vanstraelen - 2000 - impact of renewable long-term audit mandates on audit quality

The European Accounting Review 2000, 9:3, 419 442
Impact of renewable long-term audit
mandates on audit quality
Ann Vanstraelen
University of Maastricht
ABSTRACT
Anglo-American countries like the US and the UK allow companies to switch
auditors every year. In contrast, so me continental European countries restrict
auditor switching by allowing only renewable long-term audit mandates. This paper
aims to analyse the impact of renewable long-term audit mandates on audit quality.
Audit quality is considered from the viewpoint of the external users o f the ® nancial
statements. It is questioned whether renewable long-term audit mandates have an
impact on the auditor’s reporting behaviour and on auditor independence.
This research is motivated by the lack of consensus in the literature on the impact
of the length of the auditor client relationship on audit quality. Moreover, few
empirical studies use publicly available secondary data in order to determine
whether perceived threats to auditor independence actually compromise auditor
independence. Therefore, our research methodology consists in the development of
a logistic regression model in which the explanator y variables are measured using
publicly available data. The results of the study suggest that long-term auditor
client relationships signi® cantly increase th e likelihood of an unquali® ed opinion
or signi® cantly reduce the auditor’s willingness to qualify audit reports. A signi® cant
diŒerence was also found between the auditor’s re porting behaviour in the ® rst two
years versus the last year of the audit mandate. Auditors are more willing to issue
an unquali® e d audit rep ort in the ® r st two years of their o cial mandate than in
the last year of their mandate. This could be an indication that the decision to
renew the auditor’s mandate is alre ady taken and known to the auditor before he
has issued his last audit report within his current mandate. The policy implications
of these ® ndings could be in favour of mandatory auditor rotation to maintain the
value of an audit for the external users. However, given recent theoretic evidence
on the adverse eŒects of mandatory auditor rotation, there is a need to develop

alternative measures to safeguard auditors’ independence.
1. INTRODUCTION
Anglo-American countries like the US and the UK allow companies to switch
auditors every year. In contrast, some continental European countries restrict
Address for correspondenc e
Ann Vanstraelen, University of Maastricht, Department of Acco unting and Auditing,
P.O. Box 616, 6200 MD Maastricht, The Netherlands. E-mail: a.vanstraelen
ber® n.unimaas.nl
Copyright 2000 European Accounting Association
ISSN 0963-8180 print/1468-4497 online DOI: 10.1080/0963818002001714 0
Published by Routledge Journals, Taylor & Francis Ltd on behalf of the EAA
420 The European Accounting Review
auditor switching by allowing only renewable long-term audit mandates (e.g.
Belgium, three years; France, six years). This paper aims to analyse the impact
of renewable long-term audit mandates on audit quality. Audit quality is
considered from the viewpoint of the external users of the ® nancial statements.
It is questioned whether renewable long-term audit mandates have an impact
on the auditor’s reporting behaviour and on auditor independence.
This research is motivated by the lack of consensus in the literature on the
impac t of the length of the auditor client relationship on audit quality.
Moreover, empirical evidence concerning the auditor client relationship and
on the extent to which perceived threats to independence are present in current
auditing practice is limited. Indeed, few empirical studies use publicly available
secondary data on the variables of interest in order to determine whether
the perceived threats to auditor independence actually compromise auditor
independence. Therefore, our research methodology consists in the devel-
opment of a logistic regression model in which the explanatory variables are
measured using publicly available secondary data.
This paper is organized as follows. In the ® rst part, audit quality is discussed
from the viewpoint of the external users of the ® nancial statements. Second,

we w il l focus on the auditor client relationship. The third section gives a brief
description of th e audit services market in Belgium where the study was
carried out. In the fourth section the research questions will be formulated.
Sectio n 5 describes the research design and methodology, followed by the
presentation and analysis of the results. Finally, conclusions are drawn and
suggestions are give n for further research .
2. AUDIT QUALITY: COMPETENC E AND INDEPENDENCE
Audit quality is one of the mos t important issues facing the auditing
profession. A higher quality audit reduces the uncertainty associated w ith
® nancial reports prepared by managers (Wallace, 1980). De Angelo (1981)
de® ned audit quality as `the market-assessed joint probability that a given
auditor will both discover a breach in the client’s accounting system and
report the breach’. Obviously, the ® rst condition depends on the auditor’s
technological capabilities whereas the second depends on th e auditor’s inde-
pendence. In other words, `an audit re port is deemed of value if it results from
both a technically competent and independent audit process’ (Citron and
Ta‚ er, 1992). Unfortunately, unlike most goods and services, it is di cult to
directly assess the quality of audited ® nancial statements (Wilson and Grim-
lund, 1990). Since the quality of the auditor’s work is di cult to observe,
surrogate measures have been developed to evaluate audit quality.
Researchers have attempted to evaluate audit quality by means of the fol-
lowing proxies (Carcello et al., 1995):
1 litigation against audit ® rms (e.g. St Pierre and Anderson, 1984;
Palmrose, 1987; Stice, 1991; Carcello and Palmrose, 1994);
Impact of renewable long-term audit mandates 421
2 auditor selection, auditor changes and ® rm size (e.g. DeAngelo , 1981;
Nichols and Smith, 1983; Menon and Williams, 1991; Beattie and
Fearnley, 1995; Simunic and Stein, 1996);
3 nature of auditors’ opinions (e.g. DeAngelo, 1981; Hopwood et al., 1994;
Carcello et al., 1995);

4 pricing of audit services (e.g. Simunic, 1980; Simon, 1985; Palmrose,
1986; Francis and Simon, 1987; Gist, 1994); and
5 perceptions of users (e.g. Shockley and Holt, 1983; ImhoŒ, 1988; Knapp,
1991; Carcello et al., 1992).
Most behavioural studies, based on experiments, have shown that, in general,
auditors are capable to discover errors or breaches in the accounting system
of a company (e.g. Kida, 1980; Campisi and Trotman, 1985; Barnes and
Den Huan, 1993). However, several audit failures gave rise to questions and
concer n about auditors’ independence.
Auditor independence is crucial for the eŒectiveness and success of the
auditing profession (Mautz and Sharaf, 1961). As a result, auditor inde-
pendence has been a key ethical issue for the profession from its beginning.
Although there is no generally accepted precise meaning for the term inde-
pendence (Antle, 1984), there appears to be some consensus in the recognition
of the di chotomy between independence in fact; i.e., substantive independence
and independence in appearance. Independence in fact is generally addressed
in the conceptual discussions on independence, while empirical analysis has
generally attempted to measure independence through an analysis of user
perceptions. Factors that can aŒect perceived independence appear to be: size
of audit ® rm; competition; tenure ; provision of management advisory services;
threat of professional sanctions and legal liability; fear of losing clientele; fear
of loss of reputation and accounting ¯ exibility (Shockley, 1982; Schilder,
1994). However, `this perceived independence is not necessarily the same as
independence in fact, which must be monitore d as carefully’ (Schilder, 1996).
Ultimately, independence in fact is the most important and renders additional
credibility to the audited ® nancia l statements (Falk et al., 1998).
During the past years, the audit quality issue has received increased atten-
tion due to the pressures created by litigation. Litigation against audit ® rms
has indeed increased signi® cantly in recent years (C loyd et al., 1996; Krishnan
and Krishnan, 1997). This is especially true for the US, UK and Australia.

