lin and liu - 2009 - the determinants of auditor switching from the perspective of corporate governance in china

The Determinants of Auditor Switching from
the Perspective of Corporate Governance in
China
Z. Jun Lin and Ming Liu*
ABSTRACT
Manuscript Type: Empirical
Research Question/Issue: This paper reports on the association between the internal corporate governance mechanism of
firms and their auditor switching types in the Chinese context. Two types of auditor switching – namely switching to a larger
auditor or switching to a smaller auditor – are identified and examined.
Research Findings/Insights: Controlling shareholders have the incentive to seek opaqueness gains. The empirical results
demonstrate that to realize opaqueness gains, firms with weaker corporate governance generally are more likely to switch
to a smaller auditor rather than to a larger one.
Theoretical/Academic Implications: The empirical results demonstrate that firm-specific corporate governance devices will
affect a firm’s auditor switching decision. An effective corporate governance mechanism may inhibit the controlling
shareholder from switching to a smaller auditor to exploit the minority shareholders. The emerging economies are usually
featured with concentrated ownership and insufficient legal protection of minority shareholders. Compared with prior
studies, this paper generates findings more applicable to the emerging economies.
Practitioner/Policy Implications: This study may facilitate market regulators and participants to maintain close monitoring
of the structural arrangement of corporate governance of the listed firms, the independent auditing process and the
credibility of financial reporting in an emerging market like China. The findings also suggest that in order to bolster the
confidence of the market participants, the Chinese government should promote the reform of the corporate governance
system and enforce effective regulations, in particular on firms’ auditor switches.
Keywords: Corporate Governance, Supervisory Board, Government Ownership, State-Owned Enterprise (SOE),
Auditor Switching, China
INTRODUCTION
T
he purpose of this study is to investigate the association
between a firm’s internal corporate governance mecha-
nism and its auditor switching decisions in the Chinese
context. An independent auditing function can detect and
disclose earnings management and other types of miscon-

duct by business managers or controlling shareholders. In
general, if a firm has established a sound corporate gover-
nance mechanism, the firm’s management or its controlling
shareholders will not have a free hand in making decisions
on auditor selection, and vice versa. Hence, there should be
an association between a firm’s corporate governance and its
auditor selection decision. This study empirically investi-
gates the relationship between a firm’s internal corporate
governance mechanism (proxied by ownership concentra-
tion, effectiveness of the supervisory board [SB] monitoring
and shared board of directors [BoD] chairman and CEO)
and their auditor switching decisions in the context of cor-
porate governance practice in China.
This study was motivated by several factors. First, as cor-
porate governance has a positive impact on corporate finan-
cial reporting and auditing processes, a study of auditor
switching with respect to the internal corporate governance
mechanism may assist analysis of auditing quality and the
auditor’s roles in ensuring the credibility of corporate finan-
cial disclosures. Second, for the controlling owners, there is
a tradeoff between hiring a high-quality auditor to lower the
*Address for correspondence: Department of Accounting and Information Manage-
ment, Faculty of Business Administration, University of Macau, Macau, China. E-mail:

476
Corporate Governance: An International Review, 2009, 17(4): 476–491
© 2009 Blackwell Publishing Ltd
doi:10.1111/j.1467-8683.2009.00759.x
costs of raising capital and hiring a low-quality auditor to
maintain the gains from the opaqueness of corporate gover-

nance (called “opaqueness gains,” such as tunneling behav-
iors to transfer resources from a listed firm to its controlling
shareholder). The ongoing bear market in China during the
2001–2004 period provides a good opportunity to disen-
tangle these two incentives, thus allowing us to pinpoint the
association between a firm’s internal corporate governance
mechanism and its auditor switching decisions. Third, the
Chinese Institute of Certified Public Accountants (CICPA)
began to rank auditors in China in the early 2000s in order to
improve the transparency of the Chinese auditing market,
thus allowingthe possibilityof identifyinghigh-quality audi-
tors in the Chinese market. An investigation of the determi-
nants of auditor switching from the perspective of corporate
governance should contribute to an understanding of the
necessity and utility of independent audits in China.
Our regression results show that firms with larger control-
ling shareholders (a higher degree of ownership concentra-
tion) or firms where the positions of board of directors’
chairman and CEO are held by the same person are more
likely to switch to a smaller auditor rather than to a larger
one. However, the supervisory board monitoring strength is
not a significant factor underlying auditor switching deci-
sions. The findings suggest that firms with a weak internal
corporate governance mechanism generally tend to switch
to smaller or more pliable auditors to sustain the opaqueness
gains. The finding that the supervisory board function
(proxied by its size to represent monitoring effectiveness)
does not have a significant influence on auditor switching
decisions may imply that the supervisory board’s monitor-
ing function is not consistently effective in practice.

The remainder of the paper is arranged as follows. The
second section reviews the relevant literature. The third
section develops the hypotheses to examine the association
between the internal corporate governance mechanism and
auditor switching decisions. The fourth section introduces a
regression model to test the hypotheses. The fifth section
presents and discusses the empirical results. Sensitivity tests
are presented in the sixth section. Finally, the seventh section
provides some conclusions.
LITERATURE REVIEW
The Development of Auditing Profession in China
Shortly after the founding of the People’s Republic of China
in 1949,the auditing profession in China disappeared entirely
due to the public (state) ownership of all production means.
An independent auditing function was virtually nonexistent
in the planned economy before the 1980s when the state
owned and ran enterprises directly. But the mushrooming of
Sino-foreign joint ventures, brought about by the govern-
ment’s adoption of the “open-door” policy in the early 1980s,
led to the emergence of independent auditing. Due to the
involvement of non-state equity interests in joint-ventures, it
became necessary to have independent professionals, or
certified public accountants (CPAs), to verify capital contri-
butions and audit annual financial statements and income
tax returns (Lin, Tang, & Xiao, 2003). Thus, the CICPA,
a quasi-governmental organization in charge of national
administration of CPAs and auditing firms, was established
in the early 1980s. Following the business restructuring cam-
paign, shareholding (stock) companies reappeared in the
Chinese economy at the turn of 1990s, resulting in a further

sharp increase in the demand for external audits. The estab-
lishment of the Shanghai and Shenzhen stock exchanges and
the promulgation of new accounting and auditing standards
also played an important role in this process. The China
Securities Regulatory Commission (CSRC) required that the
annual reports of all listed firms be audited by registered
Chinese CPAs.
The Role of the Auditing Function
In contemporary market economies, business incorporation
leads to the separation of ownership and management.
Professional managers, rather than owners (shareholders),
are directly involved in daily business operations. Due to
various self-interests and information asymmetries, busi-
ness managers are able to pursue their own interests at the
expense of those of the owners and other stakeholders
(Jensen & Meckling, 1976). One of the binding mechanisms
over management operations and information disclosure is
the auditing function performed by independent profession-
als (Watts and Zimmerman, 1986).
Nonetheless, the utility of the auditing function depends
upon quality of the audit which is determined by the inde-
pendence and expertise of the auditors (DeAngelo, 1981;
Watkins, Hillison, & Morecroft, 2004). Audit quality is con-
sidered to be commensurate with the size of the auditors,
i.e., larger auditors should have a higher degree of indepen-
dence, possess more industrial expertise and resources, and
bear higher reputation costs, so they can provide higher-
quality auditing services (DeAngelo, 1981; Lennox, 2005).
Investors perceive the accounting numbers (e.g., earnings
and book values) audited by large auditors to have better

information quality, and therefore attach greater market
value to them (Lennox, 2005; Watkins et al., 2004).
DeFond and Subramanyam (1998) argue that there are
incentives for the controlling shareholders of firms to
pursue their own interests by manipulating the accounting
numbers or transferring resources through “tunneling”
behaviors. Thus, the controlling shareholders will weigh
their own self-interests when making auditor selection deci-
sions (Johnson, La Porta, Lopez-de-Silanes, & Shleifer, 2000;
La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 2002). On the
one hand, selecting a large auditor will signal to the market
that the financial statements are more reliable, thus the firms
may benefit from lower capital-raising costs on the equity or
debt market. On the other hand, large auditors may be more
stringent in detecting and reporting “tunneling behavior”
and hence may deprive the controlling shareholders of their
opaqueness gains (Johnson & Lys, 1990). In particular, when
firms receive unfavorable audit reports, they might initiate
an auditor switch, searching for a more pliable auditor with
the goal of “opinion shopping” (DeFond & Subramanyam,
1998; Johnson & Lys, 1990; Watkins et al., 2004). Auditor
switching may take different forms, including switching to a
smaller auditor and switching to a larger auditor. Nonethe-
less in the extant literature there is a general lack of research
differentiating the two types of auditor switching.
THE DETERMINANTS OF AUDITOR SWITCHING FROM THE PERSPECTIVE OF CORPORATE GOVERNANCE IN CHINA 477
Volume 17 Number 4 July 2009© 2009 Blackwell Publishing Ltd
Corporate Governance and the Auditing Function
Corporate governance evolved with the separation of own-
ership and management underlying the modern corporation

