casterella and johnston - 2013 - can the academic literature contribute to the debate over mandatory audit firm rotation [mafr]

Research Report
Can the academic literature contribute to the debate over
mandatory audit firm rotation?
Jeffrey R. Casterella
a,b,
⇑
, Derek Johnston
a
a
Colorado State University, Fort Collins, CO, USA
b
University of Auckland, Auckland, New Zealand
article info
Article history:
Available online 11 January 2013
Keywords:
Audit firm rotation
Independence
Audit quality
abstract
Recently, the Public Company Accounting Oversight Board (PCAOB) issued a concept
release soliciting public recommendations to improve auditor independence and audit
quality (PCAOB, 2011). The focus of the release is on mandatory audit firm rotation (MAFR)
with a request for commentaries addressing the advantages and disadvantages of MAFR. In
this paper, we briefly summarize the recent literature on mandatory audit firm rotation
and suggest how it can be useful to regulators as they consider the implementation of man-
datory rotation. We find that the conclusions reached about the possible effectiveness of
MAFR appear to depend on the type of data used (voluntary vs. mandatory auditor
changes), suggesting that regulators should exercise care when drawing inferences from
past audit firm rotation research.
Ó 2012 Elsevier Ltd. All rights reserved.

Introduction
Recently, the Public Company Accounting Oversight
Board (PCAOB) issued a concept release soliciting public
recommendations to improve auditor independence and
audit quality (PCAOB, 2011). The focus of the release is
on mandatory audit firm rotation (MAFR) and the PCAOB
requests commentaries addressing the advantages and dis-
advantages of MAFR.
1
In March of 2012, the PCAOB con-
ducted 2 days of hearings on the pros and cons of MAFR.
The hearings featured several former regulators who ad-
dressed the costs and benefits of MAFR. There was little con-
sensus. While former Federal Reserve chairman Paul Volcker
said his experience ‘‘does suggest to me the importance of
requiring rotation’’ (Tysiac, 2012), former SEC chairman
Breeden seemed less convinced saying that ‘‘mandatory
rotation would benefit some companies and it would prob-
ably harm others’’ (Chasen, 2012). Charles Bowsher, former
U.S. comptroller general, suggested a limited MAFR arrange-
ment that would apply to only the largest 25 or 40 publicly
traded companies. Finally, former SEC chairman Harvey Pitt
expressed concern that with MAFR, ‘‘the cure could turn out
to be worse than the disease, depending on the amount of
time people would be required to rotate off’’ (Cohn, 2012).
Given the opposing views on MAFR, the purpose of this pa-
per is to provide a critical summary of recent research and
suggest how it might be useful to regulators as they consider
the implementation of mandatory rotation in the U.S.
In an extensive review of the research examining the

causes and consequences of auditor switching, Stefaniak,
Robertson, and Houston (2009) note that most studies look
at the association of auditor tenure and various measures
of audit quality. In general, they find that audit quality is
1052-0457/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
/>⇑
Corresponding author at: College of Business, 253 Rockwell Hall,
Colorado State University, Fort Collins, CO 80523, USA. Tel.: +1 970 217
0947.
E-mail address: (J.R. Casterella).
1
This represents the second time in 10 years that regulators at the
federal level have formally considered the implementation of MAFR. In
2003, the U.S. General Accounting Office (GAO) concluded that the
Securities and Exchange Commission (SEC) and the PCAOB should not
consider mandating audit firm rotation until the full effects of the
Sarbanes–Oxley Act of 2002 could be assessed. The GAO report also noted
that, in the future, the PCAOB will likely need to ‘‘ evaluate whether
further enhancements or revisions, including mandatory audit firm rota-
tion, may be needed to further protect the interest and to restore investor
confidence’’ (GAO, 2003, p. 5).
Research in Accounting Regulation 25 (2013) 108–116
Contents lists available at SciVerse ScienceDirect
Research in Accounting Regulation
journal homepage: www.elsevier.com/locate/racreg
higher when there is a longer auditor–client relationship, a
finding that would seem to mitigate against a policy of
MAFR (DeFond & Francis, 2005; Stefaniak et al., 2009).
The problem with this inference is that most of these stud-
ies use data from a regulatory regime in which changing

auditors is voluntary (as is currently the case in the United
States). As a result, it is unclear that these results would
extend to a regulatory regime where audit firm rotation
is mandatory.
2
In this paper we compare studies that are based on vol-
untary auditor changes with those that are based on man-
datory or quasi-mandatory auditor changes
3
since it
appears that the PCAOB is especially interested in studies
involving the latter.
4
We find that the conclusions reached
about the possible effectiveness of MAFR depend on the type
of data used, suggesting that policy makers exercise caution
when drawing inferences from academic research.
The remainder of the paper is organized as follows. In
the next section, we provide a brief summary of the MAFR
debate, and discuss how the academic community has re-
sponded to calls for research concerning the potential ben-
efits of MAFR. In Section ‘Evidence on mandatory audit
firm rotation’, we briefly summarize the evidence from
the academic research, with a focus on differentiating
those academic studies that use voluntary changers data
from those that employ mandatory rotators data. Sec-
tion ‘Summary and discussion’ concludes with a discussion
of the implications that these two subsets of the audit firm
rotation research have on the MAFR debate.
The debate and the academic response