1
In continental Europe, litigation rates are still much lower than in Anglo-
American countries (Kinney, 1994; Gietzmann and Quick, 1998). Farmer et
al. (1987) found that the threat of litigation aŒects auditors’ willingness to take
stands opposing client’s proposed accounting treatments. Various analytical
models were developed to study the impact of liability rules on audit quality
(e.g. Melumad and Thoman, 1990; Dye, 1993; Narayanan, 1994; Schwartz,
1997). Recently, Acemoglu and Gietzmann (1997) showed that if the legal
liability is set too high, the audit market may collapse because it will become
422 The European Accounting Review
too expensive to hire auditors who must insure against potentially very high
liability claims. If the legal liability is set too low, the audit market may
collapse as well because at a certain point shareholders do not value audits any
longer since there is no clear reaso n anymore to believe in the independence of
auditors and the value of audited ® nancial statements.
2
Empirical research
will have to provide evidence on what minimu m limits should be placed on
auditors’ legal liability, in order to make the threat of litigation, controlling
auditors’ self-interested behaviour, trustworthy fo r the users of the audited
® nancial statements (Acemoglu and Gietzmann, 1997).
3. AUDITOR CLIENT RELATIONSHIP
3.1. Regulations of the auditor client relationship in the EU and
the US
Even after introducing the Eighth Directive that aimed to achieve har-
monization within the EU, there still exist diŒerences in regulation. DiŒerences
in regulations with respect to auditor independence are one example. These
diŒerences are possible given the fact that the Member States have the freedom
to ® ll in the broad concepts in thei r own way . Table 1 illustrates some
diŒerences in independence regulations between the US and a number of EU

countries.
As can be seen from Table 1, there are diŒerences in independence regu-
lations between various countries. It should be noted that policy-making must
be seen in a national context. It may be that one country regulates auditor
independence more `eŒectively’ compared with another country. However,
this does not necessarily mean that one system can be imported into another
country (Vieten, 1995). The law is in fact caught in the same dilemma as
privately set accounting standards: whether to regulate through speci® city or
generality (McBarnet and Whelan, 1992).
`Fundamental to al l of the questions about audi tor independence, is an
inherent scepticism about how close the relationship between the auditor and
the management of the client can be withou t creating, in fact or in perception,
a mutuality of interest that could impair the audi tor’s independence’ (Sutton,
1997). The present study will focus on one aspect of the auditor client
relationship, namely the length of the auditor client relationship.
3.2. Impact of the length of the auditor client relationship on audit
quality: no consensus
In the literature, there is no consensus on the impact of the length of the
auditor client relationship on auditor independence.
On the one hand, it is argued that a newly appointed auditor might fail
because of a lack of a thorough understanding of the client. Incumbent
Impact of renewable long-term audit mandates 423
auditors can pro® t from their learning curve eŒect in the detection of a
material error or breach (DeAngelo, 1981). New auditors would make more
mistakes. This idea is re¯ ected in the fact that there appear to be more
litigation cases against auditors with a relative short relationship with their
client. However, it should be noted that no strong evidence for the existence
of a learning curve has been found (O’Keefe et al., 1994).
On the other hand, it is argued that a long auditor client relationship can
decrease audit quality and auditor independence. According to Shockley

(1982), a long auditor client relationship can have the following eŒects:
`Complacency, lack of innovation, less rigorous audit procedures and a
learne d con® dence in the client may arise after long association with the
client’. DeAngelo (1981) assumed that incumbent auditors have economic
incentives not to disclos e material errors or breaches in view of retaining their
client. This practice results from the need of the incumbent auditor to protect
his investment in client-speci® c expertise that is gradually built up during the
years of co-operation. In a similar way, it was suggested that long auditor
tenure is not desirable because it gives `the audit ® rm time to develop a close
relationship with the auditee’ (Whittington et al., 1995). Thus, the auditor’s
incentive to preserve independence declines over time.
Levinthal and Fichman (1988) foun d the following pattern in auditor
client attachments. First, there is a period in which both parties have trust in
each other, the so-called honeymoon period. Soon after this initial period,
the likelihood of termination increases signi® cantly. However, in continuing
relationships the probability of auditor switching becomes increasingly less
likely. Levinthal and Fichman (1988) further found that the more complex an
audit task, the longer an auditor client relationship will be continued. This
is due to the client-speci® c knowledge, developed by the auditor, which is
di cult and costl y to achieve. Furthermore, it appeared that long auditor
client relationships are characterized by lower levels of con¯ ict. Indeed, Lev -
inthal and Fichman (1988) found that the likelihood of a quali® ed opinion
increases just after the so-called honeymoon period, but decreases in con-
tinuing relationships. This is consistent with Knapp (1991) who believes that
after a number of years there is some ki nd of a turning point in the auditor
client relationship which can be detrimental to the auditor’ s independence.
Deis and Giroux (1992) also found a negative relationshi p betw een audit
quality and the length of the auditor client relationship.
Craswell et al. (1995) argued that an audit ® rm has a strong incentive
to provide superior client service in the initial years of the auditor client

relationship. This is due to the uncertainty on the part of client management
about the quality of services to be provided by the ® rm. Moreover, research
has shown that auditors are willing to substantially reduce their fees in order
to acquire new clients (Francis and Simon, 1987). Consequently, new clients
receiving special attention may perceive that they are gettin g outstanding
value for their fee, resulting in a higher overall satisfaction (Behn et al., 1997).
Table 1 Regulations for statutory auditor in various countries
US UK NL GER BEL SP IT FR
Length of the ® rst mandate n/r 1 year
a
n/r n/r 3 years min. 3 yea rs, 3 years 6 yea rs
max. 9 years
Length of the renewed mandate n/r 1 year
a
n/r n/r 3 years 1 year 3 years 6 yea rs
Mandatory external auditor rotation n/r
b
n/r
b
n/r n/r
c
n/r abolished 9 years
d
n/r
MAS allowed for statutory auditors yes yes
e
yes yes
f
no
g

yes
f
no
h
no
Advertisin g allowed yes yes yes no
i
no
j
no no
k
no
Continge nt fees allowed for non-audit work yes yes yes no no no no
l
no
Disclosure of audit fees no yes
m
no no no no yes no
Disclosure of non-audit fees no no no no yes no no
n
no
Limits to ® nancial dependence of client yes yes, 10% yes
o
yes, 30% yes yes yes yes
or 15% during 5
consecutive
years
n/r: not regulated.
a
Private companies have the right to elect to do away with the requirement to re-elect statutory auditors ea ch year. In that case the statutory auditor remains