system (Ang, Cole, & Lin, 2000). There are various types of
principal-agent relations, e.g., between owners and manage-
ment, between creditors and owners/management, and
between controlling shareholders and minority sharehold-
ers. A primary objective of corporate governance is to
monitor the behavior of the various interested parties and to
reduce the agency costs underlying the various principal-
agent relations (Karpoff, Malatesta, & Walkling, 1996). Thus,
corporate governance can be defined as “a set of mecha-
nisms, both industrial and market-based, that induce the
self-interested parties of a company to make decisions that
maximize the value of the company to its owners” (Denis &
McConnell, 2003).
An audit provides external monitoring over a firm’s finan-
cial reporting by independent professionals (auditors) and
therefore serves a fundamental role in reinforcing informa-
tion credibility. However, the true effectiveness of external
auditing is subject to the actuality and the development of the
corporate governance environment (Holm & Laursen, 2007).
A sound corporate governance mechanism should ensure
that firms appoint qualified auditors and that the auditors
exercise independent and effective monitoring over the
financial reporting process and attest to the financial state-
ments’ conformity with the Generally Accepted Accounting
Principles (GAAP). Thus, corporate governance plays a role
in enhancing the effectiveness of the audit function.
Francis and Wilson (1988) examined the relationship
between a firm’s agency costs and its demand for audit
quality in the US market. They find that, concentrated own-
ership, a proxy for close alignment of interest between a

controlling shareholder and the listed firm, can substitute
for good audit quality and therefore is associated with
choice of a low-quality auditor. However, in China, as own-
ership is highly concentrated and legal protection of minor-
ity shareholders is insufficient, ownership concentration
may lead to severe entrenchment problems and therefore
represent poor corporate governance.
The association between corporate governance and exter-
nal auditing is an important issue worthy of serious study. In
particular, as the agency relationship, the business adminis-
trative system, and corporate governance practices in emerg-
ing markets such as China differ substantially from those in
the developed economies (Tam, 2000), we are interested in
finding out whether empirical results are similar in the
Chinese market as in the United States. This study will not
only enrich the extant auditing literature, but will also further
promote the development of corporate governance and inde-
pendent auditing practices in the emerging economies.
DEVELOPMENT OF THE HYPOTHESES
Auditing plays two major functions: internally, auditing
can be used to monitor management behavior and reduce
agency costs (Jensen & Meckling, 1976); and externally,
auditing can be used by potential creditors and investors to
evaluate the cost of capital of the company, as described by
the results of a survey of members of the American Institute
of Certified Public Accountants (AICPA) (Carpenter &
Strawser, 1971): “Almost universally, the reason expressed
[for the change in auditors] was that the underwriters
informed the client that a ‘national known auditor’ was nec-
essary to sell their offerings at the highest possible price.”

In China, on average two-thirds of the shares are held by
block shareholders (Lin, Liu, & Zhang, 2007). The block
shareholder has the power to appoint a management team,
and through the management team has access to inside
information. However, because the controlling shareholders
also have the power to select the auditor,
1
the minority
shareholders do not expect to rely on auditing to effectively
monitor management. Hence, the internal function of audit-
ing to reduce owner-manager agency costs does not exist in
China and for this function there is insufficient demand for
audit quality.
As there is no significant bond market in China, listed
firms are primarily financed by equity rather than debt. After
the establishment of Shanghai Stock Exchange and the
Shenzhen Stock Exchange, Chinese listed firms achieved
cumulative financing of RMB 1.16 trillion (RMB 6.80 = U$1)
between 1992 and 2004. In 2000 the total market capitaliza-
tion hit RMB 1.61 trillion, however the bear market thereaf-
ter resulted in the slumping of the market value by RMB .44
trillion (CSRC, 2005). During the weak market period of
2001 to 2004 listed firms were not enthusiastic about offering
new equity securities to the public, therefore the external use
of audits was also substantially diminished.
In contrast, controlling shareholders are motivated to hire
low-quality auditors to seek potential opaqueness gains at all
times, especially if they are operating in a weak corporate
governance environment where there are only loose monitor-
ing and binding contracts (Felo, Krishnamurthy, & Solieri,

2003). To address the research question empirically, we set
out to test the association between the firms’ internal corpo-
rate governance mechanism and their auditor switching deci-
sions, i.e., whether firms with a weak internal corporate
governance mechanism are more likely to switch to a lower-
quality auditor.
If such an association does not exist, internal corporate
governance mechanism may not impact the type of auditor
switches. Alternatively, if auditors of different size offer
monitoring services with varied levels of quality, the firms’
internal corporate governance mechanism may impact the
switch type, in respect of the varied monitoring functions on
the firms’ opaqueness gains. The association may be evi-
dence of the cost-benefit based demand for audit monitor-
ing: when the opaqueness gains outweigh the benefits of
lowering capital raising costs, lower-quality auditors would
be preferred by Chinese firms, and especially by firms with
weak internal corporate governance mechanisms, as these
firms have more opaqueness gains to protect (DeFond,
Wong, & Li, 2000; Lin et al., 2007). By switching to a smaller
auditor, the controlling shareholder (the agent) may com-
fortably exploit the wealth of the minority shareholders
without being watched closely by the auditor. In contrast,
switching to a larger auditor leads to more rigid audit moni-
toring and hence tunneling behaviors will be confined.
We used three proxies to measure a firm’s internal cor-
porate governance mechanism: ownership concentration
(shareholding of the largest owner); the effectiveness of
478 CORPORATE GOVERNANCE
Volume 17 Number 4 July 2009 © 2009 Blackwell Publishing Ltd

supervisory board (size of the supervisory board); and
shared CEO-Chair (whether the board of directors’ chair-
man and CEO positions are held by the same person).
A high ownership concentration is a distinct feature of
listed firms in China. A Chinese listed firm usually has a
large controlling shareholder,
2
who often is the government
or the parent state-owned enterprise.
3
Nonetheless, the own-
ership structure affects corporate governance and corporate
value in many different ways. Johnson et al. (2000) argue that
more narrowly held firms may face greater agency costs
because the controlling shareholders will have a dominant
influence on corporate affairs and they can easily bypass
monitoring by other shareholders. La Porta, Lopez-de-
Silanes, and Shleifer (1999) and La Porta et al. (2002) show
that in the emerging transitional economies, the controlling
shareholders may expropriate the minority shareholders
through aggressive “tunneling” behaviors. They further
argue that “the central agency problem in large corporations
around the world is that of restricting expropriation of
minority shareholders by controlling shareholders” (La
Porta et al., 1999). This is particularly true for Chinese listed
firms in which the controlling shareholders usually hold a
very high percentage of the equity shares.
In China, the controlling shareholders have frequently
intervened in the operations of the listed firms to benefit the
parent companies, e.g., using the listed firms as guarantors

for loan applications for the parent and related companies
and therefore exposing the listed firms to extra financial and
operating risks. In fact, the controlling shareholders of many
listed firms are keen to raise funds only on the stock market.
They frequently engage in benefit transfers through the mis-
appropriation of funds and related-party transactions to
expropriate the interests of the minority shareholders,
which, if detected, may invite external intervention by
minority shareholders and other stakeholders (Lin et al.,
2007). The desire to maximize self-interest through “tunnel-
ing” behaviors leads the listed firms to avoid being moni-
tored by a high-quality auditor. The more concentrated the
ownership structure (i.e., with a larger controlling share-
holder), the weaker the internal corporate governance
mechanism. Therefore, firms with larger controlling share-
holders are expected more likely to switch to pliable
auditors to realize opaqueness gains through tunneling
behaviors or other types of misconduct, as stated below:
Hypothesis 1 (H1): All other things being equal, a Chinese firm
with a higher percentage of total shares held by its controlling
shareholder will more likely switch to a smaller auditor.
Pursuant to the Chinese Company Law, all Chinese firms
adopted a German-style dual-board governance system,
thus each listed firm has both a board of directors and a
supervisory board. The supervisory board is composed of
the shareholders’ representatives (including Chinese Com-
munist Party officials) and an appropriate proportion of
employee representatives, who are nominated by the firm’s
employee union. The Company Law specifically defines the
supervisory board as a monitoring mechanism to carry out a

