. Empirical research should help to un-
derstand the impact of these two regulatory aspects and indicate their effectiveness in
maintaining auditor independence. Thus, we outline the newest empirical research re-
lated to audit partner rotation and non-audit services and independence in fact or in
appearance. Overall, we conclude that prior research does not point to one particu-
lar requirement that would most effectively restore trust in the audit function. Rather
the existence of multiple threats to auditor independence might demand a combina-
tion of several requirements to maintain auditor independence. Thus, more research
C. Pott (✉)
Institut f
¨
ur Unternehmensrechnung und -besteuerung, Westf
¨

alische Wilhelms-Universit
¨
at M
¨
unster,
Universit
¨
atsstr. 14–16, 48143 M
¨
unster, Germany
e-mail:
T. Mock
Anderson Graduate School of Management, University of California, 900 University Avenue,
Riverside, California 92521, USA
e-mail:
C. Watrin
Institut f
¨
ur Unternehmensrechnung und -besteuerung, Westf
¨
alische Wilhelms-Universit
¨
at M
¨
unster,
Universit
¨
atsstr. 14–16, 48143 M
¨
unster, Germany

e-mail:
13
J Betriebswirtsch (2009) 58: 209–239
210 C. Pott et al.
is needed to investigate the joint effects of different threats to auditor independence,
e. g., non-audit fees and audit partner tenure.
Keywords Auditor independence · European Union · Regulation
Zusammenfassung Durch Bilanzskandale in den USA und Europa wurde das
Vertrauen in die Unternehmens-berichterstattung und die Abschlusspr
¨
ufung stark
ersch
¨
uttert. Das Ziel von Gesetzesinitiativen wie der Sarbanes-Oxley Act in den
USA und der 8. EU-Richtlinie ist daher die Wiederherstellung des Vertrauens in die
Abschlusspr
¨
uferfunktion. Neben anderen regulatorischen Maßnahmen betreffen die
gesetzliche
¨
Anderungen die Rotation des verantwortlichen Pr
¨
ufungspartners und das
Verbot, bestimmte Nicht-Pr
¨
ufungsleistungen f
¨
ur Pr
¨
ufungsmandanten zu erbringen.

Diese Regelungen wurden implementiert, um ,,auditor independence in fact“ und
,,auditor independence in appearance“ zu erhalten. Die Einf
¨
uhrung der gesetzlichen
Anforderungen impliziert ex ante, dass die jeweiligen gesetzlichen Bestimmungen
in der Lage sind, die deutlich in Kritik geratene Abschlusspr
¨
uferfunktion zu stabi-
lisieren. Empirische Forschungsergebnisse k
¨
onnen in diesem Zusammenhang dazu
dienen, den Einfluss der drei oben genannten gesetzlichen
¨
Anderungen im Hinblick
auf die Erhaltung der Abschlusspr
¨
uferunabh
¨
angigkeit zu beurteilen. Daher stellt die-
ser Beitrag die neueste empirische Forschung zur Abschlusspr
¨
uferrotation und zu
Nicht-Pr
¨
ufungsleistungen in Bezug auf ,,auditor independence in fact“ und ,,auditor
independence in appearance“ vor. Insgesamt lassen die empirischen Befunde keine
Aussage dar
¨
uber zu, welcher spezifische Regulierungsaspekt zur Wiederherstellung
des Vertrauens in die Abschlusspr

¨
ufung beitr
¨
agt. Vielmehr k
¨
onnte das gleichzeitige
Vorliegen verschiedener Bedrohungen f
¨
ur die Abschlusspr
¨
uferunabh
¨
angigkeit eine
Kombination verschiedener regulatorischer Maßnahmen erforderlich machen. Um
diesen Zusammenhang zu untersuchen, sind jedoch weitere empirische Untersuchun-
gen n
¨
otig, die sich gezielt mit der gleichzeitigen Auswirkung der Rotation des verant-
wortlichen Pr
¨
ufungspartners und dem Verbot, bestimmte Nicht-Pr
¨
ufungsleistungen
zu erbringen, auseinandersetzen.
Schl
¨
usselw
¨
orter Wirtschaftspr
¨

uferunabh
¨
angigkeit ·Europ
¨
aische Union ·
Regulierung
JEL Klassifikationen M42
1 Introduction
The role, position, and liabilities of the statutory auditor in the European Union are
regulated individually by each Member State. However, the increasing number of
large-scale financial failures has led to calls for harmonization of the audit function.
Increasingly, there is a general belief that the lack of common practices or standards
negatively affects audit quality (European Commission 1996). Worldwide account-
ing scandals have been followed by harsh criticism of the auditing profession. In
13
Review of empirical research on rotation and non-audit services 211
the United States, the Enron collapse had a major impact on relevant regulations.
However, almost every European country also experienced an accounting scandal
around 2001. SAirGroup in Switzerland had to acknowledge corporate governance
violations in 2001, the Belgian company Lernout & Hauspie was accused of fraud
and market price manipulation in 2001, and by 2003, the Netherlands was unset-
tled by Ahold’s exaggerated earnings forecasts (Peem
¨
oller and Hofmann 2005). The
list of companies facing similar issues is quite long, including Parmalat in Italy in
2003, the Austrian company Yline in 2003, and Philipp Holzmann and Comroad
in Germany in 2002. In response, regulating bodies in the United States and the
European Union each issued major pieces of legislation: in the United States, the
Sarbanes-Oxley Act in 2002, and in the European Union, the revised 8th Direc-
tive (2006). These regulations include more severe penalties and larger enforcement

budgets to protect financial markets from fraud (see, e. g., Krishnamurthy, Zhou, and
Zhou 2006).
According to Bolkestein (2003, p. 1), the “Directive aims to reinforce the statu-
tory audit function in the Euroepan Union, which is one of the crucial elements for
underpinning the trust in the functioning of the European capital market.” The re-
vised 8th Directive was intended to become law by mid-2005, but it was not finally
issued until the end of June 2006. The Member States had to incorporate the minimal
requirements of the Directive into national law by June 2008. Among other things,
the 8th Directive (2006) establishes principles concerning auditor independence, yet
the requirements of the Directive do not directly address the auditor (at least in this
stage of regulation). Moreover, this regulation addresses Member States because they
must first incorporate the requirements into their own national law before the rules be-
come binding on auditors and other involved parties. Member States often add local
requirements to the minimum requirements issued, and thus each State differs in how
it establishes its requirements.
Here, we build on the existing literature by providing insights into the new and
fast-growing research area of the impact of regulation on auditors. Dissociating
themselves from the institutional system of self-regulation, auditors are faced with
a continuously growing set of additional requirements. For example, auditors are now
charged with understanding and applying the relevant local regulations to ensure
compliance. The auditor is now required to rotate from his or her audit client to avoid
audit bias.
Prior literature reviews on auditor independence were done several years ago and,
consequently, do not include important empirical findings from the post-SOX period
(see, e. g., Ewert 2003). Also, many literature reviews focus on empirical findings re-
lated to one independence issue, such as non-audit services (Schneider, Church, and
Ely 2006) or audit firm rotation (Cameran, Di Vincenzo, and Merlotti 2005). Prior
literature is the only evidence available by which we can study the real effects of regu-
lation. Therefore, we present a structured review of the most important empirical and
experimental literature on auditor independence. This analysis will lead to an assess-