appointed until positive action is taken to terminate the appointment.
b
Rotation for partners for publicly traded companies is mandatory after seven years.
c
An auditor of a listed public limited company is no longer allowed to sign the audit report if he has signed more than six times within a period of ten years.
In that case, an auditor partner rotation is mandatory.
d
For audits regulated by CONSOB, the Stock Exchange Regulatory Authority, the length of the a ppointm ent of the Regulated Audit Firm is restricted
under Italian Law to a ma ximum of two renewals of the initial period of three ® rm years.
e
The provision of bookkeeping and accounting services is forbidden in the case of listed or public-interest compa nies except if this is of a routine clerical
nature. Audit ® rms within the same legal entity cannot provide legal services.
f
The provision of bookk eeping and accounting services is forbidden.
g
Tax, legal services, consulting, investment and ® nancial advisory serv ices and corporate recovery a re not allowed to be given by an audit ® rm to a statutory
audit client within the same legal entity. The provision of bookkeeping and accounting services is allowe d on an ad hoc and non-recurring basis.
h
Regulated audit ® rms are allowed to provide consulting services in relation to the accounting profession.
i
Ethical advertising (general information on the auditor or audit ® rm) is allowed. However, unethical advertising (advertising that is directl y aimed to g ain
new clients) is not allowed.
j
All forms of advertising to the public in general are forbidden. Advertising on a local scale, if factual and objective, is permitted.
k
Restrictions on advertising only apply to statutory auditors who are members of one of the two professional bodies (Ordine dei Dottori Commercialisti
and Ordine dei Ragionieri) and to the Regulated Audit Firms.
l
Scale fees for certain engagements.
m

Audit ® rms need to disclose fees for statutory audit services.
n
Annual accounts need to disclose fees of the statutory auditors but not the fees of the Regulated Audit Firms.
o
There is no quanti® cation of the maximum share an audit fee of an individual client may take on the total revenues of the audit ® rm.
Source:
Based on Schilder (1994) and Buijink et al. (1996), updated.
426 The European Accounting Review
In a recent analytical model, developed by Acemoglu and Gietzmann
(1997), it was shown tha t long-term relationships enable collusion to develop
between the manager and the auditor due to the fact that the auditor is
concerned about his reappointment. More speci® cally, if the manager can
credibly threaten to dismiss the auditor, then the auditor will choose a low
duty of care and will not report discovered errors or breaches in the client’s
accounting system. By making auditor switching more di cult and securing
a certain minimum engagement period an incumbent auditor is relieved of the
fear of being dismissed during a certain minimum period of time (Falk et al.,
1998). Therefore, advocates of mandatory auditor rotation claim that the
mandatory change of auditor can increase the quality of the audit because
one gets periodically a `fresh look’ at a new auditor. Moreover, it would
decrease the possibilities of the client to in¯ uence the auditor. Opponents of
mandatory auditor rotation claim that this regulation will increase the costs
of audit without creating a signi® cant improvement in audit qualit y (Copley
and Doucet, 1994). Moreover, a long auditor client relationship makes it
possible for the audit ® rm to do the audit more e ciently, which results in
lower costs for the client. Opponents of mandatory rotation also argue that
the decision to switch auditors should be left in the hands of the client .
DaPalma and Deneckere (1995) were among the ® rst who developed a game-
theoretical model of mandatory rotation. It appeared from their analysis that
auditor rotation creates addi tional switching costs, making reputation less

valuable and therefore possibly compromising independence. In line with
these results, Gietzmann and Sen (1996) also found possible adverse eŒects
of mandatory auditor rotation on auditor independence via the eŒects on
reputation. Arrunada and Paz-Ares (1997) provided a detailed overview of
the mandatory rotation debate an d showed that mandatory rotation increases
audit costs and reduces competition in the marketplace. Moreover, Arrunada
and Paz-Ares (1997) argue that it would hamper the two main determinants
of audit quality: auditor technical competence and auditor independence. The
game-theoretical model of Summer (1998) showed that mandatory rotation
undermines the incentives for building up a reputation for independence by
destroying quasi-rents from an ongoing relation. Moreover, mandatory audit
rotation will have the adverse eŒect of strengthening collusive incentives.
4. AUDIT MARKET REGULATION IN BELGIUM
3
Company Law governs the statutory audit of companies. Large companies
4
are required to have their ® nancial statements audited by a member of the
Institute of Auditors.
5
Auditors are appointed by the general meeting of
shareholders on the recommendation of the Board of Directors. In case the
company has a works’ council, they have the right to refuse the appointment
of the nomi nee auditor and defend this position in court. The term of appoint-
ment is three years, which can subsequently be renewed without limitation
Impact of renewable long-term audit mandates 427
for further three-yearly periods. During his mandate, the auditor can only be
dismissed under very exceptional circumstances (e.g. physical incapacity or
negligence resulting in a loss of con® dence). The same applies for the res-
ignation of the auditor during his mandate. Barring grave personal reasons,
the statutory auditor is not permitted to resign during his assignment, except

before the general meeting after having informed the members in writing of
the reasons of the resignation. The resignation of the auditor is required to
be approved by the works’ council if established and for companies under
prudential control by the supervisory organ. There are various audit regu-
lations that should protect the auditor’s independence. For example, all forms
of advertising and unsolicited oŒerings of services to the public in general are
forbidden. It is also prohibited for an auditor to become full-time employed
outside the auditing profession. In that case, the auditor will los e his licence.
There exist strict rules concerning auditor independence with respect to a
particular client. Auditors are not allowed to accept an engagement if they
have been a director or manager of that client in the past three years . More-
over, they cannot accept the engagement neither in case of a personal or
commercial relationship with the client, no r in case of a ® nancial interest in
the client or fee-dependence. Furthermore, audit ® rms are not allowed to
provide non-audit services to an audit client within the same legal entity.
6
Finally, the audit profession has created some mechanisms to monitor its
members. First of all, in order to prohibit the phenomenon of lowballing,
auditors are required to report to th e Institute of Auditors the number of
hours worked for each of their clients and the fee charged to these clients.
Second, peer reviews are organized at least once eve ry ® ve years for each audit
® rm.
The auditor’s report has to be lodged with the ® nancial statements at the
Belgian National Bank and is as such publicly available. In case of a low-
quality audit, legal action against the auditor ca n be undertaken by the client
company, its shareholders or any interested third party. The liability allocation
regime adopted in Belgium is the proportional liability system.
7
There is no
possibility to reduce the auditor’ s liability since there is no legal liability cap

and it is not allowed to arrange a contractual liability cap. It should be noted
that litigation rates in Belgium are low. This is typical for countries which
have government-prescribe d accounting standards that are rather conserva-
tive, while banks or the government are the major providers of capital (Mueller
et al., 1994). In this sense, Belgium is a typical continental European country:
laws govern accounting; banks and other ® nancial institutions play a central
role in corporate ® nancing; ® nancial reporting is strongly in¯ uenced by tax
considerations and is creditor oriented.
5. RESEARCH QUESTIONS
Empirical evidence on the extent to which potential threats to independence
are actually compromising independence is still limited. Indeed, the majority
428 The European Accounting Review
of studies consider independence in terms of perceptions of independence.
These studies do not measure independence in fact but only beliefs about
independence. The present study focuses on auditor client relationships and
audit quality in practice.
It is questioned whether renewable long-term audit mandates can aŒect
auditor independence. It is assumed that long-term audit mandates result on
average in a longe r auditor’ s tenure compared to one-year audit mandates.
This assumption is based on the fact that in case of renewal of the auditor
mandate, it is again for a ® xed number of years (e.g. Belgium, three years;
France, six years). Moreover, auditor switching in countries with long-term
audit mandates appears to be rather exceptional. Two research questions are
put forward. The ® rst research question is the following:
RQ1: Are auditors which have a long relationship with their client less
willing to issue quali® ed audit reports compared to auditors with a
shorter tenure?
It could be expected that the longer the relationship between the auditor and
the client, the more likely the auditor will issue an unquali® ed opinion. This
expectation is based on the belief that long tenure could potentially aŒect

