Description
Changestotheauditenvironmenthaveledtosuggestedchangestotheregulatoryframeworkforeval-uatingauditors’judgmentsincludingtheintroductionofanAuditJudgmentRule(AJR),wherebycourtsandinspectorswillnotsecond-guessauditors’reasonedjudgmentsprovidedtheyaremadeingoodfaithandinarigorousmanner.WeexaminethepotentialeffectoftheAJRontheskepticismofAuditCom-mitteeMembers(ACMs)intermsoftheextenttowhichtheyaskprobingquestionstoexternalauditors,CFOsandHeadsofInternalAuditconcerninganaccountingestimate.ThislevelofprofessionalskepticismisacriticalelementofthedutiesofanACMintheoversightofthefinancialreportingandauditingpro-cesses,especiallyforcomplexandfutureorientatedaccountingestimates.BecauseanAJRwouldlikelyencourageadoptionofinnovativeauditprocedures,wefurtherexaminetheeffectoftheseprocedures,ascomparedtostandardprocedures,onACMs’skepticismgivenanAJR.
Accounting, Organizations and Society 46 (2015) 59–76
Contents lists available at ScienceDirect
Accounting, Organizations and Society
journal homepage: www.elsevier.com/locate/aos
The effect of an Audit Judgment Rule on audit committee members’
professional skepticism: The case of accounting estimates
Yoon Ju Kang
a
, Andrew J. Trotman
b
, Ken T. Trotman
c,?
a
University of Massachusetts at Amherst, United States
b
Northeastern University, United States
c
UNSW Australia, Australia
a r t i c l e i n f o
Article history:
Received 1 June 2014
Revised 28 February 2015
Accepted 9 March 2015
Available online 7 April 2015
a b s t r a c t
Changes to the audit environment have led to suggested changes to the regulatory framework for eval-
uating auditors’ judgments including the introduction of an Audit Judgment Rule (AJR), whereby courts
and inspectors will not second-guess auditors’ reasoned judgments provided they are made in good faith
and in a rigorous manner. We examine the potential effect of the AJR on the skepticism of Audit Com-
mittee Members (ACMs) in terms of the extent to which they ask probing questions to external auditors,
CFOs and Heads of Internal Audit concerning an accounting estimate. This level of professional skepticism
is a critical element of the duties of an ACM in the oversight of the ?nancial reporting and auditing pro-
cesses, especially for complex and future orientated accounting estimates. Because an AJR would likely
encourage adoption of innovative audit procedures, we further examine the effect of these procedures,
as compared to standard procedures, on ACMs’ skepticism given an AJR. Our ?ndings show that an AJR
increases ACMs’ perceived accountability in ensuring the reasonableness of the ?nancial statements, and
that a movement towards more innovative audit procedures under an AJR framework increases ACMs’
perceived overall comfort regarding the treatment of the accounting estimate. On average, these factors
do not affect the overall level of ACMs’ skepticism in terms of the number of questions asked or the
extent to which the questions are probing. However, these results differ depending on the demographic
background of the ACM participants.
© 2015 Elsevier Ltd. All rights reserved.
Introduction
Internationally, regulatory inspectors have reported audit de?-
ciencies with respect to the audit of estimates, and there have
been calls for improved audit quality (ASIC, 2012; European Com-
mission, 2010; FRC, 2013a; IFIAR, 2014; PCAOB, 2013). In turn, there
is uncertainty about what is expected of an auditor given the com-
plexity and future orientation of accounting estimates and the dif-
?culty of inspectors concluding on the appropriateness of the audi-
tors’ judgments (Peecher, Solomon, & Trotman, 2013). In response
to these and related issues, there have been suggestions for change
to the regulatory framework (e.g., Peecher et al., 2013; Pozen,
2008). One such change is the introduction of an Audit Judgment
Rule (AJR), which Peecher et al. (2013) suggest should be applied
by regulators in evaluating auditors’ professional judgments and
motivating auditors to improve audit quality. This AJR is based on
the Business Judgment Rule (BJR), which applies to directors in the
?
Corresponding author at: School of Accounting, School of Business, UNSW Aus-
tralia, Sydney, NSW 2052, Australia. Tel.: +61 2 9385 5831.
E-mail address: [email protected] (K.T. Trotman).
USA, where directors cannot have their judgments second-guessed
by courts and cannot be held responsible for related third party
losses provided they act in good faith and in the best interests of
the company (O’Connell and Boutros, 2002; Peecher et al., 2013).
Under an AJR courts and inspectors would not second-guess audi-
tors’ judgments provided the auditor has used ‘reasoned evaluation
made in good faith and in a rigorous, thoughtful and deliberate
manner’.
We conduct an experiment to examine the effect of the imple-
mentation of an AJR on the level of skepticism of Audit Commit-
tee Members (ACMs) as measured by their questioning behavior
in overseeing the ?nancial reporting and auditing process. Because
one purpose of an AJR would be to discourage defensive auditing,
and instead encourage a focus on audit effectiveness, we consider
how a change to innovative (as compared to standard) audit pro-
cedures for those in the AJR treatment procedures further affects
ACMs’ level of skepticism. We chose ACMs because they have a
?duciary duty to shareholders and are accountable for effectively
overseeing the ?nancial reporting and audit process (CAQ, 2014,
FEE, 2014, FRC, 2013b, Kang, 2014) by exercising professional skep-
ticism and asking probing questions about the judgments and as-http://dx.doi.org/10.1016/j.aos.2015.03.001
0361-3682/© 2015 Elsevier Ltd. All rights reserved.
60 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
sumptions underlying management’s signi?cant estimates and au-
ditors’ judgments of those estimates (Beasley, Carcello, Hermanson,
& Neal, 2009; Cohen, Gaynor, Krishnamoorthy, & Wright, 2007;
IAASB, 2014; NACD, 2013).
Examining the potential effects of alternative regulatory frame-
works such as the AJR is important given the suggestion that the
existing framework for auditors is unlikely to motivate them to tar-
get high quality audits or invest in research and development ac-
tivities aimed at improving long term audit quality (Peecher et al.,
2013, p. 597). In addition, the adoption of an AJR is consistent
with US Treasury recommendations (Pozen, 2008, p. 93) that both
the SEC and the PCAOB should issue policy statements articulat-
ing how the reasonableness of accounting and auditing judgments
be evaluated including factors to be considered when making this
evaluation. An AJR would also make it more di?cult for inspec-
tors to second-guess auditors’ judgments as long as the judgments
were made using ‘good faith and in a rigorous, thoughtful and de-
liberate manner’. We examine the AJR in a setting with accounting
estimates given that such a judgment rule is most likely to apply
in a setting involving subjective judgments. These accounting esti-
mates are future orientated and are increasingly important to the
?nancial statements (Gri?th, Hammersley, & Kadous, 2014; IFIAR,
2014; Peecher et al., 2013; Treasury, 2008). The uncertainty inher-
ent in estimates provides an opportunity for management bias as
well as di?culties for auditors in assessing their reasonableness,
resulting in potential quality implications for ?nancial statements
(Gri?th, Hammersley, Kadous, & Young, 2015).
In our study we consider both the existing regulatory frame-
work without an AJR, and another where an AJR is implemented.
Under the existing framework, even if auditors’ judgments are
made in good faith, and irrespective of how reasonable the au-
ditors’ judgment process was, courts and regulatory inspectors
can determine that alternative judgments should have been made.
While there are many ways to operationalize an AJR framework,
we chose one where under an AJR a reasonableness test is applied
and even if an inspector develops an estimate that differs from an
auditor’s estimate, uses a different method to reach an estimate
than used by the auditor, or believes that his/her own estimation
process is superior to the auditor’s process, the inspector still can
conclude that the auditor’s estimate is reasonable.
An additional characteristic of the audit environment is that the
auditor faces uncertainty about what is expected of them because
it is di?cult to make a conclusion on the appropriateness of audi-
tor judgments related to complex estimates that are future orien-
tated. Inspectors have extensive authority to determine when au-
ditor judgments are inappropriate, and de?cient judgments have
been documented by inspection agencies around the world (e.g.,
ASIC in Australia; FRC in the UK; PCAOB in the US). One conse-
quence of this situation is the prevalence of “overly cautious audits
or ‘defensive’ auditing” (Treasury, 2008, p. VII: 28),
1
which result
in auditor accountabilities primarily related to auditors’ judgment
outcomes and penalties for not reaching a minimum threshold
(Peecher et al., 2013). This leads to compliance-focused behaviors
with auditors potentially operating at or close to the minimum re-
quired in order to avoid sanctions (Bell, Peecher, & Solomon, 2005;
Peecher et al., 2013). As a result, innovative audit procedures are
unlikely to be adopted for fear of second-guessing by inspectors,
despite suggestions for the need for incentives for auditors “to go
1
There have been criticisms of second-guessing by inspection agencies (see
Treasury, 2008). These views can be seen in the responses to PCAOB inspection re-
ports. For example, “While we believe that the PCAOB should continue to challenge
judgments and documentation during the inspection process, we do not believe
that, in the end, reasonable judgments should be criticized and second-guessed
(Grant Thornton, 2009).”
beyond the ?oor and compete on the basis of quality” (Palmrose,
2006). The adoption of an AJR would provide greater incentives for
audit ?rms to adopt more innovative procedures (Peecher et al.,
2013). Consequently, we restrict our examination of how the adop-
tion of more innovative procedures would affect ACMs’ skepticism
regarding a signi?cant estimate to the situation where an AJR is
present. If an AJR is likely to encourage adoption of innovative pro-
cedures, an understanding of how ACMs will react to these proce-
dures is important in order to gain a fuller understanding of the
impact of the introduction of an AJR.
Our ?ndings show that introducing an AJR increases ACMs’ per-
ceived accountability in ensuring the reasonableness of the ?nan-
cial statements. Further analysis shows that this is related to ACMs
believing that accounting estimates become less conservative, and
auditors’ due diligence to be negatively impacted, with the intro-
duction of an AJR. We also ?nd that a move towards more inno-
vative audit procedures under an AJR framework increases ACMs’
perceived overall comfort regarding the accounting treatment at
hand. This result appears to stem from ACMs believing innova-
tive procedures lead to higher audit quality. Despite the greater ac-
countability associated with the introduction of an AJR and more
comfort with a move towards more innovative audit procedures,
our ?ndings show that this does not necessarily lead ACMs to be-
come more skeptical and ask more probing questions when an AJR
is introduced or ask less probing questions when innovative pro-
cedures are used. However, further analysis suggests a signi?cant
background effect where former audit partners show greater skep-
ticism by questioning the external auditor more when an AJR is
implemented than when it is absent, while other ACMs question
the auditor less in the presence of an AJR. We also found that the
types of questions asked vary with ACM background and, in partic-
ular, the questioning behavior of former audit partners is different
to other ACMs.
Our paper makes four major contributions to the audit liter-
ature. First, there have been calls for changes to the regulatory
framework for auditing including the need for a professional judg-
ment framework (e.g., Pozen, 2008) such as an AJR as proposed by
Peecher et al. (2013). By examining how audit committee skepti-
cism in overseeing the ?nancial reporting process is affected under
an AJR, we answer the call for research on how the adoption of an
AJR would affect the execution of ACMs’ ?duciary duties (Peecher
et al., 2013). Also, given ACMs work closely with auditors, exam-
ination of the questions asked and the judgments made by ACMs
provides a ‘window’ to observe how ACMs expect auditors to re-
act to the implementation of an AJR (see Nelson, Elliott, & TarpIey,
2002 for a similar approach).
Second, one of the major criticisms of the present audit reg-
ulatory framework is that it results in defensive auditing because
of the likely negative consequence of moving to more innovative
procedures (Bell et al., 2005; Peecher et al., 2013; Treasury, 2008).
One of the proposed bene?ts of an AJR is the opportunity to re-
duce defensive auditing and encourage auditors to use more inno-
vative audit techniques. We provide insight on how such a move
towards more innovative audit procedures would affect ?nancial
reporting quality under an AJR by examining how it in?uences au-
dit committee skepticism when overseeing the ?nancial reporting
process of accounting estimates. This question is important given
statements in the professional literature that the way standards
are applied “may be seen as an impediment to innovation” and
the need for standard setters to ensure that audit standards do not
inhibit innovation and developments in practice (FEE, 2014, p. 6).
Third, we contribute to the literature on accounting estimates.
Previous research has suggested that estimates are one of the likely
cited accounts for auditor errors (ASIC, 2012; Church & Shefchik,
2012; Gri?th et al., 2015) with implications for ?nancial report-
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 61
ing quality, and there have been suggestions of low audit qual-
ity in this area (European Commission, 2010; IFAC, 2011; PCAOB,
2010). Increasingly the audit committee is expected to exercise
professional skepticism and ask probing questions about these es-
timates (CAQ, 2014; FRC, 2013b). We address the following re-
search question posed by Beasley et al. (2009): “What is the nature
of the audit committee’s involvement in reviewing the company’s
accounting policies and accounting estimates/judgments assump-
tions?” (Beasley et al., 2009) by examining the detailed questions
our ACM participants ask about a signi?cant estimate to other key
corporate governance stakeholders: external auditors, the CFO, and
the Head of Internal Audit.
Fourth, previous research has suggested that the judgments
and questioning behavior of ACMs can depend on ACMs’ previ-
ous background including experience (Beasley et al., 2009; De-
Zoort, 1998) and ?nancial expertise (McDaniel, Martin, & Maines,
2002). We split our ACM participants based on their demographic
background: former audit partners, other quali?ed accountants and
non-accountant directors, which allows us to further explore differ-
ences with respect to our hypotheses and the variations in types
of questions asked by ACMs with different backgrounds. Follow-
ing our experiment, we also ask a series of survey questions to
gain insights into the perceptions of extremely experienced ACMs
about an AJR, and the use of more innovative audit procedures on
?nancial statement preparation, auditor judgments and inspector
judgments.
Background
Accounting estimates form the basis for most ?nancial state-
ment information and recent research shows there is an increase
in their use and complexity (Bratten, Gaynor, McDaniel, Mon-
tague, & Sierra, 2013; Christensen, Glover, & Wood, 2012; Grif-
?th et al., 2015). Professional literature also acknowledges “…
that the increasing complexity of global business operations are
compelling a growing use of judgments and estimates” (Treasury,
2008, p. VII: 17) and has been seen as a signi?cant issue for
regulators (Christensen, Glover, & Wood, 2013; SEC, 2011). When
auditing accounting estimates, auditors are required to conclude
on their ‘reasonableness’, which is a matter of professional judg-
ment. However, the uncertainty involved in accounting estimates
increases the challenges for auditors and inspection reports glob-
ally have identi?ed audit de?ciencies related to accounting esti-
mates (e.g., European Commission, 2010; IFAC, 2011; IFIAR, 2014;
PCAOB, 2013).
Along with auditors, ACMs also play a critical role related to
accounting estimates as they are extensively involved in reviewing
the estimates as well as the underlying judgments and assump-
tions made by management (Beasley et al., 2009). Moreover, the
role of the audit committee in monitoring the ?nancial reporting
process has expanded with substantial attention to its roles in ex-
ercising professional skepticism and supporting investor interests
(Deloitte, 2014; NACD, 2013).
Professional skepticism is a concept that is discussed frequently
both in academic research and in professional standards with
varying de?nitions such as an attitude that includes a question-
ing mind and a critical assessment of evidence (AICPA, 2007, ISA
240), the need for a larger and/or more persuasive sets of evi-
dence (Hurtt, Brown-Liburd, Earley, & Krishnamoorthy, 2013; Nel-
son, 2009), and the need for critical thinking and looking for
contradictory evidence (Gri?th et al., 2015). Gri?th et al. sug-
gest that improved critical thinking is more important than in-
creased doubt, or increased demand for evidence, in improving
audit quality. The literature has also moved from a ‘neutral’ to
‘presumptive doubt’ view of skepticism (Bell et al., 2005; Nelson,
2009) and differentiates between skeptical judgment and skepti-
cal action (Hurtt et al., 2013; Nelson, 2009). Center to this dis-
tinction is the idea that professional skepticism must reach some
level of threshold before it is translated into an action. Fur-
ther, given a constant level of skeptical judgment, skeptical ac-
tion can be affected by various factors such as one’s traits, incen-
tives, and knowledge (Nelson, 2009). Here, we adopt a “presump-
tive doubt” perspective of professional skepticism with a focus on
skeptical action rather than judgments, where being more skep-
tical relates to showing a greater need to collect a more persua-
sive set of evidence before concluding that an assertion is correct
(Nelson, 2009).
We characterize an ACM who is more skeptical as one who
is more questioning of management’s assertion related to an ac-
counting estimate as well as the auditors’ judgments and deci-
sions regarding management’s assertions. This is consistent with
the academic and regulatory emphasis on ACMs asking challenging
questions of external auditors, management and internal auditors
in their oversight of the ?nancial reporting and auditing process.
Speci?cally, in their in-depth interviews with 42 executives serv-
ing on US public company audit committees, Beasley et al. (2009)
?nd that the most important trait an ACM needed to possess is
to be skeptical by asking probing and challenging questions, sup-
porting earlier interview data by Gendron and Bédard (2006) who
?nd that an ACM’s skill in asking challenging questions is central
to their effectiveness as an ACM. In a survey of Australian audit
partners and managers, Contessotto and Moroney (2013) ?nd that
external auditors perceive “asking challenging questions of man-
agement and the auditor” to be by far the most important charac-
teristic of ACMs. The above results are consistent with PCAOB an-
nouncements (PCAOB, 2011; NACD, 2013) suggestions for the need
for audit committees to be more skeptical of both internal and ex-
ternal auditors as well as ?nancial executives in terms of “insisting
on logical responses to relevant questions” (NACD, 2013, p. 36).
Nelson (2009) developed a model of how audit evidence com-
bines with pre-existing knowledge, traits, and incentives to affect
the amount of professional skepticism in auditor judgments and
actions. Hurtt et al. (2013) extend that model and provide a com-
prehensive summary of research studies related to auditor skep-
ticism. We suggest that the Nelson (2009) model of professional
skepticism can be applied to ACMs by considering evidential input,
knowledge, traits and incentives of ACMs.
For ACMs the evidential input will be any information collected
and considered by the ACMs in respect of a particular accounting
treatment (Pomeroy, 2010). ACMs receive information from man-
agement, internal auditors and external auditors and the reliance
they place on these different sources are likely to vary with ACMs’
knowledge which in turn will be affected by background and ex-
perience. ACMs have a wide array of different backgrounds (e.g.,
former audit partners, CFOs, non-executive directors with limited
accounting knowledge). Previous research on ACMs has suggested
important differences based on ACM background and experience
(Beasley et al., 2009; DeZoort, 1998; DeZoort, Houston, & Her-
manson, 2003; DeZoort & Salterio, 2001; McDaniel et al., 2002;
Pomeroy, 2010). The most common measure of traits used in the
auditor professional skepticism literature relates to trait skepticism
as measured by the Hurtt (2010) scale. While there is no research
that has examined this scale for ACMs, we expect the range of
trait skepticism scores for ACMs to be wider than for auditors be-
cause of their more varied backgrounds. Finally, previous research
on ACMs has suggested a range of key incentives for ACMs: es-
tablish individual effectiveness as an ACM and overall effectiveness
of the audit committee (Gendron & Bédard, 2006; Gendron, Bé-
dard, & Gosselin, 2004), avoid loss of reputation for ACMs (Peecher
et al., 2013), constrain aggressive and inappropriate ?nancial
62 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
reporting (DeZoort & Salterio, 2001) and vigilantly monitor ?nan-
cial reporting risks to protect shareholders (Beasley et al., 2009).
