Auditing: A Journal of Practice & Theory American Accounting Association

Vol. 31, No. 2 DOI: 10.2308/ajpt-10240

May 2012
pp. 167–188

The Effect of Corporate Governance on

Auditor-Client Realignments
Cory A. Cassell, Gary A. Giroux, Linda A. Myers, and Thomas C. Omer
SUMMARY: Events leading up to the implementation of the Sarbanes-Oxley Act of 2002
(SOX) increased the public’s focus on corporate governance and increased regulatory
scrutiny of corporate governance mechanisms. These events also contributed to a
massive restructuring in the audit market that resulted in the transfer of a large number of
clients from Big N to non-Big N audit firms. We extend prior research examining the
determinants of auditor-client realignments by investigating the effect of corporate
governance on downward (i.e., from Big N to non-Big N auditors) switching activity. We
develop a corporate governance index comprised of governance characteristics that we
expect auditors to find more desirable in their clients (specifically, board and audit
committee independence, diligence, and expertise). The results suggest that Big N
auditors consider client corporate governance mechanisms when making client portfolio
decisions. Specifically, downward auditor-client realignments are more likely for clients
that score lower on our corporate governance index. However, the influence of audit
committee-related corporate governance components on downward auditor-client
realignments decreased post-SOX. The reduced effect of audit committee-related
corporate governance components is consistent with what would be expected if the audit
committee-related rules imposed by SOX reduced the variation in audit committee
quality across clients.
Keywords: corporate governance; auditor-client realignments; audit risk; financial risk;
litigation risk; earnings manipulation risk; discretionary accruals; Sarbanes-
Oxley Act of 2002.
Data Availability: The data used are publicly available from the sources cited in the
text.

Cory A. Cassell is a Professor at the University of Arkansas, Gary A. Giroux is a Professor at Texas A&M
University, Linda A. Myers is a Professor at the University of Arkansas, and Thomas C. Omer is a Professor at
Texas A&M University.

We gratefully acknowledge the helpful comments and suggestions offered by the associate editor (Chris Hogan) and two
anonymous reviewers. We also thank Jean Bedard, Mike Drake, Jackie Hammersley, James Myers, Gary Peters,
Stephanie Rasmussen, Jaime Schmidt, Susan Scholz, Jian Zhou, and workshop participants at Texas A&M University
and the University of Arkansas for helpful comments, discussions, and suggestions. Gary Giroux gratefully
acknowledges financial support from the Mays Business School, Linda Myers gratefully acknowledges financial support
from the PricewaterhouseCoopers Faculty Fellowship while at Texas A&M University and from the Garrison/Wilson
Chair at the University of Arkansas, and Thomas Omer gratefully acknowledges the financial support of Ernst & Young.
Editor’s note: Accepted by Chris Hogan.

Submitted: May 2010

Accepted: December 2011
Published Online: February 2012

167
168 Cassell, Giroux, Myers, and Omer

INTRODUCTION

E
vents of the last decade, including a number of large accounting scandals, the collapse of
Arthur Andersen LLP (Andersen), and the subsequent passage of the Sarbanes-Oxley Act
of 2002 (SOX), resulted in a massive restructuring of the audit market. During this period,
auditor-client realignment activity resulted in a net loss of more than 2,200 clients for the surviving
Big N audit firms.1,2 In response to heavy downward switching activity (i.e., switching from Big N
to non-Big N audit firms), regulators warned large audit firms against dropping smaller, less
profitable, and riskier clients. In addition, although Big N firms claimed to be dropping smaller
clients because of capacity constraints brought on by the demands of SOX, officials from the
Securities and Exchange Commission (SEC) were skeptical. For example, then SEC Chief
Accountant Donald Nicolaisen remarked, ‘‘I have expressed my view to the CEOs of the big firms
that I think it is their responsibility not to run away from the marketplace. The requirements in the
2002 law should not be a convenient tool for them to manage their business. They do have a
responsibility [to] the public trust’’ (Taub 2004).
Nevertheless, anecdotal and empirical evidence suggests that client risk characteristics played a
pivotal role in the auditor-client realignment activity. Referring to the change in Ernst &Young
LLP’s client portfolio, James S. Turley, Chairman and Chief Executive Officer (CEO), stated, ‘‘Our
client acceptance and reacceptance processes have been re-engineered with an increased focus on
determining which companies we really want as audit clients and culling out those that we do not
believe have adapted to the new environment and demands of a public company’’ (Browning
2005).3
In this study, we extend a large body of research that investigates the client characteristics
associated with auditor-client realignments (see, for example, Jones and Raghunandan 1998;
Johnstone and Bedard 2003, 2004; Choi et al. 2004; Lee et al. 2004; Rama and Read 2006; Ettredge
et al. 2007, 2011; Hogan and Martin 2009; Landsman et al. 2009). With the exception of Lee et al.
(2004), this research investigates the effect of non-governance-related client characteristics (e.g.,
financial risk, litigation risk, audit fee pressure, client mismatch, etc.) on auditor-client realignment
activity. Lee et al. (2004) investigate whether corporate governance characteristics differ for clients
experiencing auditor resignations versus a matched sample of clients that dismiss their auditors in
the pre-SOX environment. They find that governance mechanisms are weaker (e.g., there are fewer
independent audit committee members, fewer independent board of director members, fewer audit
committee members who are financial experts) when auditors resign than when auditors are
dismissed. In contrast to these prior studies, we examine the determinants of downward
auditor-client realignments and focus on the role of corporate governance in these realignment
decisions in the pre- versus post-SOX environment.
Our study is motivated by the intense media and regulatory scrutiny of corporate
governance-related issues during the years leading up to and surrounding the implementation of
SOX. Discussion of sweeping changes to corporate governance regulation began in 1998, prompted
by then SEC Chairman Arthur Levitt’s landmark address at the New York University Center for

1
Big N is a generic term used in the extant literature to describe the largest audit firms. During our sample period,
Big N refers to the Big 5 audit firms (i.e., Arthur Andersen, Deloitte & Touche, Ernst & Young, KPMG, and
PricewaterhouseCoopers) prior to Andersen’s collapse, and to the surviving Big 4 audit firms thereafter.
2
From 2000 through 2007, Audit Analytics reports more than 15,000 auditor switches. Almost 19 percent of these
involved clients switching from Big N to non-Big N audit firms, but only 4 percent involved clients switching
from non-Big N to Big N audit firms.
3
Consistent with this notion, Hogan and Martin (2009) provide evidence that auditor-client realignment activity
from 2000 through 2004 resulted in a transfer of clients with heightened audit and business risk from the Big N
audit firms to smaller audit firms.

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 169

Law and Business. In response to concerns expressed by Levitt about the adequacy of director and
audit committee oversight (Levitt 1998), the New York Stock Exchange and National Association
of Securities Dealers sponsored the Blue Ribbon Committee on Improving the Effectiveness of
Corporate Audit Committees (BRC 1999; SEC 2003). The BRC report, issued in February 1999,
included a number of recommendations aimed at improving audit committee independence and
effectiveness, as well as mechanisms for improving accountability among the audit committee,
external auditors, and management (BRC 1999). Later that year, the major exchanges revised their
listing standards to incorporate many of the BRC’s recommendations (SEC 2003).
In addition, in February 2002, then SEC Chairman Harvey Pitt responded to numerous
corporate scandals, including Enron, Adelphia, and Worldcom, and the failure of Andersen by
calling on the major exchanges to review all of their governance-related listing standards, including
standards not related to audit committees (SEC 2002). Proposed changes associated with this listing
standard review, along with changes in audit committee standards mandated by SOX, were
approved by the SEC in November 2003 (SEC 2003).4 In particular, one important change is the
increased role of the audit committee with respect to appointment and dismissal decisions relating
to auditors. This change should have resulted in audit committees taking a fresh look at
auditor-client relationships.5
Media coverage of the corporate failures and corporate governance reforms was extensive (see,
for example, Lear 1998; Lublin and MacDonald 1998; Lublin and Nelson 1998; MacDonald 2000;
Lublin 2002; Hymowitz 2003). In addition, the highly publicized collapse of Big N audit firm Arthur
Andersen captured the public’s attention and provided a dramatic example of the costs of audit
failure. In sum, corporate governance issues received extensive and persistent attention from
regulators and the media. The level of scrutiny implies that regulators and others believed that certain
corporate governance mechanisms could play a key role in preventing future corporate failures.
Although prior research suggests that client risk characteristics (e.g., financial risk, litigation
risk, and earnings manipulation risk) are the primary factors influencing auditor-client realignments
(Jones and Raghunandan 1998; Johnstone and Bedard 2003, 2004; Choi et al. 2004; Rama and
Read 2006; Hogan and Martin 2009), we posit that corporate governance was an important factor
influencing the recent wave of auditor-client realignment activity, because prior work documents an
association between certain governance characteristics and the incidence of financial reporting fraud
(Beasley 1996; Dechow et al. 1996; Abbott et al. 2000; Beasley et al. 2000a; Farber 2005). Given
the observable implications of financial reporting fraud and the intense scrutiny of corporate
governance practices by regulators and the media, we expect that Big N firms placed emphasis on
their clients’ governance mechanisms when making client retention/acceptance decisions.
Consistent with this, Cohen et al. (2002) conduct semi-structured interviews of audit firm
professionals and find that corporate governance influences auditor risk assessments and client
acceptance decisions.6 These results are supported by Sharma et al. (2008) who show that, in an

4
SOX Section 301 states that: (1) the audit committee is directly responsible for the appointment, compensation,
and oversight of the work of any registered public accounting firm employed by that issuer; (2) each member of
the audit committee must be a member of the board of directors and must be independent; (3) each audit
committee must establish procedures for the receipt, retention, and treatment of complaints (including those
provided by anonymous employees) regarding accounting, internal accounting controls, or auditing matters; (4)
each audit committee must have the authority to engage independent counsel and other advisers as it determines
necessary to carry out its duties; and (5) each issuer must provide for appropriate funding (as determined by the
audit committee) for payment of compensation to the registered public accounting firm employed by the issuer for
the purpose of rendering or issuing an audit report and to any advisers employed by the audit committee.
5
We thank the reviewer for this helpful observation.
6
Cohen et al. (2002) also suggest that corporate governance will influence audit-planning decisions, and they call
for archival research to extend and corroborate their findings.