auditor independence. Therefore, it is hypothesized that :
H1: The longer the auditor client relationship, the less likely the auditor
will issue an `unclean’ audit report (quali® ed opinion, adverse opinion
or disclaimer of opinion), all other things equal.
The second research question is the following:
RQ2: Does the auditor reporting behaviour in the ® rst two years of the
auditor’s mandate diŒer from the last year of the auditor’s mandate?
It c ould be expected that auditors are more willing to make compromises with
client’s management in the last year of their o cial mandate in th e hope of
renewing their mandate. Therefore, it is hypothesized that:
H2: Auditors are more willing to give an unquali® ed opinion in the last
year of their o cial mandate compared to the previous years, all other
things equal.
6. RESEARCH DESIGN AND METHODOLOGY
6.1. Sample design
The empirical analysis is done on the basis of two samples taken from CD-
ROMs containing the annual accounts of companies submitted to the Belgian
National Bank over the period 1992 6. For each year , two samples were
drawn. The ® rst sample contains ® nancially stressed non-bankrupt large com-
Impact of renewable long-term audit mandates 429
panies and the second sample contains ® nancially non-stressed non-bankrupt
large companies. Previous research has shown the importance of controlling
for the impact of the ® nancial health of the client on the auditor’s reporting
behaviour (Hopwood et al., 1994). Based upon commo n criteria in the litera-
ture (Kida, 1980; Mutchler, 1985; Hopwood et al., 1994) a company is con-
sidered to be a ® nancially stressed ® rm if it either experienced a negative
working capital in the current year or suŒered a loss from operations, a
bottom-line loss or a de® cit in retained earnings in the current year or previous
two years. Menon and Schwartz (1985) stressed the importance of matching
control groups of companies by year, industry and size. Therefore the two

samples were matched by: year, industr y (using NACE codi® cation, four
digits) and size (based on total assets). The audit report was collected and
examined for each company of the samples. Table 2 illustrates the sample
proportions and the type of audit reports issued for the entire period
1992 6.
6.2. Model variables and measurement
Table 3 gives an overview of the dependent and explanatory variables and the
way in which they are measured. The dependent bi nary variable is an `unclean’
audit report, coded 1 in case of a quali® ed opinion, adverse opinion or
disclaimer of opinion, and coded 0 in case of an unquali® ed opinion. The
independent variables of interest for this research are the length of the auditor
client relationship (tenure) and whether or no t the auditor is in the last year
of his o cial mandate. However, other variables that could aŒect the auditor’s
reporting behaviour are also included in the model. More speci® cally, auditor
independence could be aŒected by:
The revenues received from auditing a client (measured by the variables
FEE and ADITFEES): given that the auditor acts as a rational economic
agent, it can be expected that the revenues of auditing a client have an
impac t on the auditor’s reporting behaviour.
The probability of discovery of an incorrect audit opinion (measured by
the variable FINC): ® nancially stressed companies are more likely to
attract the attention of outsiders and consequently increase the probability
of discovery of an incorrect audit opinion.
Fear of loss of the client (measured by the variable CLIENTLOSS): if
the audit ® rm has recently lost audit clients, it may be inclined to minimize
the risk of losing more. In other words, the recent loss of audit clients
may increase the economic interest of the audit ® rm in the remaining
clients (Louwers, 1998).
Fear of loss of reputation (measured by the variable B6NB6): it has been
shown that Big Six audit ® rms are to be associated with a higher quality

430 The European Accounting Review
Table 2 Sample proportions and type of audit reports issued for the entire period
1992 6
Sample Sample
Stressed Non-stressed
® rms ® rms Tot al
YEAR 1992 n 73 n 73 n 146
unquali® ed audit report: 75.3% 95.9% 86%
other than unquali® ed audit report:
quali® ed opinion: 21.9% 4.1% 13%
adverse opinion: 0% 0% 0%
disclaimer of opinion: 2.7% 0% 1%
Sample Sample
Stressed Non-stressed
® rms ® rms Tot al
YEAR 1993 n 84 n 84 n 168
unquali® ed audit report: 78.8% 95.2% 87%
other than unquali® ed audit report:
quali® ed opinion: 16.5% 3.6% 10%
adverse opinion: 0% 0% 0%
disclaimer of opinion: 4.7% 1.2% 3%
Sample Sample
Stressed Non-stressed
® rms ® rms Tot al
YEAR 1994 n 56 n 56 n 112
unquali® ed audit report: 71.9% 91.1% 81%
other than unquali® ed audit report:
quali® ed opinion: 22.8% 5.4% 15%
adverse opinion: 0% 0% 0%
disclaimer of opinion: 5.3% 3.6% 4%

Sample Sample
Stressed Non-stressed
® rms ® rms Tot al
YEAR 1995 n 91 n 91 n 182
unquali® ed audit report: 77.2% 93.5% 85%
other than unquali® ed audit report:
quali® ed opinion: 18.3% 5.4% 12%
adverse opinion: 0% 0% 0%
disclaimer of opinion: 4.3% 1.1% 3%
Impact of renewable long-term audit mandates 431
Table 2 Continued
Sample Sample
Stressed Non-stressed
® rms ® rms Tot al
YEAR 1996 n 94 n 94 n 188
unquali® ed audit report: 76% 95.7% 86%
other than unquali® ed audit report:
quali® ed opinion: 20.8% 3.2% 12%
adverse opinion: 0% 0% 0%
disclaimer of opinion: 3.1% 1.1% 2%
Sample Sample
Stressed Non-stressed
GLOBAL: ® rms ® rms Tot al
1992 6 n 398 n 398 n 796
unquali® ed audit report: 76.2% 94.5% 85%
other than unquali® ed audit report:
quali® ed opinion: 19.9% 4.3% 12%
adverse opinion: 0% 0% 0%
disclaimer of opinion: 4% 1.3% 3%
level (e.g. DeAngelo, 1981; Davidson and Neu, 1993). This could be the

result of the fear that their reputation will be aŒected, when discovery of
an unreported breach is mad e public.
6.3. Methodology
Each of the suggested explanatory variables will ® rst be teste d in a univariate
way to see whether there are signi® cant diŒerences between companies with a
clean audit report in comparison to companies with an unclean audit report.
Subsequently, to assess the incremental contribution of each variable in the
auditor’s reporting decision, the explanatory variables will be tested in a
multivariate way. Therefore a logistic regression model is developed, given the
specia l estimation problems related to binary-dependent variables (Maddala,
1989). The speci® cation of th e logistic regression model looks as follows:
UNCLEAN
i 0
+
1
FINC
i
+
2
ADITFEES
i
+
3
FEE
i
+
4
CLIENTLOSS
i
+