Given these incentives for ACMs to be skeptical and ask probing
questions, Kang (2014) examines two potential determinants of au-
dit committee skepticism in terms of their propensity to challenge
signi?cant accounting estimates. Based on the theory of account-
ability, she predicts and ?nds that ACMs are more skeptical and
ask more probing questions given a sophisticated versus unsophis-
ticated investor base, and that this difference narrows in the pres-
ence of an anticipated additional audit report disclosure regarding
such estimates where ACMs decrease their overall level of skepti-
cism and ask less probing questions.
In this study we examine the effect of perceived account-
ability on ACMs’ skepticism by studying the likely effect of an
AJR on the extent to which ACMs question and ask for fur-
ther evidence related to a signi?cant accounting estimate. Imple-
mentation of an AJR is proposed by Peecher et al. (2013) as a
mechanism to hold auditors accountable for their judgment pro-
cesses and ultimately improve audit quality. Currently, inspec-
tors document in their inspection reports examples of judgments
of auditors that they have determined are inappropriate, irre-
spective of how reasonable the auditor judgment process was
(Peecher et al., 2013). This paper examines the likely effect the
implementation of an AJR will have on ACMs’ perceived ac-
countability and ultimately their level of skepticism in oversee-
ing the ?nancial reporting and auditing process of accounting
estimates.
Prior research and development of hypotheses
As ?duciaries of shareholders, ACMs, along with management
and auditors, form the corporate governance mosaic (Cohen, Krish-
namoorthy, & Wright, 2010) in ensuring high quality ?nancial re-
porting for public companies. Speci?cally, ACMs are charged with
overseeing the ?nancial reporting and auditing process (CAQ, 2014;
IAASB, 2014; NACD, 2013; PCAOB, 2012). Findings from interviews
show that ACMs perceive themselves to be accountable to share-
holders and believe themselves to be monitoring agents of man-
agement (Beasley et al., 2009). This implies that factors that in-
?uence ACMs’ perceived accountability towards shareholders will
likely affect their level of skepticism in overseeing the ?nancial re-
porting and audit process.
Theory holds that accountability pressure, the pressure to jus-
tify one’s view to others (Tetlock, 1983), leads individuals to de-
velop different social and cognitive strategies to obtain accep-
tance from, or avoid con?ict with, important interpersonal or in-
stitutional audiences (Gibbins & Newton, 1994; Hoffman & Patton,
1997; Koonce, Anderson, & Marchant, 1995; Lerner & Tetlock, 1994;
Lerner & Tetlock, 1999; Peecher, 1996; Tetlock, 1983). Studies that
examine the effect of accountability distinguish between two types
of accountability: process accountability and outcome accountabil-
ity. While outcome accountability holds decision makers account-
able for the actual outcome of their decision process, process ac-
countability holds decision makers accountable for the quality of
their decision process, regardless of the outcome of their decisions
(Lerner & Tetlock, 1999; Tetlock, 1983; Tetlock, 1985). Various stud-
ies in both the psychology and audit literatures (see Peecher et al.,
2013 for an overview) have shown process accountability, as com-
pared to outcome accountability, to more effectively increase inte-
grative complexity of thought (Tetlock, 1983; Tetlock & Boettger,
1989), increase cognitive effort, improve calibration, and reduce
decision biases (DeZoort et al., 2006; Koonce et al., 1995; Siegel-
Jacobs & Yates, 1996; Tetlock, 1992). Process accountability is likely
to be the primary type of accountability faced by ACMs as they
are aware of protection from the BJR, and therefore know they
are more accountable for the process rather than the outcome.
However, outcome accountability is also relevant as potential legal
action is more likely when there is a negative outcome (Peecher
et al., 2013) and negative outcomes can lead to reputation loss for
the ACMs as well as the auditors and managers.
In this paper, we study one potential factor of ACMs’ account-
ability, implementation of an AJR, and its effect on ACMs’ level of
skepticism. As discussed above, according to Nelson’s (2009) model
on professional skepticism, incentives play an important role in de-
termining the level of skepticism. Incentives that favor professional
skepticism are generated by potential regulatory enforcements by
various regulatory bodies and consequent reputation loss (Nelson,
2009). In the context of audit committees, the risk of litigation
and potential reputation loss will likely increase their perceived ac-
countability pressure in ensuring the reasonableness of the ?nan-
cial statements, and thus affect their level of skepticism in moni-
toring the ?nancial reporting process.
In the current audit environment, auditors’ judgments, even
those that are complex and involve di?cult future estimates, can
be second-guessed as courts can rule that alternative judgments
should have been reached. In addition, regulatory inspectors, such
as the PCAOB, currently have wide authority to determine that the
judgments of auditors are ‘de?cient’, irrespective of the reason-
ableness of the auditors’ judgment process. Peecher et al. (2013)
suggest this leads auditors to take less risks and be less willing
to develop ways to improve existing best practices in fear of po-
tential liability (p. 608). They further suggest that introducing an
AJR, fashioned after the BJR that applies to corporate directors in
the US, will help alleviate such problems as it will give auditors
greater con?dence that their use of innovative procedures would
not be second-guessed by inspectors or courts, providing greater
incentives for them to adopt more innovative procedures.
With the implementation of such an AJR, it is possible for ACMs
themselves to perceive less litigation risk and lower likelihood
of ‘inappropriate’ adverse inspection reports, decreasing the like-
lihood of reputation damage associated with an adverse inspec-
tion report. This could result in lowering both ACMs’ accountability
pressure and lead them to be less skeptical of signi?cant account-
ing estimates in the presence of an AJR. However, as discussed be-
low, we suggest that it is more likely that an AJR will increase
ACMs’ accountability due to audit committees taking a greater role
in the protection of shareholders.
The reduction in potential liability under an AJR is likely greater
for auditors than ACMs as it directly relates to how inspectors
should evaluate auditor judgments. This reduction in auditor’s risk
of potential liability may provide opportunities for auditors to be
more lenient in their audit procedures (Treasury, 2008). Concerned
that auditors will partake in such behavior, implementation of the
AJR will likely lead ACMs to feel more accountable towards share-
holders about ensuring high quality ?nancial reporting (e.g., ensure
the reasonableness of management’s signi?cant estimates) both in
terms of the outcome and process. Moreover, prior research shows
that ACMs bear signi?cant labor market penalties and reputation
damage in the event of ?nancial reporting failures (Monks & Mi-
now, 2004; Peecher et al., 2013; Srinivasan, 2005). Hence, the
possibility of auditors becoming more lenient under an AJR will
likely increase ACMs’ potential reputation damage, making them
feel more accountability to ensure reasonableness of the ?nancial
statements. Further, increases in accountability pressure have been
shown to lead ACMs to increase their level of skepticism in over-
seeing the ?nancial reporting and auditing process (Kang, 2014).
Therefore, we predict the introduction of the AJR will lead ACMs
to feel greater accountability pressure towards shareholders and
be more skeptical of management’s signi?cant estimates by ask-
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 63
ing more questions about the processes used by the external audi-
tor, CFO and the internal auditor, including the evidence they used,
how they gained comfort, their assessment of management’s in-
tegrity, etc. This leads to our ?rst set of hypotheses formalized be-
low.
H1a. Given standard audit procedures, ACMs will feel greater ac-
countability when an AJR is present than when it is absent.
H1b. Given standard audit procedures, ACMs will show higher lev-
els of professional skepticism when an AJR is present than when it
is absent.
While there is a perception that the reduction in potential li-
ability risk under an AJR may result in auditors being more le-
nient in their audit approach (Treasury, 2008), it may simultane-
ously lead auditors to increase their willingness to develop and
conduct more creative and innovative audit procedures (Peecher
et al., 2013). Such practice would address the criticism of defen-
sive auditing (Peecher et al., 2013; Treasury, 2008) and the call to
use more innovative audit procedures to detect ?nancial statement
fraud (Bell et al., 2005).
While there is very little research examining ACMs’ comfort
level, Rowe (2014) considers auditor comfort with management’s
estimates. He notes that the notion of comfort is commonly used
to describe the extent of auditors’ willingness to accept man-
agement assertions (e.g., Bell et al., 2005; KPMG, 2011; Mar-
tin, Rich, & Wilks, 2006; Pomeroy, 2010). The concept of ‘com-
fort’ extends beyond the appropriateness of management estimates
to the quality of those estimates (Rowe, 2014). Similarly, ACMs
need to make assessments of the appropriateness of the com-
bined judgments made by management and the auditors and they
rely heavily on external auditors in giving that comfort (Beasley
et al., 2009; Gendron et al., 2004). In doing so, it is possi-
ble that ACMs prefer and ?nd more comfort in “tried and true
procedures” (i.e., standard procedures) as opposed to more in-
novative procedures. We suggest, however, it is more likely for
ACMs to expect more innovative audit procedures to be more
likely to detect ?nancial statement fraud, leading them to perceive
greater comfort in the reasonableness of the ?nancial statements
when the auditor uses more innovative, as opposed to standard,
procedures.
H2a. Given an AJR, ACMs will have more comfort in the ?nancial
statements when the auditor uses innovative versus standard pro-
cedures.
Rowe (2014) suggests that auditor comfort affects auditors’ ex-
pectation of the most likely adjustment to the ?nancial statements
and that this in turn leads to skeptical actions (cf. Nelson, 2009)
where auditors focus their questions on the uncertainty of man-
agement estimates. With respect to ACMs, Kang (2014) ?nds that
greater comfort regarding accounting estimates lead ACMs to ask
less questions. Similarly, the potential greater comfort ACMs feel
from innovative audit procedures in our study may result in ACMs
being less skeptical. However, the fact that the procedures are in-
novative means that ACMs are likely to be less familiar with these
procedures and, therefore, to be more skeptical regarding these
procedures. Consequently we test the null hypothesis.
H2b. Given an AJR, the extent of ACMs’ professional skepticism
will not differ depending on whether the auditor uses innovative
or standard audit procedures.
Research questions
In addition to our main hypotheses, we examine three research
questions based on qualitative analysis of the questions developed
by our ACM participants. We speci?cally focus on how the demo-
graphic backgrounds of the participants affect their level of skepti-
cism in terms of asking probing questions regarding an accounting
estimate.
Previous research has emphasized the need for ACMs to ask
probing questions (Kang, 2014; Pomeroy, 2010), the need for ACMs
to be skeptical (FEE, 2014; IAASB, 2014), and that there is differ-
ent levels of performance across ACMs (Beasley et al., 2009; Cohen,
Krishnamoorthy, & Wright, 2004; Gendron & Bédard, 2006). There
has also been a long history of audit judgment research related to
knowledge and experience (Bonner, 2008; Libby & Luft, 1993; Nel-
son & Tan, 2005; Peecher, Schwartz, & Solomon, 2007). DeZoort
(1998) found that both general domain and task speci?c expe-
rience affect ACMs’ internal control assessments, and that ACMs
with experience make judgments more like an auditor than ACMs
without experience. In this paper we extend the literature on ACM
experience to consider prior experience as a former audit partner,
which is particularly important because the appointment of former
audit partners as ACMs has become far more common with the in-
troduction of requirements related to designated ?nancial experts
(Vera-Muñoz, 2005). The use of these designated ?nancial experts
has been found to be associated with improved ?nancial reporting
quality (Dhaliwal, Naiker, & Navissi, 2010; Krishnan & Visvanathan,
2008). McDaniel et al. (2002) also ?nd ?nancial expert ACMs differ
from ?nancial literate ACMs in their judgments and evaluations of
?nancial reporting quality. They call for research to see how “vari-
ous types of ?nancial experts differ in their reporting quality judg-
ments” and subsequently there have been suggestions of the need
for research on the composition of audit committees, including the
effect of different backgrounds of ACMs (Glover & Prawitt, 2013;
NACD, 2013).
We address the following research questions.
1. RQ1: Does ACM background affect the types of questions asked
and the level of probing by ACMs to external auditors, CFOs and
Heads of Internal Audit?
2. RQ2: Are there differences across treatments in the extent of
questions asked and type of questions asked between former
audit partners and other ACMs?
Given the seniority of the participants in our study we also
asked a series of survey questions related to our key independent
variables in the experiment. Peecher et al. (2013) suggest potential
bene?ts for audit quality with the adoption of an AJR and the use
of more innovative/less defensive audit procedures. However, there
have been views on potential adverse effects of implementing an
AJR (Treasury, 2008). Further, one of the suggested reasons that in-
novative procedures have not been introduced in the present regu-
latory framework is because of a potential negative effect on in-
spectors’ assessments (Peecher et al., 2013). Hence, we examine
the views of ACMs on these issues.
1. RQ3: What is the perceived impact of an AJR on management’s
estimates and the external auditor’s due diligence and how is a
change to more innovative audit procedures perceived by ACMs
to affect audit quality and audit inspectors?
Research methods
Participants
Forty-nine ACMs from Australia participated in the study. All
had served on at least one audit committee and on average had
served on a mean of 6.77 audit committees (2.86 public compa-
nies, 2.04 private companies and 1.88 other audit committees). On
average, they had served as ACMs for 10.33 years. Their average
64 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
Table 1
Participant demographics.
Mean Median SD
Panel A: Audit committee service and experience (n = 49)
Public company audit committees served on 2.86 2.00 3.27
Private company audit committees served on 2.04 2.00 2.19
Other company audit committees served on 1.88 2.00 1.72
Total audit committees served on 6.77 5.00 5.06
Number of years as an Audit Committee Member 10.33 10.0 7.94
Experience (years since undergraduate degree) 36.02 37.0 11.76
Panel B: Number of ACMs with the following attributes (n = 49) n
Number who had served on at least one public company audit committee 35/49
Number of quali?ed accountants (CPA/CA quali?ed) 35/49
Number with external audit experience 25/49
Position held: Partner 15/25
Other 10/25
Panel C: Self-rated level of knowledge (n = 49)
Financial accounting 79.27 80.80 16.73
Financial statement analysis 79.72 81.40 18.15
Audit committee best practice 77.40 80.50 17.79
External auditing 78.13 82.00 19.08
Internal audit best practice 73.27 75.00 18.77
CPA/CA quali?ed means they are a member of CPA Australia or Chartered Accountants Australia and New Zealand.
ACM= Audit Committee Member.
business experience was 36.02 years (measured as the time since
they ?nished their undergraduate degree). Of the 49 participants,
35 had served on the audit committee of at least one publicly
listed company (public company ACMs), 35 were quali?ed accoun-
tants being CPAs, chartered accountants or equivalent (CPA/CA),
and 14 were not accountants (non-accountants). Of the CPA/CAs,
25 had audit experience (15 of which are former audit partners).
2
For further analysis we classify the 49 ACMs into three categories
(referred to as ACM backgrounds): former audit partners, accoun-
tants who were not former audit partners (other accountants), and
non-accountants. In sum, this was a very senior group of ACMs.
Table 1 provides demographic details.
Participants were obtained through personal emails to alumni
and business contacts in Australia.
3
Potential participants were in-
formed that this was an on-line research study, which provides a
set of facts based on a hypothetical company and asks participants
what questions they would raise as an ACM of the company and
to make a series of professional judgments. Given the seniority of
the participants, con?dentiality of the data was emphasized and
participants were told that they would not be asked to provide
either their name or the organization(s) of which they were an
ACM when completing the research instrument. Individuals who
responded to our invitation to participate were then emailed the
link to the on-line study.
4
Research design
To test our hypotheses, we used a 2 × 1 + 1 between subjects
design. We ?rst manipulate the presence/absence of an AJR be-
tween subjects. Both these treatments hold constant the fact that
the auditor uses standard audit procedures. Our ‘plus one’ treat-
ment is the same as the AJR present treatment except that the au-
2
The audit experience of those that were not former audit partners was low with
all except one leaving auditing before reaching the manager level.
3
The emails were sent from the researchers or from a senior practitioner who is
on a university advisory board. In addition, three participants were recruited via
a LinkedIn alumni broadcast but this method was relatively unsuccessful so we
moved to the personal contact approach.
4
Four research instruments were completed on-line by individuals who had not
served on an audit committee. Their responses were not analysed as they did not
meet the criteria of having served on an audit committee. Six other participants
opened the link but did not complete the study.
ditor uses innovative procedures instead of standard procedures.
Participants were randomly allocated to one of three treatments:
(1) AJR absent/standard procedures, (2) AJR present/standard pro-
cedures, or (3) AJR present/innovative procedures. Our design al-
lows us to ?rst address the potential effect of an AJR while holding
constant the type of audit procedures presently used in practice.
Present audit procedures have been described as defensive auditing
(Bell et al., 2005; Treasury, 2008), and the introduction of more in-
novative audit procedures are unlikely to be introduced under the
present regulatory framework (Peecher et al., 2013). However, one
of the potential advantages of the introduction of an AJR is that
the changed regulatory environment would likely lead to more in-
novative audit procedures, potentially leading to higher audit qual-
ity. Consequently, we include an AJR present/innovative procedures
treatment to compare to our AJR present/standard procedures but
we did not include an AJR absent/innovative procedures treatment.
In the AJR present treatments, participants are told to assume
that new regulation has been introduced in Australia on an AJR un-
der which courts and regulatory inspectors will not second guess
auditor judgments provided the auditor meets certain key tests.
They are informed that the key test is whether the auditor has
used “reasoned evaluation made in good faith and in a rigorous,
thoughtful and deliberate manner”. In the AJR absent treatment,
participants were provided with information on the ‘Current Reg-
ulation of Auditors in Australia’ where “auditors’ judgments, even
those that are complex and involve di?cult future estimates, can
be second guessed as courts can rule that alternative judgments
should have been reached.” Appendix A illustrates the different
language that was used in the AJR present versus absent treat-
ments.
We also manipulate the type of audit procedures used (stan-
dard vs. innovative). While there have been a number of refer-
ences to the bene?ts of a move away from defensive auditing
to more innovative procedures (Bell et al., 2005; Peecher et al.,
2013), ACMs could interpret the terms standard and innovative in
a range of ways and, therefore, we set out descriptions of these
terms. For the standard audit procedures treatment, participants
were told that the audit partner prefers standard audit proce-
dures (e.g., well tried and accepted procedures commonly found
in audit manuals) over innovative audit procedures (e.g., inno-
vative use of external data outside the control of management),
and the partner notes that audit inspectors expect standard pro-
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 65
cedures and these standard procedures are helpful in reducing in-
spectors’ criticism of the audit, especially when there is poten-
tial for ?nancial statement fraud. Alternatively, for the innovative
audit procedures treatment, participants are told that the audit
partner prefers innovative audit procedures (e.g., innovative use of
external data outside the control of management) over standard
audit procedures (e.g., well tried and accepted procedures com-
monly found in audit manuals), and that the partner notes that
innovative audit procedures are helpful in adding a surprise el-
ement to the more predictable standard audit procedures, espe-
cially when there is potential for ?nancial statement fraud. This
operationalization of innovative procedures drew on suggestions
from Bell et al. (2005) where the audit procedures suggested in-
cluded types of evidence collection that would be more effec-
tive in detecting ?nancial statement fraud and would be seen as
innovative.
Case
We adapted the case materials developed by Kang (2014) with
major changes. We added new manipulations and extensively
changed the questions asked in order to test our hypotheses and
research questions. In addition, we shortened the case and changed
some terminology to make it more consistent with the Australian
environment.