series of responsibilities, including: (1) monitoring the per-
formance of the directors and senior managers to ensure
compliance with the laws, regulations, and the articles of
incorporation; (2) reviewing the financial affairs of the firm;
(3) requesting the directors and senior managers to alter
and/or rectify their personal activities if deemed in conflict
with the firm’s objectives; (4) proposing specific shareholder
meetings whenever deemed necessary; (5) fulfilling any
other duties that are stipulated in the articles of incorpora-
tion of the firm; and (6) submitting a supervisory board
report to the shareholders’ annual general meeting. The
Standard Code of Corporate Governance for Listed Compa-
nies in China issued by the CSRC and the State Economic
and Trade Commission in 2002 further requires that super-
visory board members should have some professional
knowledge or work experience in the areas of law and
accounting (CSRC, 2002).
Based on the requirements of the Company Law, the
supervisory board shall independently and effectively carry
out supervision over the activities of the directors and the
management as well as examine the financial affairs of the
firm. Such a German-style two-tiered board system with the
co-existence of a board of directors and a supervisory board
has become the backbone of corporate governance in most
Chinese listed firms since the mid-1990s. Using an event
study, Dahya, Karbhari, Xiao, and Yang (2003) report that
investors consider the supervisory board to be an important
device of corporate governance in China. Chen (2005) finds
that there is a positive association between the size of the
supervisory board and the level of corporate governance,

suggesting that a larger supervisory board should be more
effective in carrying out its legitimate monitoring responsi-
bilities. We use the number of supervisory board members
as a proxy for the monitoring effectiveness of the supervi-
sory board and have the second hypothesis:
Hypothesis 2 (H2): All other things being equal, a Chinese firm
with fewer supervisory board members will more likely switch
to a smaller auditor.
Within a sound corporate governance structure, the board
of directors must ensure that the management acts in the
best interests of the shareholders. The board of directors is
responsible for execution of the resolutions passed by the
shareholders’ meetings and for appointing, removing, and
remunerating senior managers. Traditionally, the sharing of
the position of board of directors’ chairman and CEO has
been common in the United States. However, in most Euro-
pean, British, and Canadian businesses, in an effort to
ensure better corporate governance, these two positions are
often split. Combining the two positions does have its
advantages, allowing the CEO multiple perspectives on the
firm as a result of his/her multiple roles and empowering
him/her to act with determination. Nonetheless, this prac-
tice results in less transparency of the CEO’s activities, and
as such his/her actions can go unmonitored, which paves
the way for scandals and corruptions. To the contrary, sepa-
ration of the two positions allows the board of directors’
chairman, on behalf of the stockholders, to be more impar-
tial in overseeing the work of the CEO and the overall per-
formance of management (La Porta et al., 2002; Petra, 2006).
Investors, researchers, and government officials have

gradually accepted the view that the best practices of corpo-
rate governance require the separation of the roles of board
of directors’ chairman and CEO. Such a corporate gover-
nance device has received a boost since 2003. In practice,
THE DETERMINANTS OF AUDITOR SWITCHING FROM THE PERSPECTIVE OF CORPORATE GOVERNANCE IN CHINA 479
Volume 17 Number 4 July 2009© 2009 Blackwell Publishing Ltd
market regulators and professional bodies in many devel-
oped countries require that the two important positions be
separated (Jiraporn, Young, & Davidson, 2005). In 2002 the
CSRC also adopted this requirement in its Standard Code of
Corporate Governance for Listed Companies in China. Con-
sistent with the association between internal corporate gov-
ernance and the auditing function, we have the third
hypothesis stated as the following:
Hypothesis 3 (H3): All other things being equal, a Chinese firm
with the board of directors’ chairman and CEO positions held
by the same person is more likely to switch to a smaller auditor.
RESEARCH METHODOLOGY
Model Specification
We intend to examine the determinants of audit switching
from the perspective of the internal corporate governance
mechanism of Chinese listed firms. Our sample includes
firms that switched auditors from 2001 to 2004. We classified
all firms that switched auditors only once during the 4-year
test period into two types – those switching to a larger
auditor (upward switching, or US firms) and those switch-
ing to a smaller auditor (downward switching, or DS firms)
according to the ranking order of auditors in China, which
was compiled by the CICPA in terms of the CPA firms’
annual audit revenue (see Appendix). As elaborated earlier,

the size of the auditing firm is regarded as an effective
surrogate for the independence and monitoring strength of
the auditors (Copley & Douthett, 2002; DeAngelo, 1981).
Thus we construct a model to test whether the firm’s internal
corporate governance mechanism (proxied by ownership
concentration, supervisory board, and shared CEO-Chair)
is associated with the different types of auditor switch-
ing (namely, upward switching or downward switching).
Downward switching (DS) can be expressed as a function of
the three corporate governance variables in which we are
interested and the related control variables: DS = f (owner-
ship concentration, supervisory board, shared CEO-Chair,
control variables, error terms).
As we classify all auditor switches into two types (upward
switches or downward switches), there are only two values
possible (0 or 1) for the dependent variable. Hence, we run
logit regression, which is used to predict the probability of
an occurrence of an event by fitting the data to a logistic
curve. It makes use of several predictor variables that may be
either numerical or categorical. The logistic function is
useful because it can take as an input any value from nega-
tive infinity to positive infinity, whereas the output is con-
fined to values of either 0 or 1. The dependent variable
represents the exposure to some set of risk factors, whereas
f(dependent variable) = 1/(1 + exp (-dependent variable))
represents the probability of a particular outcome, given that
set of risk factors. The following logit model is used to test
Hypotheses 1 to 3.
DS LSH SB CEOCHR GOV OPI
LNASSET LEV MB

=+ + + + +
+++
ββ β β β β
βββ
01 2 3 4 5
678
+++
+∗+∗+∗
+
ββ
βββ
β
910
11 12 13
14
02 03 04
LOSS NWISS
Yr LSH Yr LSH Yr LSH
YrrSB YrSB YrSB
Yr CEOCHR Yr CEOCH
02 03 04
02 03
15 16
17 18
∗+ ∗+ ∗
+∗ +∗
ββ
ββ
RR
Yr CEOCHR+∗ +

βε
19
04
(1)
Please see Table 1 for a description of the variables in the
model.
Three groups of variables, as discussed below, help
explain why certain factors trigger a certain type of auditor
switching. The first group consists of corporate governance
variables, namely ownership concentration (LSH), super-
visory board (SB), and shared CEO-Chair positions
(CEOCHR). Although the controlling shareholders may seek
to influence auditor selection so as to facilitate their tunnel-
ing behaviors, they are subject to the constraints of the cor-
porate governance structure in place. The second group
consists of company-specific variables that have been tested
or are considered helpful to explain auditor switching in
TABLE 1
Description of Variables
DS = 1 if the firm switches to a smaller auditor; 0 otherwise
LSH = largest owner’s shareholding as a percentage of total shares
SB = number of members of the SB
CEOCHR = 1 if the CEO also holds the position of chairman of BoD; 0 otherwise
GOV = 1 if the largest shareholder is a government agency; 0 otherwise
OPI = 1 if the firm receives an unclean auditor’s opinion for the previous year; 0 otherwise
LNASSET = log of total assets at the end of the previous year
LEV = long-term liabilities divided by total assets at the end of the previous year
MB = market to book ratio at the end of the previous year, calculated as the market value of stocks divided by
the book value
LOSS = 1 if the firm experiences a loss for the previous year; 0 otherwise

NWISS = 1 if there is a new equity issue in the two years immediately after auditor switching; 0 otherwise
Yr02 = 1 if the auditor switching occurs in Year 2002; 0 otherwise
Yr03 = 1 if the auditor switching occurs in Year 2003; 0 otherwise
Yr04 = 1 if the auditor switching occurs in Year 2004; 0 otherwise
480 CORPORATE GOVERNANCE
Volume 17 Number 4 July 2009 © 2009 Blackwell Publishing Ltd
prior studies, including government control (GOV), audi-
tor’s opinion (OPI), size (LNASSET), financial leverage
(LEV), market-to-book ratio (MB), profitability (LOSS), and
new issues (NWISS). As there were new requirements in the
Standard Code of Corporate Governance during the test
period, we incorporate the third group of interactions vari-
ables to capture the yearly impact of the gradual adoption of
these new requirements, namely Yr02, Yr03, and Yr04. With
the interactions terms, we intend to test whether the impact
of corporate governance mechanism on auditor switching
varies with the progress of corporate governance practice.
In the regression, the dependent variable is defined by the
types of auditor switching, thus it is coded 1 if a firm
switched to an auditor that was smaller than its predecessor.
For the independent variables, we expect b
1
(for ownership
concentration) and b
3
(for shared CEO-Chair) to have posi-
tive signs as firms with a high ownership concentration and
a shared board of directors’ chairman and CEO are more
likely to switch to a smaller auditor. But b
2