5
B6NB6
i
+
6
TENURE
i
+
7
MANDATE
i
+
i
432 The European Accounting Review
Table 3 Model variables and measurement
Variables Description Measurement
Dependent
UNCLEAN
i
Unclean audit report, i.e. Binary variable
quali® ed opinion, adverse UNCLEAN
i
1 in case of
opinion or disclaimer of unclean audit report, else
opinion; a clean audit report is UNCLEAN
i
0
an unquali® ed opinion
Independent
FINC

i
Financial condition of company i General discriminant score (D-
score) of a standard bankruptcy
model developed for Belgian
companies
a
ADITFEES
i
Additional audit fees
b
received Natural logarithm of the amount
by client i paid in Belgian francs by the
client for additional se rvices
provided by his auditor/audit
® rm
FEE
i
The natural logarithm of the Natural logarithm of: total
® rm audit fee for client i assets+ operational
revenues+ ® nancia l revenues
c
CLIENTLOSS
ij
Clients lost by audit ® rm j of Number of clients (scaled by the
company i during the previous annual number of ® rm clients)
year lost during the previous year
(loss  ; gain + )
B6NB6
i
Company i has a Big Six auditor Binary variable

or a Non-Big Six auditor B6NB6
i
1 in case of a Big Six
auditor, else B6NB6
i
0
TENURE
ij
Length of the auditor client The number of years that audit
relationship irm j has been engaged by client i
MANDATE
i
Indicates in which year of his Binary variable
engagement period (mandate) MANDATE
i
1 in case auditor
the auditor is (® rst two years or is in the last year of his
last year) engagement period, else:
MANDATE
i
0
Notes:
a
The D-score is calculated from the general multiple linear discriminant model (developed for
Belgian companies by Ooghe and Verbaere, 1982) consisting of the following ratios: accumulated
pro® t (loss) and reserves/total liabilities; taxes and social security charges/short-term external
liabilities; cash/restricted current assets; work in progress and ® nished goods/restricted current
assets; short-term ® nancial debts/short-term extern al liabilities. The D-score of the general bank-
ruptcy prediction model has a prediction accuracy of 82.8% for failing companies when using
the optimal cut-oŒpoint of D-score 0.1304.

b
The provision of management advisory services by the statutory auditor to his client is very
restricted in Belgium. However, additional audit fees can be obtained when the client asks the
auditor to perform other control assignments which ar e recognized by Company Law (e.g.
contribution in kind, quasi-contribution). The fees received for these additional control assign-
ments must be published in the Report of the Board of Directors.
c
In Belgium, scales for audit fees, developed by the Belgian Institute for Auditors, are based on
the average number of audit working hours, which is considered to depend on the sum of total
assets, operational and ® nancial revenues. It should be noted, though, that these scales are only
indicative and may not correspond with the actua l fees charged in practice.
Impact of renewable long-term audit mandates 433
7. RESULTS AND A NALYSIS
7.1. Univariate analysis
In order to obtain a preliminary look at between-group diŒerences, each of
the suggested explanatory variables is ® rst tested in a univariate way to see
whether there are signi® cant diŒerences between companies with a `clean’
audit report (i.e. an unquali® ed audit report) and companies with an `unclean’
audit report (i.e. a quali® ed opinion, adverse opinion or disclaimer of opinion).
The results of the Kolmogrov Smirnov test clearly showed that the explana-
tory variables do not follow a normal distribution. Therefore the non-para-
metric Mann Whitney test is used for the interval- and ratio-scaled variables
to compare the means of the two independent groups. For the nominal and
ordinal independent variables the chi-square test is used. Tables 4a and 4b
present respectively the results of the univariate analysis for the entire sample
of ® nancially stressed companies and the entire sample of ® nancially non-
stressed companies over the period 1992 6. These tables highlight some basic
diŒerences between companies that received an `unclean’ audit report and
companies that received a `clean’ audit report.
Table 4a Univariate analysis by grouping variable UNCLEAN (Unclean audit report)

for the entire sample ® nancially stressed companies over the period 1992 6
UNCLEAN audit CLEAN audit Mann Whitney U
report report (Asymptotic
Mean rank Mean rank signi® cance
(Sum of ranks) (Sum of ranks) 1-tailed)
FINC 150.64 211.17 9,750.5
(14,310.5) (62,717.5) (0.000)
FEE 185.67 199.96 13,079
(17,639) (59,389) (0.142)
CLIENTLOSS 197.54 194.19 13,626.5
(18,568.5) (57,286.5) (0.40)
ADITFEES 194 197.3 13,870
(18,430) (58,598) (0.102)
TENURE 180.61 200.30 12,597.5
(17,157.5) (59,087.5) (0.0665)
UNCLEAN audit CLEAN audit Pearson
2
report report (Asymptotic
signi® cance
1-tailed)
B6NB6 B6: 30.9% B6: 32.2% 0.060
NB6: 69.1% NB6: 67.8% (0.403)
MANDATE First 2 years: 67.4% First 2 years: 76.3% 2.968
Last year: 32.6% Last year: 23.7% (0.042)
434 The European Accounting Review
Table 4b Univariate analysis by grouping variable UNCLEAN (Unclean audit report)
for the entire sample ® nancially non-stressed companies over the period 1992 6
UNCLEAN audit CLEAN audit Mann Whitney U
report report (Asymptotic
Mean rank Mean rank signi® cance

(Sum of ranks) (Sum of ranks) 1-tailed)
FINC 158.82 198.74 3,241
(3,494) (73,534) (0.000)
FEE 205.73 195.95 3,867
(4,526) (72,502) (0.347)
CLIENTLOSS 191.73 193.08 3,965
(4,218) (70,087) (0.478)
ADITFEES 188.5 196.98 3,894
(4,147) (72,881) (0.16)
TENURE 117.30 198.11 2327.5
(2,580.5) (72,110.5) (0.0005)
UNCLEAN audit CLEAN audit Pearson
2
report report (Asymptotic
signi® cance
1-tailed)
B6NB6 B6: 13.6% B6: 27.2% 1.962
NB6: 86.4% NB6: 72.8% (0.08)
MANDATE First 2 years: 68.2% First 2 years: 76.3% 0.805
Last year: 31.8% Last year : 23.7% (0.185)
First of all, it appears in both samples that `tenure’ , one of the main
variables of interest for the purpose of this study, seems to be a signi® cant
factor in the reporting behaviour of the auditor. Companies that received a
clean audit report have a signi® cantly longer relationship with the incumbent
auditor compared with companies that received an unclean audit report.
The variable MANDATE, which is the other main variable of interest for
this study, appears to diŒer as well between companies with a `clean’ audit
report and companies with an `unclean’ audit report, though only signi® cantly
in the sample ® nancially stressed companies. The direction of the relation is
in contrast to our expectations. Indeed, it appears that auditors are more