Participants were asked to assume they are an ACM for a hy-
pothetical manufacturing company. They were then provided with
information on the audit environment including the nature of ac-
counting estimates, the relevant audit standards, what is required
of the auditor and the complexity/di?culty of auditor judgments
related to estimates. This was followed by our manipulation of
the presence/absence of an AJR described in the Research Design
section.
Participants were then given background information on the
company including an overview of the company’s operations, per-
formance, strategy, investor base, and internal audit function. Fur-
ther information was provided on the company’s external audit en-
gagement partner (including the partner’s preference for standard
or innovative audit procedures) and that the engagement team
used innovative/standard audit procedures to collect evidence re-
lated to the accounting estimates. After reading the background in-
formation, all participants read a written communication provided
by the external auditor to the audit committee regarding “a signif-
icant accounting issue related to obsolete inventory”. This commu-
nication outlined details about the nature of the accounting issue,
management’s initial and revised (more favorable) estimate,
5
and
the auditor’s assessment of management’s ?nal estimate where
they note that a ‘clean’ opinion is to be issued. The communica-
tion was followed by ?nancial statements re?ecting the effects of
the alternate (initial and revised) estimates.
Dependent variables
We measure our main construct of ACM skepticism by exam-
ining the extent of questioning exercised by our ACM participants
regarding the accounting estimate described in the case. Speci?-
cally, participants were asked to develop questions they would ask
the audit partner, CFO and the Head of Internal Audit regarding
the accounting issue at hand, in the order they came to mind. Par-
ticipants were told to provide as much detail as possible to these
open ended questions.
5
The revised write-down of $970,000 was only 24 percent of the originally pro-
posed write-down of $4 million.
In order to examine how an AJR affects ACMs’ perceived ac-
countability and test H1a, we also ask participants to indicate the
extent to which they feel accountable to ensure the reasonable-
ness of the ?nancial statements given the audit environment pre-
sented to them on a 7-point scale ranging from ?3 to 3 (‘sig-
ni?cantly not accountable’ to ‘signi?cantly accountable’). Lastly, in
order to examine how innovative versus standard procedures af-
fect ACMs’ perceived comfort over the accounting estimate, we ask
participants to rate their comfort level on a 7-point scale (?3 to
3) from ‘completely uncomfortable’ to ‘completely comfortable’ on
the following three items: the change in management’s estimate of
obsolete inventory; the external auditor’s decision to allow man-
agement’s updated, smaller write-down of inventory; and the dif-
ference in net pro?t that results from the different write-down
amounts for obsolete inventory.
Survey questions asked
In our post-experiment questionnaire, we asked the following
four survey questions to gain insights from the very senior ACMs
participating in the study. First, on a 7-point scale ranging from ?3
(‘extremely low’) to 3 (‘extremely high’) participants were asked
the effect on audit quality if auditors move from standard audit
procedures to more innovative audit procedures when there is po-
tential for ?nancial statement fraud. Second, on a 7-point scale
ranging from ‘extremely negative’ to ‘extremely positive’, partici-
pants were asked how they believed audit inspectors would re-
act to a move from standard audit procedures to more innova-
tive audit procedures. Third, they were asked to assume “an envi-
ronment where courts and inspection agencies (e.g., ASIC, PCAOB)
cannot second-guess ‘reasoned evaluation made in good faith and
in a rigorous, thoughtful and deliberate manner’ by the auditor”.
They were then asked “how do you think management’s estimates
would be affected” on a 7-point scale from ?3 (‘estimates will be-
come extremely less conservative’) to 3 (‘estimates will become ex-
tremely more conservative’) and “to what extent would your per-
ception of the external auditor’s due diligence be affected” on a
7-point scale ranging from ?3 (‘extremely negatively affected’) to
3 (‘extremely positively affected’).
Coding of qualitative analysis
The questions developed by the participants were ?rst read
by two of the researchers, blind to treatments and demographic
details, who developed a set of question type categories (listed
in Table 5) and then allocated questions provided by the partic-
ipants to the question categories. The level of intercoder agree-
ment was 88.4 percent and differences were resolved by discus-
sion. Three themes evolved: reports fairly, evidence collection and
gaining comfort (see Table 5). Under ‘Reports Fairly’ most ques-
tions related to whether the reported earnings are fairly presented
and not overstated, for example: ‘tone at the top’ issues, incentives,
debt covenants, pressure exerted on the auditor, the potential for
bias, the control environment, and the impact on shareholders. The
‘Evidence Collection’ category referred to questions about the use
of external evidence, general questions about what audit testing
was conducted, and more speci?c questions asking for more detail
on tests performed or reference to what further tests could/should
be done. The evidence collection questions asked to auditors of-
ten referred to audit evidence whereas questions to CFOs referred
to the data used by management. The third theme related to the
ACM gaining comfort for themselves and asking how the auditor,
CFO and internal auditor gained comfort. Often there was men-
tion of the term ‘comfort’, whether/how the auditor was satis?ed
66 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
or stronger questions asking the auditor/CFO to justify the large
change.
A different independent coder assessed the extent of prob-
ing of the questions asked, level of skepticism of the ACM, and
the evaluation of performance of the ACM from the perspective
of an audit committee Chairman.
6
The coder received a sepa-
rate page for each participant in random order showing a partic-
ipant number (i.e., no demographic data or experimental treat-
ment) and the questions asked by that participant directed to
the external auditor (EA), the CFO and the Head of Internal Au-
dit (IA) shown in three separate columns. The coder made four
ratings of probing – one for each of the three sets of questions
(EA, CFO and IA) and one overall assessment of the extent of
probing (see Table 4). These ratings were made on a 9-point Lik-
ert scale with end points of 1 (‘not at all probing’) and 9 (‘ex-
tremely probing’). Similar scales were used for the level of skep-
ticism shown by the ACM in their questioning and for when the
coder was asked to assume he was Chairman of the audit commit-
tee and rate the performance of the ACM on the task (‘bottom 10%
of ACMs’ to ‘top 10% of ACMs’). We refer to this coding as ‘level of
probing’.
The above coding refers to the package of questions asked by
each participant. Gri?th et al. (2015) suggest the importance of ap-
proaching evidence in a manner conducive to the critical thinking
aspect of skepticism and the importance of a targeted collection
of additional evidence (Hammersley, Johnstone, & Kadous, 2011).
This emphasis also appears bene?cial for ACMs. Therefore, ques-
tions were also coded by an author and a graduate research assis-
tant (who was blind to the research questions and experimental
treatments) on the basis of whether the question was (a) a general
request for additional evidence, or (b) was a targeted collection of
additional evidence by. The two coders achieved high reliability of
89.2 percent (Cohen et al., 2010).
7
We refer to this coding as ‘gen-
eral versus targeted evidence collection’.
Results
Manipulation checks
Manipulation checks for both independent variables indicate
that participants understood the manipulations. Speci?cally, 42 out
of 49 participants (85.6%) correctly identi?ed the information they
were provided about the audit environment in the case materials
related to an AJR.
8
Also, participants were asked to rate the in-
novativeness of the audit procedures on a 7-point scale (?3 = ‘ex-
tremely non-innovative’ and 3 = ‘extremely innovative’). The mean
response in the innovative treatment was 0.394, which was signif-
icantly greater than the mean response of ?1.715 in the standard
treatments (t
47
= 5.324; p < 0.001; one-tailed).
Hypothesis testing
Effect of an AJR given standard audit procedures (H1a & b)
H1a predicts that ACMs will feel greater accountability when
an AJR is present than when it is absent, given standard audit
procedures. Panel A of Table 2 summarizes the descriptive statis-
6
The independent coder is a present ACM, present Chairman of an audit com-
mittee and a senior member of the accounting profession (including being a former
audit partner from a Big 4 ?rm). The coder had previously participated in the study
and consequently he was familiar with the materials participants received and the
time it took to complete the study.
7
Disagreements on the application of a code to a question were resolved be-
tween the coders; where an agreement could not be reached, another author made
the ?nal decision, which was necessary for only six questions.
8
All seven participants who responded incorrectly to the AJR manipulation check
question were those in the AJR present treatments.
Table 2
Descriptive statistics of perceived accountability and overall comfort.
AJR
Absent Present
Panel A: Perceived accountability (standard error)
1
Procedures
Standard 1.90 2.68
(1.81) (0.37)
n = 17 n = 16
Innovative 2.53
(0.70)
n = 16
H1a: 2.68 versus 1.90 (t
31
= ?1.679; p = 0.052; one-tailed)
Panel B: Perceived overall comfort (standard error)
2
Procedures
Standard ?3.99 ?4.77
(2.87) (2.55)
n = 17 n = 16
Innovative ?3.17
(3.39)
n = 16
H2a: ?3.17 versus ?4.77 (t
30
= ?1.508; p = 0.071; one-tailed)
1
To elicit the participants’ accountability, we ask on a scale of ?3 (signi?cantly
not accountable) to +3 (signi?cantly accountable), “Given the audit environment
presented to you in the case material to what extent did you feel accountable to
ensure the reasonableness of the ?nancial statements?”.
2
The overall comfort scores were computed based on the factor loadings that
resulted from a factor analysis conducted on the responses provided by the partic-
ipants about their comfort level with respect to (1) the change in management’s
estimate of obsolete inventory, (2) the external auditor’s decision to allow man-
agement’s updated, smaller write-down of inventory, and (3) the difference in the
net pro?t that results from the different write-down amounts for obsolete inven-
tory. The responses were based on a scale of ?3 (completely uncomfortable) to +3
(completely comfortable). The responses to the three questions all loaded on one
factor that had an eigenvalue of 2.64.
tics regarding ACMs’ perceived accountability in ensuring the rea-
sonableness of the ?nancial statements. Consistent with our pre-
diction, given standard audit procedures, participants in the AJR
present treatment indicated that they felt signi?cantly greater
accountability (2.68 vs. 1.90) in ensuring the reasonableness of
the ?nancial statements than those in the AJR absent treatment
(t
31
= ?1.679; p = 0.052; one-tailed). Analysis of our additional
survey questions further suggest that this increase in accountabil-
ity is related to ACMs believing management’s estimates becom-
ing less conservative and auditors’ due diligence being negatively
impacted with the implementation of an AJR. Speci?cally, on a
7-point scale (?3 = ‘estimates will become extremely less conser-
vative’ and 3 = ‘estimates will become extremely more conserva-
tive’), the mean response of ?0.837 is signi?cantly less than 0
“not affected at all” (t
48
= ?8.452; p < 0.001; one-tailed). Partici-
pants were also asked to rate how their perception of the exter-
nal auditor’s due diligence would be affected on a 7-point scale
(?3 = ‘extremely negatively affected’ and 3 = ‘extremely positively
affected’). The mean response of ?0.306 is signi?cantly less than
0 ‘not affected at all’ (t
48
= ?2.399; p = 0.010; one-tailed). These
?ndings, together, suggest that ACMs feel greater accountability in
ensuring the reasonableness of the ?nancial statements when an
AJR is present due to the potential for management’s estimates
becoming more aggressive and auditors becoming less diligent in
auditing the estimates.
9
9
Analysis of the effect of an AJR given standard audit procedures on ACMs’ level
of perceived comfort (untabulated) is directionally consistent (but not statistically
signi?cant at p = 0.413; two-tailed) with our ?ndings for H1a where an AJR leads
to greater accountability. Speci?cally, implementation of an AJR appears to decrease
ACMs’ overall level of comfort from a mean of ?3.99 to ?4.77 (see Table 2, Panel
B).
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 67
Table 3
Quantity of questions developed by ACMs by experience and treatment.
Former Audit Partners Other Accountants Non-Accountants All ACMs
AJR
Absent/
Std
AJR
Present/
Std
AJR
Present/
Innov Total
AJR
Absent/
Std
AJR
Present/
Std
AJR
Present/
Innov Total
AJR
Absent/
Std
AJR
Present/
Std
AJR
Present/
Innov Total
AJR
Absent/
Std
AJR
Present/
Std
AJR
Present/
Innov
Grand
Total
n 5 5 5 15 6 6 8 20 6 5 3 14 17 16 16 49
To the EA
Total 22 29 24 75 34 25 30 89 24 14 11 49 80 68 65 213
Average 4.40 5.80 4.80 5.00 5.67 4.17 3.75 4.45 4.00 2.80 3.67 3.50 4.71 4.25 4.06 4.35
To the CFO
Total 20 26 20 66 30 23 39 92 24 20 9 53 74 69 68 211
Average 4.00 5.20 4.00 4.40 5.00 3.83 4.88 4.60 4.00 4.00 3.00 3.79 4.35 4.31 4.25 4.31
To the IA
Total 20 23 17 60 35 14 22 71 20 13 10 43 75 50 49 174
Average 4.00 4.60 3.40 4.00 5.83 2.33 2.75 3.55 3.33 2.60 3.33 3.07 4.41 3.13 3.06 3.55
All questions
Total 62 78 61 201 99 62 91 252 68 47 30 145 229 187 182 598
Average 12.40 15.60 12.20 13.40 16.50 10.33 11.38 12.60 11.33 9.40 10.00 10.36 13.47 11.69 11.38 12.20
H1b: 13.47 versus 11.69 (t
31
= 0.884; p = 0.384; two-tailed).
H2b: 11.69 versus 11.38 (t = 0.161; p = 0.873; two-tailed).
Observed background effect: Former audit partners ask more questions under an AJR (from 62 to 78) while other ACMs (other accountants and non-accountants) tend to
ask less questions when an AJR is implemented, decreasing from 167 (99 + 68) to 109 (62 + 47) (Fisher Exact Test; p = 0.002; two-tailed).
Former Audit Partners = previously held the position of an audit partner.
Other Accountants = other quali?ed accountants but not a former audit partner.
Non-Accountants = not quali?ed accountants.
All ACMs = Former Audit Partners plus Other Accountants plus Non-Accountants.
AJR Absent/Std = AJR absent/standard procedures treatment.
AJR Present/Std = AJR present/standard procedures treatment.
AJR Present/Innov = AJR present/innovative procedures treatment.
EA = External Auditor.
CFO = Chief Financial O?cer.
IA = Internal Auditor.
Total = total number of questions asked by participants in each treatment.
Average = total divided number of participants in the treatment group
.
H1b predicts that the greater accountability in the presence of
an AJR, given standard audit procedures, will lead ACMs to be more
skeptical and ask more questions. Table 3 summarizes the total and
average number of questions directed to the three different par-
ties by the three experimental treatments and background.
10
We
?nd no signi?cant effect of an AJR (absent vs. present) given stan-
dard procedures on the total number of questions asked by ACMs
(13.47 vs. 11.69; t
31
= 0.884; p = 0.384; two-tailed, Table 3).
11
Fur-
ther, there was also no signi?cant effect of an AJR given standard
procedures in terms of the number of questions directed to the
external auditor (4.71 vs. 4.25; t
31
= 0.638; p = 0.528; two-tailed),
the CFO (4.35 vs. 4.31; t
31
= 0.058; p = 0.954; two-tailed) and the
internal auditor (4.41 vs. 3.13; t
31
= 1.517; p = 0.139; two-tailed).
Interestingly, we ?nd a signi?cant background effect, where for-
mer audit partners react signi?cantly differently compared to oth-
ers. Speci?cally, former audit partners become more skeptical and
ask more questions under an AJR (from 62 to 78) while other ACMs
tend to become less skeptical and ask less questions when an AJR
is implemented, decreasing from 167 (99 + 68) to 109 (62 + 47)
(Fisher Exact Test; p = 0.002; two-tailed; see Table 3).
10
We also used the independent coder’s rating of how probing the total package
of questions were to test H1b, but ?nd no signi?cant results (t
31
= 0.013; p = 0.990;
two-tailed). See Table 4 for the independent coder’s ratings.
11
Untabulated results show that ACMs with external audit experience asked sig-
ni?cantly more questions in total than those without external audit experience
(p = 0.091). Other demographic factors, such as experience as an ACM of a pub-
lic ?rm, being a CPA or CA, or having been a former audit partner did not have
signi?cant effects on the total number of questions asked (p > 0.10).
Effect of innovativeness of procedures given an AJR (H2a & b)
H2a predicts that given an AJR, ACMs will have more com-
fort in the ?nancial statements when the auditor uses innovative
rather than standard procedures. Panel B of Table 2 summarizes
the descriptive statistics regarding ACMs’ perceived overall level of
comfort.
12
Consistent with our predictions, given an AJR, we ?nd
that ACMs in the innovative treatment felt a higher level of com-
fort regarding the issues surrounding the accounting estimate than
those in the standard treatments (?3.17 vs. ?4.77; t
30
= ?1.508;
p = 0.071; one-tailed) providing marginally signi?cant support for
H2a.
13
Analysis of our additional survey questions suggest that the
higher level of comfort associated with the conduct of innovative
audit procedures rather than standard audit procedures is likely
due to ACMs believing that a move from standard to innovative
audit procedures will lead to higher audit quality. Speci?cally, par-
ticipants were asked to indicate the extent to which they believe
audit quality would improve if auditors move from standard audit
procedures to more innovative audit procedures (?3 = ‘extremely
12
The overall level of comfort was computed based on the factor loadings that
resulted from a factor analysis conducted on the responses provided by the partic-
ipants about their comfort level with respect to (1) the change in management’s
estimate of obsolete inventory, (2) the external auditor’s decision to allow manage-
ment’s updated, smaller write-down of inventory, and (3) the difference in the net
pro?t that results from the different write-down amounts for obsolete inventory.
The responses to the three questions all loaded on one factor that had an eigen-
value of 2.64. The factor loadings were 0.911, 0.960, and 0.943, respectfully. Analysis
results based on the three separate measures are consistent with the analysis based
on the overall level of comfort.
13
Analysis of the effect of a move towards more innovative audit procedures given
an AJR on ACMs’ perceived accountability (untabulated) is not statistically signi?-
cant (2.68–2.53; p = 0.457; two-tailed) (see Table 2, Panel A).
68 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
low’ and 3 = ‘extremely high’). The mean response of 1.078 was
signi?cantly higher than 0 ‘neutral’ (t
48
= 7.584; p < 0.001; one-
tailed).
H2b tests whether given an AJR, the greater comfort level that
follows from the external auditors’ use of innovative procedures
rather than standard procedures will lead ACMs to question the ac-
counting estimate to a lesser/greater extent when the auditor uses
innovative procedures. We ?nd no signi?cant effect of the inno-
vativeness of the procedures, given an AJR, on the total number of
questions asked by the ACMs (standard: 11.69 vs. innovative: 11.38;
t
30
= 0.161; p = 0.873; two-tailed).
14
Hence, we do not reject the
null hypothesis for H2b. Such results are likely due to the fact that
although ACMs feel more comfort when the audit partner adopts
more innovative audit procedures, their unfamiliarity with the pro-
cedures lead them to be at least as questioning of the estimate as
when standard audit procedures are used.