(for supervisory
board) is expected to be negative as a large or strong super-
visory board may inhibit the firm from switching to a
smaller auditor.
Government controlled and non-government controlled
(privately owned or privately controlled) firms may have
differing corporate governance structures and may also have
different considerations when making auditor switching
decisions. In general, government agencies have a stronger
influence over government-controlled firms and therefore
can more easily access the firm’s financial information
for their decision making (Chan, Lin, & Mo, 2006). There-
fore, government-controlled firms may have less demand for
high-quality independent audits and may have a greater
propensity, compared with non-government controlled
firms, to switch to smaller auditors. Hence we add the vari-
able GOV to capture the effect of government control over
the firm’s auditor switching decisions. This variable is coded
as 1 if the largest owner of the firm is a government agency
and 0 otherwise. It is expected to be positively associated
with a downward switch of auditors.
We also control for the effects the auditor’s opinion, firm
size, financial leverage (risk), growth potential, profitability,
and whether there is new issue of equity after the auditor
switching. Prior auditor switching studies focus mainly on
the auditing markets in the Western countries. Nonetheless,
the basic theories and findings in prior research on auditor
switching should be applicable to this study as well because
the Chinese auditing profession in recent years has gradually
adopted international accounting and auditing standards.

One very common reason cited for auditor switching is
the qualifications of the auditor’s opinions. Prior research
has found that firms receiving unfavorable audit reports are
more likely to switch auditors (DeFond & Subramanyam,
1998). We expect that the auditor’s opinion (receiving an
unfavorable auditor opinion in the prior year = 1) to be posi-
tively related to a downward switch of auditors. Large firms
may be less likely to switch to a smaller auditor, as financial
analysts and the financial press will more closely scrutinize
their auditor switches. Following Friedlan (1994), we use the
log of the total assets to control for the size effect of the firms
and we expect it to be negative in the regression model.
Reed, Trombley, and Dhaliwal (2000) find that firms select-
ing Big 6 auditors tend to be highly leveraged, whereas
Titman and Trueman (1986) predict otherwise. As there are
opposing arguments and findings regarding the association
between a firm’s leverage and its auditor switching, we do
not predict the sign of the coefficient for financial leverage.
We also include the market-to-book ratio to control for the
propensity of growing firms to switch to less conservative
auditors (DeFond & Subramanyam 1998). Moreover, Sainty,
Taylor, and Williams (2002) document that profitability may
affect the selection of auditors. In our model, we expect a net
loss in the prior year (LOSS) to be positively associated with
a downward switch of auditors.
Firms may change auditors (especially from a low-quality
auditor to a high-quality auditor) to increase the marketabil-
ity of new securities (Carpenter & Strawser, 1971). Pae and
Yoo (2001) document a negative relationship between audit
quality and the cost of raising capital, i.e., a firm can reduce

its costs of raising capital by hiring a quality auditor. We
therefore include the variable of new issues (NWISS) in our
regression model, which equals 1 if the listed firm issues
new equity to the public in the two years after its auditor
switching and 0 otherwise. The new issue is used to proxy
for the firm’s intentions to issue new equity at the time of the
auditor switching. Although the firms in our sample gener-
ally did not intend to issue new equity to the public during
our test period, for accuracy and completeness we still
include the variable. Since 2002, new corporate governance
requirements have been announced and implemented,
therefore we add the year dummy variables for 2002 to 2004
to capture the potential impact of these new corporate gov-
ernance devices on the firms’ auditor switching decisions
during the test period.
Sampling
Our sample covers A-share firms that switched auditors
from the beginning of 2001 to the end of 2004.
4
There are two
main reasons to limit the sample firms that made auditor
switches to this time period. The first one is the availability of
the ranking of Chinese auditors, which has been compiled
by the CICPA since 2002. Thus it provides the possibility of
identifying or classifying the different types of auditor
switches (i.e., an upward switch or a downward switch). The
second reason is that during this time period, firms had little
intention to offer equity to the public, therefore the opaque-
ness gains from weak corporate governance significantly
outweighed the benefits from lowering the costs of raising

capital. Hence, the bear market period from 2001 to 2004 is
appropriate to test the association between the firms’ inter-
nal corporate governance mechanism and their auditor
switching decisions. Data were collected from the China
Stock Market and Accounting Research (CSMAR) Database,
the TEJ database (carrying financial information and stock
market data compiled by the Taiwan Economic Journal), and
authoritative national newspapers or magazines designated
by the CSRC to publish financial reports of listed firms, such
as China Securities Daily, Shenzhen Securities Times, and
Shanghai Securities News. The collected data on the sample
firms were cross-checked and verified by different data
sources to ensure their reliability. A description of the data is
provided in Table 2.
THE DETERMINANTS OF AUDITOR SWITCHING FROM THE PERSPECTIVE OF CORPORATE GOVERNANCE IN CHINA 481
Volume 17 Number 4 July 2009© 2009 Blackwell Publishing Ltd
TABLE 2
Description of Data
Panel A1: Sample Selection
Firms switched auditors during 2001–2004 316
Less: Firms with missing data 47
Financial, transportation, and utility firms 11
Firms switched auditors for more than once during 2001–2004 25
Final sample 233
Panel A2: Sample Distribution by Sector and Year
Sector 2001 2002 2003 2004 Total
Industry 49 42 32 17 140
Commerce 10 8 6 5 29
Property 756422
Conglomerate 13

10 9 10 42
Total 79 65 53 36 233
Panel B: Descriptive Statistics of Variables
Variable N Mean Median MIN MAX STD
DS 233 .42 0 0 1 .50
LSH (%) 233 48.94 48.38 .43 89.51 18.11
SB (#) 233 4.21 3 1 12 1.73
CEOCHR 233 .06 0 0 1 .25
GOV 233 .79 1 0 1 .41
OPI 233 .16 0 0 1 .37
LNASSET 233 21.04 20.94 19.03 24.60 .89
LEV 233 .08 .06 .00 .51 .09
MB 233 5.88 3.88 1.27 230.39 15.65
LOSS 233 .12 0 0 1 .33
NWISS 233 .03 0 0 1 .18
Yr02 233 .28 0 0 1 .45
Yr03 233 .23 0 0 1 .42
Yr04 233 .15 0 0 1 .36
482 CORPORATE GOVERNANCE
Volume 17 Number 4 July 2009 © 2009 Blackwell Publishing Ltd
Table 2.
Continued
Panel C: Correlation Coefficient Matrix of Variables
Variable DS LSH SB CEOCHR LNASSET LEV MB LOSS OPI GOV NWISS Yr02 Yr03 Yr04
DS 1.00***
LSH .11† 1.00***
SB .09† 05 1.00***
CEOCHR .13** .07 .01 1.00***
LNASSET 18*** .20*** 07 15** 1.00***
LEV .04 16*** .05 16*** .25*** 1.00***

MB .13** 04 02 03 10† 01 1.00***
LOSS .07 .04 03 05 21*** 08 .01 1.00***
OPI .05 04 .05 07 14** .06 .19*** .44*** 1.00***
GOV 00 .21*** 03 08 .09† .02 01 .00 12** 1.00***
NWISS 07 08 06 05 04 .03 04 07 08 13** 1.00***
Yr02 .05 .02 .06 .03 .04 00 05 .03 04 01 01 1.00***
Yr03 16*** 09† .01 .03 .10†
02 08 .04 04 05 .01 34*** 1.00***
Yr04 .11** 01 00 11** 01 .05 04 .09† .01 .02 02 27*** 23*** 1.00***
***, **, and † denote significance at the 1%, 5%, and 10% levels, respectively.
The variables are defined as below:
DS = 1 if the firm switches to a smaller auditor; 0 otherwise
LSH = the largest owner’s shareholding as a percentage of total shares
SB = number of SB members
CEOCHR = 1 if the CEO also holds the position of the BoD chairman; 0 otherwise
LNASSET = log of total assets at the end of the previous year
LEV = long-term liabilities divided by total assets at the end of the previous year
MB = market-to-book ratio at the end of the previous year, calculated as the market value of stocks divided by the book value
LOSS = 1 if the firm experiences a net loss for the previous year; 0 otherwise
OPI = 1 if the firm receives an unclean (non-standard) auditor opinion for the previous year; 0 otherwise
GOV = 1 if the largest shareholder is a government agency; 0 otherwise
NWISS = 1 if there is a new equity issue in the two years immediately after auditor switching; 0 otherwise
Yr02 = 1 if the switching occurs in Year 2002; 0 otherwise
Yr03 = 1 if the switching occurs in Year 2003; 0 otherwise
Yr03 = 1 if the switching occurs in Year 2004; 0 otherwise
THE DETERMINANTS OF AUDITOR SWITCHING FROM THE PERSPECTIVE OF CORPORATE GOVERNANCE IN CHINA 483
Volume 17 Number 4 July 2009© 2009 Blackwell Publishing Ltd
At the end of 2004, there were 1,387 A-share firms listed
on the two stock exchanges in China, among which 316
firms (22.7 per cent) switched auditors during the four-year