willing to issue an `unclean’ audit report in the last year of their o cial
audit mandate than in the previous years. However, it was hypothesized that
auditors would be more willing to issue an unquali® ed audit report in the last
year of their o cial mandate compared with the previous years in the hope
of renewing their mandate. This result could imply that the decision to renew
Impact of renewable long-term audit mandates 435
the mandate of the auditor is already taken and known to the auditor before
he issues his last audit report within his current mandate.
A signi® cant control variable in both samples is the ® nancial condition of
a company. Indeed, a bad ® nancial condition is a signi® cant factor in increas-
ing the auditor’s likelihood to issue an `unclean’ audit report. The control
variable Big Six/Non-Big Six audit ® rm is only signi® cant in the sample
® nancially non-stressed ® rms. No signi ® cant diŒerences were found for the
other control variables in either sample.
7.2. Logistic regression results
Logistic regression analysis facilitates the assessment of th e incremental con-
tribution of each variable in the auditor’s decision to issue a `clean’ or `unclean’
audit report. Table 5 presents the logistic regression result s for the entire
period 1992 6, both for the two samples separately as well as for the pooled
sample. The model’s chi-square values are all signi® cant indicating that the
observations are well ® t by the model. The relatively high association of
predicted probabilities with observed responses also illustrates this.
The results provide evidence for the main variable of interest for thi s study,
i.e. tenure. Consistent with the results of the univariate analysis, a long
auditor client relationship seems to signi® cantly increase the likelihood of
the auditor to issue an unquali® ed audit report. The results in Table 5 show
that the tenure variable is signi® cant not only in the pooled sample, but also
in the separate samples. Therefore, there is little reason to believe that the
result of long tenure decreasing the likelihood of an unclean audit report is
confounded by the sample selection.

Consisten t with the results of the univariate analysis, the results of the
logistic regression analysis for the sample of ® nancially stressed ® rms show
that auditors are more willing to issue a `clean’ audit report in the ® rst two
years of their o cial mandate than in the last year of their mandate. As argued
above, this could be an indication that the decision of renew al of the auditor’s
mandate is already taken and known to the auditor before he has issued his
last audit report within his current mandate. The same result is obtained for
the pooled sample. It is noted that, consistent with the univariate analysis, the
coe cient of the variable mandate is not signi® cant for the sample ® nancially
non-stressed ® rms. One explanation could be that auditor switching rates for
® nancially non-stressed ® rms are rather low. In contrast, one would expect
higher auditor switching rates for ® nancially stressed ® rms, engaging in so-
called opinion shopping.
The results of the logistic regression analysis further show that a signi® cant
control variable is a bad ® nancial condition of the client. Finally, as can be
seen in Table 6, the results are not biased due to correlations among the
covariates.
436 The European Accounting Review
Table 5 Logistic regression results for the entire period 1992 6
Dependent variable: UNCLEAN (Unclean audit report 1)
Sample ® nancially stressed ® rms
Parameter Standard Wald Signi® canceVariable
estimate error chi-square p
CONSTANT  1.485 8 1.2430 1.4287 0.2320
FINC  0.2638 0.0611 18.6519 0.0000***
ADITFEES  0.3756 0.7361 0.2604 0.6098
FEE 0.0473 0.0968 0.2389 0.6250
CLIENTLOSS 0.0021 0.0040 0.2737 0.6009
B6NB6  0.1703 0.2751 0.3836 0.5357
TENURE  0.1459 0.0731 3.9857 0.0459**

MANDATE 0.5689 0.2764 4.2349 0.0396**
*** p 1% ;
** p 5% ;
* p 10% .
 2 log likelihood: 392.390 Model chi-square: 34.971
Degrees of freedom: 7 Signi® cance: 0.000
Prediction accuracy: 76.29%
Sample ® nancially non-stressed ® rms
Parameter Standard Wald Signi® canceVariable
estimate error chi-square p
CONSTANT  1.635 2 2.5688 0.4052 0.5244
FINC  0.3025 0.1728 3.0624 0.0801*
ADITFEES  0.5481 1.9627 0.0780 0.7800
FEE 0.0920 0.1994 0.2129 0.6445
CLIENTLOSS  0.0017 0.0072 0.0583 0.8093
B6NB6  0.7182 0.6501 1.2205 0.2693
TENURE  0.5352 0.1559 11.7880 0.0006***
MANDATE 0.4885 0.5031 0.9426 0.3316
*** p 1% ;
** p 5% ;
* p 10% .
 2 log likelihood: 146.268 Model chi-square: 22.269
Degrees of freedom: 7 Signi® cance: 0.0023
Prediction accuracy: 94.27%
Impact of renewable long-term audit mandates 437
Table 5 Continued
Pooled sample (® nancially stressed+ ® nancially non-stressed ® rms)
Parameter Standard Wald Signi® canceVariable
estimate error chi-square p
CONSTANT  1.502 3 1.0950 1.8821 0.1701

FINC  0.3770 0.0561 45.1736 0.000***
ADITFEES  0.5124 1.0473 0.2394 0.6246
FEE 0.0500 0.0856 0.3410 0.5593
CLIENTLOSS 0.0013 0.0035 0.1348 0.7136
B6NB6  0.2432 0.2470 0.9690 0.3249
TENURE  0.2428 0.0649 13.9806 0.0002***
MANDATE 0.5640 0.2379 5.6187 0.0178**
*** p 1% ;
** p 5% ;
* p 10% .
 2 log likelihood: 561.443 Model chi-square: 88.438
Degrees of freedom: 7 Signi® cance: 0.000
Prediction accuracy: 85.23%
Table 6 Correlation matrix for pooled sample (® nancially stressed+ ® nancially non-
stressed ® rms)
CONST FINC CLOSS B6NB6 TEN MAN FEE ADFEE
CONST 1
FINC 0.082 1
CLIENTLOSS  0.001  0.030 1
B6NB6  0.074 0.087 0.081 1
TENURE  0.076 0.041 0.061 0.064 1
MANDATE  0.046  0.088 0.038 0.053  0.113 1
FEE  0.968  0.082  0.044  0.004  0.130  0.002 1
ADITFEES 0.002 0.002 0.002  0.003 0.000  0.004  0.002 1
8. CONCLUSION
This paper aimed to analyse the impac t of renewabl e long-term audit mandates
on audit quality. Audit quality was consi dered from the vi ewpoint of the
external users of the ® nancial statements. It was questioned whether renewable
long-term audit mandates have an impact on the auditor’s reporting behaviour
and on auditor independence.