15
Analysis of the demographic factors, however, again suggest a
potential ACM background effect. Speci?cally, for the two treat-
ments with AJR present, former audit partners as a whole decrease
the number of questions they ask from 78 to 61 when innova-
tive audit procedures are used rather than standard procedures
(see Table 3). Those without an audit partner background, how-
ever, show a slight increase in their questioning behavior from 109
(62 + 47) to 121 (91 + 30) (see Table 3) when innovative proce-
dures are used. The difference between the two groups is signif-
icant at p = 0.108 (Fisher Exact Test; two-tailed). The directional
difference in our results suggests that while the use of innovative
procedures increase ACMs’ overall perception of comfort regarding
the accounting estimate, former audit partners who are more fa-
miliar with the innovative procedures may exhibit a lower level of
skeptical behavior when innovative procedures are used, and the
unfamiliarity with innovative procedures lead those without an au-
dit partner background to keep asking probing questions. We sug-
gest the need for future research addressing this potential expla-
nation.
The overall pattern of results for H1 and H2 can be seen in
Fig. 1. From Fig. 1, Panel A it can be seen that adoption of an AJR
would increase ACM accountability and this would occur whether
the change to innovative procedures also occurred. We conjecture
that if the move increases ACM comfort, as found in H2a, this
would attenuate, or work against, an increase in accountability
from the introduction of an AJR. Overall, the combined results sug-
gest that the increase in ACM accountability from a move to an AJR
is likely to be robust to a change of the type of audit procedures
from standard to innovative. In addition, Fig. 1, Panel B suggests
that if an AJR is implemented, then innovative audit procedures
are needed if ACM comfort is considered an important priority.
Supplementary analysis
Does ACM background affect the level of probing and the types of
questions asked (RQ1)
In this section we divide participants into three background
groups: former audit partners, other quali?ed accountants and
14
Consistent with the results of H2b, the effect of the move towards more in-
novative procedures given AJR did not have a signi?cant effect on the participants’
response to the question that asked on a scale of ?3 (signi?cantly less than av-
erage amount of questioning would be warranted) to +3 (signi?cantly more than
average amount of questioning would be warranted), to what extent they would
like to question auditors and/or management about the appropriateness/su?ciency
of the write-down (Standard: 2.406 vs. Innovative: 2.244; t
30
= ?0.68; p = 0.501;
two-tailed).
15
We also test H2b using the independent coder’s rating of how probing the total
package of questions were but ?nd no signi?cant results (t
30
= ?0.222; p = 0.825;
two-tailed) (see Table 4, Panel B).
non-accountants. We then consider the differences in the level
of probing of the questions of the different background groups
(Table 4, Panel A) and the types of questions they ask (Table 5).
We outline below that while the level of probing did not vary with
ACM background, the quantity of questions asked and the type of
questions asked did vary with ACM background.
Table 4, Panel A summarizes the mean scores related to
the level of probing of the independent coder for each evalu-
ation category split across the three background groups (on a
range of 9-point scales, where higher numbers indicate higher
performance). Analysis of the scores in Table 4 Panel A show
ACMs asked probing questions of the EA (mean = 6.00), CFO
(mean = 6.08), IA (mean = 5.51), and in their total package of ques-
tions (mean = 5.92). Mean scores for all categories were signi?-
cantly above the midpoint of the scale (p a7d? 0.046; two-tailed).
Their overall level of skepticism (mean = 7.00) and performance
level (mean = 6.06) were also signi?cantly higher than the mid-
point of the scale (p a7d? 0.002; two-tailed).
16
Next we consider the types of questions asked by ACMs and the
effects of demographic background. Table 5 categorizes all ques-
tions asked by ACMs of the EAs, CFOs and IAs split by demographic
background. We incorporate columns for former audit partners
(n = 15), other accountants (n = 20) and non-accountants (n = 14).
The table classi?es questions based on our coding of questions
about reporting fairly, evidence collection and gaining comfort.
Our analysis discusses average questions asked per participant
and provides evidence that ACM background affects the type and
quantity of questions ACMs ask in their oversight process, as seen
from the following key insights from Table 5. First, all three ACM
background groups asked the most questions to EAs, next most to
CFOs and the least to IAs (former audit partners 6.5, 5.5 and 5.1
respectively; other accountants 6.0, 5.4 and 4.5 respectively; non-
accountants 4.9, 4.4 and 3.6 respectively – see average questions
per participant in Table 5, row 4C). This trend holds for questions
related to evidence collection (see Table 5, row 2B where the to-
tal for Column A of 2.5 is greater than the totals for Columns B
and C, of 1.9 and 1.6 respectively) and questions related to gain-
ing comfort (see Table 5, row 3B where the total of Column A
of 2.0 is greater than the totals of Columns B and C, 1.3 and 1.1
respectively). However, for the reports fairly category, the average
questions per participant to EAs is less (1.1, see Table 5, Column A,
row 1B) than for either CFOs or IAs (1.6 and 1.3 respectively as per
Table 5, Columns B and C, row 1B). As the evidence collection and
comfort questions
17
mainly relate to processes, while the reports
fairly questions mainly relate to outcomes, this suggests that ACMs
generally preferred to rely on EAs to obtain information about pro-
cesses that underlie the reported accounting estimate. On the other
hand, the questions directed to CFOs and IAs (see Table 5, row 1A)
by ACMs in general, and former audit partners in particular, were
more focused on outcomes. These questions, particularly those un-
der rows 1.1 and 1.2 of Table 5, related to the accuracy of the esti-
mates developed by management.
Second, the main differences between the types of questions
asked were between former audit partners (n = 15) and non-
accountants (n = 14) with former audit partners asking more than
double the number of questions on average per participant in the
‘reports fairly’ category to EAs (1.5 vs. 0.6; Column A, row 1B),
16
As noted in the testing of H2a and H2b, analysis of the scores of how prob-
ing the total package of questions were in Table 4 Panel B shows no signi?cant
difference based on the introduction of an AJR or the type of procedures used by
the auditor. The largest difference in the mean level of probing of questions asked
to external auditors by former audit partners (mean = 6.53) compared to all other
ACMs (means of 5.75 and 5.79) is signi?cant at p = 0.127, two tailed.
17
While many gaining comfort questions related to the processes used to gain
comfort, some questions speci?cally asked why the revised estimate was preferable
and thus related directly to an outcome.
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 69
Fig. 1. Observed results of perceived accountability and overall comfort.
1
To elicit the participants’ accountability, we ask on a scale of ?3 (signi?cantly not accountable) to +3 (signi?cantly accountable), “Given the audit environment presented
to you in the case material to what extent did you feel accountable to ensure the reasonableness of the ?nancial statements?”.
2
The overall comfort scores were computed based on the factor loadings that resulted from a factor analysis conducted on the responses provided by the participants about
their comfort level with respect to (1) the change in management’s estimate of obsolete inventory, (2) the external auditor’s decision to allow management’s updated, smaller
write-down of inventory, and (3) the difference in the net pro?t that results from the different write-down amounts for obsolete inventory. The responses were based on a
scale of ?3 (completely uncomfortable) to +3 (completely comfortable). The responses to the three questions all loaded on one factor that had an eigenvalue of 2.64.
to CFOs (2.5 vs. 1.1), and to IAs (1.7 vs. 0.6) (Table 5, Columns B
and C, row 1B). The difference between the former audit partners
and non-accountants was particularly evident in rows 1.1 and 1.2,
which were questions related to ?nancial statement accuracy.
Third, the emphasis on questions on evidence collection (rows
2.1–2.4 in Table 5) varies with background group. For questions to
EAs (see Table 5, Column A, row 2C), over 40% of the questions
were in the evidence collection category for all three ACM back-
ground groups (45.4%, 41.2% and 40.6% respectively, Table 5, Col-
umn A, row 2C). However, this pattern changes for questions asked
to CFOs (Table 5, Column B, row 2C); of the three ACM background
groups, former audit partners asked the least number of questions
to CFOs in the evidence collection category (20.7% of their ques-
tions) compared to other accountants (51.9% of their questions) and
non-accountants (36.1% of their questions). These differences are
particularly evident in category 2.4 (business impact and manage-
ment’s turnaround strategy).
Differences in the extent and type of questions asked by former audit
partner ACMs and other ACMs split by treatments (RQ2)
The above analysis considers differences in the three ACM back-
ground groups for the three experimental treatments combined.
In Table 6 we provide additional qualitative analysis to build on
our hypothesis testing. Given the small cell sizes when we split
our 49 participants across three experimental treatments and three
background groups from Table 5, in Table 6 we combine other ac-
countant ACMs and non-accountant ACMs into one group. Below
we consider differences in the questioning behavior of former au-
dit partners (n = 15) and other ACMs (n = 34) depending on the
70 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
Table 4
Mean evaluation scores of the extent to which questions are probing.
Mean scores
n Questions to EA Questions to CFO Questions to IA Question package Skepticism Performance
Panel A: By participant background
Former Audit Partners 15 6.53 6.07 5.93 6.13 7.07 6.33
Other Accountants 20 5.75 6.15 5.50 5.95 7.00 6.05
Non-Accountants 14 5.79 6.00 5.21 5.64 6.93 5.79
Overall 49 6.00 6.08 5.51 5.92 7.00 6.06
Panel B: By experimental treatment
AJR Absent/Standard Procedures Treatment
Former Audit Partners 5 6.40 5.60 6.00 5.60 7.00 6.00
Other Accountants 6 6.17 6.33 6.50 6.33 7.33 6.17
Non-Accountants 6 6.17 5.50 4.67 5.67 6.83 5.50
Total 17 6.24 5.82 5.71 5.88 7.06 5.88
AJR Present/Standard Procedures Treatment
Former Audit Partners 5 7.00 6.80 6.00 7.00 7.40 7.00
Other Accountants 6 5.50 5.67 5.00 5.50 6.83 5.67
Non-Accountants 5 5.20 6.20 5.00 5.20 6.80 5.80
Total 16 5.88 6.19 5.31 5.88 7.00 6.13
AJR Present/Innovative Procedures Treatment
Former Audit Partners 5 6.20 5.80 5.80 5.80 6.80 6.00
Other Accountants 8 5.63 6.38 5.13 6.00 6.88 6.25
Non-Accountants 3 6.00 6.67 6.67 6.33 7.33 6.33
Total 16 5.88 6.25 5.63 6.00 6.94 6.19
All ratings in this table are done by an independent coder with ACM experience who was blind to demographic details using 9-point Likert scales.
Questions to EA = questions directed to the External Auditor.
Questions to CFO = questions directed to the CFO.
Questions to IA = questions directed to the Head of Internal Audit.
Question Package = all questions to EAs, CFOs and IAs.
Skepticism= level of skepticism ACMs showed in their questioning behavior.
Performance = the performance of the ACM on this task from the perspective of the Chair of the Audit Committee.
Former Audit Partners = previously held the position of an audit partner.
Other Accountants = other quali?ed accountants but not a former audit partner.
Non-Accountants = not quali?ed accountants.
presence/absence of an AJR and the use of innovative versus stan-
dard procedures. Overall, these additional analyses show signi?cant
differences between former audit partners and other ACMs in the
number of questions asked, the types of questions asked, the ex-
tent to which the questions show higher levels of skepticism and
how probing the questions are. These differences are outlined be-
low and suggest the need for future research to further examine
the underlying differences in questioning behavior of ACMs of dif-
ferent backgrounds.
First, we examine some speci?c differences between former au-
dit partners and other ACMs across the AJR absent and the AJR
present treatments (for both these treatments standard procedures
are used).
18
The former audit partners in the AJR present treatment
ask more questions compared to those in the AJR absent treatment
while the opposite pattern holds for the other ACMs. The direction
of these results holds for questions asked to EAs (Table 6, Panel A),
CFOs (Table 6, Panel B) and IAs (Table 6, Panel C). The former au-
dit partners ask EAs 8.8 questions per participant on average in
the AJR present treatment versus 5.0 per participant in the AJR
absent treatment while other ACMs ask 6.3 questions per partic-
ipant in the AJR absent treatment versus 4.9 questions per partici-
pant in the AJR present treatment (see Table 6, row 5). This differ-
ence between the two groups is statistically signi?cant at p = 0.004
(Fisher Exact Test; two-tailed) showing that former audit partners
in the AJR present treatment ask more questions than those in the
AJR absent treatment but for other ACMs the reverse holds. The
18
Table 6 provides both total numbers of questions asked and average number of
questions asked. As the results do not change depending on this choice (i.e., con-
stant cell sizes for former audit partners and similar cell sizes for other ACMs), for
clarity we report per participant numbers (i.e. mean questions asked).
same pattern holds for questions directed to IAs (Table 6, Panel
C). Former audit partners in the AJR present treatment ask more
questions to IAs (6.2 per participant) than those in the AJR absent
treatment (4.8 per participant) while other ACMs in the AJR ab-
sent treatment ask more questions (5.2 per participant) than those
in the AJR present treatment (3.1 per participant; see Table 6, row
15). These differences are signi?cant (Fisher Exact Test; p = 0.017;
two-tailed). We note that the above results for former audit part-
ners for both questions to EAs and IAs are consistent with H1b,
whereas the results for other ACMs are in the opposite direction to
the predictions of H1b. However, caution is required in interpret-
ing these results given the small number of former audit partners
in each treatment group. The direction of these ?ndings is con-
sistent across the three types of questions. We illustrate this with
data in Panel A, which shows that for the former audit partners,
the average number of questions per participant is higher for the
AJR present treatment participants than the AJR absent treatment
participants across all three types of questions: reports fairly (1.8
vs. 1.2), evidence collection (4.4 vs. 2.0) and gaining comfort (2.2
vs. 1.8) (Table 6, Panel A, Column A, rows 1–3). We ?nd the oppo-
site for the three types of questions for the other ACMs with the
AJR absent treatment being larger than the AJR present treatment
in each case (1.3 vs. 0.4, 2.5 vs. 2.2 and 2.2 vs. 2.0) (Table 6, Panel
A, Column B, rows 1–3).
Second, there are both similarities and differences in the types
of questions directed to CFOs compared to those directed to
EAs and IAs discussed above. The questions directed to the CFO
(Table 6, row 10) follow a similar pattern to questions to EAs and
IAs with the former audit partners in the AJR present treatment
asking more questions (6.4 per participant) compared to those in
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 71
Table 5
Types of questions directed to the External Auditor, CFO and Head of Internal Audit.
Column A: Questions
Directed to EA
Column B: Questions
Directed to CFO
Column C: Questions
Directed to IA
Former
Audit
Partners
Other
Accts
Non-
Accts Total
Former
Audit
Partners
Other
Accts
Non-
Accts Total
Former
Audit
Partners
Other
Accts
Non-
Accts Total
Rown 15 20 14 49 15 20 14 49 15 20 14 49
Reports Fairly
1.1 Conservatism, integrity, bias and pressure exerted by
management
8 8 3 19 11 5 4 20 8 7 2 17
1.2 Loan covenants, incentives, pressure to perform, effect on
remuneration and shareholders
5 7 1 13 17 9 5 31 2 7 1 10
1.3 Management decision to use non-sales staff in the
assessment
5 0 3 8 4 3 3 10 5 6 2 13
1.4 The control environment for inventory; surprises 4 6 2 12 5 8 4 17 10 9 3 22
1A Total Questions: Reports Fairly 22 21 9 52 37 25 16 78 25 29 8 62
1B Reports Fairly Questions per Participant 1.5 1.1 0.6 1.1 2.5 1.3 1.1 1.6 1.7 1.5 0.6 1.3
1C % of Grand Total Questions 22.7 17.6 13.0 18.2 45.1 23.1 26.2 31.1 32.5 32.6 16.0 28.7
Evidence Collection
2.1 Use of external evidence 7 10 6 23 1 9 2 12 2 5 0 7
2.2 What testing was done (general question) 1 6 1 8 1 1 1 3 4 3 2 9
2.3 Asking for further details of audit testing done, reference
to further tests that could be done, and the extent of
reliance on the internal auditor
31 26 20 77 5 12 6 23 14 13 15 42
2.4 Business impact including management’s turnaround
strategy
5 7 1 13 10 34 13 57 8 6 7 21
2A Total Questions: Evidence Collection 44 49 28 121 17 56 22 95 28 27 24 79
2B Evidence Collection Questions per Participant 2.9 2.5 2.0 2.5 1.1 2.8 1.6 1.9 1.9 1.4 1.7 1.6
2C % of Grand Total Questions 45.4 41.2 40.6 42.5 20.7 51.9 36.1 37.8 36.4 30.3 48.0 36.6
Gaining Comfort
3.1 Terminology related to comfort/comfortable/con?dent 3 9 3 15 4 5 7 16 5 9 2 16
3.2 Question as to whether/how the auditor/CFO was satis?ed 9 8 6 23 4 2 0 6 3 5 1 9
3.3 Questions asking for justi?cation by the auditor/CFO
including why is the revised estimate preferable
8 13 9 30 15 12 10 37 9 10 8 27
3.4 Whether the accounting treatment and audit procedures
met accounting/audit standards, were consistent
industry/other clients and internally reviewed
9 11 11 31 1 1 2 4 0 1 2 3
3A Total Questions: Gaining Comfort 29 41 29 99 24 20 19 63 17 25 13 55
3B Gaining Comfort Questions per Participant 1.9 2.1 2.1 2.0 1.6 1.0 1.4 1.3 1.1 1.3 0.9 1.1
3C % of Grand Total Questions 29.9 34.5 42.0 34.7 29.3 18.5 31.1 25.1 22.1 28.1 26.0 25.5
4A Other Questions 2 8 3 13 4 7 4 15 7 8 5 20
4B Grand Total of Questions 97 119 69 285 82 108 61 251 77 89 50 216
4C Average Questions per Participant 6.5 6.0 4.9 5.8 5.5 5.4 4.4 5.1 5.1 4.5 3.6 4.4
The total number of classi?cations for EAs of 285 is greater than the total number of questions asked (213 questions) shown in Table 3 as some questions ?t under more
than one category. Similarly, classi?cations of 251 for CFOs and 216 for IAs is greater than the number of questions asked (211 and 174 questions respectively) shown in
Table 3.
Reports Fairly = questions related to whether the reported earnings is fairly presented and not overstated.
Evidence Collection = questions related to the nature and extent of evidence collected by the auditors or management.
Gaining Comfort = questions related to the ACM gaining comfort about the ?nancial statements.
Questions Directed to EA = questions asked by ACMs to the external auditor.
Questions Directed to CFO = questions asked by ACMs to the CFO.
Questions Directed to IA = questions asked by ACMs to the Head of Internal Audit.
Former Audit Partners = previously held the position of an audit partner.
Other Accountants = other quali?ed accountants but not a former audit partner.
Non-Accountants = not quali?ed accountants.
the AJR absent treatment (4.8 per participant). In comparison, the
other ACMs asked less questions in the AJR present treatment (4.6
per participant) compared to the AJR absent treatment (5.0 per
participant). While these results are in the same direction as the
results for questions to EAs and IAs, they are not statistically signif-
icant (Fisher Exact Test; p = 0.192; two-tailed). However, one sig-
ni?cant difference to the above trends relates to questions about
evidence collection asked to CFOs compared to EAs (Table 6, rows
2 and 7). While former audit partners on average asked 2.0 and
4.4 evidence collection questions to EAs (for AJR absent and AJR
present respectively) (Table 6, Column A, row 2), they asked con-
siderably fewer evidence collection questions to CFOs across these
treatments (1.0 and 1.2 questions) (Table 6, Column A, row 7). Also,
while the above trends show that other ACMs ask more ques-
tions to CFOs overall in the AJR absent treatment versus the AJR
present treatment (5.0 vs. 4.6, row 10), this was not the case
for the evidence collection questions to CFOs (AJR absent = 1.8;
AJR present = 2.3; Table 6, Column B, row 7). In fact, this is the
only case where other ACMs asked less questions in the AJR ab-
sent treatment compared to AJR present treatment. It appears that
while former audit partners did not see evidence collection ques-
tions to CFOs to be very important (i.e., they relied on EAs and
IAs), this was not the case for other ACMs who increased evidence
collection questions to CFOs when an AJR was present.