ment of the probable effectiveness of auditor independence regulation. Therefore, we
try to answer the following question: Is the regulation of auditor independence ef-
fective and necessary? We arrive at the answer through the findings of the existing
empirical literature.
13
212 C. Pott et al.
Our paper proceeds as follows: Section 2 defines auditor independence in both
fact and appearance and explains different theoretical approaches to auditor inde-
pendence. The regulation related to auditor independence is presented in Sect. 3.
Section 4 presents and discusses prior empirical research on auditor independence.
Section 5 concludes with perspectives on future research.
2 Auditor independence
The ex ante value of an audit to consumers of audit services (including current and
potential owners, managers, shareholders and consumers of the firm’s products) de-
pends on the auditor’s perceived ability to: (1) Discover errors or breaches in the
accounting system, and (2) withstand client pressure to disclose selectively in the
event a breach is discovered. In this context, the level of auditor independence can be
defined as the conditional probability that – given that a breach has been discovered –
the auditor will report the breach (DeAngelo 1981a, pp. 115–116; 1981b).
For the purposes of many regulatory frameworks, independence is separated into
two related concepts. First, independence requires independence in mind, defined as
a state of mind that (1) is unaffected by influences that might compromise profes-
sional judgment, and that (2) allows an individual to act with integrity and to exer-
cise objectivity and professional skepticism (International Federation of Accountants
2004, p. 17).
1
Regulatory frameworks also often use the phrase “independence in
fact” when referring to independence in mind (Securities and Exchange Commission
2001; European Commission 2002). Second, independence requires independence in
appearance, which is described as the avoidance of significant facts and/or circum-

stances that would reasonably cause a rational and informed third party to conclude
that a firm’s (or a member of the assurance team’s) integrity, objectivity or profes-
sional skepticism had been compromised (International Federation of Accountants
2004, p. 17). The appearance of an independence failure is enough to undermine con-
fidence in auditing and financial reporting. This is due to the suspicion (belief) that
independence is impaired, because independent behavior (independence in fact) is
unobservable. The result of evidence (or simple belief) of auditor failure leads to
a loss of trust in the audit process and, more generally, in financial reporting, thus
destabilizing markets (Fearnley and Beattie 2004, p. 121).
Originally deserved trust, defined as the expectation people have of each other
(Barber 1983), can also be unsettled by other means. Aside from financial statements
and audit reports, users and auditors in today’s world are flooded with information
from a number of sources (e. g., business magazines, newspapers, internet, television,
etc.). According to extant research, this kind of information influences decision-
makers’ perceptions, attitudes, decisions, and behaviors (e. g., Barber and Odean
1
The International Auditing and Assurance Standards Board (IAASB) of the International Federation of
Accountants (IFAC) serves the public interest by setting – independently and under its own authority –
high quality standards dealing with auditing, review, other assurance, quality control and related services,
and facilitating the convergence of national and international standards. This enhances the quality and
uniformity of practice in these areas throughout the world, as well as strengthened public confidence in
financial reporting generally.
13
Review of empirical research on rotation and non-audit services 213
2008). Additionally, this information can influence decision-makers’ ex post evalu-
ations of other professionals’ behaviors and responsibilities (e. g., Bonner 2008). In
summary, several prior studies have found that users have disproportionate expecta-
tions of auditors, whereas the auditor’s performance does not (have to) meet those
expectations. Extensive media reporting on audit scandals and inappropriate auditor
behavior leads users to experience the gap between their expectations and the ac-

tual behavior of the auditor (e. g., Porter 1993), which cannot be directly attributed to
a failure of independence either in fact or in appearance.
Explanations for the loss of trust in audit quality are largely based on the devel-
opment of the auditing sector over the last four decades. Following their geographic
expansion and increased competition, audit firms have been driven towards prof-
itability and growth to maintain competitiveness. The industry has had to generate
more revenue by both securing new audit clients and retaining existing ones. Due
to the demand of globally-operating clients, other services became alternative profit
generators. Thus, a growing percentage of public accounting firms’ total revenues
began to come from non-audit services, such as tax advice, information systems de-
sign and implementation, human resource management and general consulting (see,
e. g., Palmrose 1986; DeFond, Raghunandan, and Subramanyan 2002). This increase
reflects an absolute growth in non-audit activities coupled with stagnation in the
audit-service market. However, the joint provision of audit and non-audit services to
a client implies the existence of different contractual relationships between the audi-
tor and the client (Antle 1982, 1984).
Agency theory explains this phenomenon in terms of the separation of ownership
by investors and control by management (Jensen and Meckling 1976). The agency re-
lationship is specified by the contract under which the owners (principals) engage the
management (agent) to manage the firm on their behalf. Since both parties to the re-
lationship are utility maximizers, the agent will not always act in the best interest of
the principals (Antle 1984, p. 2). This may result in divergent interests and investor
losses, due to opportunistic agent behavior.
The use of an external auditor is one control mechanism for reducing the risk of
managers’ opportunism. However, the external auditor is still hired by the owner as an
agent. The auditor-agent must produce information to be used in contracting with the
manager (Antle 1984, p. 2). Therefore, the auditor is required to maintain indepen-
dence from the client management. In considering the impact of non-audit services
on auditor independence, it is important to point out that where audit and non-audit
services are provided to the same client, two different contractual relationships exist

(Beattie, Fearnley, and Brandt 1999). The non-audit services contractual relationship
is with the client. In the auditing contractual relationship, however, the auditor owes
an additional duty to the stakeholders. Moreover, the audit is subject to regulatory
oversight. However, the audit firm may perceive the purchase of an audit in the same
light as that of any other service.
The provision of non-audit services to audit clients increases the economic bond
between the audit firm and the client. This bond could lead to impaired independence
because (1) the audit firm is unwilling to criticize the work done by its non-audit ser-
vice department, (2) the audit firm does not want to lose lucrative non-audit services
provided to the audit client, and (3) the audit firm does not want to lose the audit en-
13
214 C. Pott et al.
gagement. In such a situation, the auditor may be inclined to agree with management
interpretations of accounting issues. In this respect moral hazard could lead the audi-
tor to agree with a certain level of dependency on the client in response to receiving an
additional amount of non-audit service fees. In such a situation moral hazard would
be covered legally under the additional non-audit service fees. Thus, the auditor has
her or his own opportunistic incentives, which may lead to biased judgments about
the nature and scope of the examination (audit work) made and biased reporting as to
whether the financial statements fairly represent the financial condition of the audited
firm (financial reporting) (Goldman and Barlev 1974).
2
As long as client-specific quasi-rents can be earned, the auditor has an economics
incentive to keep this client. Quasi-rents arise naturally in longer audit-client relation-
ships, when the auditor invests in the current period with the expectation of return
in future periods (Lee and Gu 1998). Substantial future quasi-rents are generated as
the result of the start-up or learning costs of the initial audits and the rotation costs
that all clients must incur when changing auditor. Furthermore, the economies of
scale resulting from combining auditing and other services are an additional source
of quasi-rents. It is true that the provision of service increases auditor quasi-rents that

are specific to the client.
With a positive initial investment in a new client, the future revenues must be
higher than future costs in order that the auditor is interested in keeping the relation-
ship with the client. Thus, the auditor has no incentives to tell the “truth” when the
truth is “bad news” from the client’s perspective (DeAngelo 1981a, p. 116). However,
if no client-specific quasi-rents are expected from a given client relationship, an au-
ditor is indifferent to the termination of that relationship. Consequently; he has no
economic incentive to conceal a discovered breach. In this case, the auditor is per-
fectly independent with respect to that particular client. Thus, if contracting among
the auditor and the client were costless, then auditors would always be perfectly inde-
pendent from their clients. It follows that a necessary condition for a positive demand
for costly auditing and less than perfect auditor independence is that contracting
among auditor and clients be costly.
However, when contracting is costly, incumbent auditors need to possess a com-
parative advantage over their competitors in future periods, in order to expect to
earn quasi-rents. Competition for initial auditing contracts leads many auditors to
set the fees for their initial audit engagement below the total current costs (called
“low-balling”) (DeAngelo 1981a, 1981b). Therefore, the competition for the “prop-
erty rights” of incumbency forces auditors to “low ball” in the initial period. Thus,
“low balling” is a competitive response to the expectation of future-quasi rents, and
does not itself impair independence. In fact, it is the client-specific quasi-rents that are
the necessary condition for impaired independence.
2
However, a parallel performance of audit and non-audit service to a client could also lead to information
and cost benefits, which could increase the overall audit quality (e. g., Quick 2002).
13
Review of empirical research on rotation and non-audit services 215
3 Auditor independence regulation
The need for financial regulation is often justified by the potential for market fail-
ure that can result when managers possess inside information while investors do not