The results of the study sugge st that long-term auditor client relationships
signi® cantly increase the likelihood of the auditor to issue an unquali® ed audit
438 The European Accounting Review
report. Or, long tenure signi® cantly reduces the auditor’s willingness to qualify
audit reports. Given the fact that th e tenure variable is signi® c ant in the
sample of ® nancially stressed and ® nancially non-stressed ® rms separately,
there is little reason to believe that the result of long tenure decreasing the
likelihoo d of an unclean audit report is confounded by the sample selection.
A signi ® cant diŒerence was also found between the auditor reporting behav-
iour in the ® rst years versus the last year of the audit mandate, at least for the
entire sample and the subsample of ® nancially stressed ® rms. Auditors are
more willing to issue a `clean’ audit report in the ® rst two years of their o cial
mandate than in the last year of their mandate. This could be an indication
that the decision of renewal of the auditor’s mandate is already taken and
known to the auditor before he has issued his last audit report within his
current mandate. In other words, if the auditor already knows that his man-
date will not be renewed, he might be m ore willing to issue an `unclean’ audit
report if this would be appropriate.
The policy implications of these ® ndings could be in favour of mandatory
auditor rotation to maintain the value of an audit for the external users.
However, given recent theoretical evidence on the adverse eŒects of mandatory
auditor rotation, there is a need to develop alternative measures to safeguard
auditors’ independence. In this respect, empirical research on the impact of
competition in the audit market on audit qualit y and the impact of a capital
market that can learn about the trustworthiness of auditors on audit quality
could be promising (Summer, 1998).
Further research could analyse the impact of renewable long-term audit
mandates on audit quality from the viewpoint of the management of a
company. In this respect, a qualitative analysis on the basis of interviews might
be promising. Considering the viewpoint of both company’s management and

the external users of the ® nancial statements can lead to a more rigorous
conclusion. Furthermore, the impact on audit quality of other aspects of the
auditor client relationship, which are regulated diŒerently within the EU, is
interesting enough to be researched.
ACKNOWLEDGEMENTS
The author wishes to acknowledge the helpful comments received from Ann
Jorissen, Steven Maijoor and Chrystelle Richard as well as from two anony-
mous referees.
NOTES
1 For discussions of the litigation environment, see Albrecht and Willingha m (1993),
Kinney (1994) and Krishnan and Krishnan (1997).
2 This statement is valid under the assumption that independence cannot be en sured
by means of private contracts (Acemoglu, 1994).
Impact of renewable long-term audit mandates 439
3 The following description is based on: Flower and Lefebvre (1994); Block and
Jorissen (1995); Buij ink et al. (1996); Meuwissen (1999).
4 A company is considered to be large if it either exceeds more than one of the
following criteria: (i) average number of persons employed on annual basis is ® fty;
(ii) annual turnover, exclusive of VAT, 6,250,000 euros; (iii) balance sheet total,
3,125,000 euros, or whose average number of employees during the period exceeds
100.
5 The Institute of Auditors, created in 1953, is a public institute under the authority of
the Ministry of Economic AŒairs and is assisted by the High Council for Economic
Professions. Belgian Company Law requires that statutory audits of entities under
prudential control (e.g. listed companies, banks and other ® nancial institutions,
insurance companies and hospitals) can only be performed by auditors approved
by the various o cial bodies regulating those industries.
6 In Belgium audit ® rm networks provide non-audit service s through separate legal
entities.
7 This syste m is characterized by the fact that liability is placed upon the defendants

according to their contribution to the damage.
REFERENCES
Acemoglu, D. (1994) `A dynamic model of collusion’, Mimeo, MIT, Department of
Economics.
Acemoglu, D. and Gietzmann, M. (1997) `Auditor independence, incomplete contracts
and the role of legal liability’, European Accounting Review, 6(3): 355 75.
Albrecht, W. and Willingham, J. (1993) `An evaluation of SAS No. 53, the auditor’s
responsibility to detect and report errors and irregularities’, in The Expectation Gap
Standards: Proceedings of the Expectatio n Gap Roundtable. New York: AICPA.
Antle, R. (1984) `Auditor independence’ , Journal of Accounting Research, Spring: 1
20.
Arrunada, B. and Paz-Ares, C. (1997) `Mandatory rotation of company auditors: a
critical examination’, International Review of Law and Economics, 17: 31 61.
Barnes, P. and Den Hu an, H. (1993) `The auditor’s going-concern decision: some UK
evidence concern in independence and competence’, Journal of Business Finance and
Accounting, 20(2) January: 213 28.
Beattie, V. an d Fearnley, S. (1995) `The importance of audit ® rm characteristics and
the drivers of auditor change in UK listed companies’, Accountancy and Business
Research, 25(100): 227 39.
Behn, B., Carcello, J., Hermanson, D. and Hermanson, R. (1997) `The determinants
of audit client satisfaction among clients of Big 6 ® rms’, Accounting Horizons, 11(1):
7 14.
Block, H. and Jorissen, A. (1995) `Belgium: individual accounts’, in Ordelheide, D.
and KPMG (eds) Transnational Accounting. New York: Macmillan.
Buijink, W. F., Maijoor, S., Meuwissen, R. and Van Witteloostuijn, A. (1996) `The
role, position, and liability of the statutory auditor within the European Union’,
MARC, University of Maastricht.
Campisi, S. and Trotman, K. T. (1985) `Auditor consensus in going concern judg-
ments’, Accounting and Business Research, Winter: 23 31.
Carcello, J. V. and Palmrose, Z. (1994) `Auditor litigation and modi® ed reporting on

bankrupt clients’, Journal of Accounting Research, Supplement: 1 30.
Carcello, J. V., Hermanson, R. H. and McGrath, N. T. (1992) `Audit quality attributes:
the perceptions of audit partners, preparers, and ® nancial statement users’, Auditing:
A Journal of Practice and Theory, 11 (Spring): 1 15.
440 The European Accounting Review
Carcello, J. V., Hermanson, D. R. and Huss, H. F. (1995) `Audit quality research in
the United States’, Maandblad voor Accountancy en Bedrijfskunde, 69 (June): 35 9
64.
Citron, D. B. and Ta‚ er, R. J. (1992) `The audit report under going concern uncer-
tainties: an empirical analysis’, Accounting and Business Research, 22(88): 337
45.
Cloyd, C., Fr ederickson, J. and Hill, J. (1996) `Motivatin g factors against independent
auditors: experimental evidence on the impo rtance of causality’, Journal of Account-
ing and Public Policy, 15(3): 185 218.
Copley, P. A. and Doucet, M. S. (1994) `Improving public sector audit quality’,
Accountants’ Journal, 73(2) March: 57 8.
Craswell, A. , Francis, J. and Taylor, S. (1995) `Auditor brand name reputations and
industry sp ecializations’, Journal of Accounting and Economics, 20(3): 297 322.
DaPalma, A. and Deneckere, R. (1995) `The market for audit services and mandatory
rotation’, Working Paper, HEC, Universite
Â
de Gene
Á
ve.
Davidson, R. A. and Neu, D. (1993) `A note on the association between quali® ed
audit opinions and auditor switches’, Contemporary Accounting Research, 9(2): 479
88.
De Angelo, L. E. (1981) `Auditor size and audit quality’, Journal of Accounting and
Economics, 3: 183 99.
Deis, D. R. and Giroux, G. A. (1992) `Det erminants of audit quality in the public