The arguments for and against mandatory audit firm
rotation are well-documented.
5
Proponents of MAFR be-
lieve that long-tenure auditor–client relationships increase
the likelihood of audit failures. This belief is based on the
assumption that a long-tenure relationship leads to an in-
creased level of familiarity between the two parties that nat-
urally erodes auditor independence and professional
skepticism.
6
Hence, proponents believe that audit firms
should be required to roll-off engagements after a fixed per-
iod of time. Such a policy would prevent the existence of any
long-tenure auditor–client relationships which would, they
argue, decrease the number of audit failures. Conversely,
opponents of MAFR point out that audit firms gain valuable
knowledge about their clients over time. As a result, oppo-
nents express concern that a mandatory rotation policy
would result in a lack of client-specific knowledge, causing
short-tenure auditor–client relationships to be more prone
to audit failures.
Table 1 summarizes the various types of auditor–client
relationships that would exist with and without a MAFR
policy. In particular, the table describes audits in terms of
mandatory rotation regime (yes/no) and auditor tenure
(short/medium/long). For purposes of this table, we define
short-tenure relationships as those between 1 and 3 years,
medium between 4 and 6 years, and long as 7 years or
longer.

7
Since the U.S. does not have a MAFR requirement,
auditor–client relationships in the U.S. can be either short,
medium, or long (cells 3, 4 or 5). Conversely, in a regime
with mandatory rotation, no long-term auditor–client rela-
tionships would exist, leaving only short and medium rela-
tionships (cells 1 and 2).
If the United States were to adopt a rotation policy,
long-tenure audits (cell 5) would be forced into short-ten-
ure audits (cell 1). Proponents argue that such a shift
would result in higher audit quality due to enhanced audi-
tor independence. In contrast, opponents believe that audit
quality would decline because the new auditor would lack
experience with the client. Put simply, proponents of
MAFR argue that short-tenure auditor–client relationships
will result in higher quality audits, while opponents of
MAFR argue that short-tenure auditor–client relationships
will result in lower quality audits.
The goal of research in this area is to provide evidence
regarding the potential effect of mandatory rotation on
the quality of audits conducted on U.S. companies. Ideally,
one would compare measures of audit quality from cell 5
audits with those from cell 1 audits. Because the U.S. does
not have a mandatory rotation policy in place, the neces-
sary data from cell 1 audits for U.S. public companies is
not available. To circumvent this data issue, most studies
substitute cell 3 data for cell 1 data. Generally speaking,
these studies analyze the differences in various measures
of audit quality between cell 5 and cell 3, and find that
the quality of cell 5 audits are, on average, significantly

better than the quality of cell 3 audits. Based on that evi-
dence alone, one would conclude that a mandatory rota-
tion policy would have a negative effect on audit quality,
resulting in a higher occurrence of audit failures.
This is where the data is issue is important. By replacing
cell 1 audits with cell 3 audits, these studies rely on the
assumption that the characteristics of companies with
auditor relationships that fall into cell 3 are similar to
those that would have short-term auditor relationships in
Table 1
Types of auditor–client relationships by mandatory rotation regime.
Mandatory
rotation
regime?
Short-tenure
(e.g., 63 years)
Medium-tenure
(e.g., 4–6 years)
Long-tenure
(e.g., P7 years)
Yes Cell 1 Cell 2 Not applicable
No Cell 3 Cell 4 Cell 5
2
In fact, the PCAOB points out this limitation of the academic literature
on p. 16 of its concept release.
3
We use the term ‘‘quasi-mandatory data’’ to characterize the data
gathered from controlled experiments that are designed to mimic a MAFR
environment.
4

Throughout the manuscript, we refer to new audits created in a MAFR
regime as ‘‘rotators’’ and new audits stemming from voluntary switching
decisions as ‘‘changers’’.
5
See Stefaniak et al. (2009) for a review of the advantages and
disadvantages of MAFR.
6
See Previts (1998) for a review of auditor independence, its origins and
the potential for auditor conflicts of interest. In addition, Kleinman, Palmon,
and Anandarajan (1998) provide an extensive review of studies that
investigate auditor independence.
7
Of course, the specific number of years used by the regulator to classify
each relationship could be different from those that we use. Related to this
point, the PCAOB asks for input regarding the ‘‘appropriate term length’’ for
rotation in its concept release.
J.R. Casterella, D. Johnston /Research in Accounting Regulation 25 (2013) 108–116
109
a mandatory rotation regime. Since cell 3 audits represent
new auditor–client relationships that have been created
voluntarily, rather than via a mandatory rotation policy,
this assumption is likely invalid. For example, Walker, Le-
wis, and Casterella (2001) show that companies that vol-
untarily change auditors differ significantly from a
random sample of companies. More specifically, their anal-
ysis suggests that voluntary changers are more likely to
have fraud and be financially distressed. Moreover, Carey,
Geiger, and O’Connell (2008) find that companies are more
apt to voluntarily change auditors following qualified audit
opinions.