(called “information asymmetry”), as well as from the inappropriate behavior of au-
ditors, (called “moral hazard”
3
). New laws that attempt to address these issues in the
United States (Sarbanes-Oxley Act of 2002) and in the European Union (Council of
the European Communities 2006) indicate that the predominant view among legisla-
tors is that we need more regulation, more severe penalties, and larger enforcement
budgets to protect financial markets from fraud.
The Sarbanes-Oxley Act (SOX) of 2002 has been the most significant legislative
response to corporate scandals such as Enron. The stated purpose of the act “is to pro-
tect investors by improving the accuracy and reliability of corporate disclosures made
pursuant to the securities laws, and for other purposes” (SOX 2002). As a correc-
tive measure in light of significant accounting failures, SOX strongly reinforces the
position of investors first defined by the Securities and Exchange Commission. This
Act is applicable to publicly traded companies, and it created the Public Company
Accounting Oversight Board under the supervision of the Securities and Exchange
Commission (Title I). It also imposes greater restrictions and requirements in terms
of auditor independence (Title II).
The role, position, and liabilities of the statutory auditor within the European
Union are inconsistently regulated by each Member State (Quick, Turley, and
Willekens 2008).
4
However, the increasing number of important financial failures has
led to a call for a minimum level of harmonization of the auditing function. There is
an increasing belief that the lack of common practices and/or standards has a nega-
tive impact on auditing quality and on the freedom to establish and provide services
in the auditing field (European Commission 1996). Therefore, the European Commis-
sion proposed a revised Directive, to ensure that investors and other interested parties
could fully rely on the accuracy of audited accounts, and also to enhance common
protection within the European Union. This might be an indication that harmoniza-

tion on a voluntary basis was not fully effective. The 8th European Directive will be
mandatory by June 2008 for the entire European Union (Council of the European
Communities 2006). The revised 8th Directive states that Member States shall ensure
that the auditor (or the key audit partner responsible for carrying out the audit on be-
half of the audit firm) rotates from the audit engagement within a maximum period of
five years (Article 40). SOX requires auditor and review partner rotation within the
same timeframe.
SOX makes it unlawful for a registered public accounting firm (or any person as-
sociated with the firm, to the extent determined appropriate by the Securities and
Exchange Commission) to provide to a client – contemporaneous with the audit – any
non-audit service. This service may be in the form of bookkeeping, or other services
related to accounting records or financial statements, financial information systems
3
For a detailed discussion on the need for regulation concerning financial reporting and accounting, see
Watrin (2001).
4
For a review on auditor independence regulation is provided by Quick (2006).
13
216 C. Pott et al.
design and implementation, appraisal or valuation services and fairness opinions; ac-
tuarial services, internal audit services, management functions or human resources;
broker-dealer services, legal services, and any other service that the board of direc-
tors determines to be impermissible by regulation. However, a public accounting firm
may engage in any non-audit service for an audit client (including tax services not
described above) if the activity is approved in advance by the client’s corporate audit
committee (SOX Section 201(a)).
When a statutory audit is carried out in a Member State, the 8th Directive requires
that the statutory auditor (or audit firm) be independent from the audited entity and
not be in any way involved in management decisions of the audited entity. Further-
more, a statutory auditor (or audit firm) is prohibited from conducting a statutory

audit if it has any financial, business, employment or other relationships with the
audited entity (including the provision of additional, non-audit services) that might
compromise the independence of the statutory auditor (or audit firm) (Article 23.1).
Likewise, as an example of how the profession has the ability to regulate itself
and does not shy away from adopting a tougher stance on its members, the Inter-
national Federation of Accountants (IFAC) issued a Code of Ethics for Professional
Accountants (revised 2005). The Code is divided into three sections. The first in-
cludes principles that are applicable to all professional accountants, while the second
and third parts distinguish between those principles that affect professional accoun-
tants in public practice and those that are applicable to other accountants employed
in business and industry. These new rules of independence set out a conceptual
framework that focuses on the factors that pose a threat to independence for all as-
surance engagements, and the safeguards auditors should put in place to preserve
their independence. In addition, the updated Code provides concrete examples of how
the conceptual approach to independence is to be applied to specific circumstances
and relationships. Thus we find situations of conflict being addressed, such as em-
ployment with assurance clients, provision of non-assurance services to assurance
clients including temporary staff assignments, recruiting of senior management, pro-
vision of tax and legal services, and preparation of accounting records and financial
statements.
4 Empirical research on auditor independence
The taxonomy used to structure the empirical literature on auditor independence fol-
lows the two major requirements of both SOX and the EU Directive: Research on
auditor rotation and non-audit services. Furthermore, in contrast to other literature
reviews (e. g., Schneider et al. 2006), we distinguish between empirical research re-
lated to independencein fact and independencein appearance. Independence in fact is
measured by the actual behavior of auditors, audit opinions issued, earnings manage-
ment and quality, and restatements of financial data (Ewert 2003, p. 531). Following
Ewert (2003), studies related to independence in appearance can be compared de-
pending on whether the direct assessments of stakeholders or the association between

the client’s earnings and capital market reaction and behavior related to the choice
of auditor is measured as the dependent variable. We also indicate whether the stud-
13
Review of empirical research on rotation and non-audit services 217
ies were conducted in the US, Europe or elsewhere, in order to consider potential
differences in findings related to jurisdictional differences.
4.1 Auditor rotation
With increasing audit partner tenure, the relationship between auditor and her/his
client might introduce a so-called familiarity threat to auditor independence (IFAC
2005). The IFAC Code of Ethics states that a “familiarity threat occurs when, by
virtue of a close relationship with an assurance client, its directors, officers or em-
ployees, a firm or a member of the assurance team becomes too sympathetic to the
client’s interests” (IFAC 2005, p. 32). This threat may result in a decline in the audit
partner’s ability to judge the company’s performance correctly.
5
A long relationship between audit partners and their clients might further influence
the auditor’s independence and objectivity. In the early years of tenure an audit part-
ner is most certain to be independent and to apply creative audit-testing approaches
(AICPA 1978;Hoyle1978). With an increasing period of engagement, the audit
programs may become routine, as a result of the auditors’ deeper knowledge of the
client’s systems and control procedures (Hoyle 1978). Shockley (1981) asserts that
“complacency, lack of innovation, less rigorous audit procedures and a developed
confidence in the client may arise after a long association with the client.” The major
concern then is that the auditors are anticipating the results on the basis of prior
audit years, instead of being alert to detect material misstatements (Arrunadaand Paz-
Ares 1997;Quick2004). Third, over time, the auditor’s incentives shift toward main-
taining and profiting from the client and the audit. The prospect of “client-specific
rents” that the auditor can extract only over time may create an economic dependency
on the client, which impairs the auditor’s independence (DeAngelo 1981a, 1981b).
The auditor expertise hypothesis is based on the information asymmetry between