Third, comparison is also be made between the innovative au-
dit procedures treatment and the standard procedures treatment
(for both these treatments AJR is present). The trend for former
audit partners was to ask less questions in the innovative proce-
dures treatment compared to the standard procedures treatment to
72 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
Table 6
Total questions (average per participant) from all ACMs split by treatments.
Column A: Former Audit Partner ACMs Column B: Other ACMs
AJR Absent/
Standard
AJR Present/
Standard
AJR Present/
Innovative Total
AJR Absent/
Standard
AJR Present/
Standard
AJR Present/
Innovative Total
Grand
Total
Row n 5 5 5 15 12 11 11 34 49
Panel A: Questions directed to the External Auditor (EA)
1 Reports Fairly: Total Questions
(Average)
6(1.2) 9(1.8) 7(1.4) 22(1.5) 15(1.3) 4(0.4) 11(1.0) 30(0.9) 52(1.1)
2 Evidence Collection: Total Questions
(Average)
10(2.0) 22(4.4) 12(2.4) 44(2.9) 30(2.5) 24(2.2) 23(2.1) 77(2.3) 121(2.5)
3 Gaining Comfort: Total Questions
(Average)
9(1.8) 11(2.2) 9(1.8) 29(1.9) 26(2.2) 22(2.0) 22(2.0) 70(2.1) 99(2.0)
4 Other 0 2 0 2 4 4 3 11 13
5 Grand Total per Treatment 25(5.0) 44(8.8) 28(5.6) 97(6.5) 75(6.3) 54(4.9) 59(5.4)
188(5.5)
285(5.8)
Panel B: Questions directed to the CFO
6 Reports Fairly: Total Questions
(Average)
12(2.4) 15(3.0) 10(2.0) 37(2.5) 18(1.5) 12(1.1) 11(1.0) 41(1.2) 78(1.6)
7 Evidence Collection: Total Questions
(Average)
5(1.0) 6(1.2) 6(1.2) 17(1.1) 21(1.8) 25(2.3) 32(2.9) 78(2.3) 95(1.9)
8 Gaining Comfort: Total Questions
(Average)
6(1.2) 9(1.8) 9(1.8) 24(1.6) 16(1.3) 10(0.9) 13(1.2) 39(1.1) 63(1.1)
9 Other 1 2 1 4 5 4 2 11 15
10 Grand Total per Treatment 24(4.8) 32(6.4) 26(5.2) 82(5.5) 60(5.0) 51(4.6) 58(5.3)
169(5.0)
251(5.1)
Panel C: Questions directed to the Internal Auditor (IA)
11 Reports Fairly: Total Questions
(Average)
8(1.6) 11(2.2) 6(1.2) 25(1.7) 15(1.3) 9(0.8) 13(1.2) 37(1.1) 62(1.3)
12 Evidence Collection: Total Questions
(Average)
9(1.8) 11(2.2) 8(1.6) 28(1.9) 26(2.2) 12(1.1) 13(1.2) 51(1.5) 79(1.6)
13 Gaining Comfort: Total Questions
(Average)
5(1.0) 7(1.4) 5(1.0) 17(1.1) 17(1.4) 7(0.6) 14(1.3) 38(1.1) 55(1.1)
14 Other 2 2 3 7 4 6 3 13 20
15 Grand Total per Treatment 24(4.8) 31(6.2) 22(4.4) 77(5.1) 62(5.2) 34(3.1) 43(3.9) 139(4.1) 216(4.4)
Former Audit Partner ACMs = previously held the position of an audit partner.
Other ACMs = other quali?ed accountants who are not former audit partners (20) and not quali?ed accountants (14).
AJR Absent/Std = AJR absent treatment/standard procedures.
AJR Present/Std = AJR present treatment/standard procedures.
AJR Present/Innov = AJR present treatment/innovative procedures.
Reports fairly = questions related to whether the reported earnings is fairly presented and not overstated.
Evidence collection = questions related to the nature and extent of evidence collected by the auditors or management.
Gaining comfort = questions related to the ACM gaining comfort about the ?nancial statements.
EAs (5.6 vs. 8.8, Table 6, Panel A, row 5), CFOs (5.2 vs. 6.4; Table 6,
Panel B, row 10) and IAs (4.4 vs. 6.2, Table 6, Panel C, row 15), sug-
gesting former audit partners are more comfortable with the use
of innovative procedures. The comparisons for other ACMs show
the opposite trend where the mean number of questions asked
per participant was greater in the innovative procedures treatment
than the standard procedures treatment (5.4 vs. 4.9, 5.3 vs. 4.6, 3.9
vs. 3.1 for questions to EAs, CFOs and IAs respectively; Table 6, Col-
umn B, rows 5, 10 and 15 respectively). These differences between
former audit partner ACMs and other ACMs would have the oppo-
site effect on the results for H2b with former audit partners ask-
ing less questions when innovative audit procedures are used, and
other ACMs asking more questions when innovative procedures are
used. One explanation for these effects is that the other ACMs are
likely to be less familiar with innovative procedures and, therefore,
ask more questions.
Fourth, all of the above analyses have related to the number of
questions asked to EAs, CFOs and IAs. Here we return to Table 4,
Panel B where the independent coder’s evaluation of the level of
probing is categorized by both participant background and experi-
mental treatment. Across all columns in Table 4, Panel B the scores
in row AJR present/standard procedures for the former audit part-
ners (7.00, 6.80, 6.00, 7.00, 7.40 and 7.00) were higher than the
equivalent scores for the other two treatments. Speci?cally, we
observe the scores for former audit partners were higher in the
AJR present/standard procedures while the opposite pattern held
for ‘other accountant’ ACMs (5.50, 5.67, 5.00, 5.50, 6.83 and 5.67).
We also note that when comparing the two treatments with AJR
present, the scores for former audit partners decreased with a
move towards more innovative procedures, while we observe the
opposite trend for other ACMs. These results are consistent with
our earlier ?ndings of higher skepticism by former audit partners
in the AJR present/standard procedures treatment.
Fifth, as described in the research methods, we categorized
questions on the basis of general versus targeted evidence collec-
tion, with the latter being more conducive to higher levels of skep-
ticism (Gri?th et al., 2015). Comparing the two treatments with
standard procedures we found that the former audit partners as a
group showed an increase for the need for more targeted evidence
given an AJR (AJR absent: 4.40 vs. AJR present: 5.80) while other
ACMs showed a decrease (4.83–3.55) in the need for more targeted
evidence (p = 0.059; Fisher Exact Test; two-tailed; untabulated).
The impact of a change to an AJR and more innovative audit
procedures on ACMs’ perceptions (RQ3)?
Here we consider the responses of our participants to a number
of post-experiment questions. Our results show that ACMs believe
that an AJR will lead management estimates to become somewhat
less conservative. The mean response of ?0.837 is signi?cantly less
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 73
than 0 ‘not affected at all’ (p < 0.001; two-tailed).
19
This provides
a potential explanation for our results for H1a where we ?nd the
introduction of an AJR increases ACMs’ perceived accountability to-
wards ensuring the reasonableness of the ?nancial statements. We
suggest that the belief that management estimates will become
somewhat less conservative with an AJR is likely to increase ACMs’
perceived accountability about ensuring the reasonableness of the
accounting estimate.
There have been suggestions that external auditor due diligence
could be adversely affected by an AJR (Pozen, 2008). Consistent
with such argument, we ?nd that ACMs who participated in our
study believe an AJR will lead auditor due diligence to be some-
what negatively affected. The mean response of ?0.306 is signi?-
cantly lower than 0 ‘not affected at all’ (p = 0.020; two-tailed). The
different demographic factors do not have a signi?cant effect on
the results (p > 0.100).
We also ?nd that ACMs believe that a change to more innova-
tive procedures by auditors would increase audit quality. The mean
response, 1.078, is signi?cantly greater than 0 ‘neutral’ (p < 0.001;
two-tailed). This provides a potential explanation for our results
for H2a where we ?nd auditors’ use of more innovative proce-
dures under an AJR framework increases ACMs’ perceived comfort
regarding the accounting estimate. The different demographic fac-
tors do not have a signi?cant effect on how ACMs perceive this is-
sue (p a7e? 0.100; two-tailed). Also, ACMs believe that the change
to more innovative procedures will not have a signi?cant effect on
inspectors. The mean response of -0.071 is not signi?cantly differ-
ent from 0 ‘neutral’ (p = 0.688; two-tailed).
Discussion and conclusions
Changes to the audit environment have led to suggestions for
the need to change the regulatory framework for auditors includ-
ing the introduction of an AJR and relatedly the use of more in-
novative audit procedures (Peecher et al., 2013). However, such
changes are not without their critics with some stakeholders sug-
gesting that they could negatively affect the judgments of man-
agement and auditors as well as the oversight process of the au-
dit committee, resulting in decreased ?nancial statement quality
(Treasury, 2008). Here we focus on the effect of an AJR on the
oversight process of ACMs given the increased role audit commit-
tees play in monitoring the ?nancial reporting and audit process,
especially those that relate to complex accounting estimates (CAQ,
2014; FRC, 2013a).
Our key ?ndings show that introducing an AJR increases ACMs’
perceived accountability in ensuring the reasonableness of the ?-
nancial statements. Further analysis of our survey questions shows
that this result is related to ACMs believing accounting estimates
are likely to become less conservative and auditors’ due diligence
to be negatively impacted with the introduction of an AJR. To gain
a greater understanding of the impact of an AJR we consider the
effects of its introduction both with standard and innovative au-
dit procedures. The latter is important because an AJR is likely to
encourage the adoption of innovative procedures. We ?nd that a
move towards more innovative procedures under an AJR frame-
work increases ACMs’ perceived overall comfort regarding the ac-
counting issue at hand. Results from our survey questions suggest
19
This analysis is conducted for all participants in the three conditions together
as the questions speci?cally ask the participants to assume an AJR rule exists
or a move towards innovative procedures. Thus, we did not expect any differ-
ence between treatments. For all questions examined under RQ3, t-tests were
conducted between our three treatments and no signi?cant results were found
(p a7e? 0.165; two-tailed). We also obtain participants’ perceptions of the internal
audit quality and ?nd no signi?cant difference between our experimental condi-
tions (p a7e? 0.269; two-tailed).
that this result stems from ACMs believing innovative procedures
lead to higher audit quality than standard procedures. However,
despite the greater accountability and comfort associated with the
introduction of an AJR and a move towards more innovative proce-
dures, our ?ndings show that this does not necessarily lead ACMs
to be more skeptical and ask more probing questions when an AJR
is introduced or ask less probing questions when innovative proce-
dures are used by the auditor.
We note that while an AJR did not lead to more questions be-
ing asked than without an AJR, it also did not lead ACMs to ask less
questions either. Despite the concern that introducing some form
of an AJR will lead to a potential decrease in ?nancial statement
quality due to less conservative ?nancial statement preparation or
lower audit quality due to lower auditor diligence or less audit
committee scrutiny, we did not ?nd any evidence of the latter. In
fact, our supplementary analysis, using an independent coder, on
the level of probing of ACM questions shows ACMs under all treat-
ments exercised a high level of skepticism and asked very probing
questions. We ?nd no evidence that the extent to which ACMs are
probing is adversely affected by the introduction of an AJR either
with standard audit procedures or innovative audit procedures.
Hence, despite the concern for reduced ?nancial reporting quality
with the introduction of an AJR (Pozen, 2008; Treasury, 2008), our
results suggest ACMs will continue to ask probing questions with a
change to an AJR. Moreover, given the stated advantages of an AJR
combined with innovative audit procedures (Peecher et al., 2013),
we suggest further consideration to be given to such changes un-
less future research shows negative consequences on other parts of
the ?nancial reporting chain such as preparers or auditors.
The large number of questions asked and the high level of prob-
ing by ACMs noted above is consistent with the results of inter-
views by Cohen et al. (2010) that in the post-SOX period, ACMs are
asking probing and di?cult questions. This is in contrast to earlier
literature (e.g., Cohen, Krishnamoorthy, & Wright, 2002; Turley &
Zaman, 2007) suggesting ACMs were not effective questioners. Re-
cent professional literature suggests the importance of audit com-
mittees seeking to understand whether or not auditors have exer-
cised an appropriate degree of skepticism on issues where there
is a disagreement between auditors and management as well as
issues where the two parties have agreed (FRC, 2012). The FRC
(2012) further suggests that ACMs should require an explanation
of the auditor’s rationale for particular conclusions, the alternative
that was considered and the reasons for speci?c judgments be-
ing considered the most appropriate. The participants in our study
asked numerous such questions as illustrated in Table 5. These
questions related to such factors as the accuracy of the ?nancial
statements, the processes used by the external auditors, CFOs and
internal auditors in collecting evidence and how each of these
groups gained comfort in the accounting numbers.
Our paper also provides insights on the likely effects of an AJR
on managers and auditors from the viewpoint of ACMs. Looking at
the views of these ACMs is important as they work closely with
both managers and auditors, and will have informed opinions of
the likely effects of an AJR. While there was no evidence of an AJR
negatively affecting the performance of ACMs, ACMs in general be-
lieve that an AJR may lead to management estimates that are less
conservative and that the introduction of an AJR may lead to less
auditor diligence. If these perceptions of ACMs are accurate, this
could potentially lead to overall lower ?nancial statement qual-
ity. However, in a multi-period setting, as management and audi-
tors become aware that ACMs are likely to continue to ask probing
questions, there are incentives for auditors to further improve au-
dit quality with consequent incentives for management to become
more conservative in their estimates.
Supplementary analysis of the questions developed by our ACM
participants show that ACM background (particularly, background
74 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
as a former audit partner) leads to signi?cant differences in how
ACMs are affected by our speci?c manipulations of an AJR and the
nature of audit procedures. In particular, former audit partners be-
came more skeptical and asked more questions when an AJR is
present versus absent, while other ACMs decrease their question-
ing behavior. These results were consistent across the three types
of question categories (reports fairly, evidence collection and gain-
ing comfort). We suggest that a likely explanation for this effect
is that former audit partners better understand the relationship
between auditors and inspectors, and the reasons for suggestions
of second-guessing in the inspection process. Other differences be-
tween former audit partners and other ACMs is that the former
audit partners place a greater reliance on asking ‘evidence collec-
tion’ questions to EAs rather than CFOs, they appear more comfort-
able with innovative audit procedures, they looked for more tar-
geted evidence and their questions were most probing when they
were part of the AJR present/standard audit procedures treatment.
The differences in questioning behavior between ACMs of differ-
ent backgrounds have the potential for cognitive stimulation (Chen,
Trotman, & Zhou, 2015) when the audit committee works as an
interacting group. Future research should examine how and when
these bene?ts will occur. The above ?ndings also support the em-
phasis placed on the importance of ?nancial expertise (e.g., Glover
& Prawitt, 2013; IAASB, 2014) and on diversity among members of
the audit committee (Beasley et al., 2009; Cohen et al., 2010; Ernst,
2013).
Sometimes there are unintended consequences of new controls
and regulations (see Bell et al., 2005) and as a result the intended
consequences are not realized. For example, one potential bene?t
of an AJR is that it is likely to encourage auditors to use less de-
fensive/more innovative audit procedures. A negative consequence
would be if this leads ACMs to ask less probing questions when
more innovative procedures were used. Our results suggest that
this is not likely to be a problem as we ?nd no differences in the
number of questions asked or the level of probing of those ques-
tions when innovative rather than standard procedures were used.
If an AJR is implemented, this may imply that auditors should con-
sider engaging in more innovative procedures as our study ?nds
that these procedures provide ACMs with greater comfort. Also,
as ACMs prefer innovative procedures, and see them as providing
higher audit quality, regulators could consider the potential for en-
couragement and greater acceptance of such procedures.
The introduction of an AJR together with innovative procedures
presents a cost-bene?t tradeoff for ACMs. Previous research has
suggested that an AJR will result in greater use of innovative proce-
dures (Peecher et al., 2013), and if this occurs, there are bene?ts for
ACMs as we ?nd that they are more comfortable with innovative
procedures and see them as providing higher audit quality. How-
ever, for ACMs there is also a cost as they need to be concerned
that the introduction of an AJR may result in potentially greater
reporting aggressiveness by management and potentially less dili-
gence by auditors.
A number of limitations to our study should be noted. First,
we examined the individual judgments of ACMs. However, audit
committees are teams usually consisting of three or more mem-
bers and there can be both process gains and losses from group
interaction. Future research could begin to examine process gains
and losses in audit committees by considering the outputs of both
nominal and interacting teams. Second, we selected a case where
the preparer had decreased the amount of an inventory obsoles-
cence account from their original estimate, and the auditors had
accepted the reduced estimate of the write-down. Using a case
where the auditor had disagreed with the client may affect the
questioning behavior of ACMs and should be considered in future
research. Third, our manipulation of standard versus innovative
procedures was based on previous literature but there are other
alternative descriptions that could have been used and differences
in the nature of the innovative procedures could be considered in
future research. Fourth, as discussed earlier, we did not include
an AJR absent/innovative procedures condition in our design. How-
ever, we note that existing audit standards, both in the US and in-
ternationally, do not preclude auditors from using innovative pro-
cedures and, therefore, future research could examine the effects of
innovative procedures when an AJR rule is absent but other incen-
tives for the use of innovative procedures exist. Fifth, our qualita-
tive analysis of questions asked by ACMs suggests potential inter-
action effects between participant background and the treatment
received. Deeper analysis of issues related to the background of
ACMs will require selection of participants from two diverse back-
grounds (e.g., former audit partners versus non-accountant ACMs)
with larger cell sizes. Sixth, we examined the effect of an AJR on
ACMs’ judgments in a one-period experimental setting. Future re-
search could consider examining the effect in a multi-period set-
ting to see if ACMs adjust their accountability levels after they
see how auditors actually behave in terms of being more or less
diligent with the implementation of an AJR and how auditors and
management react to the level of questioning by ACMs.
Acknowledgements
We thank the participants at the 5th Audit Quality Conference
in Venice, the 2014 Accounting, Organizations and Society Confer-
ence on Accounting Estimates (October 2014) and the Northeast-
ern University Research Workshop. We also thank Sandra Vera-
Muñoz (our commentator), Lisa Koonce (editor), two anonymous
reviewers and the helpful comments of Michael Coleman, Brian
Fitzgerald, Gary Monroe, Mark Nelson, Mark Peecher, Steve Salte-
rio, Roger Simnett, Ira Solomon and Arnie Wright. We acknowledge
a research grant from Chartered Accountants Australia and New
Zealand and the ?nancial support of an ARC Australian Professo-
rial Fellowship to Ken Trotman. The ?nancial support of Deloitte
& Touche LLP for their support of the AOS Estimates Conference is
gratefully acknowledged.
Appendix A
AJR Absent
Current regulation of auditors in Australia
In the current audit environment, auditors’ judgments, even
those that are complex and involve di?cult future estimates, can
be second guessed as courts can rule that alternative judgments
should have been reached. In addition, irrespective of how reason-
able the auditors’ judgment process was, regulatory inspectors (e.g.
ASIC in Australia; PCAOB in the USA) currently have wide authority
to determine that the judgments of auditors are ‘de?cient’.