8
Given the ample empirical evidence that suggests
that companies that voluntarily change auditors are already
‘troubled’ companies, it is not surprising that prior research
finds that the quality of cell 5 audits is higher than cell 3
audits, even after attempting to control for the determinants
of the decision to voluntarily change auditors. For this rea-
son, Carcello and Nagy (2004) caution that it is unclear
whether the empirical results obtained from studies based
on voluntary changers would extend to a MAFR setting.
In summary, the MAFR literature seems to rely heavily
on the use of cell 3 audits (i.e., voluntary changers); never-
theless, some studies exploit settings that may be more
representative of a mandatory rotation regime. In the next
section, we briefly summarize the findings of prior rotation
research, with a focus on differentiating studies based on
the type of data employed.
Evidence on mandatory audit firm rotation
The Appendix summarizes, in chronological order, the
key findings from 24 audit firm rotation studies from the
year 2001 to present. For brevity’s sake, we limit the
Appendix to empirical archival and experimental research
studies published in the top 25 accounting journals as
determined by Glover, Prawitt, and Wood (2006).
9
For each
study, the Appendix provides information on the relation
examined, sample used, and primary results.
DeFond and Francis (2005) and Stefaniak et al. (2009)
conclude that the majority of extant research does not pro-

vide evidence to support the introduction of MAFR in the
U.S. as a means to enhance audit quality. One can reach a
similar conclusion based on a cursory review of the Appen-
dix. In 16 of the 24 studies, there is no evidence to support
the introduction of mandatory rotation. On the contrary,
they bolster the argument that new audits may be prob-
lematic because of a learning curve effect. Most of these
studies estimate regression models that associate audit
tenure with various measures of audit quality using data
from regimes where changing auditors is a voluntary,
rather than a mandatory, action. In such settings, the inclu-
sion of a variable such as auditor tenure likely introduces
self-selection bias into the model. Although there are
instrumental variable methods that one can use to correct
for this issue, they require assumptions about the data that
may be unrealistic (Larcker & Rusticus, 2010).
We strongly suggest that policy makers recognize
explicitly the difference between voluntary and mandatory
auditor changes. To this end, in Section ‘Evidence from
studies using voluntary changers data’, we briefly summa-
rize the key findings from the 13 studies that rely on data
from voluntary changers to answer their research ques-
tions. In Section ‘Evidence from studies using mandatory
rotators data’, we review the 11 studies that examine the
potential benefits of audit firm rotation using mandatory
rotators data.
Evidence from studies using voluntary changers data
Panel A of Table 2 lists the 13 research studies that
investigate the potential benefits of MAFR using voluntary
changers (cell 3) data. All of the studies are archival (rather

than experimental), and most of these studies regress var-
ious measures of audit quality on auditor tenure and a host
of control variables. For example, Geiger and Raghunandan
(2002) examine the association of auditor tenure with
audit quality using a sample of 117 stressed companies
that eventually filed for bankruptcy. After controlling for
factors such as firm size and the probability of bankruptcy,
they find that as auditor tenure increases, the probability
of an audit failure decreases. In addition, Johnson, Khurana,
and Reynolds (2002) investigate the relation between
auditor tenure and two measures of financial reporting
quality (the absolute value of unexpected accruals and
the persistence of accruals), and find that companies in
short-term relationships (defined as 2–3 years) with their
auditors have lower quality financial reports relative to
companies that have been with their auditor for 4–8 years.
Similarly, Myers, Myers, and Omer (2003) document a neg-
ative association between auditor tenure and the absolute
value of discretionary accruals using a sample of U.S. com-
panies, a finding that was later confirmed by Chen, Lin, and
Lin (2008) based on a sample of Taiwanese companies. Fi-
nally, prior research using voluntary changers data finds
that auditor tenure is: (a) negatively related to fraudulent
financial reporting and firm risk (Carcello & Nagy, 2004;
Crabtree, Brandon, & Maher, 2006; Mansi, Maxwell, & Mill-
er, 2004); and (b) positively associated with investors’ per-
ceptions of earnings quality and conservative financial
reporting (Ghosh & Moon, 2005; Jenkins & Velury, 2008).
Of the 13 studies that use voluntary changers data, only
Li (2010) and Chi, Ling Lei, and Mikhail (2011) provide