the client and the auditor, which is reduced over time as auditors acquire client-
specific knowledge. Because increased client-specific knowledge provides a compar-
ative advantage in detecting material misstatements in financial reports, the lack of
this knowledge in the early years of an audit engagement may result in a lower quality
audit (e. g., Beck, Frecka, and Solomon 1988; Geiger and Raghunandan 2002).
4.1.1 Research studies
The rotation-related empirical literature, divided into studies on independence in fact
and in appearance, is presented in Table 1. Furthermore, the research is classified
based on whether the studies investigate audit firm rotation and audit firm tenure
and/or audit partner rotation and audit partner tenure.
Dopuch, King, and Schwartz (2001) observed that the highest frequencies of au-
ditors’ selections of favorable reports, as a measure of the actual behavior of the
auditor, occurred in regimes without mandatory audit firm rotation or retention. This
5
Opponents argue that a long audit partner association with a particular client results in higher audit qual-
ity. Myers et al. (2003) emphasize the higher audit costs in the early periods of audit tenure and the
increase in client and industry knowledge gained over repeated audits.
13
218 C. Pott et al.
Table 1 Research related to rotation and tenure
Jurisdiction Author(s) (Year) Measure Results
Supportive Not supportive
Auditor Independence in Fact
Audit Firm Rotation
Measure of auditor independence in fact: actual behavior of auditor
USA Dopuch et al.
(2001)
Auditors‘ willing-
ness to issue reports
biased in favour of

management
Rotation requirements decreased auditor-subjects’
willingness to issue biased reports, relative to the
two regimes on which rotation was not imposed
Managers also made higher investments than predicted in
these non-retention, non-rotation regimes, thereby raising
the probabilities they would have high assets
Audit Firm Tenure
Measure of auditor independence in fact: audit opinion issued
USA Geiger &
Raghunandan
(2002)
Audit Report preceding
bankruptcy
Indication that audit reporting failures are more likely to
happen in the earlier years of the auditor-client relationship
than when the auditors had served these clients for longer
tenures
Belgium Knechel &
Vanstraelen
(2007)
Going-concern audit
opinions
Auditors do not become less independent over time
Measure of auditor independence in fact: earnings management and quality
USA Myers et al.
(2003)
Abnormal accruals
and current accruals
Longer auditor tenure results in auditor placing greater

constraints on extreme management decisions in the re-
porting of financial performance
USA Davis et al.
(2007)
Abnormal accruals Firms with both short (two to three years) and
long (13–15 years or more) tenure are more likely
to report levels of discretionary accruals that allow
them to meet or beat earnings forecasts
USA Carcello &
Nagy (2004)
Companies’ cita-
tion for fraudulent
financial reporting
Fraudulent financial reporting is more likely to occur In
the first three years of the auditor-client relationship; no
evidence that fraudulent financial reporting is more likely
given long auditor tenure
13
Review of empirical research on rotation and non-audit services 219
Table 1 continued
Jurisdiction Author(s) (Year) Measure Results
Supportive Not supportive
USA Johnson et al.
(2002)
Abnormal accru-
als and earnings
persistence
Relative to medium audit-firm tenures of four to eight
years, short audit-firm tenures of two to three years are as-
sociated with lower-quality financial reports; no evidence

of reduced financial-reporting quality for longer audit-firm
tenures of nine or more years
Auditor Independence in Appearance
Audit Firm Rotation
Measure of auditor independence in appearance: association between client’s earnings and capital market reactions
USA Gates et al.
(2007)
Assessments on
company earnings
Individuals’ confidence in financial statements does not
improve through audit firm rotation
Audit Firm Tenure
Measure of auditor independence in appearance: association between client’s earnings and capital market reactions
USA Mansi et al.
(2004)
Return investors re-
quired on corporate
bonds
Investors require lower rates of return as the length of
tenure increases
USA Gosh &
Moon (2005)
Perceptions of in-
vestors and information
intermediaries
Positive association between investor perceptions
of earnings quality and tenure
Audit Firm Rotation and Tenure
Measure of auditor independence in appearance: direct assessments of stakeholders
Australia Azizkhani et al.

(2007)
Analysts’ perception of
financial data credibil-
ity
Audit firm tenure is significantly associated with
alowerex ante cost of equity capital for non-Big
4 audit firm clients
Audit firm rotation is not significantly associated with
changes in the ex ante cost of equity capital
Auditor Independence in Fact
Audit Partner Rotation
Measure of auditor independence in fact: earnings management and quality
Taiwan Chietal.
(2005)
Abnormal working cap-
ital accruals and abnor-
mal accruals
Audit quality is higher for companies subject to
the mandatory rotation regime in 2004 compared
to firms not subject to rotation in the same year
13
220 C. Pott et al.
Table 1 continued
Jurisdiction Author(s) (Year) Measure Results
Supportive Not supportive
Australia Hamilton et al.
(2005)
Abnormal accruals Audit partner changes most likely reflecting part-
ner rotation (i.e., they are not due to a switch of
audit firm) are associated with lower signed unex-

pected accruals
Germany Watrin et al.
(2008)
Abnormal accruals and
current accruals
More income-increasing earnings management
when the review engagement partner switches
Audit Partner Tenure
Measure of auditor independence in fact: audit opinion issued and earnings management and quality
Australia Carey &
Simnett (2006)
Audit opinion, abnor-
mal working capital
accruals and beating
(missing) earnings
benchmarks
Long tenure observations lead to a lower propen-
sity to issue a going-concern opinion and indi-
cates some evidence of just beating (missing)
earnings benchmarks, consistent with deteriora-
tion in audit quality associated with long audit
partner tenure
No evidence of an association between long audit tenure
and abnormal working capital accruals
Audit Partner Tenure and Rotation
Measure of auditor independence in fact: earnings management and quality
Italy Cameran et al.
(2008)
Abnormal accruals and
current accruals

Audit quality shows the tendency to improve
rather than worsen over time
Not support that mandatory audit rotation positively affects
audit quality
Auditor Independence in Appearance
Audit Partner Rotation
Measure of auditor independence in appearance: association between client’s earnings and capital market reactions
USA Gates et al.
(2007)
Assessments on
company earnings
Individuals’ confidence in financial statements
does improve through audit partner rotation
Audit Partner Tenure and Rotation
Measure of auditor independence in appearance: direct assessments of stakeholders
Australia Azizkhani et al.
(2007)
Analysts’ perception of
financial data credibil-
ity
Engagement partner tenure is significantly associ-
ated with a lower ex ante cost of equity capital
Audit partner rotation is not significantly associated with
changes in the ex ante cost of equity capital
13
Review of empirical research on rotation and non-audit services 221
result is consistent with the notion that auditors in these regimes react to economic in-
centives to bias their reports in favor of management. But managers also made higher
investments than predicted in these non-retention, non-rotation regimes, thereby rais-
ing the probability they would have high assets. These higher investments reduced the

overall risk to the auditors of having liability penalties imposed on them. In effect, the
combination of high investments and high favorable reports increased the welfare of
both parties.
Two studies proxy independence in fact by audit opinions issued. Geiger and
Raghunandan (2002) studied the relationship between audit firm tenure and audit re-
porting as a proxy for audit failure. The authors present evidence that auditors may
be more likely to be less objective in the initial years of an audit engagement, be-
cause reporting failures were more likely to occur in those initial years. Thus, this
research could imply that with mandatory auditor rotation in place, the risk of audit
failure (e. g., issuance of an incorrect audit opinion) might increase. However, since
tenure is measured as audit-firm tenure rather than individual-auditor tenure, the re-
sults cannot be generalized to the effectiveness of audit partner rotation. Addressing
the dominance of United States evidence on public firms, Knechel and Vanstraelen
(2007) studied the effect of audit firm tenure on audit quality for private companies
in Belgium. The authors use the likelihood of an auditor issuing a going-concern au-
dit opinion as an indicator of audit quality. Similar to that of Geiger and Raghunandan
(2002), their sample consists of stressed bankrupt and non-bankrupt companies.
They presume that a decrease in audit quality is indicated by an increase in the
likelihood that an auditor does not issue a going-concern opinion on a company that
subsequently goes bankrupt, or an increase in the likelihood that an auditor issues
a going-concern opinion to a company that survives. Knechel and Vanstraelen (2007)
found that auditors do not become less independent over time, nor do they become
better at predicting bankruptcy. Thus, their results refute the contention that long-term
auditor-client relationships undermine audit quality.
Studies using different measures of earnings management and quality as the de-
pendent variable only investigate audit firm tenure. Myers, Myers, and Omer (2003)
examined the relation between auditor firm tenure and earnings quality to calcu-
late abnormal or discretionary accruals, using the Jones Model
6
and absolute current