sector’, Accounting Review, July: 462 79.
Dye, R. (1993) `Discussion: limiting auditors’ liability’, Journal of Economics and
Management Strategy, 2(3) Fall: 435 43.
Falk, H., Frucot, V. and van Zijl, T. (1998) `Auditor independence: an international
perspective’, paper presented at the IAAER/CIERA conference, Chicago, October.
Farmer, T. A. , Rittenberg, L. E. and Trompeter, G. M. (1987) `An investigation of
the impact of economic and organizational factors on auditor independence’, Aud-
iting: A Journal of Practice and Theory, 6 (Fall): 1 14.
Flower, J. and Lefebvre, C. (1994) European Financial Reporting: Belgium. London:
Routledge.
Francis, J. and Simon, D. (1987) `A test of audit pricing in the small-client segment of
the US audit market’ , Accounting Review, 62(1): 145 57.
Gietzmann, M. B. and Sen, P. K. (1996) `Auditor rotation: an economic analysis
of the European Union’s proposed Fifth Directive’, Mimeo, London School of
Economics.
Gietzmann, M. B. and Quick, R. (1998) `Capping auditor liability: the German experi-
ence’, Accounting, Organizations and S ociety, 23(1): 81 103.
Gist, W. (1994) `Empirical evidence on the eŒect of audit structure on audit pricing’,
Auditing: A Journal of Practice and Theory, 13(2) Fall: 25 40.
Hopwood, W., McKeown, J. C. and Mutchler, J. (1994 ) `A reexamination of auditor
versus model accuracy within the context of the going-concern opinion decision’,
Contemporary Accounting Research, 10(2) Spring: 409 31.
ImhoŒ, E. A. Jr (1988) `A comparison of analysts’ accounting quality judgments
among CPA ® rms’ clients’, Auditing: A Journal of Practice and Theory , Spring: 182
91.
Kida, T. (1980) `An investigation into auditor’ s continuity and related quali® cation
judgements’, Journal of Accounting Research, Autumn: 506 23.
Kinney, W. (1994) `Auditor litigation research: profess ional help is needed’, Accounting
Horizons, 8(2) June: 80 6.
Knapp, M. C. (1991) `Factors that audit committee members use as surrogates for

audit quality’, Auditing: A Journal of Practice and Theory, 10(1) Spring: 35 52.
Krishnan, J. and Krishnan, J. (1997) `Litigation risk and auditor resignations’,
Accounting Review, 72(4) October: 539 60.
Impact of renewable long-term audit mandates 441
Levinthal, D. A. and Fichman, M. (1988) `Dynamics of interorganizational attachments:
auditor client attachments’, Administrative Science Quarterly, 33(3): 345 69.
Louwers, T. (1998) `The relation between going concern opinions and th e auditor’s
loss function’, Journal of Ac counting Research, 36(1) Spring: 143 56.
Maddala, G. S. (1989) Limited Dependent and Qualitative Variables in Econometrics.
Cambridge: Cambridge University Press.
Mautz, R. K. and Sharaf, H. A. (1961) The Philosophy of Auditing. Sarasota, FL:
American Accounting Association.
McBarnet, D. and Whelan, C. (1992) `Regulating accounting: limits in the law’ , in
Bromwich, M. and Hopwood, A. (eds) Accounting and the Law. Hemel Hempstead:
Prentice-Hall.
Melumad, N. and Thoman, L. (1990) `On auditors and the courts in an adverse
selection setting’, Journal of Accounting Research, Spring: 77 120.
Menon, K . and Schwartz, K. B. (1985) `Auditor switches by failing ® rms’, Accounting
Review, April: 248 61.
Menon, K. and Williams, D . (1991) `Auditor credibility and initial public oŒering’,
Accounting Review, 66(2): 313 32.
Meuwissen, R. (1999) `The economics of auditor careers and audit markets’, Ph.D.
dissertation, University of Maastricht.
Mueller, G., Gernon, H. an d Meek, G. (1994 ) Accounting: An International Perspective.
Burr Ridge, IL: Irwin.
Mutchler, J. (1985 ) `A multivariate analysis of the auditor’s going concern opinion
decision’, Journal of Accounting Research, Autumn: 668 82.
Narayanan, V. (1994) `An analysis of auditor liability rules’, Journal of Accounting
Research, 32 (Supplement): 39 59.
Nichols, D. and Smith, D. (1983) `Auditor credibility and auditor ch anges’, Journal

of Accounting Research, Autumn: 534 44.
O’Keefe, T., Simunic, D. and Stein, M. (1994) `The production of audit services:
evidence from a major public accounting ® rm’, Journal of Accounting Research,
32(2) Autumn: 241 61.
Ooghe, H. and Verbaere, E. (1982) `Determinanten van faling: verklaring en predictie’,
University of Ge nt.
Palmrose, Z. (1986) `Audit fees and auditor size: further evidence’, Journal of Account-
ing Research, 24(1) Spring: 97 110.
Palmrose, Z. (1987) `Litigation and independent auditors: the role of business failures
and management fraud’, Auditing: A Journal of Practice and Theory, 6 (Spring): 90
102.
Schilder, A. (1994) `Auditor independence: an exploratory study on some core issues
of accountants’ ethics’, Ph.D. dissertation, University of Groningen.
Schilder, A. (1996) `Research opportunities in auditing in the European Union’,
Accounting Horizons, 10(4): 98 108.
Schwartz, R. (1997) `Legal regimes, audit quality an d investment’, Accounting Review,
72(3) July: 385 406.
Shockley, R. A. (1981 ) `Perceptions on auditor’s independence: an empirical analysis’ ,
Accounting Review, 56(4) October: 785 800.
Shockley, R. A. (1982) `Per ceptions of audit independence: a conceptual model’,
Journal of Accounting, Auditing and Finance, 5 (Winter): 126 43.
Shockley, R. A. and Holt, R. N. (1983) `A b ehavioural investigation of supplier
diŒerentiation in the market for audit services’, Journal of Accounting Research,
21(2) Autumn: 545 64.
Simon, D. T. (1985) `The audit services market: empirical evidence’, Audit ing: A
Journal of Practice and Theory, Fall: 71 8.
Simunic, D. A. (1980 ) `The pricing of audit ser vices: theory and evidence’, Journal of
Accounting Research, Spring: 161 90.
442 The European Accounting Review
Simunic, D. and Stein, M. (1996) `The impact of litigation risk on audit pricing: a

review of the economics and the evidence’, Auditing: A Journal of Practice and
Theory, 15 (Supplement): 119 34.
Stice, J. (1991) `Using ® nancial and market in formation to identify pre-engagement
factors associated with lawsuits against auditors’, Accounting Review, 66(33): 516
33.
St Pierre, K. and Anderson, J. (1984) `An analysis of t he factors associated with
lawsuits against public accountants’, Accountin g Review, 59(2): 242 63.
Summer, M. (1998 ) `Does mandatory rotation enhance auditor independence’, Zeit-
schrift fur Wirtschafts und Sozialwissenschaften, 118(3): 327 59.
Sutton, M. ( 1997) `Auditor independence: the challenge of fact and appearance’,
Accounting Horizons, 11(1): 86 91.
Vieten, H. (1995) `Auditing in Brit ain and Germany compared: professions, knowledge
and the state’ , European Accounting Review, 4(3): 485 511.
Wallace, W. (1980) The Economic Role of the Audit in Free and Regulated Markets.
London: Touche Ross & Co.
Whittington, G., Grout, P. and Jewitt, I. (1995) `Is total auditor independence a good
thing?’, Accountancy, February: 75.
Wilson, T. E. and Grimlund, R. A. (1990) `An examination of the importance of an
auditor’s reputation’ , Auditing: A Journal of Practice and Theory, 9 (Spring): 43
59.