Auditing: A Journal of Practice & Theory

May 2012
170 Cassell, Giroux, Myers, and Omer

experimental setting, client acceptance decisions, risk assessments, and the extent and timing of
audit procedures are sensitive to clients’ corporate governance. Moreover, prior research (e.g.,
Cohen and Hanno 2000; Bedard and Johnstone 2004) suggests that audit firms increase audit effort
when faced with inadequate client corporate governance structures. Thus, we expect corporate
governance characteristics to be associated with auditor-client realignment activity because: (1)
corporate governance characteristics are associated with audit-relevant outcomes (e.g., the
incidence of fraud); (2) the additional audit effort necessary for clients with inadequate corporate
governance is problematic when resources are constrained; and (3) corporate governance
mechanisms were heavily scrutinized leading up to the implementation of SOX.
We investigate the factors influencing auditor-client realignments using a sample of clients that
switched from Big N to non-Big N audit firms and a matched sample of Big N clients that did not
switch audit firms.7,8
We create a corporate governance index that incorporates governance characteristics that we
expect auditors to find more desirable in their clients (e.g., board and audit committee
independence, diligence, and expertise). Our models also include three audit risk factors that are
known from prior literature to affect realignments: (1) financial (or client business) risk, measured
as the Altman (1968) z-score as modified by Shumway (2001); (2) litigation (or auditor) risk,
measured based on the model developed by Stice (1991); and (3) earnings manipulation (or audit)
risk, measured using the absolute value of discretionary accruals (Teoh et al. 1998a, 1998b; Xie et
al. 2003; Bedard and Johnstone 2004; DuCharme et al. 2004; Park and Shin 2004).
Our results suggest that Big N auditors consider client corporate governance mechanisms when
making client portfolio decisions. Specifically, we document a significant association between our
corporate governance index and downward (Big N to non-Big N) auditor-client realignments in
both the pre-and post-SOX periods, suggesting a movement of clients with governance
characteristics that we expect auditors to find less desirable in their clients from the Big N audit
firms to non-Big N audit firms throughout our sample period. Although our results provide some
evidence suggesting that the influence of corporate governance on auditor-client realignments
decreases in the post-SOX period (relative to the pre-SOX period), we note that this decrease is
primarily related to the effect of audit committee-related components of corporate governance. This

7
We do not examine non-Big N to Big N auditor switches because we follow prior literature in assuming that Big N
audit firms provide higher quality audits. Here, upward auditor-client realignments should generate positive
changes in audit quality and, thus, should be of less concern. Moreover, upward auditor-client realignments are
relatively rare events. Specifically, for the period 2000 through 2007, upward auditor-client realignments
represent less than 5 percent of the total number of realignment observations available in Audit Analytics.
8
In (untabulated) analyses, we investigate whether the effect of corporate governance differs for auditor-client
realignments that are initiated by the auditor (i.e., resignations) versus those that are initiated by the client (i.e.,
dismissals). Although we find that the magnitude of the coefficient on our corporate governance index is larger for
resignations (relative to dismissals), the difference in the coefficient estimates is not significant at conventional
levels. We expect that our inability to detect a significant difference is due, in part, to a relatively small sample of
resignation observations as well as noise in the classification of switching observations as resignations versus
dismissals. Among the switching observations in our sample, there are 384 observations where the client is
recorded in Audit Analytics as having dismissed the auditor and only 114 observations where the auditor is
recorded as having resigned from the engagement (and the small sample size issue is exacerbated when we
separate these observations into the pre- and post-SOX periods). Regarding noise in the classification of
resignations versus dismissals, our discussions with practitioners and with other audit researchers suggest that the
8-K classification is not always accurate. That is, in some cases, audit firms may ‘‘let clients dismiss them’’ (rather
than resign) or may increase their audit fees for less desirable clients to the point where these clients dismiss them.
This behavior (while unobservable empirically) could be especially pervasive in our setting because regulators
were explicitly warning large audit firms not to use capacity constraints to ‘‘walk away’’ from less desirable
clients. Supporting this notion, Ettredge et al. (2007) provide evidence that suggests that some dismissals are
‘‘implicit resignations.’’

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 171

result is consistent with what would be expected if the audit committee-related disclosure rules
imposed by SOX reduced the cross-sectional variation in the quality of audit committees.
Our study makes a number of contributions to the literature. First, we add to the literature
on auditor-client realignments by demonstrating the importance of corporate governance in
these decisions. Second, we provide evidence suggesting that Big N audit firms react
expeditiously to pending changes in the regulatory environment by altering their client
portfolios. Third, we document a potentially negative outcome of the downward realignment
activity observed in the years surrounding the implementation of SOX if the corporate
governance characteristics included in our index are positively associated with financial
reporting quality (as suggested by some prior research)9 and if Big N audit firms provide higher
quality audits (as also suggested by prior research).10 Finally, we add support to prior literature
(e.g., Simunic and Stein 1990; Jones and Raghunandan 1998; Johnstone and Bedard 2003,
2004; Choi et al. 2004; Rama and Read 2006; Hogan and Martin 2009; Landsman et al. 2009)
that suggests Big N audit firms alter their client portfolios in response to client risk
characteristics and changes in regulatory scrutiny.
In the next section, we describe the measures in our corporate governance index and client risk
variables, and we develop our auditor-client realignment model. The following section describes
our empirical method, including our sample selection, and presents the results of our analyses. The
final section concludes.

MODEL DEVELOPMENT
We focus on the effect of corporate governance on auditor-client realignments while
controlling for financial risk, litigation risk, and earnings manipulation risk. In the remainder of this
section, we discuss each of these measures, beginning with the corporate governance index used in
our analysis.

Corporate Governance Index

Prior corporate governance literature typically considers various characteristics of board and
audit committee members to assess the quality of a company’s governance mechanisms (e.g.,
Menon and Williams 1994; Dechow et al. 1996; Beasley et al. 2000b; Carcello et al. 2002). Our
corporate governance index (GOVSCORE_HC) is comprised of attributes that capture the
independence, diligence, and financial expertise of board and audit committee members and is
formed using data that are hand collected from company proxy statements. We selected the
corporate governance attributes included in GOVSCORE_HC because these attributes are
negatively associated with audit-relevant outcomes (e.g., the incidence of fraud, restatements,
etc.). Thus, we posit that auditors find these governance characteristics desirable in their clients
because they should reduce the likelihood of audit failure on average. GOVSCORE_HC is
constructed as follows:
GOVSCORE HC ¼ R ðBOARDIND; CEOCHAIR; ATTENDANCE75; ACDILIGENCE;
ACFINEXPERT; NOTSTAGGEREDÞ:

9
See, for example, Beasley (1996); Dechow et al. (1996); Abbott et al. (2000); Klein (2002); Imhoff (2003); Xie
et al. (2003); Abbott et al. (2004); Bedard et al. (2004); Farber (2005); Vafeas (2005); Krishnan and Visvanathan
(2008); and Dechow et al. (2010).
10
See, for example, DeAngelo (1981); Becker et al. (1998); Palmrose (1988); Beatty (1989); Teoh and Wong
(1993); Francis and Krishnan (1999); Khurana and Raman (2004); Pittman and Fortin (2004); Blokdijk et al.
(2006); and Krishnamurthy et al. (2006).

Auditing: A Journal of Practice & Theory

May 2012
172 Cassell, Giroux, Myers, and Omer

where:
BOARDIND ¼ 1 if the percentage of independent directors is greater than the median
percentage of independent directors, 0 otherwise;11
CEOCHAIR ¼ 1 if the CEO is not the Chairman of the Board, 0 otherwise;
ATTENDANCE75 ¼ 1 if all board members attended at least 75 percent of the board meetings,
0 otherwise;
ACDILIGENCE ¼ 1 if the number of audit committee meetings is greater than the median audit
committee meeting frequency, 0 otherwise;
ACFINEXPERT ¼ 1 if at least one of the audit committee members is a financial expert, 0
otherwise; and
NOTSTAGGERED ¼ 1 if the board is not staggered, 0 otherwise.
By construction, auditors should find clients with a higher score on GOVSCORE_HC more
appealing. We briefly summarize the rationale for each of the GOVSCORE_HC inputs below.