accruals
7
as proxies for earnings quality. Lower accrual levels are associated with
greater auditor conservatism, which has been interpreted as being suggestive of
higher-quality audits. They concluded that longer auditor tenure results in auditors
placing greater constraints on extreme management decisions in the reporting of fi-
nancial performance. Opposing these findings, Davis, Soo, and Trompeter (2007)
find that firms with both short (two to three years) and long (13–15 years or more)
tenure are more likely to report levels of abnormal accruals that allow the client to
meet or beat earnings forecasts. The results suggest that while regulatory mandates
6
Abnormal accruals are calculated from the total accruals, which are calculated using the statement of
cash flow as income before discontinued operations and extraordinary items from the statement of cash
flows.
7
Current accruals are measured as (∆CA −∆CASH) −(∆CL −∆STDEBT), where CA denotes total
current assets; CASH, cash and cash equivalents; CL, total current liabilities; STDEBT, the current
maturities of long-term debt and other short-term debt included in current liabilities.
13
222 C. Pott et al.
for periodic auditor turnover have negative effects, sustained long-term auditor-client
relationships may also be detrimental to audit quality.
Carcello and Nagy (2004) examined the relation between audit firm tenure and
fraudulent financial reporting. Comparing firms that were cited for fraudulent report-
ing between 1990 and 2001 with both: (1) A matched set of non-fraud firms and
(2) the available population of non-fraud firms, they found that fraudulent financial
reporting is more likely to occur in the first three years of the auditor-client rela-
tionship. The authors failed to find any evidence that fraudulent financial reporting
is more likely after a long (nine years or more) auditor tenure. Using two proxies
for financial-reporting quality and a sample of Big Six clients (matched by industry

and size), Johnson, Khurana, and Reynolds (2002) found that relative to medium au-
dit firm tenures of four to eight years, short audit-firm tenures of two to three years
are associated with lower quality financial reports. They measure the absolute value
of unexpected accruals as a proxy for financial reporting quality. The magnitude of
absolute value of unexpected accruals measures a company’s success in managing
earnings either up or down, as needed, depending on year-specific situations. In gen-
eral, using unexpected accruals as a proxy for financial-reporting quality is a joint test
of earnings management and the model of expected accruals used. As a second proxy
for financial-reporting quality, they examine the relationship between current-period
accruals and future income. However, they find no evidence of reduced financial-
reporting quality for longer audit-firm tenures of nine or more years.
Research on the relationship between audit firm rotation and auditor independence
in appearance, measured as the association between client’s earnings and capital
market reactions, also produces mixed results. Gates, Lowe, and Reckers (2007)re-
ported that even in an environment of strong corporate governance controls, audit
firm rotation does not incrementally influence individuals’ confidence in reported fi-
nancial statements, measured as the assessed quality of a company’s earnings.
Little research studies audit firm tenure and its effects on auditor independence
in appearance (as the association between client’s earnings and capital market re-
actions). Mansi, Maxwell, and Miller (2004) found that the requirement of investors
ford lower rates of return as the length of tenure increased provides direct evidence of
the value that investors attach to audit tenure. These data suggest that mandatory au-
ditor rotation may not be uniformly beneficial and could, in fact, be viewed negatively
by the capital market for riskier firms. Gosh and Moon’s (2005) results are consis-
tent with the hypothesis that investors and information intermediaries perceive longer
auditor tenure as improving audit quality.
Azizkhani, Monroe, and Shailer (2007) use the client-specific ex ante cost of
equity capital to measure whether audit firm tenure and rotation affect analysts’
assessments as direct assessments of stakeholders of financial data credibility. The
authors use a sample of Australian based companies with data from 1995 to 2005 to

test these relationships. They determined that for non-Big Four audit firms, audit firm
tenure is significantly associated with a lower ex ante cost of equity capital. Audit
firm rotation is not significantly associated with changes in the ex ante cost of equity
capital.
The varied results of rotation and tenure studies may be due to the use of the noisy
measure of audit firm rotation when audit partner rotation and audit firm tenure when
13
Review of empirical research on rotation and non-audit services 223
audit partner tenure should be measured instead. Therefore, the results on the ef-
fect of audit firm rotation and tenure are difficult to generalize. However, researchers
have only recently investigated the change in regulations on audit partner changes by
collecting individual signatures, which have to be provided in audit reports in some
countries as an indication of audit engagement partner rotation, e. g., in Australia and
Germany.
Most research investigates the effects of audit partner rotation using abnormal ac-
cruals following Jones (1991), abnormal working capital accruals following DeFond
and Park (2001)
8
and/or current accruals, as measures of earnings management and
quality. Chi, Huang, Liao, and Xie (2005) investigated whether the mandatory audit
partner rotation implemented in 2004 in Taiwan has had a positive influence on audit
quality. They find that audit quality is higher for companies subject to the manda-
tory rotation regime in 2004, compared to firms not subject to rotation in the same
year. The authors use absolute and signed abnormal accruals and abnormal working
capital accruals as proxies for audit quality. Hamilton, Ruddock, Stokes, and Tay-
lor (2005) use a sample of 3,621 firm-years between 1998 and 2003 from Australian
based companies, and report that audit partner changes most likely to reflect partner
rotation (when the company is not due for a switch of audit firm) are associated with
lower signed abnormal accruals in Australia. They conclude that this result is con-
sistent with more conservative reporting following an auditor rotation. Furthermore,

this explanation is supported by evidence that suggests a significant increase in the
asymmetrically timed recognition of economic losses when firms have a change of
audit partner. Using unique data from Germany, where both the audit engagement
and review partner can be identified, Watrin, Lindscheid, and Pott (2008) investigated
the effect of partner switches. Using exact, absolute, and signed values of abnormal
working capital accruals and current accruals as proxies for audit quality, they find
evidence of more income-increasing earnings management when the audit review
partner switches. Thus, mandatory audit review partner rotation might lead to more
aggressive accounting practices. However, they do not find an effect for any of our ac-
crual measures when the audit engagement partner switches. Therefore, their findings
do not support a mandatory audit engagement partner rotation.
Also, some studies have investigated audit partner tenure. Depending on the meas-
ure of audit quality used, Carey and Simnetts’ (2006) results imply different effects.
The three measures of audit quality examined are the auditor’s propensity to issue
a going-concern audit opinion for distressed companies, the direction and amount of
abnormal working capital accruals, and just beating (missing) earnings benchmarks;
all as measures of earnings management and quality. Their sample contains 1,021
Australian listed (since 1995) companies. First, when studying the propensity of an
auditor to issue ongoing-concern audit opinions for distressed companies, their re-
sults indicate that such opinions diminish over the audit partner’s tenure. However,
this result is associated with non-Big Six audit firms. For long tenure observations,
8
Abnormal working capital accruals =[WC
t
−(WC
t−1
/SALES
t−1
) ×SALES
t