AJR Present
New regulation of auditors in Australia
Please assume that new legislation in Australia has intro-
duced an Audit Judgment Rule whereby courts and regulatory
inspectors (e.g. ASIC in Australia; PCAOB in the USA) will not
second guess auditor judgments provided the auditor meets
certain key tests. Under an Audit Judgment Rule, a reasonableness
test is applied and even if an inspector develops an estimate that
differs from an auditor’s estimate, uses a different method to reach
an estimate than used by the auditor, or believes that his/her own
estimation process is superior to the auditor’s process, the inspec-
tor still can conclude that the auditor’s estimate is not unreason-
able. The key test is whether the auditor has used ‘reasoned eval-
uation made in good faith and in a rigorous, thoughtful and delib-
erate manner’. In summary, unless the auditor’s judgment process
is unreasonable, the auditor’s judgments cannot be said to be ‘de-
?cient’.
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 75
References
American Institute of Certi?ed Public Accountants (AICPA) (2007). Consideration of
fraud in a ?nancial statement audit. AU Section 316. New York, NY: AICPA.
Australian Securities & Investment Commission (ASIC) (2012). Report 317: Au-
dit inspection program report for 2011-12. Accessed May 2014.
Beasley, M. S., Carcello, J. V., Hermanson, D. R., & Neal, T. L. (2009). The audit com-
mittee oversight process. Contemporary Accounting Research, 26(1), 65–122.
Bell, T. B., Peecher, M. E., & Solomon, I. (2005). The 21st century public company
audit: Conceptual elements of KPMG’s global audit methodology. Montvale, New
Jersey: KPMG.
Bonner, S. E. (2008). Judgment and decision making in accounting. Upper Saddle River,
NJ: Pearson Prentice Hall.
Bratten, B., Gaynor, L. M., McDaniel, L., Montague, N. R., & Sierra, G. E. (2013). The
audit of fair values and other estimates: The effects of underlying environmen-
tal, task, and auditor-speci?c factors. Auditing: A Journal of Practice & Theory,
32(Supplement 1), 7–44.
Centre for Audit Quality (CAQ) (2014). CAQ approach to audit quality indicators.
Accessed May 2014.
Chen, X., Trotman, K. T., & Zhou, F. (2015). Nominal versus interacting electronic
fraud brainstorming in hierarchical audit teams? The Accounting Review, 90(1),
175–198.
Christensen, B. E., Glover, S. M., & Wood, D. A. (2012). Extreme estimation uncer-
tainty in fair value estimates: Implications for audit assurance. Auditing: A Jour-
nal of Practice & Theory, 31(1), 127–146.
Christensen, B. E., Glover, S. M., & Wood, D. A. (2013). Extreme estimation uncer-
tainty and audit assurance. Current Issues in Auditing, 7(1), 36–42.
Church, B. K., & Shefchik, L. B. (2012). PCAOB inspections and large accounting ?rms.
Accounting Horizons, 26(1), 43–63.
Cohen, J., Gaynor, L. M., Krishnamoorthy, G., & Wright, A. M. (2007). Auditor com-
munications with the audit committee and the board of directors: Policy rec-
ommendations and opportunities for future research. Accounting Horizons, 21(2),
165–187.
Cohen, J., Krishnamoorthy, G., & Wright, A. M. (2002). Corporate governance and the
audit process. Contemporary Accounting Research, 19(4), 573–594.
Cohen, J., Krishnamoorthy, G., & Wright, A. M. (2004). The corporate governance
mosaic and ?nancial reporting quality. Journal of Accounting Literature, 23, 87–
152.
Cohen, J., Krishnamoorthy, G., & Wright, A. M. (2010). Corporate governance in the
post-Sarbanes-Oxley era: Auditors’ experiences. Contemporary Accounting Re-
search, 27, 751–786.
Contessotto, C., & Moroney, R. (2013). The association between audit committee ef-
fectiveness and audit risk. Accounting & Finance, 54(2), 393–418.
Deloitte, L. L. P. (2014). Audit committee brief: A New Era in audit committee reporting.
Accessed May 2014.
DeZoort, F. T. (1998). An analysis of experience effects on audit committee members’
oversight judgment. Accounting, Organizations and Society, 23(1), 1–21.
DeZoort, F. T., Harrison, P., & Taylor, M. (2006). Accountability and auditors’ mate-
riality judgments: The effects of differential pressure strength on conservatism,
variability, and effort. Accounting, Organizations and Society, 31(4–5), 373–390.
DeZoort, F. T., Houston, R. W., & Hermanson, D. R. (2003). Audit committee mem-
ber support for proposed audit adjustments: A source credibility perspective.
Auditing: A Journal of Practice & Theory, 22(2), 189–205.
DeZoort, F. T., & Salterio, S. E. (2001). The effects of corporate governance experi-
ence and ?nancial-reporting and audit knowledge on audit committee mem-
bers’ judgments. Auditing: A Journal of Practice & Theory, 20(2), 31–47.
Dhaliwal, D., Naiker, V., & Navissi, F. (2010). The association between accruals qual-
ity and the characteristics of accounting experts and mix of expertise on Audit
Committees. Contemporary Accounting Research, 27(3), 787–827.
Ernst & Young (2013). Insights for European Audit Committees: Greater Business
Challenges Call for Stronger Audit Committees. Issue 18. Accessed May 2014.
European Commission (2010). Green paper audit policy: Lessons from the Crisis.
Federation of European Accountants (FEE) (2014). Opening a discussion: The fu-
ture of audit and assurance. Accessed May 2014.
Financial Reporting Council (FRC) (2012). Auditing practices board: Professional scep-
ticism – establishing a common understanding and rea?rming its central role in
delivering audit quality. London: Financial Reporting Council.
Financial Reporting Council (FRC) (2013a). Audit quality inspections an-
nual report 2012/2013.
Accessed May 2014.
Financial Reporting Council (FRC) (2013b). Audit quality thematic review: Ma-
teriality. Accessed May 2014.
Gendron, Y., & Bédard, J. (2006). On the constitution of audit committee effective-
ness. Accounting, Organizations and Society, 31(3), 211–239.
Gendron, Y., Bédard, J., & Gosselin, M. (2004). Getting inside the black box: A ?eld
study of practices in “effective” audit committees. Auditing: A Journal of Practice
& Theory, 23(1), 153–171.
Gibbins, M., & Newton, J. (1994). An empirical exploration of complex accountability
in public accounting. Journal of Accounting Research, 32, 165–186.
Glover, S. M., & Prawitt, D. F. (2013). Enhancing auditor professional skepticism.
Accessed October 2014.
Grant Thornton (2009). Response Letter in Release No. 104-2009-117: Report account-
ing oversight board.
Gri?th, E. E., Hammersley, J. S., & Kadous, K. (2014). Audits of complex estimates as
veri?cation of management numbers: How institutional pressures shape prac-
tice. Contemporary Accounting Research, in press.
Gri?th, E. E., Hammersley, J. S., Kadous, K., & Young, D. (2015). Auditor mindsets
and audits of complex estimates. Journal of Accounting Research, 53(1), 49–77.
Hammersley, J. S., Johnstone, K. M., & Kadous, K. (2011). How do audit seniors re-
spond to heightened fraud risk? Auditing: A Journal of Practice & Theory, 30(3),
81–101.
Hoffman, V., & Patton, J. (1997). Accountability, the dilution effect, and conservatism
in auditor’s fraud judgments. Journal of Accounting Research, 35(2), 227–237.
Hurtt, R. K. (2010). Development of a scale to measure professional skepticism. Au-
diting: A Journal of Practice & Theory, 29(1), 149–171.
Hurtt, R. K., Brown-Liburd, H., Earley, C. E., & Krishnamoorthy, G. (2013). Research
on auditor professional skepticism: Literature synthesis and opportunities for
future research. Auditing: A Journal of Practice & Theory, 32(Supplement 1), 45–
97.
International Auditing and Assurance Standards Board (IAASB) (2014). A framework
for audit quality: Key elements that create an environment for audit quality.
Accessed May 2014.
International Forum of Independent Audit Regulators (IFIAR) (2014).
2013 annual report & accounts. Accessed May 2014.
International Federation of Accountants (IFAC) (2011). Audit quality: An IAASB per-
spective. New York: IFAC.
International Standards of Auditing (ISA) 240. The Auditor’s responsibilities relating to
fraud in an audit of ?nancial statements. 2013 Handbook of International Qual-
ity Control, Auditing, Review, Other Assurance, and Related Services Pronounce-
ments, IFAC.
Kang, Y. J. (2014). Are audit committees more challenging given a sophisticated investor
base? Does the answer change given anticipation of additional mandatory audit
report Disclosure? University of Massachusetts-Amherst, Working Paper.
Koonce, L., Anderson, U., & Marchant, G. (1995). Justi?cation of decisions in auditing.
Journal of Accounting Research, 33(2), 369–384.
KPMG (2011). Elevating professional judgment in auditing. KPMG LLP.
Krishnan, G. V., & Visvanathan, G. (2008). Does the SOX de?nition of an accounting
expert matter? The association between audit committee directors’ account-
ing expertise and accounting conservatism. Contemporary Accounting Research,
25(3), 827–857.
Lerner, J. S., & Tetlock, P. E. (1994). Accountability and social cognition. In V. S. Ra-
machandran (Series Ed.), Encyclopedia of human behavior: Vol. 1 (pp. 3098–3121).
San Diego, CA: Academic Press.
Lerner, J. S., & Tetlock, P. E. (1999). Accounting for the effects of accountability. Psy-
chological Bulletin, 125(2), 255–275.
Libby, R., & Luft, J. (1993). Determinants of judgment performance in accounting
settings: Ability, knowledge, motivation, and environment. Accounting, Organi-
zations and Society, 18(5), 425–450.
Martin, R. D., Rich, J. S., & Wilks, T. J. (2006). Auditing fair value measurements: A
synthesis of relevant research. Accounting Horizons, 20(3), 287–303.
McDaniel, L., Martin, R. D., & Maines, L. A. (2002). Evaluating ?nancial reporting
quality: The effects of ?nancial expertise vs. ?nancial literacy. The Accounting
Review, 77(s-1), 139–167.
Monks, R., & Minow, N. (2004). Corporate governance. Malden, MA: Blackwell Pub-
lishing Ltd..
National Association of Corporate Directors (NACD) (2013). Honing skepticism. NACD
Directorship. . January/February 2013, 29-37.
Nelson, M. W. (2009). A model and literature review of professional skepticism in
auditing. Auditing: Journal of Practice & Theory, 28(2), 1–34.
Nelson, M. W., Elliott, J. A., & TarpIey, R. L. (2002). Evidence from auditors about
managers’ and auditors’ earnings management decisions. The Accounting Review,
77(4), 175–202.
Nelson, M., & Tan, H.-T. (2005). Judgment and decision making research in auditing:
A task, person, and interpersonal interaction perspective. Auditing: A Journal of
Practice & Theory, 24, 41–71.
O’Connell, J., & Boutros, A. S. (2002). Treating medical malpractice claims under a
variant of the business judgment rule. Notre Dame Law Review, 77, 373–436.
Palmrose, Z. (2006). Maintaining the value and viability of auditors as gatekeepers
under SOX: An auditing master proposal. In U. Fuchita, & R. Litan (Eds.), Finan-
cial gatekeepers: Can they protect investors (pp. 103–135). Baltimore, MD: The
Brookings Institute.
Public Company Accounting Oversight Board (PCAOB) (2010). Report on observations
of PCAOB inspectors related to audit risk areas affected by the economic crisis. Re-
lease No. 2010-006. Washington, D.C.: PCAOB.
76 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
Public Company Accounting Oversight Board (PCAOB) (2011). Concept release on au-
ditor independence and audit ?rm rotation. Release No. 2011-006. Washington,
D.C.: PCAOB.
Public Company Accounting Oversight Board (PCAOB) (2012). Auditing standard no.
16 communications with audit committees (AS 16). Washington, D.C.: PCAOB.
Public Company Accounting Oversight Board (PCAOB) (2013). Report on 2007–2010
inspections of domestic ?rms that audit 100 or fewer public companies. Accessed
May 2014.
Peecher, M. E. (1996). The in?uence of auditor’s justi?cation processes on their de-
cisions: A cognitive model and experimental evidence. Journal of Accounting Re-
search, 34(1), 125–140.
Peecher, M. E., Schwartz, R., & Solomon, I. (2007). It’s all about audit quality: Per-
spectives on strategic-systems auditing. Accounting, Organizations and Society,
32(4–5), 463–485.
Peecher, M. E., Solomon, I., & Trotman, K. T. (2013). An accountability framework for
?nancial statement auditors and related research questions. Accounting, Organi-
zations and Society, 38(8), 596–620.
Pomeroy, B. (2010). Audit committee investigation of signi?cant accounting deci-
sions. Auditing: A Journal of Practice & Theory, 29(1), 173–205.
Pozen, R. (2008). Final report of the advisory committee on improvements to ?nan-
cial reporting to the United States securities and exchange commission. Washington
DC: Securities and Exchange Commission Advisory Committee.
Rowe, S. P. (2014). Auditor Comfort Under Uncertainty. Tulane University Working
Paper.
Securities and Exchange Commission (SEC) 2011. Measurement uncertainty in ?nan-
cial reporting: How much to recognize and how best to communicate it. Brie?ng
Paper for the Financial Reporting Series Roundtable Discussion. Accessed Oc-
tober 2014.
Siegel-Jacobs, K., & Yates, J. F. (1996). Effects of procedural and outcome accountabil-
ity on judgment quality. Organizational Behavior and Human Decision Processes,
65(1), 1–17.
Srinivasan, S. (2005). Consequences of ?nancial reporting failure for outside direc-
tors: Evidence from accounting restatements and audit committee members.
Journal of Accounting Research, 43(2), 291–334.
Tetlock, P. E. (1983). Accountability and complexity of thought. Journal of Personality
and Social Psychology, 45(1), 74–83.
Tetlock, P. E. (1992). The impact of accountability on judgment and choice: Toward
a social contingency model. Advances in Experimental Social Psychology, 25, 331–
376.
Tetlock, P. E., & Boettger, R. (1989). Accountability: A social magni?er of the dilution
effect. Journal of Personality and Social Psychology, 57(3), 388–398.
Tetlock, P. E. (1985). Accountability: The neglected social context of judgment and
choice. In B. M. Staw, & L. Cummings (Eds.), Research in organizational behavior,
7 (pp. 297–332). Greenwich, Connecticut: JAI Press.
The Department of the Treasury (Treasury) (2008). Advisory committee on the audit-
ing profession. Final Report, October 6, 2008.
Turley, S., & Zaman, M. (2007). Audit committee effectiveness: Informal processes
and behavioural effects. Accounting, Auditing & Accountability Journal, 20(5),
765–788.
Vera-Muñoz, S. C. (2005). Corporate governance reforms: Rede?ned expectations
of audit committee responsibilities and effectiveness. Journal of Business Ethics,
62(2), 115–127.
doc_117372679.pdf
Changestotheauditenvironmenthaveledtosuggestedchangestotheregulatoryframeworkforeval-uatingauditors’judgmentsincludingtheintroductionofanAuditJudgmentRule(AJR),wherebycourtsandinspectorswillnotsecond-guessauditors’reasonedjudgmentsprovidedtheyaremadeingoodfaithandinarigorousmanner.WeexaminethepotentialeffectoftheAJRontheskepticismofAuditCom-mitteeMembers(ACMs)intermsoftheextenttowhichtheyaskprobingquestionstoexternalauditors,CFOsandHeadsofInternalAuditconcerninganaccountingestimate.ThislevelofprofessionalskepticismisacriticalelementofthedutiesofanACMintheoversightofthefinancialreportingandauditingpro-cesses,especiallyforcomplexandfutureorientatedaccountingestimates.BecauseanAJRwouldlikelyencourageadoptionofinnovativeauditprocedures,wefurtherexaminetheeffectoftheseprocedures,ascomparedtostandardprocedures,onACMs’skepticismgivenanAJR.
Accounting, Organizations and Society 46 (2015) 59–76
Contents lists available at ScienceDirect
Accounting, Organizations and Society
journal homepage: www.elsevier.com/locate/aos
The effect of an Audit Judgment Rule on audit committee members’
professional skepticism: The case of accounting estimates
Yoon Ju Kang
a
, Andrew J. Trotman
b
, Ken T. Trotman
c,?
a
University of Massachusetts at Amherst, United States
b
Northeastern University, United States
c
UNSW Australia, Australia
a r t i c l e i n f o
Article history:
Received 1 June 2014
Revised 28 February 2015
Accepted 9 March 2015
Available online 7 April 2015
a b s t r a c t
Changes to the audit environment have led to suggested changes to the regulatory framework for eval-
uating auditors’ judgments including the introduction of an Audit Judgment Rule (AJR), whereby courts
and inspectors will not second-guess auditors’ reasoned judgments provided they are made in good faith
and in a rigorous manner. We examine the potential effect of the AJR on the skepticism of Audit Com-
mittee Members (ACMs) in terms of the extent to which they ask probing questions to external auditors,
CFOs and Heads of Internal Audit concerning an accounting estimate. This level of professional skepticism
is a critical element of the duties of an ACM in the oversight of the ?nancial reporting and auditing pro-
cesses, especially for complex and future orientated accounting estimates. Because an AJR would likely
encourage adoption of innovative audit procedures, we further examine the effect of these procedures,
as compared to standard procedures, on ACMs’ skepticism given an AJR. Our ?ndings show that an AJR
increases ACMs’ perceived accountability in ensuring the reasonableness of the ?nancial statements, and
that a movement towards more innovative audit procedures under an AJR framework increases ACMs’
perceived overall comfort regarding the treatment of the accounting estimate. On average, these factors
do not affect the overall level of ACMs’ skepticism in terms of the number of questions asked or the
extent to which the questions are probing. However, these results differ depending on the demographic
background of the ACM participants.
© 2015 Elsevier Ltd. All rights reserved.
Introduction
Internationally, regulatory inspectors have reported audit de?-
ciencies with respect to the audit of estimates, and there have
been calls for improved audit quality (ASIC, 2012; European Com-
mission, 2010; FRC, 2013a; IFIAR, 2014; PCAOB, 2013). In turn, there
is uncertainty about what is expected of an auditor given the com-
plexity and future orientation of accounting estimates and the dif-
?culty of inspectors concluding on the appropriateness of the audi-
tors’ judgments (Peecher, Solomon, & Trotman, 2013). In response
to these and related issues, there have been suggestions for change
to the regulatory framework (e.g., Peecher et al., 2013; Pozen,
2008). One such change is the introduction of an Audit Judgment
Rule (AJR), which Peecher et al. (2013) suggest should be applied
by regulators in evaluating auditors’ professional judgments and
motivating auditors to improve audit quality. This AJR is based on
the Business Judgment Rule (BJR), which applies to directors in the
?
Corresponding author at: School of Accounting, School of Business, UNSW Aus-
tralia, Sydney, NSW 2052, Australia. Tel.: +61 2 9385 5831.
E-mail address: [email protected] (K.T. Trotman).
USA, where directors cannot have their judgments second-guessed
by courts and cannot be held responsible for related third party
losses provided they act in good faith and in the best interests of
the company (O’Connell and Boutros, 2002; Peecher et al., 2013).