period from 2001 to 2004. This implies that generally a firm
is not willing to switch auditors because of the potential high
costs associated with auditor switching, such as the costs of
searching for and renegotiating with a new auditor and the
potentially unfavorable market responses to an auditor
switch (Reed et al., 2000; Watkins et al., 2004). Panel A of
Table 2 presents the sample size for this study. Financial,
transportation, and utility firms are excluded because the
nature of their operations is very different from that of other
types of firms. We also delete firms that switched auditors
more than once during the four-year period. The frequent
switching of auditors may indicate some serious underlying
reasons that are beyond the scope of this study. Further-
more, firms that switched twice or more may have switched
to a larger auditor at one time and to a smaller auditor the
other time, making it difficult to categorize the type of
switching. The final sample consists of 233 firms.
EMPIRICAL RESULTS
Panel B of Table 2 presents the basic statistics on the tested
variables. Among the 233 sample firms, 134 firms switched
to larger auditors and 99 firms switched to smaller auditors,
based on the rankings prepared by the CICPA. On average,
the largest controlling owners held 48.94 per cent of the total
shares of the sample firms, indicating a high ownership
concentration in the Chinese listed firms. About 79 per cent
of our sample firms were directly owned by the government
or governmental agencies, reflecting that most Chinese
listed firms were originally carved out from state-owned
enterprises and that various government agencies remain
the largest owners of the listed firms. In the sample, 16 per

cent (37/233) of the firms received unfavorable auditor opin-
ions before their auditor switches. This high percentage may
support the assertion that there is an association between
receiving unfavorable auditor reports and switching audi-
tors. The average size of the supervisory board was about
4.21, with a minimum of 1 member and a maximum of 12
members. In 6 per cent of the sample firms (15/233), the
CEO also held the position of board of directors’ chairman.
Very few firms intended to issue equity during the bear
market, as indicated in the table that only 3 per cent (8/233)
of the sample firms issued equity in the two-year period
after their auditor switching.
Panel C of Table 2 presents the correlation coefficient
matrix for the variables used in the regression model. A
downward switch is significantly and positively correlated
with ownership concentration, shared CEO-Chair, and
market-to-book ratio; and it is significantly and negatively
related to firm size. The correlation coefficients among the
independent variables are moderate, with no value exceed-
ing .50 (the largest is .44).
Table 3 provides the empirical results from the regression
which tests whether a firm with a weak internal corporate
governance mechanism is inclined to switch to a smaller
auditor. Since there are only two values possible for the
dependent variable (1 for a downward switch and 0 for an
upward switch), we use a logit regression. With a Chi-square
of 40.03, p < .01, and a pseudo R-square of .21, the regression
model is satisfactory in differentiating firms switching to
smaller auditors from firms switching to larger auditors at
an acceptable level of significance. The probability of a down-

ward switch can be expressed as f(DS) = 1/(1 + exp(-DS)),
i.e., the larger the value of the DS, the higher the probability
of a downward switch.
The coefficient for ownership concentration is positively
significant at the 5 per cent level (Coeff. = .02, Wald =
4.78, p < .05), which supports H1. If under the current situa-
tion the probability of a downward switch f(DS) is exactly .50
(equal probability of an upward switch or a downward
switch), then an increase of 10 per cent of the total sharehold-
ing will cause f(DS) to increase to .55. This suggests that firms
with a higher degree of ownership concentration (i.e., higher
percentage of equity shares held by the largest shareholder)
are more likely to switch to a smaller auditor. Although the
test results seem to be the same as Francis and Wilson 1988
findings, our interpretation is completely different. The rela-
tively high internal ownership in upward switches should
better align the interests between managers and sharehold-
ers, but may not be significant enough to expropriate the
minority shareholders, and hence proxies for effective corpo-
rate governance (Francis & Wilson, 1988; Sabherwal & Smith,
2008). Effective corporate governance can substitute for a
quality auditor, and therefore is associated with switching to
a lower-quality auditor in the United States. In contrast, in
China, due to the awfully high level of concentration, the
divergence between cash flow rights and control rights, and
the ineffective legal protection, entrenchment problems
become detrimental, hence ownership concentration proxies
for poor corporate governance. Poor corporate governance
allows the controlling shareholders more space to hire a
lower-quality auditor to realize their opaqueness gains,

therefore leading to our results – a high ownership concen-
tration, a proxy for poor corporate governance, is associated
with switching to a lower-quality auditor.
The coefficient for the size of the supervisory board is
insignificant (Coeff. = .03, Wald = .05, p > .10), therefore H2
is not supported, indicating that the effectiveness of super-
visory board monitoring may not be related to whether a
firm switches to a larger or a smaller auditor. Another pos-
sible interpretation is that in practice the monitoring role of
the supervisory board is dubious, thus the supervisory
board presently does not have a significant impact on a
firm’s auditor switching decision. In fact, as supervisory
board members are mainly from inside the firm, whether the
supervisory board can effectively play a monitoring role is
controversial. Some researchers contend that the supervi-
sory board is mainly decorative in China (Dahya et al., 2003)
and our findings seem to support this viewpoint.
Consistent with H3, the coefficient for the shared CEO-
Chair is positive and significant at the 5 per cent level
(Coeff. = 1.28, Wald = 3.85, p < .05). If the current status is the
separation of the CEO and the board of directors’ chairman
and the probability of a downward switch f(DS) is exactly
.50, then a change to the combination of the two key roles
will lead f(DS) to increase to .78. Thus, a firm is more likely
to switch to a smaller auditor if its CEO also holds the
position of board of directors’ chairman, as expected.
484 CORPORATE GOVERNANCE
Volume 17 Number 4 July 2009 © 2009 Blackwell Publishing Ltd
TABLE 3
Internal Corporate Governance Mechanism and Switching to a Smaller Auditor – Main Test

DS LSH SB CEOCHR GOV OPI LNASSET LEV MB=+ + + + + + + +
ββ β β β β β β β
01 2 3 4 5 6 7 8
+++ +∗+∗
+∗+
ββ β β
ββ
910 11 12
13 14
02 03
04
LOSS NWISS Yr LSH Yr LSH
Yr LSH YrrSB YrSB YrSB YrCEOCHR YrCEOCH02 03 04 02 03
15 16 17 18
∗+ ∗+ ∗+ ∗ + ∗
βββ β
RRYrCEOCHR+∗ +
βε
19
04
(1)
Intercept LSH SB CEOCHR GOV OPI LNASSET LEV MB LOSS
b
0
b
1
b
2
b
3

b
4
b
5
b
6
b
7
b
8
b
9
Prediction ? +- + ++ - +++
Coefficient 6.42 .02 .03 1.28 .03 11 43 3.56 .13 .18
Wald 2.25 4.78** .05 3.85** .01 .06 4.56** 3.95** 3.70† .12
NWISS Yr02*LSH Yr03*LSH Yr04*LSH Yr02*SB Yr03*SB Yr04*SB Yr02*CEOCHR Yr03*CEOCHR Yr04*CEOCHR
b
10
b
11
b
12
b
13
b
14
b
15
b
16

b
17
b
18
b
19
Prediction - ?????????
Coefficient 31 .01 01 .01 .03 .09 .13 .26 .09 .01
Wald .12 1.18 2.21 2.24 .03 .24 .38 .53 .20 .04
Notes: (1) The Wald-Wolfowitz test examines whether the mean (median) of each variable for Top 10 clients equals to that of non Top-10 clients. ***, **, and † denote
significance at the 1%, 5%, and 10% levels, respectively.
(2) N = 233, Chi-square = 40.03, and Pseudo R-square = .21.
DS = 1 if the firm switches to a smaller auditor; 0 otherwise
LSH = the largest owner’s shareholding as a percentage of total shares
SB = number of SB members
CEOCHR = 1 if the CEO also holds the position of the BoD chairman; 0 otherwise
GOV = 1 if the largest shareholder is a government agency; 0 otherwise
OPI = 1 if the firm receives an unclean (non-standard) auditor opinion for the previous year; 0 otherwise
LNASSET = log of total assets at the end of the previous year
LEV = long-term liabilities divided by total assets at the end of the previous year
MB = market-to-book ratio at the end of the previous year, calculated as the market value of stocks divided by the book value
LOSS = 1 if the firm experiences a net loss for the previous year; 0 otherwise
NWISS = 1 if there is a new equity issue in the two years immediately after auditor switching; 0 otherwise
Yr02 = 1 if the switching occurs in Year 2002; 0 otherwise
Yr03 = 1 if the switching occurs in Year 2003; 0 otherwise
Yr04 = 1 if the switching occurs in Year 2004; 0 otherwise
THE DETERMINANTS OF AUDITOR SWITCHING FROM THE PERSPECTIVE OF CORPORATE GOVERNANCE IN CHINA 485
Volume 17 Number 4 July 2009© 2009 Blackwell Publishing Ltd
The coefficient of government control is positive but
insignificant, so government-controlled firms, compared to