]/TA
t−1
,whereWC
t
de-
notes non-cash working capital computed as (current assets minus cash and short term investments)
minus (current liabilities minus short term debt), SALES
t
equals total sales for the year t,andTA
t−1
denotes total assets for the year t −1.
13
224 C. Pott et al.
we find a lower propensity to issue a going-concernopinion and some evidence of just
beating (missing) earnings benchmarks, consistent with deterioration in audit quality
associated with long audit partner tenure. There is no evidence of an association of
long audit tenure with abnormal working capital accruals.
Also, one study investigated both audit partner rotation and audit partner tenure.
Cameran, Prencipe, and Trombretta (2008) provide evidence from Europe on the
effect of mandatory auditor rotation on audit quality. In Italy, a mandatory auditor ro-
tation regime has been in place for over twenty years. Cameran et al. (2008) found
that audit quality, measured as abnormal working capital accruals and current accru-
als as proxies for earnings management and quality, shows a tendency to improve
rather than worsen over time. The authors also study the effects of voluntary auditor
changes compared to mandatory auditor changes. Overall, the results do not support
the notion that mandatory audit rotation positively affects audit quality.
Lastly, studying auditor independence in appearance Gates et al. (2007) reported
that in an environment of strong corporate governance controls, audit firm rotation
incrementally influenced individuals’ confidence in reported financial statements,
measured as the association between the client’s earnings and capital market reac-

tions.
Azizkhani et al. (2007) use the client-specific ex ante cost of equity capital to
measure whether audit partner tenure and rotation affect analysts’ perceptions, cap-
turing the assessments of stakeholders of financial data credibility. The authors use
a data from 1995 to 2005 on a sample of Australian based companies to test these
relationships. They determined that for non-Big 4 audit firms, engagement partner
tenure is significantly associated with a lower ex ante cost of equity capital. Audit
partner rotation is not significantly associated with changes in the ex ante cost of
equity capital.
4.1.2 Implications
Research on ,,rotation“ has to be separated into studies that use the tenure of the au-
dit firm as a proxy for the level of dependency between the audit firm and the client,
and studies that use a change in audit firm (actual rotation) as a proxy for contempora-
neous change in the quality of audited financial data. In fact, only a small number of
countries (Brazil, India, Italy, Singapore and South Korea) have a legal requirement
for audit firms to be rotated after a maximum specified period. Based on the academic
literature collected, the present analysis supports the idea that the rotation rule pro-
vides doubtful benefits at best. Moreover, most of the empirical papers present data
that do not support the introduction of this rule. Thus, the overall empirical evidence
cannot confirm the alleged benefits of mandatory rotation; however, this is possibly
related to the use of the noisy measure of audit firm rotation when audit partner
rotation should be measured instead.
Many studies investigate audit firm rotation instead of audit partner rotation be-
cause of the lack of available data. Thus, the results are difficult to generalize with
respect to the effect of mandatory audit partner rotation. However, the relevance of
audit partner rotation in maintaining audit quality is still an unsolved issue world-
wide. Proponents of mandatory audit partner rotation argue that with an increase of
13
Review of empirical research on rotation and non-audit services 225
audit tenure, the responsible audit partners’ independence might be impaired due to

familiarity threats. Opponents of audit partner rotation believe that a sufficiently long
tenure is required to acquire the requisite knowledge about the client to provide high
quality audits. While regulatory bodies tend to believe that audit engagement partner
and other key audit partner rotation is beneficial (e. g., SOX, 8th EU-Directive), the
overall evidence related to audit partner tenure and rotation is mixed, and likely varies
with cultural disparities (e. g., Taiwan versus Australia). Furthermore, even studies
from the same jurisdiction (e. g., Australia alone) provide mixed results, which are
most likely derived from the use of different measures of earnings management be-
havior. However, the results seem to indicate that there are benefits to audit partner
tenure as well as rotation. A generalization of such benefits requires more research
from different jurisdictions in the post-rotation requirement period, and more robust
measures for capturing auditor independence.
Several caveats must be remembered when interpreting the current findings. First,
the samples used in audit partner tenure and rotation research are rather small and of-
ten consist of only large corporations. Thus, the results are limited to this group of
companies. Also, some studies only cover a very short time period, pointing to dis-
advantages of short-term examination of mandatory audit partner rotation and tenure.
Specifically, it may take some time for mandatory audit partner rotation to affect the
audit quality due to the time required for new audit partners to accumulate client-
specific knowledge and experience (Cameran et al. 2008). Future research should use
data over a longer period of time to examine the effect of mandatory audit partner
rotation on audit quality.
Furthermore, even if different studies claim to investigate the same setting, differ-
ences between studies are very likely making it difficult to compare findings from
different jurisdictions. Depending on data availability, different studies cover com-
pletely different time periods, which were affected by different regulatory initiatives,
economic shocks and other society specific differences. Also, most available research
is based on Australian data, which therefore comes from a highly developed industrial
country. However, Australia is the only country where data on audit partner rotation
includes only the signature of the responsible audit partner, whereas information on

the review partner is not provided. German findings, on the other hand, provide ev-
idence related to audit engagement partner and review partner rotation. Taiwanese
researchers have also studied the effects of dual signatures on audit reports, where
a particular audit function cannot be attributed to one of the two signers. Also, re-
sults from Taiwan may not be universally applicable, since it is an underdeveloped
industrial nation.
Studies of the relationships between audit rotation, as well as tenure and indepen-
dence in appearance, do not indicate a clear picture as to whether audit firm and audit
partner rotation is beneficial. Independence in appearance can lead to negative effects,
such as confidence losses in audit quality and destabilizing market effects, similar to
those of independence in fact. However, it is questionable to what degree indepen-
dence issues are perceived because intervals are affected by external information on,
for example, news on rotation requirements being implemented in the future. When
collecting data from individuals, their perceptions are most likely distorted by pre-
vious knowledge and experience with the related research topic, which might affect
13
226 C. Pott et al.
the results. Even if a controlled experiment is designed, it is most likely impossible
to hold these outside effects completely constant.
4.2 Non-audit service fees
4.2.1 Research studies
The literature on non-audit service fees was first separated into studies related to au-
ditor independence in fact and in appearance. Within these two subgroups, we further
classify the studies based on the classification provided by Ewert (2003), which has
been mentioned before. Table 2, summarizing the research studies, is structured in the
same way.
Studies from different jurisdictions have investigated auditor independence in fact,
measured as earnings management and quality. Reynolds, Deis, and Francis (2004)
report a significant positive association between the relative level of non-audit fees
and abnormal accruals based on Jones (1991). They argue that this result is due to

small-to-medium-sized high-growth firms, especially firms having initial public of-
ferings and in the e-commerce, biomedical, telecommunication, and pharmaceutical
industries. After factoring these variables into their analysis, they found no evidence
that the relative level of non-audit service fees impairs auditor independence. Also,
negative results were supported by Larcker and Richardson (2004) reporting a sig-
nificant negative association between the non-audit service and the total fee ratio and
Jones-type absolute value of accruals.
Farag (2006) studied the relation between the absolute value of abnormal accru-
als and the ratio of non-audit service to audit fees. She found this relation to be
only significant for firms with above-median persistence of earnings. Ferguson, Seow,
and Young (2004) found three measures of non-audit services to be positively re-
lated to abnormal accruals, following Jones (1991). Other results show that accrual
quality, measured as abnormal accruals following Dechow and Dichev (2002), has
a significant negative association with the magnitude of non-audit fees, and a signifi-
cant positive association with audit fees (Srinidhi and Gul 2007). The Dechow and
Dichev model uses firm-wise time-series regressions with the total current accruals as
the dependent variable and the cash flows of previous, current, and subsequent years
as the independent variables. This latter result is consistent with the proposition that
higher audit fees reflect higher audit effort and better judgments about the propriety
of accruals. However, this is not consistent with the proposition that audit fees are
associated with economic bonding. Ruddock, Taylor, and Taylor (2006) investigated
whether non-audit services provided by incumbent auditors are associated with a re-
duction in the extent to which earnings reflect bad news on a timely basis (news-based
conservatism). Reduced conservatism is to be expected only if relatively high lev-
els of non-audit fees result in reduced auditor independence and lower quality audits.
They find that higher-than-expected levels of non-audit services are not associated
with reduced conservatism.
Lim and Tan (2008) found evidence that audit quality (measured by increased
propensity to issue a going-concern opinion) increased the propensity of the audi-
tors to miss analysts’ forecasts. They also conclude that audit quality led to higher