Board Independence
We include the percentage of independent directors (BOARDIND) because independent
directors should be better monitors of management than non-independent directors. Consistent with
this, Beasley (1996), Dechow et al. (1996), Klein (2002), and Dechow et al. (2010) provide
evidence suggesting that earnings quality is higher when boards are more independent. Prior work
also documents a positive association between an increase in the proportion of independent board
members and internal control material weakness remediation (Johnstone et al. 2011) and stock
returns of companies previously charged with fraud (Farber 2005). We also include the CEO-
Chairman duality indicator (CEOCHAIR) because prior research suggests that directors’ ability to
monitor management is reduced when the CEO-Chairman influences board meeting agendas
(Jensen 1993; Bebchuk et al. 2002; Bebchuk and Fried 2003; Davila and Penalva 2006) and that
financial reporting quality is lower when CEO-Chairman duality exists (Dechow et al. 1996; Imhoff
2003; Farber 2005). Finally, we include an indicator for non-staggered board member elections
(NOTSTAGGERED) to represent the effect of board structure on overall governance quality. Prior
research (e.g., Gompers et al. 2003; Bebchuk and Cohen 2005; Faleye 2007; Zhao and Chen 2008)
suggests that managerial (shareholder) power is higher (lower) when board member elections are
staggered.

Board and Audit Committee Diligence

Board and audit committee diligence is beneficial because directors play an important advisory
role to management (Adams and Ferreira 2007) and because directors presumably access
information during board and audit committee meetings that helps them to perform their monitoring
role more effectively (Adams and Ferreira 2011). Consistent with this, prior research finds a

11
For those governance characteristics where the ‘‘score’’ on the characteristic depends on the relation between the
company-specific characteristic and the sample median for that characteristic, we re-compute the company-
specific characteristic and the sample median for the characteristic each year. As a result, each of the governance
indices provides a corporate governance ‘‘score’’ for a given company relative to the other companies in the
sample in that year. This allows us to avoid setting an arbitrary cutoff for a given characteristic and provides us
with a measure of corporate governance that is independent of the year in which the measure is computed.
Importantly, if there is a trend in certain corporate governance characteristics over time (e.g., an increase in the
frequency of audit committee meetings), this method allows us to avoid classifying all of the post-SOX
observations as ones with higher scores on our corporate governance index. This is important because of SOX-
related mandates pertaining to corporate governance that were enacted during our sample period.

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 173

negative association between board and audit committee meeting frequency and the magnitude of
discretionary current accruals (Xie et al. 2003) and a negative association between audit committee
meeting frequency and financial statement restatements (Abbott et al. 2004), fraudulent financial
reporting (Abbott et al. 2000; Farber 2005), and the propensity to meet earnings benchmarks
(Vafeas 2005). Thus, we include ATTENDANCE75 and ACDILIGENCE to capture aspects of
board and audit committee diligence.

Audit Committee Financial Expertise

We include an audit committee financial expert indicator (ACFINEXPERT) because financial
expertise should enable audit committee members to perform their monitoring role better
(Krishnamoorthy et al. 2002). Bedard et al. (2004) suggest that audit committees that include a
financial expert are better able to constrain earnings management, and Abbott et al. (2004) find that
companies are less likely to restate financial statements when their audit committees include a
financial expert. Similarly, Johnstone et al. (2011) find a positive association between internal
control material weakness remediation and improvements in audit committee financial expertise.
Prior research also finds a positive association between audit committee financial expertise and
audit fees, suggesting that audit committees that include a financial expert are more likely to
demand increased audit work (Abbott et al. 2003).
Our definition of audit committee financial expertise is based on results in DeFond et al. (2005)
and Krishnan and Visvanathan (2008). DeFond et al. (2005) document a positive market reaction to
the appointment of a financial expert to the audit committee only when the newly appointed audit
committee member has prior accounting-specific experience. Krishnan and Visvanathan (2008) find
that financial reporting conservatism is greater when the audit committee includes a financial expert
with prior accounting-specific experience. As such, we adopt a definition of audit committee
financial expertise that requires prior accounting-specific experience. Specifically, ACFINEXPERT
is set equal to 1 if any member of the audit committee has prior work experience as a public
accountant, auditor, principal or chief financial officer, controller, or principal or chief accounting
officer,and 0 otherwise.

Audit Committee-Related and Non-Audit Committee-Related Governance Components

In supplemental tests, we disaggregate our corporate governance index (GOVSCORE_HC) into
two measures. The first measure (GOVSCORE_HCAC ) includes the two audit committee-related
components of GOVSCORE_HC and the second measure (GOVSCORE_HCNon-AC ) includes the
remaining components of GOVSCORE_HC. Specifically, GOVSCORE_HCAC and GOVSCORE_
HCNon-AC are constructed as follows:
GOVSCORE HCAC ¼ RðACDILIGENCE; ACFINEXPERTÞ; and

GOVSCORE HCNonAC ¼ RðBOARDIND; CEOCHAIR; ATTENDANCE75;

NOTSTAGGEREDÞ:
This approach enables us to investigate whether characteristics of the audit committee have effects
on auditor-client realignment activity different from other governance components.

Financial Risk
Financial risk, defined as the probability that a client’s business conditions will deteriorate, is
usually measured based on financial distress. For example, Choi et al. (2004) examine changes in

Auditing: A Journal of Practice & Theory

May 2012
174 Cassell, Giroux, Myers, and Omer

Big 6 audit portfolio risk during periods of changing auditor litigation liability using alternative
measures of financial distress that are meant to capture financial risk.12 Following Choi et al.
(2004), we use the Shumway-weighted Altman model (MOD_ALTMAN) to measure financial risk
(Shumway 2001). However, our results are qualitatively similar when we use the other measures in
Choi et al. (2004).13

Litigation Risk
Litigation risk can be defined as the potential loss from an engagement due to litigation
(Johnstone and Bedard 2003). Stice (1991) models litigation against an auditor based on client
characteristics (including relative levels of accounts receivable and inventory, sales growth, and
financial condition), auditor characteristics (including auditor size, independence, and tenure),
and stock market characteristics (including the variance of abnormal returns and market value).
He finds that the financial condition of the client, the variance of abnormal returns, and market
value are all significant determinants of auditor litigation. We use scores from the Stice (1991)
model as a proxy for litigation risk. Specifically, we use the year-matched analysis in Table 6 of
Stice (1991) to construct a summary litigation measure for each company-year observation in our
sample.14

Earnings Manipulation Risk

Earnings manipulation risk can be defined as the probability that the financial statements are
materially misstated. Our use of discretionary accruals to proxy for earnings manipulation risk is
consistent with Bedard and Johnstone’s (2004) measure of manipulation risk, which was based, in
part, on studies that use discretionary accruals to investigate earnings management. Following
Kothari et al. (2005), we use a performance-adjusted modified Jones model to measure
discretionary accruals, and we drop industries without at least ten firm-year observations.15
Because we have no a priori expectation on the direction of unexpected accruals, we focus on the
absolute value of discretionary accruals (Warfield et al. 1995; Bartov et al. 2000; Johnson et al.
2002; Klein 2002; Myers et al. 2003). According to Klein (2002), the absolute value of
discretionary accruals captures managers’ intervention in reporting accounting earnings. Therefore,
we suggest that this is a good proxy for earnings manipulation risk.

Empirical Model
Our analysis of auditor-client realignments follows Johnstone and Bedard (2003, 2004). We
extend this earlier work with a focus on corporate governance, which we posit to be an important
factor in auditor-client realignments. Specifically, we estimate the following conditional logistic
regression model of the likelihood of a client switching from a Big N audit firm to a non-Big N

12
Choi et al. (2004) find that increased regulatory scrutiny of the financial reports results in Big N audit firms
altering their client portfolios to reduce their levels of financial risk and, thus, adjusting their overall risk.
13
For brevity, we refer the reader to Table 1 for a detailed description of the variable measurement procedures for
our financial, litigation, and audit risk variables.
14
We also performed analyses using the Shu (2000) measure of litigation risk rather than the measure based on
Stice (1991). However, because the Shu (2000) measure is more data intensive, analyses using this measure are
based on a sample that is approximately 27 percent smaller than those reported in the tables. Nevertheless, the
results of these supplemental analyses are qualitatively similar to those reported in the tables.
15
We define industries at the two-digit SIC code level, and we use performance-adjusted discretionary accruals
because Kothari et al. (2005) demonstrate that inferences are more reliable when this measure is used.