modest support for a move to MAFR. Specifically, using
Big N audits over the period 1980–2004, Li (2010) finds
that the reporting of conservative earnings is adversely af-
fected by long-tenure audits of small companies and
weekly monitored companies. Moreover, Chi et al. (2011)
examine, among other things, the association of auditor
tenure with earnings management, and find conflicting re-
sults. Although they find that auditor tenure is negatively
related to accrual earnings management, they document
a positive association between auditor tenure and real
earnings management. The latter result provides some
support for mandatory audit firm rotation.
8
See Stefaniak et al. (2009) for a detailed discussion of the studies that
explore the determinants of firms’ decisions to change audit firms.
9
In addition we include published work from two journals dedicated to
auditing and/or the regulation of auditing and accounting: Research in
Accounting Regulation and International Journal of Auditing. We also
include two recent working papers that specifically focus on MAFR.
110 J.R. Casterella, D. Johnston /Research in Accounting Regulation 25 (2013) 108–116
Interestingly, of the 13 papers summarized in Panel A,
only Li (2010) appears to take steps to explicitly control
for the potential self-selection bias that may arise through
the use of voluntary changers data. Specifically, Li (2010, p.
239) points out that he ‘‘ controls for an endogeneity
problem possibly arising from the auditor tenure variable
which is often ignored in previous studies.’’
10
The PCAOB

has voiced concern about studies that rely on samples of
companies that have voluntarily changed auditors, perhaps
because of the thorny econometric issues that can arise
when using such data. With this in mind, we next discuss
prior research that exploits settings that are more consistent
with a regime of mandatory rotation.
Evidence from studies using mandatory rotators data
Panel B of Table 2 lists the 11 studies that use manda-
tory rotators data (cell 1). Of the 11 studies listed, eight
provide support for MAFR. As Panel B indicates, five of
the studies use archival data, while six use experimental
data. We discuss the archival studies first, and then turn
our attention to the experimental studies.
The demise of Arthur Andersen (AA) created a unique
setting to study the effects of audit firm rotation. Although
not a perfect analogy to short-term audits that would be
obtained in a MAFR regime (i.e., cell 1 audits), we believe
that the use of former AA clients is likely more representa-
tive of mandatory rotators than are voluntary changers.
Three of the five archival studies listed in Panel B use the
AA setting to investigate the effect of rotation on audit
quality, and provide mixed results. In particular, Blouin,
Grein, and Rountree (2007) focus on former AA clients that
had extreme discretionary accruals while with AA. Among
other things, they find that moving to a new audit firm did
not significantly curtail extreme income-increasing or in-
come-decreasing discretionary accruals. In contrast, the
two other studies that use AA data provide evidence that
lends some support to MAFR. For example, Nagy (2005)
finds a significant increase in the audit quality of smaller

AA clients that were forced to find new auditors. Also, Kea-
ley, Lee, and Stein (2007) documents a positive association
between the number of years that clients were with AA
and the audit fee charged by the new auditor. This result
suggests that audit firms perceive a new client coming
from a long-term relationship with its previous auditor
to be riskier than one coming from a short-term relation-
ship with its predecessor auditor.
The remaining two archival studies use true mandatory
audit firm rotation data, and once again, provide mixed re-
sults. Lowensohn, Reck, Casterella, and Lewis (2011) exam-
ine the Florida government audit environment in which
there exists: (a) both rotation and non-rotation regimes;
and (b) an independent measure of the joint quality of
the audit and of the financial statements of the reporting
entity. They find that, after controlling for other factors
Table 2
Mandatory audit firm rotation studies partitioned by setting.
Authors and publication date Method U.S. or international data? Support for MAFR?
Panel A: Studies using voluntary changers (cell 3) data
Geiger and Raghunandan (2002) Archival U.S. No
Johnson et al. (2002) Archival U.S. No
Myers et al. (2003) Archival U.S. No
Mansi et al. (2004) Archival U.S. No
Carcello and Nagy (2004) Archival U.S. No
Ghosh and Moon (2005) Archival U.S. No
Crabtree et al. (2006) Archival U.S. No
Chen et al. (2008) Archival Taiwan No
Jenkins and Velury (2008) Archival U.S. No
Davis et al. (2009) Archival U.S. No

Ruiz-Barbadillo, Gómez-Aguilar, and Carrera (2009)
a
Archival Spain No
Li (2010) Archival U.S. Yes
Chi et al. (2011) Archival U.S. Yes
b
Panel B: Studies using mandatory rotators (cell 1) data
Dopuch et al. (2001) Experimental N/A Yes
Nagy (2005) Archival US Yes
Arel et al. (2006) Experimental N/A Yes
Jennings et al. (2006) Experimental N/A Yes
Blouin et al. (2007) Archival US No
Kealey et al. (2007) Archival US Yes
Kaplan and Mauldin (2008) Experimental N/A No
Wang and Tuttle (2009) Experimental N/A Yes
Daniels and Booker (2011) Experimental N/A Yes
c
Kwon et al. (2011) Archival Korea No
Lowensohn et al. (working paper) Archival US Yes
a
Since the mandatory rotation policy in Spain was never enforced, we consider the Ruiz-Barbadillo study as being conducted in a voluntary setting.
b
To be more precise, Chi et al. (2011) find conflicting results. On the one hand, they find that auditor tenure is negatively associated with abnormal
accruals. Conversely, they show that auditor tenure is positively related to real earnings management.
c
Daniels and Booker (2011) report inconsistent results. On the one hand, they find that the perception of independence is higher with MAFR. Conversely,
they show that perceptions of audit quality are neither higher, nor lower with MAFR.
10
Myers et al. (2003) and Davis, Soo, and Trompeter (2009) attempt to
alleviate concerns about endogeneity with the removal of ‘‘quick-auditor-