13
Review of empirical research on rotation and non-audit services 227
Table 2 Research related to non-audit service fees
Jurisdiction Author(s) (Year) Measure Results
Supportive Not supportive
Auditor Independence in Fact
Measure of auditor independence in fact: earnings management and quality
USA Reynolds et al.
(2004)
Abnormal accruals Non-audit fees affect the magnitude of discre-
tionary accruals positively
USA Larcker &
Richardson
(2004)
Abnormal accruals Significant negative association between the non-audit ser-
vice to total fee ratio and the absolute value of accruals
USA Farag (2006) Abnormal accruals Relation between the absolute value of discre-
tionary accruals and the ratio of non-audit service
to audit fees; this relation is only significant for
firms with above-median persistence of earnings
UK Ferguson et al.
(2004)
Abnormal accruals Three measures of non-audit services are posi-
tively related to discretionary accruals
USA Srinidhi &
Gul (2007)
Abnormal accruals Accrual quality has a significant negative associ-
ation with the magnitude of non-audit fees and
a significant positive association with audit fees
Australia Ruddock et al.

(2006)
News-based
conservatism
Higher-than-expected levels of non-audit services
are not associated with reduced conservatism
Measure of auditor independence in fact: audit opinions issued
USA Lim&Tan
(2008)
Auditor’s propensity to
issue going-concern au-
dit opinions
Audit quality (measured by increased propen-
sity to issue going-concern opinion) increased the
propensity to miss analysts’ forecasts and led
to higher earnings-response coefficients when the
level of non-audit services acquired from industry
specialist auditors was compared to that acquired
from non-specialist auditors
USA DeFond et al.
(2002)
Auditor’s propensity to
issue going-concern au-
dit opinions
No evidence that non-audit service fees impair auditor in-
dependence
13
228 C. Pott et al.
Table 2 continued
Jurisdiction Author(s) (Year) Measure Results
Supportive Not supportive

UK Basioudis et al.
(2008)
Going-concern
modified audit opinion
Magnitude of both audit fees and non-audit fees
are significantly associated with the issuance of
a going-concern modified audit opinion
New
Zealand
Hay et al.
(2006)
Audit report qualifica-
tion or modification
Results indicate no significant relation between audit qual-
ification or modification and non-audit fees
Australia Sharma (2001) Going-concern
modified audit opinion
Higher NAS fees were associated with a lower
likelihood of receiving going-concern modified re-
ports
Australia Sharma and
Sidhu (2001)
Going-concern
modified audit opinion
Higher NAS fees were associated with a lower
likelihood of receiving going-concern modified re-
ports
USA Firth (2002) Going-concern
modified audit opinion
High non-audit fees were associated with clean audit re-

ports
USA Craswell et al.
(2002)
Going-concern
modified audit opinion
Fee dependence does not affect propensity to issue unqual-
ified opinions
Measure of auditor independence in fact: restatement of financial data
USA Kinney et al.
(2004)
Restatements No consistent evidence of positive association between
audit firm fees for either financial information system de-
sign and implementation of internal audit services and
restatement
Auditor Independence in Appearance
Measure of auditor independence in appearance: assessments of stakeholders
USA Jenkins &
Krawczyk
(2003)
Perceptions of auditor
independence
They find the public’s perceptions are not negatively af-
fected by any of six different NAS included in the study
USA Mauldin (2003) Perceptions of auditor
independence
Professional investors perceive that independence
is impaired when the auditor provides internal au-
dit services or merger and acquisition services
13
Review of empirical research on rotation and non-audit services 229

Table 2 continued
Jurisdiction Author(s) (Year) Measure Results
Supportive Not supportive
USA Gaynor et al.
(2005)
Decision making
by corporate board
members
Investors recommend joint provision of auditing
and NAS if audit quality improves
Audit committee members are reluctant to recommend
joint provision when public disclosures are required, even
at the expense of audit quality improvement
Denmark Quick &
Warming-
Rasmussen
(2005)
Stakeholder
perceptions
Shareholders, bank loan officers, and journalists
perceive a negative effect on auditor indepen-
dence if non-audit services are provided
USA Khurana &
Raman (2006)
Investors perceptions
of financial reporting
credibility
Non-audit and total fees are perceived negatively by in-
vestors; to the extent that investors perceive total fees
negatively, recent regulatory initiatives to limit non-audit

fees may not adequately address the perceived threat
Measure of auditor independence in appearance: association between client’s earnings and capital market reactions
USA Dopuch et al.
(2003)
Asset price reaction They find that asset prices reflect participants’ be-
liefs with non-audit services resulting in lower
prices, even when the beliefs are incorrect
USA Davis &
Hollie (2008)
Asset price reaction Examine the effect of various fee ratios on in-
vestor perception of auditor independence and
market behavior over time; consistent with prior
findings, they report decreasing asset prices
USA Church &
Zhang (2005)
Assessments of asset
value and willingness
to sue the auditor
Investors’ willingness to sue the auditor increases
when non-audit services are provided
Assessments of asset value are not affected by non-audit
services
USA Frankel et al.
(2002)
Earnings surprises and
discretionary accruals
Non-audit fees affect small earnings surprises
USA Ashbaugh et al.
(2003)
Abnormal returns Fees generated from non-audit services do not affect in-

vestors’ perceptions of auditor independence
USA Francis and
Ke (2005)
Market valuation of
quarterly earnings
Market valuation of quarterly earnings, following
the initial proxy statement fee disclosure, is sig-
nificantly lower for firms with high non-audit fees
than for firms with low non-audit fees
13
230 C. Pott et al.
Table 2 continued
Jurisdiction Author(s) (Year) Measure Results
Supportive Not supportive
USA Krishnan et al.
(2005)
Earnings response co-
efficient
Negative association between non-audit fees and
the earnings response coefficient
Measure of auditor independence in appearance: behavior related to the choice of auditor
USA Raghunandan
(2003)
Proportion of share-
holder votes against
auditor ratification
Proportion of shareholder votes against auditor
ratification is positively associated with the rela-
tive magnitude of the non-audit fee ratio
13