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 175

audit firm as a function of corporate governance, the client risk variables, time period, and other
controls:16
REALIGN ¼ b1 GOVSCORE HC þ b2 GOVSCORE HCPOSTSOX þ b3 MOD ALTMAN
þ b4 MOD ALTMAN POSTSOX þ b5 STICE þ b6 STICEPOSTSOX
þ b7 ABS DA þ b8 ABS DAPOSTSOX þ b9 ABFEES þ b10 ABFEESPOSTSOX
þ b11 FORSEG þ b12 FORSEGPOSTSOX þ b13 LNASSET
þ b14 LNASSET POSTSOX þ e:
ð1Þ
17
where:
REALIGN ¼ 1 if the client switched from a Big N to a non-Big N auditor, 0 otherwise;
GOVSCORE_HC ¼ the governance score (calculated as described previously);
POSTSOX ¼ 1 if the observation is from 2003 through 2007, 0 otherwise;18
MOD_ALTMAN ¼ the modified Altman Z-score;
STICE ¼ the Stice litigation score;
ABS_DA ¼ the absolute value of discretionary accruals;
ABFEES ¼ abnormal audit fees;
FORSEG ¼ the number of foreign segments, from Compustat; and
LNASSET ¼ the natural log of total assets.
To allow for tests in which we investigate the effects of different corporate governance
components (i.e., GOVSCORE_HCAC and GOVSCORE_HCNon-AC ) on auditor-client realignment
activity, we modify Model (1) as follows:
REALIGN ¼ b1 GOVSCORE HCAC þ b2 GOVSCORE HCAC POSTSOX
þ b3 GOVSCORE HCNonAC þ b4 GOVSCORE HCNonAC POSTSOX
þ b5 MOD ALTMAN þ b6 MOD ALTMAN POSTSOX þ b7 STICE
þ b8 STICEPOSTSOX þ b9 ABS DA þ b10 ABS DAPOSTSOX þ b11 ABFEES
þ b12 ABFEESPOSTSOX þ b13 FORSEG þ b14 FORSEGPOSTSOX
þ b15 LNASSET þ b16 LNASSET POSTSOX þ e:
ð2Þ
where:
GOVSCORE_HCAC ¼ the governance score on the audit committee-related components of
GOVSCORE_HC (calculated as described previously); and
GOVSCORE_HCNon-AC ¼ the governance score on the non-audit committee-related
components of GOVSCORE_HC (calculated as described previously).

16
We thank the reviewer for suggesting that we use conditional logit for our analyses. As noted by Cram et al.
(2009), this approach estimates a no intercept logit regression of pair-wise differences in the dependent variable
(REALIGN) on pair-wise differences in the independent variables and produces consistent coefficient estimates
and standard errors that are correct for use in inferences. Note that the use of conditional logit renders the main
effect for POSTSOX meaningless because there is no within-group variation in this variable.
17
See Table 1 for details of the calculations.
18
The implementation of SOX Section 404, which led to significant resource constraints among the Big N audit
firms, provided the impetus for the observed downward switching activity during our sample period. Although
SOX was enacted in 2002, the requirements under Section 404 did not become effective until 2004.
Nevertheless, we include 2003 in the post-SOX period because many companies engaged their auditors to
perform a SOX 404 ‘‘test run’’ during 2003. In untabulated analyses, we drop all observations from 2002 (the
year SOX was passed) and find results that are qualitatively similar to those reported in the tables.

Auditing: A Journal of Practice & Theory

May 2012
176 Cassell, Giroux, Myers, and Omer

In Models (1) and (2), we include interaction terms between the risk variables
(GOVSCORE_HC, GOVSCORE_HCAC, GOVSCORE_HCNon-AC, MOD_ALTMAN, STICE, and
ABS_DA) and POSTSOX to capture changes in auditor-client realignment decisions across the pre-
and post-SOX time periods. Because we expect auditor retention/acceptance decisions to be based
on potential billing rates as well as on client risk (Johnstone and Bedard 2003), we include
abnormal audit fees (ABFEES) in the model. Moreover, extant literature finds a relation between
audit fees (which proxy for audit effort) and corporate governance. For example, audit fees increase
with director independence and diligence (Carcello et al. 2002) and with audit committee
independence and expertise (Abbott et al. 2003). Furthermore, Bedard and Johnstone (2004) find
that planned audit hours and billing rates increase when corporate governance risk is greater. Given
these relations, it is important to control for audit fees in our model. We include an interaction term
between abnormal audit fees and the post-SOX indicator variable to capture differences in fee
structure resulting from the implementation of SOX (Ghosh and Pawlewicz 2008). Following
Larcker and Richardson (2004) and Omer et al. (2006), we use the fee model residual divided by
the log of actual fees to proxy for abnormal audit fees.19 The remaining variables, FORSEG and
LNASSET, are included to control for the effects of complexity and firm size on the probability that
a given client will switch from a Big N to a non-Big N auditor.

EMPIRICAL METHOD AND RESULTS

Sample Selection
For the period 2000 through 2007, we collect auditor-client realignment observations from the
Audit Analytics database. The initial sample is comprised of 2,855 unique Big N to non-Big N
auditor change observations. We exclude 308 auditor changes where Arthur Andersen was the
predecessor or successor auditor as well as an additional 1,898 observations with missing data items
needed to construct the non-governance related variables included in Model 1.20 For the remaining
649 auditor change observations, we construct a control sample of non-switching Big N client
observations, matched on year, size (total assets), and industry (two-digit SIC code).
As discussed above, we hand-collect corporate governance data for the GOVSCORE_HC
variable from client proxy statements.21 We exclude 73 observations because a proxy statement was
unavailable for either the sample company and/or an acceptable control company. Finally, we
exclude 6 observations for clients that appear in the auditor change sample more than once, and 72
observations where either: (1) total assets for the closest control company were more than two times
larger than for the sample company; or (2) total assets for the closest control company were more
than two times smaller than for the sample company. Table 2 summarizes the derivation of our

19
Table 1 provides additional details on the model used to construct the ABFEES measure. The model is estimated
using the population of available observations (i.e., all observations with sufficient data to estimate the model).
20
For these variables, we collect financial data from Compustat, audit fee data from AuditAnalytics, and market
data from the Center for Research on Security Prices (CRSP). We delete observations for foreign companies,
observations with invalid or missing auditor information, observations with missing SIC codes, and observations
with assets of less than $1 million. Following prior research, we drop observations in regulated (SIC codes 4000–
4999) and financial (SIC codes 6000–6999) industries prior to estimating the abnormal accruals model used to
construct ABS_DA. For the financial risk measure (MOD_ALTMAN), we winsorize observations at the top and
bottom 1 percent of the distribution for each of the five summary measures used to calculate the modified Altman
z-score. We winzorize the litigation risk measure (STICE) and the earnings manipulation measure (ABS_DA) at
the top and bottom 1 percent of the distribution to control for possible outliers.
21
Specifically, we collect corporate governance data from the client proxy statement filed on the date closest to the
auditor change date over the window 12 months to þ1 month.

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 177

TABLE 1
Variable Definitions

REALIGN ¼ 1 if the client switched from a Big N to a non-Big N auditor, 0 otherwise;

GOVSCORE ¼ Governance risk measures based on characteristics of the board of directors
and of the audit committee. We use the following measures in our tests:
GOVSCORE_HC ¼ This score measures the strength of corporate governance based on data hand
collected from proxy statements. The score is the sum of the following
items, measured by year:
ACFINEXPERT ¼ 1 if at least one of the audit committee members is a financial expert, 0
otherwise;
ACDILIGENCE ¼ 1 if the number of audit committee meetings is greater than the median audit
committee meeting frequency, 0 otherwise;
BOARDIND ¼ 1 if the percentage of independent directors is greater than the median
percentage of independent directors, 0 otherwise;
CEOCHAIR ¼ 1 if the CEO is not the Chairman of the Board, 0 otherwise;
ATTENDANCE75 ¼ 1 if all board members attended at least 75 percent of the board meetings, 0
otherwise; and
NOTSTAGGERED ¼ 1 if the board is not staggered, 0 otherwise.
GOVSCORE_HCAC ¼ sum of the audit committee-related components of GOVSCORE_HC:
ACFINEXPERT and ACDILIGENCE;
GOVSCORE_HCNon-AC ¼ sum of the non-audit committee-related components of GOVSCORE_HC:
BOARDIND, CEOCHAIR, ATTENDANCE75, and NOTSTAGGERED
MOD_ALTMAN (a ¼ Estimated, following Shumway (2001), as:
measure of financial 1:2ðCA  CLÞ=TA þ 0:6ðRE=TAÞ þ 10:0ðEBITAÞ þ 0:05ðMVEQ=TLÞ þ 0:47
risk) ðSALES=TAÞ;
where:
CA ¼ current assets [ACT];
CL ¼ current liabilities [LCT];
TA ¼ total assets [AT];
RE ¼ retained earnings [RE];
EBITA ¼ earnings before interest and tax divided by total assets ([PI] þ [XINT])/[AT];
MVEQ ¼ market value [PRCC_F]  [CSHO];
TL ¼ total liabilities [LT]; and
SALES ¼ total sales [SALE].
STICE (a measure of ¼ Estimated, following Stice (2001, Table 6), as:
litigation risk) 315:74  0:273ðRAÞ þ 0:423ðIAÞ þ 1:053ðGROWTHÞ  0:18ðFINCONDÞ þ
2:276ðBIGNÞ  1:517ðTENUREÞ  323:44ðINDEPÞ þ 2725:8ðVARÞ þ
0:269ðLMVÞ;
where:
RA ¼ receivables divided by total assets [RECT]/[AT];
IA ¼ inventory divided by total assets [INVT]/[AT];
GROWTH ¼ ([SALE: current year]  [SALE: prior year])/[SALE: prior year];
FINCOND ¼ Altman’s (1968) z-score;
BIGN ¼ 1 if the auditor is a Big N auditor, 0 otherwise;
TENURE ¼ 1 if the auditor tenure is more than three years, 0 otherwise;
INDEP ¼ 1  (client sales/total sales of all clients of the auditor);
VAR ¼ variance of abnormal returns for the client; and
LMV ¼ natural logarithm of market value ([PRCC_F]  [CSHO]).
(continued on next page)