changer’’ companies from their analyses. ‘‘Quick-auditor-changers’’ are
auditor–client relationships that last less than 5 years.
J.R. Casterella, D. Johnston /Research in Accounting Regulation 25 (2013) 108–116
111
believed to be associated with reporting quality, govern-
ments that rotate issue higher quality reports. Kwon, Lim,
and Simnett (2010) investigate the impact of rotation on
audit quality in Korea, which began requiring audit firm
rotation in 2006. Their analysis suggests that while audit
fees increased upon adoption of the mandatory rotation
policy, the implementation did not have a statistically sig-
nificant effect on audit quality.
Due to their controlled nature, experimental studies can
be particularly useful when it comes to examining the ef-
fect that mandatory rotation may have on auditor indepen-
dence and audit quality. Of the six experimental studies
listed in Panel B, three investigate whether perceptions of
auditor independence are incrementally influenced by
MAFR. Specifically, Jennings, Pany, and Reckers (2006) used
49 professional judges and find that auditor independence
perceptions are enhanced in a MAFR regime compared to
a regime that only rotates audit partners. Similarly, Daniels
and Booker (2011) surveyed bank loan officers and find that
MAFR improves the perception of auditor independence
even though MAFR does not affect their perception of audit
quality. In contrast, Kaplan and Mauldin (2008) show that
non-professional investors’ perceptions of independence
are not significantly enhanced in a MAFR regime (compared
to a regime that only rotates audit partners).
The remaining three experimental studies examine the

impact of mandatory rotation on the quality of the subse-
quent audits. For example, Arel, Brody, and Pany (2006)
find that auditors in a MAFR regime are more likely to issue
modified audit opinions when a departure from GAAP is
present. Likewise, Dopuch, King, and Schwartz (2001) find
that participants in a MAFR condition are less likely to is-
sue biased reports. Finally, Wang and Tuttle (2009) inves-
tigate how MAFR affects auditor–client negotiations.
Specifically, they find that in a mandatory rotation regime
auditors adopt less cooperative negotiation strategies and
are more likely to adopt ‘‘obliging’’ strategies in situations
when MAFR does not exist.
Summary and discussion
Table 3 employs a simple 2 Â 2 contingency table to
summarize the aforementioned 24 studies based on: (a)
whether the key findings provide support for or against a
mandatory rotation policy; and (b) whether the study re-
lied on mandatory rotators (cell 1) or voluntary changers
(cell 3) data. Table 3 reveals that eight of the 11 studies
conducted in a mandatory rotation setting provide evi-
dence that lends support to a MAFR policy. In stark con-
trast, only two of the 13 studies using voluntary changers
data provide support for mandatory audit firm rotation.
Crude statistical analysis consisting of a chi-square test
for independence reveals that support for MAFR depends
on whether voluntary or mandatory data was used in the
study (
v
2
= 8.07, p-value < 0.01). Further, a simple odds-ra-

tio computation reveals that that studies conducted in
mandatory settings are 14.661 times more likely to find
support for MAFR. For numerous reasons, including the
small sample size and lack of control variables, we realize
that this analysis is far from rigorous.
11
Nevertheless, it
does lend some support to the concern expressed by the
PCAOB and some academics regarding whether the empiri-
cal results obtained from studies based on voluntary chang-
ers would extend to a MAFR setting.
The purpose of this manuscript is to provide a critical
summary of recent mandatory audit firm rotation research
and suggest how it might be useful to regulators as they
consider the implementation of mandatory rotation in
the U.S. While Stefaniak et al. (2009) conclude that ‘‘a
majority of the evidence does not support the implementa-
tion of a MAFR regime in the U.S.’’ (p. 108, Section 5.2.5),
we show that if one separates the literature into two sub-
sets, the results found in the mandatory settings are much
more supportive of MAFR. In general, the first subset
examines the effect of auditor tenure on audit quality in
settings where changing auditors is a voluntary, rather
than mandatory, action. More often than not, the conclu-
sion from this stream of the literature is that short-tenure
audits have inferior audit quality compared to long-tenure
audits. However, it is questionable whether the findings of
this stream of literature provide much useful guidance to
the debate on MAFR because these studies rely on samples
of short tenure audits that are not representative of com-

panies that would be forced to rotate under a MAFR
regime.
The second subset of the audit firm rotation literature
also examines the association between auditor tenure
and audit quality; however, these studies deliberately
avoid the use of voluntary short-tenure audits and, instead,
use short-tenure audits more likely to represent that which
would exist under a mandatory rotation regime. The
majority of the studies from this subset of the literature
are supportive of a move to MAFR, which is in stark con-
trast to the conclusions reached from the studies that rely
on voluntary changers data. Consequently, it appears that
the conclusions reached about the possible effectiveness
of MAFR appear to depend on the type of data (voluntary
vs. mandatory) used. For this reason, we believe regulators
should focus heavily on the type of data used in past audit
Table 3
Summary 2 Â 2 contingency table. Support for MAFR by setting.
Setting
Support for
MAFR?
Mandatory (use
cells 1 and 2)
Voluntary (use cells
3, 4, and 5)
Totals
Yes 8 2 10
No 3 11 14
Totals 11 13 24
11