non-government-controlled firms, are more likely to switch
to smaller auditors, but the difference is only marginal. In
addition, the coefficient for the auditor opinion variable
(OPI) is positively related to switching to smaller auditors,
but not at a conventional level of significance. Therefore,
there is insufficient evidence to support the assertion that
Chinese firms receiving unfavorable audit opinions are
inclined to switch to smaller auditors for the purpose of
“opinion shopping.” This is not surprising as other studies
in the United States (Watkins et al., 2004) and China (Chan et
al., 2006) report similar findings.
Consistent with our prediction, firm size is negatively
related to switching to a smaller auditor at the .05 level
of significance (Coeff. = 43, Wald = 4.56, p < .05). This
finding confirms that larger firms are less likely to make
downward switching decisions. The reasons may be larger
firms have more complicated operations and they care more
about their reputation. In addition, both financial leverage
(Coeff. = 3.56, Wald = 3.95, p < .05) and the market-
to-book ratio (Coeff. = .13, Wald = 3.70, p < .10) have positive
and significant coefficients. Therefore, firms with greater
growth potentials and higher financial leverage ratios are
inclined to switch to smaller auditors. The profitability of
firms (proxied by the loss incurred in the prior year) does
not have a significant effect on auditor switching decisions,
nor does a new issue (NWISS). We suggest that the rare
cases of new equity issues (8 out of 233) may be the main
reason for the insignificant result.
As indicated by Table 3, the coefficients for year dummy
variables are not significant. This suggests that although

Chinese listed firms were required to adopt the independent
non-executive director and audit committee practices in
their corporate governance after 2002, auditor switches
during the test period were not significantly affected. Pos-
sible explanations are that most listed firms did not adopt
the new practices until 2004 or the enforcement of the new
practices was not satisfactory during the test period.
In summary, the empirical results support that there is an
association between a firm’s internal corporate governance
mechanism and its auditor switching decisions. Firms with
weaker internal corporate governance (proxied by owner-
ship concentration and a shared CEO-Chair) are inclined to
switch to smaller auditors so that they can shirk from more
stringent audit monitoring and realize opaqueness gains.
This result implies that the opaqueness gains derived from
self-interested activities such as tunneling behaviors might
be a crucial consideration when Chinese firms switch
auditors.
SENSITIVITY TESTS
To examine the robustness of the empirical results, we per-
formed varied sensitivity tests. First, we adopt a more strict
definition for a downward switch (DS), which is defined as
switching from a Top 10 auditor to a non-Top 10 auditor. The
empirical results are similar (Table 4-1). In order to make
sure that the largest shareholder controls the listed firm, we
limit the sample to firms in which the largest shareholders
own a significant percentage of the total shares. After delet-
ing the observations in which the largest owners hold less
than 10 per cent of the total shares, the empirical results
remain qualitatively the same (Table 4-2).

We also adopt alternative proxies to measure the control
variables. We use the log of revenues to proxy for firm size,
the total debt to asset ratio to proxy for financial leverage,
and the increase of assets for growth. After making the
changes for these control variables, the empirical results still
reveal a significantly positive relation between a downward
switch (DS) and ownership concentration (LSH) and
between a DS and a shared CEO-Chair (CEOCHR). Again,
the coefficient for the supervisory board is insignificant.
Therefore, H1 and H3 are robustly supported, and there is
insufficient evidence for H2.
CONCLUSIONS
The purpose of this paper is to investigate the determinants
of Chinese firms’ auditor switching decisions from the per-
spective of their internal corporate governance mechanism.
Three measures are used to proxy for the internal corporate
governance mechanism, including the concentration of own-
ership (shareholding of the controlling owner), the effective-
ness of the supervisory board’s monitoring (proxied by
supervisory board size), and shared CEO and board of direc-
tors’ chairman. We divided all auditor switches during the
2001–2004 period into two types: switching to a larger
auditor and switching to a smaller auditor, and empirically
examined the impact of the corporate governance variables
on the firms’ auditor switching decisions.
Three hypotheses are used to test the association between
the firms’ internal corporate governance mechanism and
their auditor switching decisions. H1 and H3 are supported,
but there is insufficient evidence to support H2. Hence firms
with larger controlling owners or firms in which the posi-

tions of board of directors’ chairman and CEO are held by
the same person are more likely to switch to a smaller
auditor. However, it is inconclusive whether firms with
smaller supervisory boards opt for a smaller auditor. So
generally, we conclude that firms with a relatively weak
internal corporate governance mechanism are more likely to
switch to smaller auditors in order to protect or realize the
opaqueness gains associated with a weak corporate gover-
nance mechanism. That the supervisory board does not have
a significant impact on firms’ auditor switching may indicate
supervisory board members’ inability to monitor effectively,
possibly because they are mainly from inside the firms
and/or do not have sufficient corporate governance
expertise.
This study contributes to the literature on auditing
research. First, although some prior studies have examined
whether there is an association between firms’ auditor
switching decisions and firm-specific characteristics, most of
these studies do not consider the different types of auditor
switching, driven by different motivations. In this study, two
major types of auditor switching are classified, namely
switching to a larger auditor and switching to a smaller one.
Second, compared with the Francis and Wilson, 1988, study
of the US market, our results are more applicable to the
486 CORPORATE GOVERNANCE
Volume 17 Number 4 July 2009 © 2009 Blackwell Publishing Ltd
TABLE 4-1
Internal Corporate Governance Mechanism and Switching to a Smaller Auditor – Sensitivity: Top 10 to Non-Top 10
DS LSH SB CEOCHR GOV OPI LNASSET LEV MB=+ + + + + + + +
ββ β β β β β β β

01 2 3 4 5 6 7 8
+++ +∗+∗
+∗+
ββ β β
ββ
910 11 12
13 14
02 03
04
LOSS NWISS Yr LSH Yr LSH
Yr LSH YrrSB YrSB YrSB YrCEOCHR YrCEOCH02 03 04 02 03
15 16 17 18
∗+ ∗+ ∗+ ∗ + ∗
βββ β
RRYrCEOCHR+∗ +
βε
19
04
(1)
Intercept LSH SB CEOCHR GOV OPI LNASSET LEV MB LOSS
b
0
b
1
b
2
b
3
b
4

b
5
b
6
b
7
b
8
b
9
Pred. ? +- + ++ - +++
Coeff. 12.80 .04 .40 1.33 08 59 -1.21 3.05 .19 22
Wald 3.20† 4.23** .40 4.00** .08 1.27 4.01** .51 .46 .01
NWISS Yr02*LSH Yr03*LSH Yr04*LSH Yr02*SB Yr03*SB Yr04*SB Yr02*CEOCHR Yr03*CEOCHR Yr04*CEOCHR
b
10
b
11
b
12
b
13
b
14
b
15
b
16
b
17

b
18
b
19
Pred. - ? ? ? ??? ? ? ?
Coeff. 18 .01 .01 .02 21 05 28 .35 34 .15
Wald .03 1.11 .58 2.19 .28 .05 .16 1.70 1.31 .60
Notes: (1) The Wald-Wolfowitz test examines whether the mean (median) of each variable for Top 10 clients equals to that of non Top-10 clients. ***, **, and † denote
significance at the 1%, 5%, and 10% levels, respectively.
(2) N = 62, Chi-square = 42.02, and Pseudo R-square = .67.
DS = 1 if the firm switches to a smaller auditor; 0 otherwise
LSH = the largest owner’s shareholding as a percentage of total shares
SB = number of SB members
CEOCHR = 1 if the CEO also holds the position of the BoD chairman; 0 otherwise
GOV = 1 if the largest shareholder is a government agency; 0 otherwise
OPI = 1 if the firm receives an unclean (non-standard) auditor opinion for the previous year; 0 otherwise
LNASSET = log of total assets at the end of the previous year
LEV = long-term liabilities divided by total assets at the end of the previous year
MB = market-to-book ratio at the end of the previous year, calculated as the market value of stocks divided by the book value
LOSS = 1 if the firm experiences a net loss for the previous year; 0 otherwise
NWISS = 1 if there is a new equity issue in the two years immediately after auditor switching; 0 otherwise
Yr02 = 1 if the switching occurs in Year 2002; 0 otherwise
Yr03 = 1 if the switching occurs in Year 2003; 0 otherwise
Yr04 = 1 if the switching occurs in Year 2004; 0 otherwise
THE DETERMINANTS OF AUDITOR SWITCHING FROM THE PERSPECTIVE OF CORPORATE GOVERNANCE IN CHINA 487
Volume 17 Number 4 July 2009© 2009 Blackwell Publishing Ltd
TABLE 4-2
Internal Corporate Governance Mechanism and Switching to a Smaller Auditor – Sensitivity: Ten Per Cent Cutoff
DS LSH SB CEOCHR GOV OPI LNASSET LEV MB=+ + + + + + + +
ββ β β β β β β β