Under an AJR courts and inspectors would not second-guess audi-
tors’ judgments provided the auditor has used ‘reasoned evaluation
made in good faith and in a rigorous, thoughtful and deliberate
manner’.
We conduct an experiment to examine the effect of the imple-
mentation of an AJR on the level of skepticism of Audit Commit-
tee Members (ACMs) as measured by their questioning behavior
in overseeing the ?nancial reporting and auditing process. Because
one purpose of an AJR would be to discourage defensive auditing,
and instead encourage a focus on audit effectiveness, we consider
how a change to innovative (as compared to standard) audit pro-
cedures for those in the AJR treatment procedures further affects
ACMs’ level of skepticism. We chose ACMs because they have a
?duciary duty to shareholders and are accountable for effectively
overseeing the ?nancial reporting and audit process (CAQ, 2014,
FEE, 2014, FRC, 2013b, Kang, 2014) by exercising professional skep-
ticism and asking probing questions about the judgments and as-http://dx.doi.org/10.1016/j.aos.2015.03.001
0361-3682/© 2015 Elsevier Ltd. All rights reserved.
60 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
sumptions underlying management’s signi?cant estimates and au-
ditors’ judgments of those estimates (Beasley, Carcello, Hermanson,
& Neal, 2009; Cohen, Gaynor, Krishnamoorthy, & Wright, 2007;
IAASB, 2014; NACD, 2013).
Examining the potential effects of alternative regulatory frame-
works such as the AJR is important given the suggestion that the
existing framework for auditors is unlikely to motivate them to tar-
get high quality audits or invest in research and development ac-
tivities aimed at improving long term audit quality (Peecher et al.,
2013, p. 597). In addition, the adoption of an AJR is consistent
with US Treasury recommendations (Pozen, 2008, p. 93) that both
the SEC and the PCAOB should issue policy statements articulat-
ing how the reasonableness of accounting and auditing judgments
be evaluated including factors to be considered when making this
evaluation. An AJR would also make it more di?cult for inspec-
tors to second-guess auditors’ judgments as long as the judgments
were made using ‘good faith and in a rigorous, thoughtful and de-
liberate manner’. We examine the AJR in a setting with accounting
estimates given that such a judgment rule is most likely to apply
in a setting involving subjective judgments. These accounting esti-
mates are future orientated and are increasingly important to the
?nancial statements (Gri?th, Hammersley, & Kadous, 2014; IFIAR,
2014; Peecher et al., 2013; Treasury, 2008). The uncertainty inher-
ent in estimates provides an opportunity for management bias as
well as di?culties for auditors in assessing their reasonableness,
resulting in potential quality implications for ?nancial statements
(Gri?th, Hammersley, Kadous, & Young, 2015).
In our study we consider both the existing regulatory frame-
work without an AJR, and another where an AJR is implemented.
Under the existing framework, even if auditors’ judgments are
made in good faith, and irrespective of how reasonable the au-
ditors’ judgment process was, courts and regulatory inspectors
can determine that alternative judgments should have been made.
While there are many ways to operationalize an AJR framework,
we chose one where under an AJR a reasonableness test is applied
and even if an inspector develops an estimate that differs from an
auditor’s estimate, uses a different method to reach an estimate
than used by the auditor, or believes that his/her own estimation
process is superior to the auditor’s process, the inspector still can
conclude that the auditor’s estimate is reasonable.
An additional characteristic of the audit environment is that the
auditor faces uncertainty about what is expected of them because
it is di?cult to make a conclusion on the appropriateness of audi-
tor judgments related to complex estimates that are future orien-
tated. Inspectors have extensive authority to determine when au-
ditor judgments are inappropriate, and de?cient judgments have
been documented by inspection agencies around the world (e.g.,
ASIC in Australia; FRC in the UK; PCAOB in the US). One conse-
quence of this situation is the prevalence of “overly cautious audits
or ‘defensive’ auditing” (Treasury, 2008, p. VII: 28),
1
which result
in auditor accountabilities primarily related to auditors’ judgment
outcomes and penalties for not reaching a minimum threshold
(Peecher et al., 2013). This leads to compliance-focused behaviors
with auditors potentially operating at or close to the minimum re-
quired in order to avoid sanctions (Bell, Peecher, & Solomon, 2005;
Peecher et al., 2013). As a result, innovative audit procedures are
unlikely to be adopted for fear of second-guessing by inspectors,
despite suggestions for the need for incentives for auditors “to go
1
There have been criticisms of second-guessing by inspection agencies (see
Treasury, 2008). These views can be seen in the responses to PCAOB inspection re-
ports. For example, “While we believe that the PCAOB should continue to challenge
judgments and documentation during the inspection process, we do not believe
that, in the end, reasonable judgments should be criticized and second-guessed
(Grant Thornton, 2009).”
beyond the ?oor and compete on the basis of quality” (Palmrose,
2006). The adoption of an AJR would provide greater incentives for
audit ?rms to adopt more innovative procedures (Peecher et al.,
2013). Consequently, we restrict our examination of how the adop-
tion of more innovative procedures would affect ACMs’ skepticism
regarding a signi?cant estimate to the situation where an AJR is
present. If an AJR is likely to encourage adoption of innovative pro-
cedures, an understanding of how ACMs will react to these proce-
dures is important in order to gain a fuller understanding of the
impact of the introduction of an AJR.
Our ?ndings show that introducing an AJR increases ACMs’ per-
ceived accountability in ensuring the reasonableness of the ?nan-
cial statements. Further analysis shows that this is related to ACMs
believing that accounting estimates become less conservative, and
auditors’ due diligence to be negatively impacted, with the intro-
duction of an AJR. We also ?nd that a move towards more inno-
vative audit procedures under an AJR framework increases ACMs’
perceived overall comfort regarding the accounting treatment at
hand. This result appears to stem from ACMs believing innova-
tive procedures lead to higher audit quality. Despite the greater ac-
countability associated with the introduction of an AJR and more
comfort with a move towards more innovative audit procedures,
our ?ndings show that this does not necessarily lead ACMs to be-
come more skeptical and ask more probing questions when an AJR
is introduced or ask less probing questions when innovative pro-
cedures are used. However, further analysis suggests a signi?cant
background effect where former audit partners show greater skep-
ticism by questioning the external auditor more when an AJR is
implemented than when it is absent, while other ACMs question
the auditor less in the presence of an AJR. We also found that the
types of questions asked vary with ACM background and, in partic-
ular, the questioning behavior of former audit partners is different
to other ACMs.
Our paper makes four major contributions to the audit liter-
ature. First, there have been calls for changes to the regulatory
framework for auditing including the need for a professional judg-
ment framework (e.g., Pozen, 2008) such as an AJR as proposed by
Peecher et al. (2013). By examining how audit committee skepti-
cism in overseeing the ?nancial reporting process is affected under
an AJR, we answer the call for research on how the adoption of an
AJR would affect the execution of ACMs’ ?duciary duties (Peecher
et al., 2013). Also, given ACMs work closely with auditors, exam-
ination of the questions asked and the judgments made by ACMs
provides a ‘window’ to observe how ACMs expect auditors to re-
act to the implementation of an AJR (see Nelson, Elliott, & TarpIey,
2002 for a similar approach).
Second, one of the major criticisms of the present audit reg-
ulatory framework is that it results in defensive auditing because
of the likely negative consequence of moving to more innovative
procedures (Bell et al., 2005; Peecher et al., 2013; Treasury, 2008).
One of the proposed bene?ts of an AJR is the opportunity to re-
duce defensive auditing and encourage auditors to use more inno-
vative audit techniques. We provide insight on how such a move
towards more innovative audit procedures would affect ?nancial
reporting quality under an AJR by examining how it in?uences au-
dit committee skepticism when overseeing the ?nancial reporting
process of accounting estimates. This question is important given
statements in the professional literature that the way standards
are applied “may be seen as an impediment to innovation” and
the need for standard setters to ensure that audit standards do not
inhibit innovation and developments in practice (FEE, 2014, p. 6).
Third, we contribute to the literature on accounting estimates.
Previous research has suggested that estimates are one of the likely
cited accounts for auditor errors (ASIC, 2012; Church & Shefchik,
2012; Gri?th et al., 2015) with implications for ?nancial report-
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 61
ing quality, and there have been suggestions of low audit qual-
ity in this area (European Commission, 2010; IFAC, 2011; PCAOB,
2010). Increasingly the audit committee is expected to exercise
professional skepticism and ask probing questions about these es-
timates (CAQ, 2014; FRC, 2013b). We address the following re-
search question posed by Beasley et al. (2009): “What is the nature
of the audit committee’s involvement in reviewing the company’s
accounting policies and accounting estimates/judgments assump-
tions?” (Beasley et al., 2009) by examining the detailed questions
our ACM participants ask about a signi?cant estimate to other key
corporate governance stakeholders: external auditors, the CFO, and
the Head of Internal Audit.
Fourth, previous research has suggested that the judgments
and questioning behavior of ACMs can depend on ACMs’ previ-
ous background including experience (Beasley et al., 2009; De-
Zoort, 1998) and ?nancial expertise (McDaniel, Martin, & Maines,
2002). We split our ACM participants based on their demographic
background: former audit partners, other quali?ed accountants and
non-accountant directors, which allows us to further explore differ-
ences with respect to our hypotheses and the variations in types
of questions asked by ACMs with different backgrounds. Follow-
ing our experiment, we also ask a series of survey questions to
gain insights into the perceptions of extremely experienced ACMs
about an AJR, and the use of more innovative audit procedures on
?nancial statement preparation, auditor judgments and inspector
judgments.
Background
Accounting estimates form the basis for most ?nancial state-
ment information and recent research shows there is an increase
in their use and complexity (Bratten, Gaynor, McDaniel, Mon-
tague, & Sierra, 2013; Christensen, Glover, & Wood, 2012; Grif-
?th et al., 2015). Professional literature also acknowledges “…
that the increasing complexity of global business operations are
compelling a growing use of judgments and estimates” (Treasury,
2008, p. VII: 17) and has been seen as a signi?cant issue for
regulators (Christensen, Glover, & Wood, 2013; SEC, 2011). When
auditing accounting estimates, auditors are required to conclude
on their ‘reasonableness’, which is a matter of professional judg-
ment. However, the uncertainty involved in accounting estimates
increases the challenges for auditors and inspection reports glob-
ally have identi?ed audit de?ciencies related to accounting esti-
mates (e.g., European Commission, 2010; IFAC, 2011; IFIAR, 2014;
PCAOB, 2013).
Along with auditors, ACMs also play a critical role related to
accounting estimates as they are extensively involved in reviewing
the estimates as well as the underlying judgments and assump-
tions made by management (Beasley et al., 2009). Moreover, the
role of the audit committee in monitoring the ?nancial reporting
process has expanded with substantial attention to its roles in ex-
ercising professional skepticism and supporting investor interests
(Deloitte, 2014; NACD, 2013).
Professional skepticism is a concept that is discussed frequently
both in academic research and in professional standards with
varying de?nitions such as an attitude that includes a question-
ing mind and a critical assessment of evidence (AICPA, 2007, ISA
240), the need for a larger and/or more persuasive sets of evi-
dence (Hurtt, Brown-Liburd, Earley, & Krishnamoorthy, 2013; Nel-
son, 2009), and the need for critical thinking and looking for
contradictory evidence (Gri?th et al., 2015). Gri?th et al. sug-
gest that improved critical thinking is more important than in-
creased doubt, or increased demand for evidence, in improving
audit quality. The literature has also moved from a ‘neutral’ to
‘presumptive doubt’ view of skepticism (Bell et al., 2005; Nelson,
2009) and differentiates between skeptical judgment and skepti-
cal action (Hurtt et al., 2013; Nelson, 2009). Center to this dis-
tinction is the idea that professional skepticism must reach some
level of threshold before it is translated into an action. Fur-
ther, given a constant level of skeptical judgment, skeptical ac-
tion can be affected by various factors such as one’s traits, incen-
tives, and knowledge (Nelson, 2009). Here, we adopt a “presump-
tive doubt” perspective of professional skepticism with a focus on
skeptical action rather than judgments, where being more skep-
tical relates to showing a greater need to collect a more persua-
sive set of evidence before concluding that an assertion is correct
(Nelson, 2009).
We characterize an ACM who is more skeptical as one who
is more questioning of management’s assertion related to an ac-
counting estimate as well as the auditors’ judgments and deci-
sions regarding management’s assertions. This is consistent with
the academic and regulatory emphasis on ACMs asking challenging
questions of external auditors, management and internal auditors
in their oversight of the ?nancial reporting and auditing process.
Speci?cally, in their in-depth interviews with 42 executives serv-
ing on US public company audit committees, Beasley et al. (2009)
?nd that the most important trait an ACM needed to possess is
to be skeptical by asking probing and challenging questions, sup-
porting earlier interview data by Gendron and Bédard (2006) who
?nd that an ACM’s skill in asking challenging questions is central
to their effectiveness as an ACM. In a survey of Australian audit
partners and managers, Contessotto and Moroney (2013) ?nd that
external auditors perceive “asking challenging questions of man-
agement and the auditor” to be by far the most important charac-
teristic of ACMs. The above results are consistent with PCAOB an-
nouncements (PCAOB, 2011; NACD, 2013) suggestions for the need
for audit committees to be more skeptical of both internal and ex-
ternal auditors as well as ?nancial executives in terms of “insisting
on logical responses to relevant questions” (NACD, 2013, p. 36).
Nelson (2009) developed a model of how audit evidence com-
bines with pre-existing knowledge, traits, and incentives to affect
the amount of professional skepticism in auditor judgments and
actions. Hurtt et al. (2013) extend that model and provide a com-
prehensive summary of research studies related to auditor skep-
ticism. We suggest that the Nelson (2009) model of professional
skepticism can be applied to ACMs by considering evidential input,
knowledge, traits and incentives of ACMs.
For ACMs the evidential input will be any information collected
and considered by the ACMs in respect of a particular accounting
treatment (Pomeroy, 2010). ACMs receive information from man-
agement, internal auditors and external auditors and the reliance
they place on these different sources are likely to vary with ACMs’
knowledge which in turn will be affected by background and ex-
perience. ACMs have a wide array of different backgrounds (e.g.,
former audit partners, CFOs, non-executive directors with limited
accounting knowledge). Previous research on ACMs has suggested
important differences based on ACM background and experience
(Beasley et al., 2009; DeZoort, 1998; DeZoort, Houston, & Her-
manson, 2003; DeZoort & Salterio, 2001; McDaniel et al., 2002;
Pomeroy, 2010). The most common measure of traits used in the
auditor professional skepticism literature relates to trait skepticism
as measured by the Hurtt (2010) scale. While there is no research
that has examined this scale for ACMs, we expect the range of
trait skepticism scores for ACMs to be wider than for auditors be-
cause of their more varied backgrounds. Finally, previous research
on ACMs has suggested a range of key incentives for ACMs: es-
tablish individual effectiveness as an ACM and overall effectiveness
of the audit committee (Gendron & Bédard, 2006; Gendron, Bé-
dard, & Gosselin, 2004), avoid loss of reputation for ACMs (Peecher
et al., 2013), constrain aggressive and inappropriate ?nancial
62 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
reporting (DeZoort & Salterio, 2001) and vigilantly monitor ?nan-
cial reporting risks to protect shareholders (Beasley et al., 2009).
Given these incentives for ACMs to be skeptical and ask probing
questions, Kang (2014) examines two potential determinants of au-
dit committee skepticism in terms of their propensity to challenge
signi?cant accounting estimates. Based on the theory of account-
ability, she predicts and ?nds that ACMs are more skeptical and
ask more probing questions given a sophisticated versus unsophis-
ticated investor base, and that this difference narrows in the pres-
ence of an anticipated additional audit report disclosure regarding
such estimates where ACMs decrease their overall level of skepti-
cism and ask less probing questions.
In this study we examine the effect of perceived account-
ability on ACMs’ skepticism by studying the likely effect of an
AJR on the extent to which ACMs question and ask for fur-
ther evidence related to a signi?cant accounting estimate. Imple-
mentation of an AJR is proposed by Peecher et al. (2013) as a
mechanism to hold auditors accountable for their judgment pro-
cesses and ultimately improve audit quality. Currently, inspec-
tors document in their inspection reports examples of judgments
of auditors that they have determined are inappropriate, irre-
spective of how reasonable the auditor judgment process was
(Peecher et al., 2013). This paper examines the likely effect the
implementation of an AJR will have on ACMs’ perceived ac-
countability and ultimately their level of skepticism in oversee-
ing the ?nancial reporting and auditing process of accounting
estimates.
Prior research and development of hypotheses
As ?duciaries of shareholders, ACMs, along with management
and auditors, form the corporate governance mosaic (Cohen, Krish-
namoorthy, & Wright, 2010) in ensuring high quality ?nancial re-
porting for public companies. Speci?cally, ACMs are charged with
overseeing the ?nancial reporting and auditing process (CAQ, 2014;
IAASB, 2014; NACD, 2013; PCAOB, 2012). Findings from interviews
show that ACMs perceive themselves to be accountable to share-
holders and believe themselves to be monitoring agents of man-
agement (Beasley et al., 2009). This implies that factors that in-
?uence ACMs’ perceived accountability towards shareholders will
likely affect their level of skepticism in overseeing the ?nancial re-
porting and audit process.
Theory holds that accountability pressure, the pressure to jus-
tify one’s view to others (Tetlock, 1983), leads individuals to de-
velop different social and cognitive strategies to obtain accep-
tance from, or avoid con?ict with, important interpersonal or in-
stitutional audiences (Gibbins & Newton, 1994; Hoffman & Patton,
1997; Koonce, Anderson, & Marchant, 1995; Lerner & Tetlock, 1994;
Lerner & Tetlock, 1999; Peecher, 1996; Tetlock, 1983). Studies that
examine the effect of accountability distinguish between two types
of accountability: process accountability and outcome accountabil-
ity. While outcome accountability holds decision makers account-
able for the actual outcome of their decision process, process ac-
countability holds decision makers accountable for the quality of
their decision process, regardless of the outcome of their decisions
(Lerner & Tetlock, 1999; Tetlock, 1983; Tetlock, 1985). Various stud-
ies in both the psychology and audit literatures (see Peecher et al.,
2013 for an overview) have shown process accountability, as com-
pared to outcome accountability, to more effectively increase inte-
grative complexity of thought (Tetlock, 1983; Tetlock & Boettger,
1989), increase cognitive effort, improve calibration, and reduce
decision biases (DeZoort et al., 2006; Koonce et al., 1995; Siegel-
Jacobs & Yates, 1996; Tetlock, 1992). Process accountability is likely
to be the primary type of accountability faced by ACMs as they
are aware of protection from the BJR, and therefore know they
are more accountable for the process rather than the outcome.
However, outcome accountability is also relevant as potential legal
action is more likely when there is a negative outcome (Peecher
et al., 2013) and negative outcomes can lead to reputation loss for
the ACMs as well as the auditors and managers.
In this paper, we study one potential factor of ACMs’ account-
ability, implementation of an AJR, and its effect on ACMs’ level of
skepticism. As discussed above, according to Nelson’s (2009) model
on professional skepticism, incentives play an important role in de-
termining the level of skepticism. Incentives that favor professional
skepticism are generated by potential regulatory enforcements by
various regulatory bodies and consequent reputation loss (Nelson,
2009). In the context of audit committees, the risk of litigation
and potential reputation loss will likely increase their perceived ac-
countability pressure in ensuring the reasonableness of the ?nan-
cial statements, and thus affect their level of skepticism in moni-
toring the ?nancial reporting process.