Review of empirical research on rotation and non-audit services 231
earnings-response coefficients, when the level of non-audit services acquired from
industry specialist auditors was compared to that acquired from non-specialist au-
ditors. DeFond et al. (2002) found no evidence that non-audit service fees impair
auditor independence, which is assumed to be caused by an auditor’s propensity to
issue going-concern audit opinions. Basioudis, Papakonstantinou, and Geiger (2008)
studied audit reports provided to financially stressed companies in the United King-
dom and the magnitude of audit and non-audit service fees paid to the company’s
auditors. They found that the magnitude of both audit fees and non-audit fees are
significantly associated with the issuance of a going-concern modified audit opinion.
In particular, financially stressed companies with high audit fees are more likely to
receive a going-concern modified audit opinion, whereas companies with high non-
audit fees are less likely to receive a going-concern modified audit opinion.
Hay, Knechel, and Li (2006) examined evidence in New Zealand about whether
auditors providing more non-audit services are less independent.The results of the lo-
gistic regression test of the relation between audit opinion and the level of non-audit
fees show that there is no significant relation between audit qualification or modifi-
cation and non-audit fees. This does not indicate a lack of independence. Addition-
ally, Sharma (2001) and Sharma and Sidhu (2001) examined a sample of forty-nine
bankrupt Australian companies and conclude that higher NAS fees were associated
with a lower likelihood of receiving going-concern modified reports. Firth (2002),
in contrast, found that high non-audit fees were associated with clean audit reports.
Similar, Craswell, Stokes, and Laughton (2002) examined the relationship between
qualified opinions and fee dependence (defined as the ratio of fees paid by a particu-
lar client to the audit firm’s total national and local office fees), and concluded that fee
dependence does not affect the propensity to issue unqualified opinions.
Kinney, Palmrose, and Scholz (2004) discovered no consistent evidence of a pos-
itive association between audit firm fees and financial information system design, or
between implementation of internal audit services and restatements of financial data;
they assumed that in case of an restatement the auditor was lax in enforcing generally

accepted accounting principles (GAAP).
We also discuss studies that examine whether the auditor’s provision of non-audit
services affects the assessments of stakeholders, reflecting independence in appear-
ance. Jenkins and Krawczyk (2003) surveyed the public as a proxy for investors, and
assessed public perceptions of auditor independence for six different non-audit ser-
vice types. They found that the public’s perceptions were not negatively affected by
any of the non-audit services included in their study. Opposing these findings for in-
ternal audit services and merger and acquisition services, Mauldin (2003) found that
professional investors perceive independence to be impaired when such services are
provided. Using corporate board members – who are placed in the position of an au-
dit committee member or investor – Gaynor, McDaniel, and Neal (2005) reported that
both groups were more likely to recommend the joint provision of auditing and non-
audit services when the audit quality improves. However, when public disclosures are
required, even at the expense of improved audit quality, audit committee members are
reluctant to recommend a joint provision, whereas investors are not.
Quick and Warming-Rasmussen (2005) provided evidence from Denmark. They
found that shareholders, bank loan officers and journalists perceived a negative effect
13
232 C. Pott et al.
on auditor independence if additional non-audit services were provided. Furthermore,
they showed that the type of non-audit service affected the degree to which auditor
independence was perceived to be impaired. Khurana and Raman (2006) report that
non-audit and total fees were perceived negatively by investors; to the extent that in-
vestors perceive total fees negatively, recent regulatory initiatives to limit non-audit
fees may not adequately address the perceived threat.
Some studies investigate independence in appearance, proxied by the associa-
tion between the client’s earnings and capital market reactions. Dopuch, King, and
Schwartz (2003) examined how the auditor’s provision of non-audit services affected
asset prices in an experimental double-auction institution. They found that asset
prices reflected participants’ beliefs, with non-audit services resulting in lower prices,

even when the beliefs are incorrect. Extending Dopuch et al. (2003), Davis and Hol-
lie (2008) examined the effects of various fee ratios on investor perceptions of auditor
independence and market behavior over time. Consistent with prior findings, they
reported decreasing asset prices. The effect of non-audit services on investors’ assess-
ment of asset value and willingness to sue the auditor was studied by Church and
Zhang (2005). The authors found that assessments of asset value were not affected
by non-audit services, though investors’ willingness to sue the auditor increased when
non-audit services were provided.
The relationship between auditor fees and market reactions – and between auditor
fees and share values on the date the fees were disclosed – were further examined by
Frankel, Johnson, and Nelson (2002). The analysis of data collected from proxy state-
ments showed evidence that non-audit fees positively affected earnings surprises. The
opposite was shown for audit fees, as well as between non-audit fees and share values
on the date the fees were disclosed. In contrast, Ashbaugh, LaFond, and Mayhew
(2003) investigated the association between auditor fees and firms that met analysts’
forecasts. No evidence was discovered that auditors violated their independence due
to clients’ purchases of additional non-audit services.
Francis and Ke (2005) found that the market valuation of quarterly earnings – fol-
lowing the initial proxy statement fee disclosure – was significantly lower for firms
with high non-audit fees than for firms with low non-audit fees. They found a 17% re-
duction in market valuation following the mandated fee disclosures for median firms
with high non-audit fees. Also suggesting that independence in appearance may be
affected, Krishnan, Sami, and Zhang (2005) reported a negative association between
non-audit fees and the earnings response coefficient.
Studying shareholder votes as a behavior related to the choice of auditor, Raghu-
nandan (2003) found that the proportion of shareholder votes against auditor rati-
fication was positively associated with the relative magnitude of the non-audit fee
ratio.
4.2.2 Implications
Research investigating different users’ perceptions has found mixed results on the

question of whether non-audit services should be provided for an annual statement
audit client. Studies indicate that the position of the user, e. g., professional in-
vestor, non-professional investor, or corporate board member, determines how the
13
Review of empirical research on rotation and non-audit services 233
user perceives auditor independence non-audit services are provided. Furthermore,
their findings are associated with the legal environments, which differ in many coun-
tries. For example, the amounts spent for non-audit services might differ, or liability
exposure could affect perceptions of auditor independence. Thus, overall results
are not generally valid and are highly dependent on what sample is used, and in
all probability on how the research instrument is constructed. Therefore, research
on auditor independence perceptions related to non-audit services has to be care-
fully interpreted. Overall non-audit service fees can impair auditor independence in
appearance.
In comparison, research on independence in fact finds mixed results on impair-
ment from non-audit services. In particular, the research does not support a negative
effect of non-audit services on earnings management effects. Earnings management
measures are often noisy and may capture many other firm-specific effects, facts that
may determine the size of non-audit service fees. Furthermore, studies differ in how
non-audit service fees are measured (e. g., ratio, total fees, or individual non-audit ser-
vice fees). Overall, the research on non-audit services and auditor independence in
fact does not indicate a benefit to prohibiting the provision of non-audit services for
a financial statement audit client.
Different studies using audit opinion issued as a proxy for auditor independence
draw different conclusions. Some studies argue that a decrease in the propensity that
a going-concern audit opinion is issued together with an increase in non-audit fees
implies that the auditor is less independent. Other researchers conclude that this nega-
tive association implies cleaner audit reports, which can be interpreted as evidence
for a more independent auditor. Firth (2002) provides two possible reasons for the
negative association between the propensity to receive a going-concern audit opinion

and non-audit service fees: (1) A lack of auditor independence, or (2) the possibil-
ity that the consultancy services clear up uncertainties and disagreements prior to the
audit. In this respect, clean conclusions cannot be drawn as to whether auditor inde-
pendence is impaired. Thus, future research needs to focus on ways to disentangle the
two possible theoretical explanations on the operational level.
The behavior of market participants does not imply impaired independence in
fact, but market participants assess the existence of non-audit service fees as a sign
of impaired independence in appearance. This leads to the conclusion that mitigat-
ing factors not captured in perception research might affect individuals’ assessments,
e. g., audit committee composition or ownership distribution. The existence of nega-
tive perceptions cannot be ignored, since negative market effects can also be derived
from users’ perceptions. Thus, banning the provision of non-audit services might be
an adequate measure to increase user confidence in auditors.
The European approach leaves significant room for discretion to Member States.
Instead of implementing a strict rules-based approach, e. g., a list of non-audit ser-
vices that are prohibited from being provided to a financial statement audit client, the
8th Directive currently outlines a principles-based approach. In any case, to improve
the transparency of the audit process, EU public interest entities are required to dis-
close fees paid to audit firms. This includes fees for audit services as well as fees for
non-audit services. In addition, audit firms have to disclose threats to their indepen-
dence and confirm their independence in writing. Furthermore, Member States are
13