01 2 3 4 5 6 7 8
+++ +∗+∗
+∗+
ββ β β
ββ
910 11 12
13 14
02 03
04
LOSS NWISS Yr LSH Yr LSH
Yr LSH YrrSB YrSB YrSB YrCEOCHR YrCEOCH02 03 04 02 03
15 16 17 18
∗+ ∗+ ∗+ ∗ + ∗
βββ β
RRYrCEOCHR+∗ +
βε
19
04
(1)
Intercept LSH SB CEOCHR GOV OPI LNASSET LEV MB LOSS
b
0
b
1
b
2
b
3
b
4

b
5
b
6
b
7
b
8
b
9
Pred. ? +- + ++ - +++
Coeff. 6.73 .02 .04 1.25 .05 13 44 3.45 .13 .16
Wald 2.48 4.17** .05 3.93** .01 .08 4.66** 3.70† 3.57† .10
NWISS Yr02*LSH Yr03*LSH Yr04*LSH Yr02*SB Yr03*SB Yr04*SB Yr02*CEOCHR Yr03*CEOCHR Yr04*CEOCHR
b
10
b
11
b
12
b
13
b
14
b
15
b
16
b
17

b
18
b
19
Pred. - ? ? ? ??? ? ? ?
Coeff. 37 .00 01 .01 .02 .08 .12 .27 .07 .02
Wald .17 .79 1.99 2.14 .02 .22 .24 .63 .19 .15
Notes: (1) The Wald-Wolfowitz test examines whether the mean (median) of each variable for Top 10 clients equals to that of non Top-10 clients. ***, **, and † denote
significance at the 1%, 5%, and 10% levels, respectively.
(2) N = 230, Chi-square = 38.88, and Pseudo R-square = .21.
DS = 1 if the firm switches to a smaller auditor; 0 otherwise
LSH = the largest owner’s shareholding as a percentage of total shares
SB = number of SB members
CEOCHR = 1 if the CEO also holds the position of the BoD chairman; 0 otherwise
GOV = 1 if the largest shareholder is a government agency; 0 otherwise
OPI = 1 if the firm receives an unclean (non-standard) auditor opinion for the previous year; 0 otherwise
LNASSET = log of total assets at the end of the previous year
LEV = long-term liabilities divided by total assets at the end of the previous year
MB = market-to-book ratio at the end of the previous year, calculated as the market value of stocks divided by the book value
LOSS = 1 if the firm experiences a net loss for the previous year; 0 otherwise
NWISS = 1 if there is a new equity issue in the two years immediately after auditor switching; 0 otherwise
Yr02 = 1 if the switching occurs in Year 2002; 0 otherwise
Yr03 = 1 if the switching occurs in Year 2003; 0 otherwise
Yr04 = 1 if the switching occurs in Year 2004; 0 otherwise
488 CORPORATE GOVERNANCE
Volume 17 Number 4 July 2009 © 2009 Blackwell Publishing Ltd
emerging economies, where there is a high degree of own-
ership concentration and the main agency relationship is
between the controlling and minority shareholders. Third,
while prior research (Jensen & Meckling, 1976) contends

that auditing can serve a major role of coping with agency
problems and reducing agency costs, this function is
not hailed by Chinese firms (usually with highly concen-
trated ownership) when financing opportunities are rare.
This study will be helpful for understanding the recent
development of auditing and corporate governance prac-
tices in emerging markets such as China.
There are also important policy implications from this
study. The quality of independent audits and corporate dis-
closure is identified as an important factor for the Chinese
stock market, which is in transition towards a market-
oriented economy. Our findings on the determinants of
auditor switching in the Chinese context will shed light on
how to improve firms’ corporate governance and audit
monitoring to enhance the credibility of corporate reporting
and to promote smooth development of the capital market.
As CSRC has granted licenses to Qualified Foreign Institu-
tional Investors (QFII) to participate directly in China’s
domestic stock market, the findings suggest that interna-
tional investors need to be aware of the structural arrange-
ment of corporate governance of the listed firms and the
effectiveness of audit monitoring in China.
To bolster the confidence of the market participants, the
Chinese government should promote the reform of corpo-
rate governance of the listed firms and enhance the regula-
tors’ surveillance over the behaviors of the controlling
shareholders. In particular, regulators should closely watch
over firms’ auditor switches, and especially those downward
switches, with an aim of preventing possible expropriation
of minority shareholders’ interests thereafter. As a result, the

auditing profession and the stock market may be able to
develop smoothly in China. In line with the continuing
progress of the economic reforms and business restructur-
ing, the Chinese accounting and auditing practices have
moved towards internationalization rapidly in recent years.
Therefore the Chinese experience involving auditor switch-
ing can be borrowed by other emerging economies in their
development of the independent audit function as well.
In future research, as more Chinese firms report their
auditing fee information, more rigorous simultaneous equa-
tion methods may be used to control both the demand and
the supply sides of the independent auditing function. In
addition, when sufficient data become available, other cor-
porate governance variables, such as the characteristics of
the independent directors and the different committees
under board of directors may be incorporated into the
regression model.
ACKNOWLEDGEMENTS
We are grateful to the editor William Judge, the Guest
Editor, two anonymous reviewers, and the participants of
the Symposium on Corporate Governance in China and
India for their helpful comments and suggestions. We
acknowledge the financial support from Faculty Research
Grants of Hong Kong Baptist University and from the
Research Committee of University of Macau.
NOTES
1. Most listed firms in China did not establish audit committees
responsible for engaging auditors until 2004 or 2005. Before the
establishment of audit committees, the controlling shareholders,
through the management, basically made the auditor selection

decisions.
2. In China, the largest owner (mainly the parent state-owned
enterprise) held, on average, around 50 per cent of the total
equity of listed firms in 2000.
3. Most Chinese listed firms were restructured from former state-
owned enterprises (SOEs). There are normally three types of
equity holding for a Chinese listed firm, namely the state-owned
shares (representing the state’s interest in the firm), the social-
legal-entity shares (mainly the interest of the parent state-owned
enterprise or other social agencies), and the public shares held
by institutional and individual investors, nevertheless, a major
part of the total equity shares is usually controlled by the
state (government agencies) and/or the parent state-owned
enterprise.
4. The stock market in China is segregated. A-share firms are for
domestic investors whereas a small number of B-share firms are
mainly for overseas investors. According to the existing regula-
tions, the financial statements of B-share firms must be audited
by international auditing firms (usually the Big 5/4). There are
few auditor switches for B-share firms because of the limited
availability of alternative auditors. In this paper, firms that issue
both A- and B-shares are excluded.
THE DETERMINANTS OF AUDITOR SWITCHING FROM THE PERSPECTIVE OF CORPORATE GOVERNANCE IN CHINA 489
Volume 17 Number 4 July 2009© 2009 Blackwell Publishing Ltd
APPENDIX
Ranking of Auditors in China
Ranking Auditor Ranking Auditor
1 PwC Zhongtian 44 Jiangsu Tianhua
2 KPMG Huazhen 45 Huayan
3 Deloitte Huayong 46 Gansu Wulian