We also estimate a logistic regression to assess whether studies
conducted in mandatory settings are more apt to find results that support
MAFR than studies conducted in voluntary settings. Our dependent variable
in the model is SUPPORT, which is a dichotomous variable that equals one if
the study’s key findings support MAFR, and zero otherwise. The three
independent variables in the model consist of: (1) MANDATORY, which
assumes the value of one if the study used mandatory rotators data, and
zero otherwise; (2) LOG_SIZE, which is the natural logarithm of the sample
size on which the key findings in the study are based; and (3) ARCHIVAL,
which equals one if the study was archival in nature, and zero otherwise.
The results (not reported) reveal that MANDATORY is positively related to
SUPPORT (p-value < 0.05).
112 J.R. Casterella, D. Johnston /Research in Accounting Regulation 25 (2013) 108–116
firm rotation research before extrapolating such research
to policy making decisions. In addition, future MAFR re-
search should avoid the use of voluntary changers that
are known, ex ante, to be different from a random sample
of companies and instead focus on settings that are more
indicative of a regime with a MAFR policy.
Acknowledgement
We appreciate the guidance and input from Barry Lewis
and Robert Knechel. We also appreciate the helpful com-
ments from Ken Bills and the research assistance of
Stephen O’Dorisio.
Appendix A. A summary of the audit firm rotation literature from 2001 to present.
Authors and date Relation examined Sample Primary results Support
for
MAFR?
Dopuch et al. (2001) Audit firm tenure and
willingness to issue biased

reports
An experiment with
72 manager subjects
and 72 auditor
subjects
Auditor independence
is higher in the
presence of
mandatory audit firm
rotation
Yes
Geiger and Raghunandan
(2002)
Audit firm tenure and
audit reporting failures
US audits of soon-to-
be-bankrupt
companies between
1996 and 1998
Longer tenure is
negatively associated
with audit failures
No
Johnson et al. (2002) Audit firm tenure and
financial reporting quality
US audits between
1986 and 1995
Longer tenure does
not reduce financial
reporting quality

No
Myers et al. (2003) Audit firm tenure and
earnings quality
US audits between
1988 and 2000
Longer tenure does
not reduce audit
quality or earnings
quality
No
Carcello and Nagy (2004) Audit firm tenure and
AAERs
US audits between
1988 and 2000
Longer tenure is not
associated with more
AAERs
No
Mansi et al. (2004) Audit firm tenure and the
cost of debt financing
US reporting years
between 1974 and
1998
Longer tenure is
associated with lower
cost of debt financing.
Therefore, mandatory
audit firm rotation
may be viewed by the
capital markets as

unbeneficial
No
Ghosh and Moon (2005) Audit firm tenure and
investor perceptions of
earnings quality
US audits between
1990 and 2000
Longer tenure does
not reduce investor
perceptions of
earnings quality
No
Nagy (2005) Forced audit firm change
and audit quality
Arthur Andersen
client audits in 2000
and 2001 compared to
former Arthur
Andersen client audits
performed by the Big
N in 2002 and 2003
Discretionary accruals
are lower for smaller,
ex-Arthur Andersen,
clients which provides
some support for
mandatory audit firm
rotation
Yes
Arel et al. (2006) Audit firm rotation and

modified audit opinions
An experiment with
105 CPA firm auditors
Mandatory audit firm
rotation is associated
with more modified
audit opinions
Yes
Crabtree et al. (2006) Audit firm tenure and
bond ratings for newly
issued bonds
New US debt issues
between 1990 and
2002
Longer tenure does
not decrease bond
ratings for newly
issued bonds
No
(continued on next page)
J.R. Casterella, D. Johnston /Research in Accounting Regulation 25 (2013) 108–116
113
Appendix A (continued)
Authors and date Relation examined Sample Primary results Support
for
MAFR?
Jennings et al. (2006) Audit firm rotation and
the perception of auditor
independence
An experiment with

49 National Judicial
College judges
Perceptions of auditor
independence are
enhanced with
mandatory audit firm
rotation
Yes
Blouin et al. (2007) Forced audit firm change
and financial reporting
quality
US audits in 2002 of
former Arthur
Andersen clients
Financial reporting
quality did not
improve for Arthur
Andersen clients that
were forced into new
[shorter] tenure audits
No
Kealey et al. (2007) Prior audit firm tenure
and audit fees paid to
successor auditors
US audits in 2002 of
former Arthur
Andersen clients
The successor audit
firm fees are positively
related to the number