Auditing: A Journal of Practice & Theory

May 2012
178 Cassell, Giroux, Myers, and Omer

TABLE 1 (continued)

ABS_DA (a measure of ¼ Estimated, following Kothari et al. (2005), as the absolute value of
earnings manipulation discretionary accruals estimated as the residual from the performance-
risk) adjusted modified Jones model:
TAt =ASSETSt1 ¼ a þ b1 1=ASSETSt1 þ b2 ðDSALESt  DARt Þ=ASSETSt1 þ
b3 PPEt =ASSETSt1 þ b4 ROA þ et :
where:
TA ¼ total accruals ([IB]  [OANCF]);
ASSETS ¼ total assets [AT];
SALES ¼ total sales [SALE];
AR ¼ accounts receivable [RECT];
PPE ¼ property, plant, and equipment [PPEGT];
ROA ¼ return on assets [IB]/[AT: prior year]; and
et ¼ discretionary accruals.
ABFEES ¼ Abnormal audit fees estimated as the residual from the following abnormal
audit fee model, estimated by year. The residual is scaled by log of total
audit fees:
LnFEES ¼ a þ b1 LNASSET þ b2 ðAR þ INVÞ=ASSETS þ b3 LEV þ b4 ROA þ
b5 SEG þ b6 BIGN þ b7 RESTATE þ b8 ACCEL FILER þ b9 GOING CONCE
RN þ b10 BUSY SEASON þ b11 LOSS þ b12111 INDUSTRY þ e:
where:
LnFEES ¼ natural logarithm of total audit fees;
LNASSET ¼ natural logarithm of total assets [AT];
AR ¼ accounts receivable [RECT];
INV ¼ total inventory [INVT];
ASSETS ¼ total assets [AT];
LEV ¼ leverage [LT]/[AT];
ROA ¼ return on assets [PI]/[AT];
SEG ¼ number of business segments;
BIGN ¼ 1 if the auditor is a Big N auditor, 0 otherwise;
RESTATE ¼ 1 if the client issues a restatement in the current year, 0 otherwise;
ACCEL_FILER ¼ 1 if the client is an accelerated filer in the current year, 0 otherwise;
GOING_CONCERN ¼ 1 if the client receives a going-concern opinion in the current year, 0
otherwise;
BUSY_SEASON ¼ 1 if the client’s fiscal year-end falls in the month of December, 0 otherwise;
LOSS ¼ 1 if net income [NI] is negative, 0 otherwise; and
INDUSTRY ¼ 2-digit SIC code industry dummy variables.
FORSEG ¼ number of foreign segments; and
LNASSET ¼ natural logarithm of total assets [AT].

Compustat data items are shown in brackets.

sample, which consists of 498 Big N to non-Big N auditor-client realignment observations and 498
non-switching Big N client observations.

Descriptive Statistics
Descriptive statistics for our corporate governance measures are presented in Table 3. We note
that, relative to sample companies, control companies have higher scores on the vast majority of our
GOVSCORE_HC components in both the pre- and post-SOX periods. Consistent with this, control

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 179

TABLE 2
Summary of Sample Construction: Auditor Change (Switching) Analysis
Company-Year
Observations
Total unique Big N to non-Big N auditor change observations available from 2,855
AuditAnalytics for the years 2000 to 2007 inclusive.
Less: observations for auditor changes where Arthur Andersen was the predecessor or (308)
successor auditor.
Less: observations missing data items needed to construct the STICE, MOD_ALTMAN, (1,898)
ABS_DA, or ABFEES measures.
Less: observations with no available match and/or no proxy available to collect the data (73)
necessary for the GOVSCORE_HC variable.
Less: observations for firms with more than one observation in the auditor change (6)
sample.
Less: observations with no suitable match (based on difference in total assets). (72)

Auditor Change Sample 498

Non-Switching Matched Sample 498

Full Sample 996

companies have higher overall corporate governance scores (GOVSCORE_HC) in both the pre- and
post-SOX periods. The results also indicate that the higher corporate governance scores for control
companies are primarily due to higher scores on audit committee-related governance components
(GOVSCORE_HCAC ) in the pre-SOX period and on non-audit committee-related governance
components (GOVSCORE_HCNon-AC ) in the post-SOX period. Finally, we note a significant pre- to
post-SOX increase in the overall corporate governance score (GOVSCORE_HC) for both sample
and control companies. These increases appear to be the result of increases in audit committee-
related governance components (GOVSCORE_HCAC ) for both sample and control companies.
Table 4 presents descriptive statistics for our risk variables and for our non-risk control
variables. We find significant pre- to post-SOX decreases in financial risk (MOD_ALTMAN),
litigation risk (STICE), and earnings manipulation risk (ABS_DA) for both sample and control
companies.22 Pre-SOX, sample companies have lower financial risk (MOD_ALTMAN) and
abnormal fees (ABFEES) than do control companies. Post-SOX, sample companies have higher
litigation risk (STICE) and lower abnormal fees (ABFEES) than do control companies. We find no
significant difference in sample and control company size (ASSET) in either period, consistent with
our matching criterion.

22
We note that the pre- to post-SOX changes in MOD_ALTMAN and STICE are quite large for both sample and
control companies. We have investigated changes in the distributions of these variables in the population of
available observations (since both variables are constructed using the population of available observations). Like
the changes noted in Table 4 for our small sample of switching observations, we note a large pre- to post-SOX
increase in MOD_ALTMAN (specifically, the mean increased from 1.37 to 0.52 and the median increased from
1.05 to 1.43) and a large pre- to post-SOX decrease in STICE (specifically, the mean decreased from 7.03 to 1.31
and the median decreased from 3.02 to 0.17) in the population. Thus, the data suggest that it may not be
reasonable to compare mean values for these variables across time. However, our focus is on contrasting the
mean values for these variables between switching and non-switching observations (where the mean values are
constructed over the same time period).

Auditing: A Journal of Practice & Theory

May 2012
180 Cassell, Giroux, Myers, and Omer

TABLE 3
Descriptive Statistics for GOVSCORE_HC, GOVSCORE_HCAC, GOVSCORE_HCNon-AC, and
Components: Pre- and Post-SOX (n ¼ 996)
Panel A: Sample Companies
Pooled Pre-SOX Post-SOX Difference p-value
ACFINEXPERT 0.498 0.375 0.548 0.173 ***
ACDILIGENCE 0.414 0.306 0.458 0.152 ***
BOARDIND 0.450 0.431 0.458 0.027
CEOCHAIR 0.492 0.479 0.497 0.018
ATTENDANCE75 0.884 0.799 0.918 0.119 ***
NOTSTAGGERED 0.496 0.604 0.452 0.152 ***
GOVSCORE_HC 3.233 2.993 3.331 0.337 ***
GOVSCORE_HCAC 0.912 0.681 1.006 0.325 ***
GOVSCORE_HCNon-AC 2.321 2.313 2.325 0.012

Panel B: Control Companies

Pooled Pre-SOX Post-SOX Difference p-value
ACFINEXPERT 0.532 0.417 0.579 0.162 ***
ACDILIGENCE 0.478 0.465 0.483 0.018
BOARDIND 0.524 0.556 0.511 0.044
CEOCHAIR 0.492 0.493 0.492 0.002
ATTENDANCE75 0.914 0.847 0.941 0.093 ***
NOTSTAGGERED 0.518 0.542 0.508 0.033
GOVSCORE_HC 3.458 3.319 3.514 0.195 *
GOVSCORE_HCAC 1.010 0.882 1.062 0.180 **
GOVSCORE_HCNon-AC 2.448 2.438 2.452 0.014

Panel C: Tests for Diff. in Sample versus Control Companies

Pooled Pre-SOX Post-SOX
ACFINEXPERT
ACDILIGENCE ** ***
BOARDIND ** **
CEOCHAIR
ATTENDANCE75
NOTSTAGGERED
GOVSCORE_HC *** ** **
GOVSCORE_HCAC ** **
GOVSCORE_HCNon-AC ** **
*, **, *** Indicate significance at the 10, 5, and 1 percent levels, respectively.
All variables are as defined in Table 1. p-values for tests of differences in sample means are based on two-tailed tests.
The pre-SOX period includes the years 2000–2002 inclusive, and the post-SOX period includes the years 2003–2007
inclusive.