4 EY Huaming 47 Zhejiang Wanbang
5 Lixin Changjiang 48 Beijing Zhongzhou
6 Yuehua 49 Huazheng
7 Xinyongzhonghe 50 Guangdong Hengxin
8 Beijing Jingdu 51 Nanjing Yonghua
9 Jiangsu Gongzheng 52 Shandong Wanlu
10 EY Dahua 53 Chongqing Tianjian
11 Zhongshen 54 Xi’an Sigma
12 Zhongruihua 55 Jiangsu Tianye
13 Tianzhizixin 56 Anhui Huapu
14 Shanghai Zhonghua 57 Xiamen Tianjian
15 Li’anda 58 Sichuan Huaxin
16 Zhejiang Tianjian 59 Shandong Huide
17 Tianjian 60 Beijing Zhongxingyu
18 Guangzhou Yangcheng 61 Beijing Zhongwei
19 Zhongtianhuazheng 62 Shenzhen Tianjian
20 Shenzhen Pengcheng 63 Fujian Huaxing
21 Shanghai Donghua 64 Beijing Zhongxing
22 Tianyi 65 Shanghai Wanlong
23 Hubei Daxin 66 Shandong Tianhengxin
24 Shanghai Gongxin 67 Yatai Group
25 Jiangsu Suya 68 Zhongtianyin
26 Zhongxi 69 Huajian
27 Zhongxingcai 70 Guangdong Kangyuan
28 Wuhan Zhonghuan 71 Shanghai Tongcheng
29 Zhejiang Dongfang 72 Zhonghengxin
30 Beijing Zhongluhua 73 Beijing Zhongzheng
31 Shanghai Shangkuai 74 Shanghai Shangshen
32 Tianjin Wuzhou 75 Liaoning Tianjian
33 Shandong Zhengyuan 76 Beijing Xinghua

34 Shenzhen Nanfang 77 Beijing Zhongtianheng
35 Shenzhen Dahua 78 Shandong Qianju
36 Guangdong Zhengzhong 79 Sichuan Hongri
37 Hu’nan Kaiyuan 80 Zhongqin Wanxin
38 Beijing Yongtuo 81 Hebei Hua’an
39 Yunnan Yatai 82 Beijing Zhongpingjian
40 Zhonglei 83 Sichuan Junhe
41 Jiangsu Tianheng 84 Shanghai Jiahua
42 Guangdong Tianhua 85 Guangxi Xianghao
43 Beijing Tianhua
(1) The rankings are based on average audit revenues of Year 2002–2004 as compiled by the CICPA.
(2) Auditors must be ranked among the top 100 based on revenues for each year of 2002–2004.
(3) An auditor ranked higher is larger than an auditor ranked lower. It is a downward switch if a listed firm switches from a higher
ranked auditor to a lower ranked one, and vice versa.
490 CORPORATE GOVERNANCE
Volume 17 Number 4 July 2009 © 2009 Blackwell Publishing Ltd
REFERENCES
Ang, J. S., Cole, R. A., & Lin, J. W. 2000. Agency costs and owner-
ship structure. Journal of Finance, 55: 81–106.
Carpenter, C. G. & Strawser, R. H. 1971. Displacement of auditors
when clients go public. Journal of Accountancy, 131(6): 55–58.
Chan, K. H., Lin, K. Z., & Mo, P. L. 2006. A political-economic
analysis of auditor reporting and auditor switches. Review of
Accounting Studies, 11(1): 21–48.
Chen, J. J. 2005. China’s institutional environment and corporate
governance. In M. Hirschey, J. Kose, & A. K. Makhija (Eds.),
Corporate governance: A global perspective, vol. 11: 75–93.
Boston: JAI.
Copley, P. A. & Douthett, E. B., Jr. 2002. The association between
auditor choice, ownership retained, and earnings disclosure by

firms making initial public offerings. Contemporary Accounting
Research, 19(1): 49–75.
CSRC. 2002. Standard code on corporate governance for publicly
listed companies (in Chinese). Beijing, China.
CSRC. 2005. An introduction to China’s securities and futures
markets (in Chinese). Beijing, China.
Dahya, J., Karbhari, Y., Xiao, J. Z., & Yang, M. 2003. The usefulness
of the supervisory board report in China. Corporate Governance:
An International Review, 11: 308–321.
DeAngelo, L. E. 1981. Auditor size and audit quality. Journal of
Accounting and Economics, 3: 183–199.
DeFond, M. L., & Subramanyam, K. R. 1998. Auditor changes and
discretionary accruals. Journal of Accounting and Economics, 25:
35–67.
DeFond, M. L., Wong, T. J., & Li, S. 2000. The impact of improved
auditor independence on audit market concentration in China.
Journal of Accounting and Economics, 28: 269–305.
Denis, D. K. & McConnell, J. J. 2003. International corporate gover-
nance. Journal of Financial and Quantitative Analysis, 38(1):
1–36.
Felo, A. J., Krishnamurthy, S., & Solieri, S. 2003. Audit committee
characteristics and the quality of financial reporting: An empiri-
cal analysis. Working paper, Penn. State University.
Francis, J. R. & Wilson, E. R. 1988. Auditor changes: A joint test of
theories relating to agency costs and auditor differentiation.
Accounting Review, 63: 663–682.
Friedlan, J. M. 1994. Accounting choices of issuers of initial public
offerings. Contemporary Accounting Research, 11(1): 1–31.
Holm, C. & Laursen, P. B. 2007. Risk and control developments in
corporate governance: Changing the role of the external auditor?

Corporate Governance: An International Review, 15: 322–333.
Jensen, M. C. & Meckling, W. H. 1976. Theory of the firm: Mana-
gerial behavior, agency costs, and ownership structure. Journal
of Financial Economics, 3: 305–360.
Jiraporn, P., Young, S. K., & Davidson, W. N., III. 2005. CEO com-
pensation, shareholder rights, and corporate governance: An
empirical investigation. Journal of Economics and Finance, 29:
242–258.
Johnson, S., La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. 2000.
Tunneling. American Economic Review, 90(2): 22–27.
Johnson, W. B. & Lys, T. 1990. The market for audit services: Evi-
dence from voluntary auditor changes. Journal of Accounting
and Economics, 12: 281–308.
Karpoff, J. M., Malatesta, P. H., & Walkling, R. A. 1996. Corporate
governance and shareholder initiatives: Empirical evidence.
Journal of Financial Economics, 42: 365–395.
La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. 1999. Corporate
ownership around the world. Journal of Finance, 54: 471–517.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. 2002.
Investor protection and corporate valuation. Journal of Finance,
57: 1147–1170.
Lennox, C. S. 2005. Audit quality and executive officers’ affiliation
with CPA firms. Journal of Accounting and Economics, 39: 201–
231.
Lin, Z. J., Tang, Q. L., & Xiao, J. 2003. An experimental study of
users’ responses to qualified audit reports in China. Journal of
International Accounting, Auditing, and Taxation, 12(1): 1–21.
Lin, Z. J., Liu, M., & Zhang, X. 2007. The development of corporate
governance in China. The Company Lawyer, 28(7): 195–203.
Pae, S. & Yoo, S. W. 2001. Strategic interaction in auditing: An

analysis of auditors’ legal liability, internal control system quality,
and audit effort. Accounting Review, 76: 333–356.
Petra, S. T. 2006. Corporate governance reforms: Fact or fiction?
Corporate Governance: An International Review, 6: 107–116.
Reed, B. J., Trombley, M. A., & Dhaliwal, D. S. 2000. Demand for
audit quality: The case of Laventhol and Horwath’s auditees.
Journal of Accounting, Auditing and Finance, 15(2): 183–198.
Sabherwal, S. & Smith, S. D. 2008. Concentrated shareholders as
substitutes for outsider analysts. Corporate Governance: An
International Review, 16: 562–577.
Sainty, B. J., Taylor, G. K., & Williams, D. D. 2002. Investor dissat-
isfaction toward auditors. Journal of Accounting. Auditing and
Finance, 17(2): 111–136.
Tam, O. K. 2000. Models of corporate governance for Chinese com-
panies. Corporate Governance: An International Review,8:
52–64.
Titman, S. & Trueman, B. 1986. Information quality and the valua-
tion of new issues. Journal of Accounting and Economics,8:
159–172.
Watkins, A. L., Hillison, W., & Morecroft, S. E. 2004. Audit quality:
A synthesis of theory and empirical evidence. Journal of
Accounting Literature, 23: 153–193.
Watts, R. L. & Zimmerman, J. L. 1986. Positive Accounting Theory.
Englewood Cliffs, NJ: Prentice-Hall.
Z. Jun Lin is a professor of accounting in the School of
Business Administration, Hong Kong Baptist University. His
research interests are financial accounting, international
accounting, auditing, and corporate governance.
Ming Liu is an assistant professor of accounting in the
Faculty of Business Administration, University of Macau.

His research interests are auditing, capital market efficiency,
and corporate governance.
THE DETERMINANTS OF AUDITOR SWITCHING FROM THE PERSPECTIVE OF CORPORATE GOVERNANCE IN CHINA 491
Volume 17 Number 4 July 2009© 2009 Blackwell Publishing Ltd