of years the company
was a client of Arthur
Andersen. This
relation is seen as
evidence that longer
tenure [with Arthur
Andersen] is a risk
factor priced in the
successor audit firm
fees
Yes
Chen et al. (2008) Audit firm and audit
partner tenure and
earnings quality
Taiwan audits
between 1990 and
2001
Above and beyond the
effects of partner
rotation on earnings
quality, longer tenure
does not reduce
earnings quality
No
Jenkins and Velury (2008) Audit firm tenure and
accounting conservatism
US audits between
1980 and 2004
Accounting
conservatism is

reduced in short-
tenure audits
No
Kaplan and Mauldin (2008) Audit firm rotation and
the perception of auditor
independence
An experiment with
55 MBA student
subjects
Above and beyond
partner rotation,
mandatory audit firm
rotation does not lead
to increased
perception of auditor
independence
No
Davis et al. (2009) Audit firm tenure and
earnings management
US audits between
1988 and 2006
Longer tenure is
associated with more
earnings management
in the pre-SOX period,
but not in the post-
SOX period
No
Ruiz-Barbadillo et al. (2009) Mandatory audit firm
rotation and going

concern opinions
Spanish audits of
financially distressed
companies between
1991 and 2000
The existence of a
mandatory rotation
rule does not lead to
an increase in going
concern opinions to
distressed companies
No
114 J.R. Casterella, D. Johnston /Research in Accounting Regulation 25 (2013) 108–116
References
Arel, B., Brody, R., & Pany, K. (2006). Findings on the effects of audit firm
rotation on the audit process under varying strengths of corporate
governance. Advances in Accounting, 22, 1–27.
Blouin, J., Grein, B. M., & Rountree, B. R. (2007). An analysis of forced
auditor change: The case of former Arthur Andersen clients. The
Accounting Review, 82(3), 621–650.
Carcello, J. V., & Nagy, A. L. (2004). Audit firm tenure and fraudulent
financial reporting. Auditing: A Journal of Practice & Theory, 23(2),
55–69.
Appendix A (continued)
Authors and date Relation examined Sample Primary results Support
for
MAFR?
Wang and Tuttle (2009) Audit firm rotation and
auditor negotiation
strategies

An experiment with
54 graduate business
students
Mandatory audit firm
rotation is associated
with auditors who
display less
cooperative
negotiation strategies
with their clients
Yes
Li (2010) Audit firm tenure and
accounting conservatism
US audits between
1980 and 2004
Longer tenure is
positively related to
conservatism, but only
for large companies
and strongly
monitored companies.
For smaller companies
and weakly monitored
companies, longer
tenure is negatively
related to
conservatism
Yes
Chi et al. (2011) Audit firm tenure and
earnings management

US audits between
2001 and 2008
Longer tenure is
negatively associated
with abnormal
accruals, but
positively related to
real earnings
management
Mixed
Daniels and Booker (2011) Audit firm rotation the
perception of auditor
independence and audit
quality
An experiment with
207 bank loan officers
Perceptions of auditor
independence are
enhanced with
mandatory audit firm
rotation. Perceptions
of audit quality are
unchanged with
mandatory audit firm
rotation
Mixed
Kwon et al. (2011) Mandatory audit firm
rotation and audit quality
Korean audits from
2000 to 2007

Audit quality was
unchanged after
mandatory audit firm
rotation was
introduced
No
Lowensohn et al. (working
paper)
Mandatory audit firm
rotation and audit quality
Florida municipal and
county audits in 2003
New audits formed by
a rotation policy have
higher audit quality as
compared to the audit
quality of long-
standing audits that
do not have a rotation
policy
Yes
J.R. Casterella, D. Johnston /Research in Accounting Regulation 25 (2013) 108–116
115
Carey, P., Geiger, M. A., & O’Connell, B. O. (2008). Costs associated with
going-concern-modified audit opinions: An analysis of the Australian
audit market. Abacus, 44(1), 61–81.
Chasen, E. (2012, March 21). Volcker backs auditor rotation. CFO Journal
(online). Available from />volcker-backs-auditor-rotation/ (2012, April 17).
Chen, C. Y., Lin, C. J., & Lin, Y. C. (2008). Audit partner tenure, audit firm
tenure, and discretionary accruals: Does long auditor tenure impair

earnings quality? Contemporary Accounting Research, 25(2), 415–445.
Chi, W., Ling Lei, L., & Mikhail, P. (2011). Is enhanced audit quality
associated with greater real earnings management? Accounting
Horizons, 25(2), 315–335.
Cohn, M. (2012, March 22). PCAOB hears pros and cons of audit firm
rotation. Accounting Today (online). Available from http://
www.accountingtoday.com/news/PCAOB-Hears-Pros-Cons-Audit-Firm-
Rotation-62079-1.html?pg=1 (2012, April 17).
Crabtree, A. D., Brandon, D. M., & Maher, J. J. (2006). The impact of auditor
tenure on initial bond ratings. Advances in Accounting, 22, 97–121.
Daniels, B. W., & Booker, Q. (2011). The effects of audit firm rotation on
perceived auditor independence and audit quality. Research in
Accounting Regulation, 23(1), 78–82.
Davis, L. R., Soo, B. S., & Trompeter, G. M. (2009). Auditor tenure and the
ability to meet or beat earnings forecasts. Contemporary Accounting
Research, 26(2), 517–548.
DeFond, M. L., & Francis, J. R. (2005). Audit research after Sarbanes–Oxley.
Auditing: A Journal of Practice & Theory, 24(Suppl.), 5–30.
Dopuch, N., King, R. R., & Schwartz, R. (2001). An experimental
investigation of retention and rotation requirements. Journal of
Accounting Research, 39(1), 93–117.
General Accounting Office (GAO). (2003). Required study on the potential
effects of mandatory audit firm rotation. Report to the Senate
Committee on Banking, Housing, and Urban Affairs and the House
Committee on Financial Services, November, Washington, DC.
Geiger, M. A., & Raghunandan, K. (2002). Auditor tenure and audit
reporting failures. Auditing: A Journal of Practice & Theory, 21(1),
67–78.
Ghosh, A., & Moon, D. (2005). Auditor tenure and perceptions of audit
quality. The Accounting Review, 80(2), 585–612.