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 181

TABLE 4
Descriptive Statistics for Control Variables Pre- and Post-SOX (n ¼ 996)

Panel A: Sample Companies

Pooled Pre-SOX Post-SOX Difference p-value
a
MOD_ALTMAN 0.160 1.604 0.427 2.030 ***
STICEb 5.025 11.276 2.483 8.793 ***
ABS_DAb 0.092 0.111 0.084 0.027 ***
ABFEES 0.010 0.014 0.009 0.005
FORSEG 1.307 0.986 1.438 0.452 **
ASSET 108.7 39.3 164.4 125.1 ***

Panel B: Control Companies

Pooled Pre-SOX Post-SOX Difference p-value
a
MOD_ALTMAN 0.972 3.457 0.038 3.495 ***
STICEb 3.898 10.868 1.063 9.804 ***
ABS_DAb 0.097 0.118 0.088 0.029 ***
ABFEES 0.008 0.010 0.008 0.002
FORSEG 1.193 1.076 1.240 0.164
ASSET 110.7 40.0 167.3 127.3 ***

Panel C: Tests for Diff. in Sample versus Control Companies

Pooled Pre-SOX Post-SOX
MOD_ALTMANa ** **
STICEb ***
ABS_DAb
ABFEES *** *** ***
FORSEG
ASSET

**, *** Indicate significance at the 5 and 1 percent levels, respectively.

a
For these measures, higher scores indicate lower risk.
b
For these measures, higher scores indicate higher risk.
ASSET is total assets (Compustat data item AT). All other variables are as defined in Table 1. p-values for tests of
differences in sample means are based on two-tailed tests. The pre-SOX period includes the years 2000–2002 inclusive
and the post-SOX period includes the years 2003–2007 inclusive.

Multivariate Results
GOVSCORE_HC
Table 5, Column (1) presents the results from estimating Model (1). The negative and
significant coefficient on GOVSCORE_HC (p ¼ 0.003) suggests that clients with lower (higher)
governance scores were more (less) likely to switch from a Big N audit firm to a non-Big N audit
firm in the pre-SOX period. In addition, the coefficient on the interaction between GOVSCORE_HC
and POSTSOX is not significant at conventional levels (p ¼ 0.117), suggesting that the effect of
corporate governance on auditor-client realignments did not change post-SOX. The joint test for the
post-SOX effect of GOVSCORE_HC (i.e., GOVSCORE_HC þ GOVSCORE_HC  POSTSOX) is

Auditing: A Journal of Practice & Theory

May 2012
182 Cassell, Giroux, Myers, and Omer

TABLE 5
Auditor-Client Realignments Pre- and Post-SOX: GOVSCORE_HC (n ¼ 996)
Columnð1Þ : REALIGN
¼ b1 GOVSCORE HC þ b2 GOVSCORE HCPOSTSOX þ b3 MOD ALTMAN
þ b4 MOD ALTMAN POSTSOX þ b5 STICE þ b6 STICEPOSTSOX
þ b7 ABS DA þ b8 ABS DAPOSTSOX þ b9 ABFEES þ b10 ABFEESPOSTSOX
þ b11 FORSEG þ b12 FORSEGPOSTSOX þ b13 LNASSET
þ b14 LNASSET POSTSOX þ e:

Columnð2Þ : REALIGN
¼ b1 GOVSCORE HCAC þ b2 GOVSCORE HCAC POSTSOX
þ b3 GOVSCORE HCNonAC þ b4 GOVSCORE HCNonAC POSTSOX
þ b5 MOD ALTMAN þ b6 MOD ALTMAN POSTSOX þ b7 STICE
þ b8 STICEPOSTSOX þ b9 ABS DA þ b10 ABS DAPOSTSOX þ b11 ABFEES
þ b12 ABFEESPOSTSOX þ b13 FORSEG þ b14 FORSEGPOSTSOX
þ b15 LNASSET þ b16 LNASSETPOSTSOX þ e:

Panel A: Regression Results

(1) (2)
Prediction Coef. Est. p-value Coef. Est. p-value
GOVSCORE_HC  0.336 0.003***
GOVSCORE_HC  POSTSOX ? 0.219 0.117
GOVSCORE_HCAC  0.568 0.006***
GOVSCORE_HCAC  POSTSOX ? 0.525 0.034**
GOVSCORE_HCNon-AC  0.240 0.043**
GOVSCORE_HCNon-AC  POSTSOX ? 0.074 0.657
MOD_ALTMAN  0.085 0.998 0.080 0.996
MOD_ALTMAN  POSTSOX ? 0.041 0.281 0.036 0.345
STICE þ 0.026 0.027** 0.028 0.019**
STICE  POSTSOX ? 0.034 0.180 0.032 0.208
ABS_DA þ 0.003 0.501 0.119 0.538
ABS_DA  POSTSOX ? 0.140 0.930 0.033 0.984
ABFEES  6.888 0.002*** 6.699 0.002***
ABFEES  POSTSOX ? 2.030 0.458 1.670 0.546
FORSEG  0.016 0.581 0.010 0.552
FORSEG  POSTSOX ? 0.072 0.427 0.079 0.382
LNASSET  1.193 0.080* 1.298 0.063*
LNASSET  POSTSOX ? 0.236 0.818 0.312 0.761
n 996 996
Pseudo R2 0.096 0.100

(continued on next page)

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 183

TABLE 5 (continued)

Panel B: Joint Tests

Prediction Chi-Square p-value Chi-Square p-value
GOVSCORE_HC  2.68 0.051*
GOVSCORE_HCAC  0.16 0.345
GOVSCORE_HCNon-AC  3.37 0.034**
MOD_ALTMAN  3.80 0.975 3.82 0.975
STICE þ 7.70 0.003*** 7.68 0.003***
ABS_DA þ 0.02 0.559 0.02 0.562
ABFEES  11.54 0.001*** 12.11 0.001***
FORSEG  3.58 0.971 3.70 0.973
LNASSET  2.79 0.048** 2.91 0.044**

*, **, *** Indicate significance at the 10, 5, and 1 percent levels, respectively.
All variables are as defined in Table 1. The pre-SOX period includes the years 2000–2002 inclusive and the post-SOX
period includes the years 2003–2007 inclusive.
Dependent variable ¼ 1 if an auditor-client realignment occurred, 0 otherwise. p-values are one- (two)-tailed when a
directional prediction is (is not) made.

negative and significant (p ¼ 0.051), indicating that corporate governance continues to influence
auditor-client realignments during the post-SOX period. The positive and significant coefficient on
STICE (p ¼ 0.002) suggests that auditor-client realignments are more likely for firms with high
litigation risk, and the negative and significant coefficients on LNASSET (p ¼ 0.080) and ABFEES
(p ¼ 0.002) suggest that auditor-client realignments are less likely when clients are larger or more
profitable. The results indicate that these relations hold in both the pre- and post-SOX periods and
that there is no change in these relations between the pre- and post-SOX periods.

GOVSCORE_HCAC and GOVSCORE_HCNon-AC

Table 5, Column (2) presents the results from estimating Model (2). The negative and
significant coefficient on GOVSCORE_HCAC (p ¼ 0.006) suggests that auditor-client realignments
were more (less) likely for clients with lower (higher) scores on audit committee-related governance
components in the pre-SOX period, and the negative and significant coefficient on GOVSCORE_
HCNon-AC (p ¼ 0.043) suggests that realignments were also more (less) likely for clients with lower
(higher) scores on non-audit committee-related corporate governance components during the same
period. However, the coefficient on the interaction between GOVSCORE_HCAC and POSTSOX is
positive and significant (p ¼ 0.034), suggesting that the effect of audit committee-related corporate
governance components on auditor-client realignments is reduced in the post-SOX period.
Moreover, the joint test for the post-SOX effect of GOVSCORE_HCAC (i.e., GOVSCORE_HCAC þ
GOVSCORE_HCAC  POSTSOX) is not significant (p ¼ 0.345), indicating that audit committee-
related corporate governance components do not affect auditor-client realignments during the post-
SOX period. However, the joint test for the post-SOX effect of GOVSCORE_HCNon-AC (i.e.,
GOVSCORE_HCNon-AC þ GOVSCORE_HCNon-AC  POSTSOX) remains negative and significant
(p ¼ 0.034), indicating that non-audit committee-related corporate governance components
continue to influence auditor-client realignments during the post-SOX period. Similar to the results
in Column (1), we find that auditor-client realignments are more likely for clients with higher
litigation risk (STICE) and less likely among larger and more profitable clients (LNASSET and
ABFEES, respectively). The results indicate that these relations hold in both the pre- and post-SOX
periods, and that there is no change in these relations between the pre- and post-SOX periods.

Auditing: A Journal of Practice & Theory

May 2012
184 Cassell, Giroux, Myers, and Omer

Collectively, the results in Table 5 indicate that corporate governance has a significant
influence on auditor-client realignments throughout our sample period. Our results also provide
some evidence suggesting that the influence of corporate governance on auditor-client realignments
diminishes in the post-SOX period (relative to the pre-SOX period). However, the results indicate
that the reduced effect of corporate governance in the post-SOX period is due to the reduced effect
of audit committee-related corporate governance components. This result is consistent with what
would be expected if the audit committee-related rules imposed by SOX and other
contemporaneous regulations reduced the variation in the quality of audit committees across
clients.23

CONCLUSIONS
This paper examines auditor-client realignments observed in the wake of corporate accounting
scandals, intense scrutiny of corporate governance issues following recommendations made by the
BRC, the collapse of Andersen, and the implementation of SOX. Prior research suggests that client
risk characteristics (e.g., earnings manipulation, financial, and litigation risk) are the primary factors
influencing auditor-client realignments. However, because corporate governance-related issues
were prominent following Levitt’s landmark ‘‘Numbers Game’’ speech, in the coverage of
Andersen’s collapse, and during the highly publicized SOX debate, we suggest that client corporate
governance may be an additional risk factor influencing auditor-client realignments.
We examine the factors influencing auditor-client realignments using a sample of clients that
switched from Big N to non-Big N audit firms and a matched sample of Big N clients that did not
switch audit firms. Our results suggest that Big N auditors consider client corporate governance
mechanisms when making client portfolio decisions. Thus, our study documents a potentially
negative outcome of the realignment activity observed in the years surrounding the implementation
of SOX. Specifically, this realignment activity could have implications for the quality of financial
statements if the corporate governance characteristics included in our index—specifically, board
and audit committee independence, diligence, and expertise—are positively associated with
financial reporting quality and if Big N audit firms provide higher quality audits.24
We acknowledge that some prior work suggests that a ‘‘one size fits all’’ approach for corporate
governance is not feasible because differences in governance characteristics could indicate ‘‘better
governance’’ or cross-sectional variation in governance characteristics that exists because optimal
structures differ across clients (see, for example, Shleifer and Vishny 1997; DeFond and Francis

23
Under SOX Section 407, all companies are required to disclose whether there is at least one financial expert
serving on the audit committee. When a financial expert serves on the audit committee, the company must
disclose the name of the expert and indicate whether s/he is independent of management. When there is no
financial expert serving on the audit committee, the company must provide an explanation (see SEC Release No.
33-8177).
24
Note that the studies we cite to support the six characteristics included in our corporate governance index provide
evidence that suggests that, on average, clients with higher scores on these characteristics are less likely to pose
significant governance risk for the auditor. Based on this evidence, we construct a corporate governance index
that makes similar assumptions about the effect of corporate governance characteristics across all clients (e.g.,
more independent boards are preferable, etc.). We acknowledge, however, that evidence on the effects of
corporate governance on financial reporting outcomes is mixed. In particular, in a recent Committee of
Sponsoring Organizations (COSO) study, Beasley et al. (2010) perform univariate tests contrasting the
governance characteristics of fraud versus no-fraud companies and find few significant differences in the
corporate governance mechanisms adopted by the two groups. If the financial reporting risks posed by certain
corporate governance mechanisms are not sufficiently large, we would not expect auditors to respond to these
risks when they make client portfolio decisions.

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 185

2005; Griffin et al. 2008).25 Thus, our paper documents the relation between specific governance
characteristics and auditor-client alignments but does not necessarily address the effect of the
‘‘strength of governance’’ on auditor-client alignments.

REFERENCES
Abbott, L., Y. Park, and S. Parker. 2000. The effects of audit committee activity and independence on
corporate fraud. Managerial Finance 26 (11): 55–67.
Abbott, L., S. Parker, G. Peters, and K. Raghunandan. 2003. The association between audit committee
characteristics and audit fees. Auditing: A Journal of Practice & Theory 22 (2): 17–32.
Abbott, L., S. Parker, and G. Peters. 2004. Audit committee characteristics and restatements. Auditing: A
Journal of Practice & Theory 23 (1): 69–87.
Adams, R. B., and D. Ferreira. 2007. A theory of friendly boards. The Journal of Finance 62 (1): 217–250.
Adams, R. B., and D. Ferreira. 2011. Does regulatory pressure provide sufficient incentives for bank
directors? Evidence from directors’ attendance records. International Review of Finance
(forthcoming).
Altman, E. 1968. Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. The
Journal of Finance 23 (4): 589–609.
Bartov, E., F. Gul, and J. Tsui. 2000. Discretionary-accruals models and audit qualifications. Journal of
Accounting and Economics 30 (3): 421–452.
Beasley, M. S. 1996. An empirical analysis of the relation between the board of director composition and
financial statement fraud. The Accounting Review 71 (4): 443–465.
Beasley, M. S., J. V. Carcello, and D. R. Hermanson. 2000a. Fraud-Related SEC Enforcement Actions
against Auditors: 1987–1997. New York, NY: AICPA.
Beasley, M. S., J. V. Carcello, D. R. Hermanson, and P. D. Lapides. 2000b. Fraudulent financial reporting:
Consideration of industry traits and corporate governance mechanisms. Accounting Horizons 14 (4):
441–454.
Beasley, M. S., J. V. Carcello, D. R. Hermanson, and T. L. Neal. 2010. Fraudulent financial reporting 1998-
2007: An analysis of U.S. public companies. Committee of Sponsoring Organizations of the
Treadway Commission. Available at: http://www.coso.org/documents/COSOFRAUDSTUDY2010_
001.pdf
Beatty, R. 1989. Auditor reputation and the pricing of initial public offerings. The Accounting Review 64
(4): 693–710.
Bebchuk, L. A., J. M. Fried, and D. I. Walker. 2002. Managerial power and rent extraction in the design of
executive compensation. The University of Chicago Law Review 39 (3): 751–846.
Bebchuk, L. A., and J. M. Fried. 2003. Executive compensation as an agency problem. Journal of Economic
Perspectives 17 (3): 71–92.
Bebchuk, L. A., and A. Cohen. 2005. The costs of entrenched boards. Journal of Financial Economics 78
(2): 409–433.
Becker, C. L., M. L. DeFond, J. Jiambalvo, and K. R. Subramanyam. 1998. The effect of audit quality on
earnings management. Contemporary Accounting Research 15 (1): 1–21.
Bedard, J., S. M. Chtourou, and L. Courteau. 2004. The effect of audit committee expertise, independence,
and activity on aggressive earnings management. Auditing: A Journal of Practice & Theory 23 (2):
13–35.
Bedard, J., and K. Johnstone. 2004. Earnings manipulation risk, corporate governance risk, and auditors’
planning and pricing decisions. The Accounting Review 79 (2): 277–304.

25
For example, Lambert et al. (1993), Boyd (1994), and Core et al. (1999) find that CEO compensation is
increasing in the percentage of outside directors in their samples.

Auditing: A Journal of Practice & Theory

May 2012
186 Cassell, Giroux, Myers, and Omer

Blokdijk, H., F. Drieenhuizen, D. A. Simunic, and M. T. Stein. 2006. An analysis of cross-sectional

differences in Big and non-Big public accounting firms’ audit programs. Auditing: A Journal of
Practice & Theory 25 (1): 27–48.
Blue Ribbon Committee on Improving the Effectiveness of Corporate Audit Committees (BRC). 1999.
Report and Recommendations of the Blue Ribbon Committee on Improving the Effectiveness of
Corporate Audit Committees. Available at: http://www.kpmg.com/aci/docs/practices/blueribbon.pdf
Boyd, B. K. 1994. Board control and CEO compensation. Strategic Management Journal 15 (5): 335–344.
Browning, L. 2005. Sorry, the auditor said, but we want a divorce. The New York Times (February 6).
Carcello, J., D. Hermanson, T. Neal, and R. Riley. 2002. Board characteristics and audit fees.
Contemporary Accounting Research 19 (3): 365–384.
Choi, J. H., R. K. Doogar, and A. R. Ganguly. 2004. The riskiness of large audit firm client portfolios and
changes in audit liability regimes: Evidence from the U.S. audit market. Contemporary Accounting
Research 21 (4): 747–785.
Cohen, J., and D. Hanno. 2000. Auditors’ consideration of corporate governance and management control
philosophy in preplanning and planning judgments. Auditing: A Journal of Practice & Theory 19 (2):
133–146.
Cohen, J., G. Krishnamoorthy, and A. M. Wright. 2002. Corporate governance and the audit process.
Contemporary Accounting Research 19 (4): 573–594.
Core, J. E., R. W. Holthausen, and D. F. Larcker. 1999. Corporate governance, chief executive officer
compensation, and firm performance. Journal of Financial Economics 51 (3): 371–406.
Cram, D. P., V. Karan, and I. Stuart. 2009. Three threats to validity of choice-based and matched-sample
studies in accounting research. Contemporary Accounting Research 26 (2): 477–516.
Davila, A., and F. Penalva. 2006. Governance structure and the weighting of performance measures in CEO
compensation. Review of Accounting Studies 11 (4): 463–493.
DeAngelo, L. E. 1981. Auditor size and audit quality. Journal of Accounting and Economics 3 (3): 183–
199.
Dechow, P. M., R. G. Sloan, and A. P. Sweeney. 1996. Causes and consequences of earnings
manipulations: An analysis of firms subject to enforcement actions by the SEC. Contemporary
Accounting Research 13 (1): 1–36.
Dechow, P. M., L. A. Myers, and C. Shakespeare. 2010. Fair value accounting and gains from asset
securitizations: A convenient earnings management tool with compensation side-benefits. Journal of
Accounting and Economics 49 (1–2): 2–25.
DeFond, M. L., and J. Francis. 2005. Audit research after Sarbanes-Oxley. Auditing: A Journal of Practice
& Theory 24 (1): 5–30.
DeFond, M. L., R. N. Hann, and X. Hu. 2005. Does the market value financial expertise on audit
committees and boards of directors? Journal of Accounting Research 43 (2): 153–193.
DuCharme, L. L., P. H. Malatesta, and S. E. Sefcik. 2004. Earnings management, stock issues, and
shareholder lawsuits. Journal of Financial Economics 71 (1): 27–49.
Ettredge, M., C. Li, and S. Scholz. 2007. Audit fees and auditor dismissals in the Sarbanes-Oxley era.
Accounting Horizons 21 (4): 371–386.
Ettredge, M., J. Heintz, C. Li, and S. Scholz. 2011. Auditor realignments accompanying implementation of
SOX 404 ICFR reporting requirements. Accounting Horizons 25 (1): 17–39.
Faleye, O. 2007. Classified boards, firm value, and managerial entrenchment. Journal of Financial
Economics 83 (2): 501–529.
Farber, D. 2005. Restoring trust after fraud: Does corporate governance matter? The Accounting Review 80
(2): 539–561.
Francis, J. R., and J. Krishnan. 1999. Accounting accruals and auditor reporting conservatism.
Contemporary Accounting Research 16 (1): 135–166.
Ghosh, A., and R. Pawlewicz. 2008. The Impact of Regulation on Auditor Fees: Evidence from the
Sarbanes-Oxley Act. Working paper, Baruch College–CUNY.
Gompers, P. A., J. L. Ishii, and A. Metrick. 2003. Corporate governance and stock prices. Quarterly
Journal of Economics 118 (1): 107–155.