Glover, S. M., Prawitt, D. F., & Wood, D. A. (2006). Publication records of
faculty promoted at the top 75 accounting research programs. Issues
in Accounting Education, 21(3), 195–218.
Jenkins, D. S., & Velury, U. (2008). Does auditor tenure influence the
reporting of conservative earnings? Journal of Accounting & Public
Policy, 27(2), 115–132.
Jennings, M. M., Pany, K. J., & Reckers, P. M. J. (2006). Strong corporate
governance and audit firm rotation: Effects on judges’ independence
perceptions and litigation judgments. Accounting Horizons, 20(3),
253–270.
Johnson, V. E., Khurana, I. K., & Reynolds, J. K. (2002). Audit-firm tenure
and the quality of financial reports. Contemporary Accounting
Research, 19(4), 637–660.
Kaplan, S. E., & Mauldin, E. G. (2008). Auditor rotation and the appearance
of independence: Evidence from non-professional investors. Journal of
Accounting & Public Policy, 27(2), 177–192.
Kealey, B. T., Lee, H. Y., & Stein, M. T. (2007). The association between
audit-firm tenure and audit fees paid to successor auditors: Evidence
from Arthur Andersen. Auditing: A Journal of Practice & Theory, 26(2),
95–116.
Kleinman, G., Palmon, D., & Anandarajan, A. (1998). Auditor
independence: A synthesis of theory and empirical research.
Research in Accounting Regulation, 12(1), 3–42.
Kwon, S. Y., Lim, Y., & Simnett, R. (2011). Mandatory audit firm rotation
and audit quality: Evidence from the Korean audit market. SSRN
eLibrary.
Larcker, D. F., & Rusticus, T. O. (2010). On the use of instrumental variables
in accounting research. Journal of Accounting and Economics, 49,
186–205.
Li, D. (2010). Does auditor tenure affect accounting conservatism? Further

evidence. Journal of Accounting and Public Policy, 29, 226–241.
Lowensohn, S. H., Reck, J., Casterella, J. R., & Lewis, B. (2011) An empirical
investigation of auditor rotation requirements. SSRN eLibrary.
Mansi, S. A., Maxwell, W. F., & Miller, D. P. (2004). Does auditor quality
and tenure matter to investors? Evidence from the bond market.
Journal of Accounting Research, 42(4), 755–793.
Myers, J. N., Myers, L. A., & Omer, T. C. (2003). Exploring the term of the
auditor–client relationship and the quality of earnings: A case for
mandatory auditor rotation? The Accounting Review, 78(3), 779–799.
Nagy, A. L. (2005). Mandatory audit firm turnover, financial reporting
quality, and client bargaining power: The case of Arthur Andersen.
Accounting Horizons, 19(2), 51–68.
Previts, G. J. (1998). Auditor independence: A perspective on its origins
and orientations. Research in Accounting Regulation, 12(1), 299–317.
Public Company Accounting Oversight Board (PCAOB). (2011). Concept
release on auditor independence and audit firm rotation. PCAOB Release
No. 2011-006, August 16.
Ruiz-Barbadillo, E., Gómez-Aguilar, N., & Carrera, N. (2009). Does
mandatory audit firm rotation enhance auditor independence?
Evidence from Spain. Auditing, 28(1), 113–135.
Stefaniak, C. M., Robertson, J. C., & Houston, R. W. (2009). The causes and
consequences of auditor switching: A review of the literature. Journal
of Accounting Literature, 28, 47–121.
Tysiac, K. (2012, March 21). PCAOB panelists say mandatory firm rotation
could be harmful, helpful. Journal of Accountancy (online). Available
from />(2012, April 21).
Walker, P. L., Lewis, B. L., & Casterella, J. R. (2001). Mandatory auditor
rotation: Arguments and current evidence. Accounting Enquiries, 10,
209–242.
Wang, K. J., & Tuttle, B. M. (2009). The impact of auditor rotation on

auditor–client negotiation. Accounting, Organizations & Society, 34(2),
222–243.
116 J.R. Casterella, D. Johnston /Research in Accounting Regulation 25 (2013) 108–116