Auditing: A Journal of Practice & Theory

May 2012
The Effect of Corporate Governance on Auditor-Client Realignments 187

Griffin, P. A., D. H. Lont, and Y. Sun. 2008. Corporate governance and audit fees: Evidence of
countervailing relations. Journal of Contemporary Accounting and Economics 4 (1): 18–49.
Hogan, C. E., and R. D. Martin. 2009. Risk shifts in the market for audits: An examination of changes in
risk for ‘‘second tier’’ audit firms. Auditing: A Journal of Practice & Theory 28 (2): 93–118.
Hymowitz, C. 2003. How to fix a broken system: A rush of new plans promise to make corporate boards
more accountable; Will they work? Wall Street Journal (February 24).
Imhoff, E. A. 2003. Accounting quality, auditing, and corporate governance. Accounting Horizons 17
(Supplement): 117–128.
Jensen, M. C. 1993. The modern industrial revolution, exit, and the failure of internal control systems.
Journal of Finance 48 (July): 831–880.
Johnson, V. E., I. K. Khurana, and J. K. Reynolds. 2002. Audit-firm tenure and the quality of financial
reports. Contemporary Accounting Research 19 (4): 637–660.
Johnstone, K., and J. Bedard. 2003. Risk management in client acceptance decisions. The Accounting
Review 18 (4): 1003–1025.
Johnstone, K., and J. Bedard. 2004. Audit firm portfolio management decisions. Journal of Accounting
Research 42 (4): 659–690.
Johnstone, K., C. Li, and K. H. Rupley. 2011. Changes in corporate governance associated with the
revelation of internal control material weaknesses and their subsequent remediation. Contemporary
Accounting Research 28 (1): 331–383.
Jones, F. L., and K. Raghunandan. 1998. Client risk and recent changes in the market for audit services.
Journal of Accounting and Public Policy 17 (2): 169–181.
Khurana, I. K., and K. K. Raman. 2004. Litigation risk and the financial reporting credibility of Big 4 versus
Non-Big 4 audits: Evidence from Anglo-American countries. The Accounting Review 79 (2): 473–
495.
Klein, A. 2002. Audit committee, board of director characteristics, and earnings management. Journal of
Accounting and Economics 33 (3): 375–400.
Kothari, S., A. Leone, and C. Wasley. 2005. Performance matched discretionary accrual measures. Journal
of Accounting and Economics 39 (1): 163–197.
Krishnamoorthy, G., A. Wright, and J. Cohen. 2002. Audit committee effectiveness and financial reporting
quality: Implications for auditor independence. Australian Accounting Review 12 (3): 3–13.
Krishnamurthy, S., J. Zhou, and N. Zhou. 2006. Auditor reputation, auditor independence, and the stock
market impact of Andersen’s indictment on its client firms. Contemporary Accounting Research 23
(2): 465–490.
Krishnan, G. V., and G. Visvanathan. 2008. Does the SOX definition of an accounting expert matter? The
association between audit committee directors’ accounting expertise and accounting conservatism.
Contemporary Accounting Research 25 (3): 827–857.
Lambert, R., D. Larcker, and K. Weigelt. 1993. The structure of organizational incentives. Administrative
Science Quarterly 38 (3): 438–461.
Landsman, W. R., K. K. Nelson, and B. R. Rountree. 2009. Auditor switches in the pre- and post-Enron
Eras: Risk or realignment? The Accounting Review 84 (2): 531–558.
Larcker, D. F., and S. A. Richardson. 2004. Fees paid to audit firms, accrual choices, and corporate
governance. Journal of Accounting Research 42 (5): 625–658.
Lear, R. W. 1998. Auditing the audit committee. Chief Executive Magazine (November).
Lee, H. Y., V. Mande, and R. Ortman. 2004. The effect of audit committee and board of director
independence on auditor resignation. Auditing: A Journal of Practice & Theory 23 (2): 131–146.
Levitt, A. 1998. The numbers game. Speech delivered at the NYU Center for Law and Business, New York,
NY, September 28.
Lublin, J. S., and E. MacDonald. 1998. Management: Scandals signal laxity of audit panels. The Wall Street
Journal (July 17).
Lublin, J. S., and E. Nelson. 1998. Cendant audit panel’s ties raise questions. The Wall Street Journal (July
24).

Auditing: A Journal of Practice & Theory

May 2012
188 Cassell, Giroux, Myers, and Omer

Lublin, J. S. 2002. NYSE considers rules to boost power of boards—Fostering the independence of
directors couple improve governance, advisors say. The Wall Street Journal (June 3).
MacDonald, E. 2000. Study finds boards of some web firms are insider-heavy. The Wall Street Journal
(January 21).
Menon, K., and D. D. Williams. 1994. The insurance hypothesis and market prices. The Accounting Review
69 (2): 327–342.
Myers, J. N., L. A. Myers, and T. C. Omer. 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.
Omer, T. C., J. Bedard, and D. Falsetta. 2006. Auditor-provided tax services: The affects of a changing
regulatory environment. The Accounting Review 81 (5): 1095–1117.
Palmrose, Z. 1988. An analysis of auditor litigation and audit service quality. The Accounting Review 63
(1): 55–73.
Park, Y. W., and H. Shin. 2004. Board composition and earnings management in Canada. Journal of
Corporate Finance 10 (3): 431–457.
Pittman, J. A., and S. Fortin. 2004. Auditor choice and the cost of debt capital for newly public firms.
Journal of Accounting and Economics 37 (1): 113–136.
Rama, D. V., and W. J. Read. 2006. Resignations by the Big 4 and the market for audit services. Accounting
Horizons 20 (2): 97–109.
Securities and Exchange Commission (SEC). 2002. Pitt Seeks Review of Corporate Governance, Conduct
Codes. Release No. 2002-23. February 13. Washington, D.C.: SEC.
Securities and Exchange Commission (SEC). 2003. NASD and NYSE Rulemaking: Relating to Corporate
Governance. Release No. 34-48745. November 4. Washington, D.C.: SEC.
Sharma, D. S., E. Boo, and V. Sharma. 2008. The impact of non-mandatory corporate governance on
auditors’ client acceptance, risk and planning judgments. Accounting and Business Research 38 (2):
105–120.
Shleifer, A., and R. Vishny. 1997. A survey of corporate governance. The Journal of Finance 52 (2): 737–
783.
Shu, S. Z. 2000. Auditor resignations: Clientele effects and legal liability. Journal of Accounting and
Economics 29 (2): 173–205.
Shumway, T. 2001. Forecasting bankruptcy more accurately: A simple hazard model. Journal of Business
74 (1): 101–124.
Simunic, D., and M. Stein. 1990. Audit risk in a client portfolio context. Contemporary Accounting
Research 6 (2): 329–343.
Stice, J. 1991. Using financial and market information to identify pre-engagement factors associated with
lawsuits against auditors. The Accounting Review 66 (3): 516–533.
Taub, S. 2004. Big Four seen shedding small clients: Sarbox rules stretch audit firms’ resources, E&Y Chief
Executive claims. CFO.com. September. Available at: http://www.cfo.com
Teoh, S. H., and T. J. Wong. 1993. Perceived auditor quality and the earnings response coefficient. The
Accounting Review 68 (2): 346–366.
Teoh, S. H., I. Welch, and T. J. Wong. 1998a. Earnings management and the underperformance of seasoned
equity offerings. Journal of Financial Economics 50 (1): 63–99.
Teoh, S. H., I. Welch, and T. J. Wong. 1998b. Earnings management and the long-run market performance
of initial public offerings. The Journal of Finance 53 (6): 1935–1974.
Vafeas, N. 2005. Audit committees, boards, and the quality of reported earnings. Contemporary Accounting
Research 22 (4): 1093–1122.
Warfield, T., J. Wild, and K. Wild. 1995. Managerial ownership, accounting choices, and informativeness
of earnings. Journal of Accounting and Economics 20 (1): 61–92.
Xie, B., W. N. Davidson, and P. J. DaDalt. 2003. Earnings management and corporate governance: The role
of the board and the audit committee. Journal of Corporate Finance 9 (3): 295–316.
Zhao, Y., and K. H. Chen. 2008. Staggered boards and earnings management. The Accounting Review 83
(5): 1347–1381.

Auditing: A Journal of Practice & Theory

May 2012
Copyright of Auditing is the property of American Accounting Association and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.