In the current audit environment, auditors’ judgments, even
those that are complex and involve di?cult future estimates, can
be second-guessed as courts can rule that alternative judgments
should have been reached. In addition, regulatory inspectors, such
as the PCAOB, currently have wide authority to determine that the
judgments of auditors are ‘de?cient’, irrespective of the reason-
ableness of the auditors’ judgment process. Peecher et al. (2013)
suggest this leads auditors to take less risks and be less willing
to develop ways to improve existing best practices in fear of po-
tential liability (p. 608). They further suggest that introducing an
AJR, fashioned after the BJR that applies to corporate directors in
the US, will help alleviate such problems as it will give auditors
greater con?dence that their use of innovative procedures would
not be second-guessed by inspectors or courts, providing greater
incentives for them to adopt more innovative procedures.
With the implementation of such an AJR, it is possible for ACMs
themselves to perceive less litigation risk and lower likelihood
of ‘inappropriate’ adverse inspection reports, decreasing the like-
lihood of reputation damage associated with an adverse inspec-
tion report. This could result in lowering both ACMs’ accountability
pressure and lead them to be less skeptical of signi?cant account-
ing estimates in the presence of an AJR. However, as discussed be-
low, we suggest that it is more likely that an AJR will increase
ACMs’ accountability due to audit committees taking a greater role
in the protection of shareholders.
The reduction in potential liability under an AJR is likely greater
for auditors than ACMs as it directly relates to how inspectors
should evaluate auditor judgments. This reduction in auditor’s risk
of potential liability may provide opportunities for auditors to be
more lenient in their audit procedures (Treasury, 2008). Concerned
that auditors will partake in such behavior, implementation of the
AJR will likely lead ACMs to feel more accountable towards share-
holders about ensuring high quality ?nancial reporting (e.g., ensure
the reasonableness of management’s signi?cant estimates) both in
terms of the outcome and process. Moreover, prior research shows
that ACMs bear signi?cant labor market penalties and reputation
damage in the event of ?nancial reporting failures (Monks & Mi-
now, 2004; Peecher et al., 2013; Srinivasan, 2005). Hence, the
possibility of auditors becoming more lenient under an AJR will
likely increase ACMs’ potential reputation damage, making them
feel more accountability to ensure reasonableness of the ?nancial
statements. Further, increases in accountability pressure have been
shown to lead ACMs to increase their level of skepticism in over-
seeing the ?nancial reporting and auditing process (Kang, 2014).
Therefore, we predict the introduction of the AJR will lead ACMs
to feel greater accountability pressure towards shareholders and
be more skeptical of management’s signi?cant estimates by ask-
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 63
ing more questions about the processes used by the external audi-
tor, CFO and the internal auditor, including the evidence they used,
how they gained comfort, their assessment of management’s in-
tegrity, etc. This leads to our ?rst set of hypotheses formalized be-
low.
H1a. Given standard audit procedures, ACMs will feel greater ac-
countability when an AJR is present than when it is absent.
H1b. Given standard audit procedures, ACMs will show higher lev-
els of professional skepticism when an AJR is present than when it
is absent.
While there is a perception that the reduction in potential li-
ability risk under an AJR may result in auditors being more le-
nient in their audit approach (Treasury, 2008), it may simultane-
ously lead auditors to increase their willingness to develop and
conduct more creative and innovative audit procedures (Peecher
et al., 2013). Such practice would address the criticism of defen-
sive auditing (Peecher et al., 2013; Treasury, 2008) and the call to
use more innovative audit procedures to detect ?nancial statement
fraud (Bell et al., 2005).
While there is very little research examining ACMs’ comfort
level, Rowe (2014) considers auditor comfort with management’s
estimates. He notes that the notion of comfort is commonly used
to describe the extent of auditors’ willingness to accept man-
agement assertions (e.g., Bell et al., 2005; KPMG, 2011; Mar-
tin, Rich, & Wilks, 2006; Pomeroy, 2010). The concept of ‘com-
fort’ extends beyond the appropriateness of management estimates
to the quality of those estimates (Rowe, 2014). Similarly, ACMs
need to make assessments of the appropriateness of the com-
bined judgments made by management and the auditors and they
rely heavily on external auditors in giving that comfort (Beasley
et al., 2009; Gendron et al., 2004). In doing so, it is possi-
ble that ACMs prefer and ?nd more comfort in “tried and true
procedures” (i.e., standard procedures) as opposed to more in-
novative procedures. We suggest, however, it is more likely for
ACMs to expect more innovative audit procedures to be more
likely to detect ?nancial statement fraud, leading them to perceive
greater comfort in the reasonableness of the ?nancial statements
when the auditor uses more innovative, as opposed to standard,
procedures.
H2a. Given an AJR, ACMs will have more comfort in the ?nancial
statements when the auditor uses innovative versus standard pro-
cedures.
Rowe (2014) suggests that auditor comfort affects auditors’ ex-
pectation of the most likely adjustment to the ?nancial statements
and that this in turn leads to skeptical actions (cf. Nelson, 2009)
where auditors focus their questions on the uncertainty of man-
agement estimates. With respect to ACMs, Kang (2014) ?nds that
greater comfort regarding accounting estimates lead ACMs to ask
less questions. Similarly, the potential greater comfort ACMs feel
from innovative audit procedures in our study may result in ACMs
being less skeptical. However, the fact that the procedures are in-
novative means that ACMs are likely to be less familiar with these
procedures and, therefore, to be more skeptical regarding these
procedures. Consequently we test the null hypothesis.
H2b. Given an AJR, the extent of ACMs’ professional skepticism
will not differ depending on whether the auditor uses innovative
or standard audit procedures.
Research questions
In addition to our main hypotheses, we examine three research
questions based on qualitative analysis of the questions developed
by our ACM participants. We speci?cally focus on how the demo-
graphic backgrounds of the participants affect their level of skepti-
cism in terms of asking probing questions regarding an accounting
estimate.
Previous research has emphasized the need for ACMs to ask
probing questions (Kang, 2014; Pomeroy, 2010), the need for ACMs
to be skeptical (FEE, 2014; IAASB, 2014), and that there is differ-
ent levels of performance across ACMs (Beasley et al., 2009; Cohen,
Krishnamoorthy, & Wright, 2004; Gendron & Bédard, 2006). There
has also been a long history of audit judgment research related to
knowledge and experience (Bonner, 2008; Libby & Luft, 1993; Nel-
son & Tan, 2005; Peecher, Schwartz, & Solomon, 2007). DeZoort
(1998) found that both general domain and task speci?c expe-
rience affect ACMs’ internal control assessments, and that ACMs
with experience make judgments more like an auditor than ACMs
without experience. In this paper we extend the literature on ACM
experience to consider prior experience as a former audit partner,
which is particularly important because the appointment of former
audit partners as ACMs has become far more common with the in-
troduction of requirements related to designated ?nancial experts
(Vera-Muñoz, 2005). The use of these designated ?nancial experts
has been found to be associated with improved ?nancial reporting
quality (Dhaliwal, Naiker, & Navissi, 2010; Krishnan & Visvanathan,
2008). McDaniel et al. (2002) also ?nd ?nancial expert ACMs differ
from ?nancial literate ACMs in their judgments and evaluations of
?nancial reporting quality. They call for research to see how “vari-
ous types of ?nancial experts differ in their reporting quality judg-
ments” and subsequently there have been suggestions of the need
for research on the composition of audit committees, including the
effect of different backgrounds of ACMs (Glover & Prawitt, 2013;
NACD, 2013).
We address the following research questions.
1. RQ1: Does ACM background affect the types of questions asked
and the level of probing by ACMs to external auditors, CFOs and
Heads of Internal Audit?
2. RQ2: Are there differences across treatments in the extent of
questions asked and type of questions asked between former
audit partners and other ACMs?
Given the seniority of the participants in our study we also
asked a series of survey questions related to our key independent
variables in the experiment. Peecher et al. (2013) suggest potential
bene?ts for audit quality with the adoption of an AJR and the use
of more innovative/less defensive audit procedures. However, there
have been views on potential adverse effects of implementing an
AJR (Treasury, 2008). Further, one of the suggested reasons that in-
novative procedures have not been introduced in the present regu-
latory framework is because of a potential negative effect on in-
spectors’ assessments (Peecher et al., 2013). Hence, we examine
the views of ACMs on these issues.
1. RQ3: What is the perceived impact of an AJR on management’s
estimates and the external auditor’s due diligence and how is a
change to more innovative audit procedures perceived by ACMs
to affect audit quality and audit inspectors?
Research methods
Participants
Forty-nine ACMs from Australia participated in the study. All
had served on at least one audit committee and on average had
served on a mean of 6.77 audit committees (2.86 public compa-
nies, 2.04 private companies and 1.88 other audit committees). On
average, they had served as ACMs for 10.33 years. Their average
64 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
Table 1
Participant demographics.
Mean Median SD
Panel A: Audit committee service and experience (n = 49)
Public company audit committees served on 2.86 2.00 3.27
Private company audit committees served on 2.04 2.00 2.19
Other company audit committees served on 1.88 2.00 1.72
Total audit committees served on 6.77 5.00 5.06
Number of years as an Audit Committee Member 10.33 10.0 7.94
Experience (years since undergraduate degree) 36.02 37.0 11.76
Panel B: Number of ACMs with the following attributes (n = 49) n
Number who had served on at least one public company audit committee 35/49
Number of quali?ed accountants (CPA/CA quali?ed) 35/49
Number with external audit experience 25/49
Position held: Partner 15/25
Other 10/25
Panel C: Self-rated level of knowledge (n = 49)
Financial accounting 79.27 80.80 16.73
Financial statement analysis 79.72 81.40 18.15
Audit committee best practice 77.40 80.50 17.79
External auditing 78.13 82.00 19.08
Internal audit best practice 73.27 75.00 18.77
CPA/CA quali?ed means they are a member of CPA Australia or Chartered Accountants Australia and New Zealand.
ACM= Audit Committee Member.
business experience was 36.02 years (measured as the time since
they ?nished their undergraduate degree). Of the 49 participants,
35 had served on the audit committee of at least one publicly
listed company (public company ACMs), 35 were quali?ed accoun-
tants being CPAs, chartered accountants or equivalent (CPA/CA),
and 14 were not accountants (non-accountants). Of the CPA/CAs,
25 had audit experience (15 of which are former audit partners).
2
For further analysis we classify the 49 ACMs into three categories
(referred to as ACM backgrounds): former audit partners, accoun-
tants who were not former audit partners (other accountants), and
non-accountants. In sum, this was a very senior group of ACMs.
Table 1 provides demographic details.
Participants were obtained through personal emails to alumni
and business contacts in Australia.
3
Potential participants were in-
formed that this was an on-line research study, which provides a
set of facts based on a hypothetical company and asks participants
what questions they would raise as an ACM of the company and
to make a series of professional judgments. Given the seniority of
the participants, con?dentiality of the data was emphasized and
participants were told that they would not be asked to provide
either their name or the organization(s) of which they were an
ACM when completing the research instrument. Individuals who
responded to our invitation to participate were then emailed the
link to the on-line study.
4
Research design
To test our hypotheses, we used a 2 × 1 + 1 between subjects
design. We ?rst manipulate the presence/absence of an AJR be-
tween subjects. Both these treatments hold constant the fact that
the auditor uses standard audit procedures. Our ‘plus one’ treat-
ment is the same as the AJR present treatment except that the au-
2
The audit experience of those that were not former audit partners was low with
all except one leaving auditing before reaching the manager level.
3
The emails were sent from the researchers or from a senior practitioner who is
on a university advisory board. In addition, three participants were recruited via
a LinkedIn alumni broadcast but this method was relatively unsuccessful so we
moved to the personal contact approach.
4
Four research instruments were completed on-line by individuals who had not
served on an audit committee. Their responses were not analysed as they did not
meet the criteria of having served on an audit committee. Six other participants
opened the link but did not complete the study.
ditor uses innovative procedures instead of standard procedures.
Participants were randomly allocated to one of three treatments:
(1) AJR absent/standard procedures, (2) AJR present/standard pro-
cedures, or (3) AJR present/innovative procedures. Our design al-
lows us to ?rst address the potential effect of an AJR while holding
constant the type of audit procedures presently used in practice.
Present audit procedures have been described as defensive auditing
(Bell et al., 2005; Treasury, 2008), and the introduction of more in-
novative audit procedures are unlikely to be introduced under the
present regulatory framework (Peecher et al., 2013). However, one
of the potential advantages of the introduction of an AJR is that
the changed regulatory environment would likely lead to more in-
novative audit procedures, potentially leading to higher audit qual-
ity. Consequently, we include an AJR present/innovative procedures
treatment to compare to our AJR present/standard procedures but
we did not include an AJR absent/innovative procedures treatment.
In the AJR present treatments, participants are told to assume
that new regulation has been introduced in Australia on an AJR un-
der which courts and regulatory inspectors will not second guess
auditor judgments provided the auditor meets certain key tests.
They are informed that the key test is whether the auditor has
used “reasoned evaluation made in good faith and in a rigorous,
thoughtful and deliberate manner”. In the AJR absent treatment,
participants were provided with information on the ‘Current Reg-
ulation of Auditors in Australia’ where “auditors’ judgments, even
those that are complex and involve di?cult future estimates, can
be second guessed as courts can rule that alternative judgments
should have been reached.” Appendix A illustrates the different
language that was used in the AJR present versus absent treat-
ments.
We also manipulate the type of audit procedures used (stan-
dard vs. innovative). While there have been a number of refer-
ences to the bene?ts of a move away from defensive auditing
to more innovative procedures (Bell et al., 2005; Peecher et al.,
2013), ACMs could interpret the terms standard and innovative in
a range of ways and, therefore, we set out descriptions of these
terms. For the standard audit procedures treatment, participants
were told that the audit partner prefers standard audit proce-
dures (e.g., well tried and accepted procedures commonly found
in audit manuals) over innovative audit procedures (e.g., inno-
vative use of external data outside the control of management),
and the partner notes that audit inspectors expect standard pro-
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 65
cedures and these standard procedures are helpful in reducing in-
spectors’ criticism of the audit, especially when there is poten-
tial for ?nancial statement fraud. Alternatively, for the innovative
audit procedures treatment, participants are told that the audit
partner prefers innovative audit procedures (e.g., innovative use of
external data outside the control of management) over standard
audit procedures (e.g., well tried and accepted procedures com-
monly found in audit manuals), and that the partner notes that
innovative audit procedures are helpful in adding a surprise el-
ement to the more predictable standard audit procedures, espe-
cially when there is potential for ?nancial statement fraud. This
operationalization of innovative procedures drew on suggestions
from Bell et al. (2005) where the audit procedures suggested in-
cluded types of evidence collection that would be more effec-
tive in detecting ?nancial statement fraud and would be seen as
innovative.
Case
We adapted the case materials developed by Kang (2014) with
major changes. We added new manipulations and extensively
changed the questions asked in order to test our hypotheses and
research questions. In addition, we shortened the case and changed
some terminology to make it more consistent with the Australian
environment.
Participants were asked to assume they are an ACM for a hy-
pothetical manufacturing company. They were then provided with
information on the audit environment including the nature of ac-
counting estimates, the relevant audit standards, what is required
of the auditor and the complexity/di?culty of auditor judgments
related to estimates. This was followed by our manipulation of
the presence/absence of an AJR described in the Research Design
section.
Participants were then given background information on the
company including an overview of the company’s operations, per-
formance, strategy, investor base, and internal audit function. Fur-
ther information was provided on the company’s external audit en-
gagement partner (including the partner’s preference for standard
or innovative audit procedures) and that the engagement team
used innovative/standard audit procedures to collect evidence re-
lated to the accounting estimates. After reading the background in-
formation, all participants read a written communication provided
by the external auditor to the audit committee regarding “a signif-
icant accounting issue related to obsolete inventory”. This commu-
nication outlined details about the nature of the accounting issue,
management’s initial and revised (more favorable) estimate,
5
and
the auditor’s assessment of management’s ?nal estimate where
they note that a ‘clean’ opinion is to be issued. The communica-
tion was followed by ?nancial statements re?ecting the effects of
the alternate (initial and revised) estimates.
Dependent variables
We measure our main construct of ACM skepticism by exam-
ining the extent of questioning exercised by our ACM participants
regarding the accounting estimate described in the case. Speci?-
cally, participants were asked to develop questions they would ask
the audit partner, CFO and the Head of Internal Audit regarding
the accounting issue at hand, in the order they came to mind. Par-
ticipants were told to provide as much detail as possible to these
open ended questions.
5
The revised write-down of $970,000 was only 24 percent of the originally pro-
posed write-down of $4 million.
In order to examine how an AJR affects ACMs’ perceived ac-
countability and test H1a, we also ask participants to indicate the
extent to which they feel accountable to ensure the reasonable-
ness of the ?nancial statements given the audit environment pre-
sented to them on a 7-point scale ranging from ?3 to 3 (‘sig-
ni?cantly not accountable’ to ‘signi?cantly accountable’). Lastly, in
order to examine how innovative versus standard procedures af-
fect ACMs’ perceived comfort over the accounting estimate, we ask
participants to rate their comfort level on a 7-point scale (?3 to
3) from ‘completely uncomfortable’ to ‘completely comfortable’ on
the following three items: the change in management’s estimate of
obsolete inventory; the external auditor’s decision to allow man-
agement’s updated, smaller write-down of inventory; and the dif-
ference in net pro?t that results from the different write-down
amounts for obsolete inventory.
Survey questions asked
In our post-experiment questionnaire, we asked the following
four survey questions to gain insights from the very senior ACMs
participating in the study. First, on a 7-point scale ranging from ?3
(‘extremely low’) to 3 (‘extremely high’) participants were asked
the effect on audit quality if auditors move from standard audit
procedures to more innovative audit procedures when there is po-
tential for ?nancial statement fraud. Second, on a 7-point scale
ranging from ‘extremely negative’ to ‘extremely positive’, partici-
pants were asked how they believed audit inspectors would re-
act to a move from standard audit procedures to more innova-
tive audit procedures. Third, they were asked to assume “an envi-
ronment where courts and inspection agencies (e.g., ASIC, PCAOB)
cannot second-guess ‘reasoned evaluation made in good faith and
in a rigorous, thoughtful and deliberate manner’ by the auditor”.
They were then asked “how do you think management’s estimates
would be affected” on a 7-point scale from ?3 (‘estimates will be-
come extremely less conservative’) to 3 (‘estimates will become ex-
tremely more conservative’) and “to what extent would your per-
ception of the external auditor’s due diligence be affected” on a
7-point scale ranging from ?3 (‘extremely negatively affected’) to
3 (‘extremely positively affected’).
Coding of qualitative analysis
The questions developed by the participants were ?rst read
by two of the researchers, blind to treatments and demographic
details, who developed a set of question type categories (listed
in Table 5) and then allocated questions provided by the partic-
ipants to the question categories. The level of intercoder agree-
ment was 88.4 percent and differences were resolved by discus-
sion. Three themes evolved: reports fairly, evidence collection and
gaining comfort (see Table 5). Under ‘Reports Fairly’ most ques-
tions related to whether the reported earnings are fairly presented
and not overstated, for example: ‘tone at the top’ issues, incentives,
debt covenants, pressure exerted on the auditor, the potential for
bias, the control environment, and the impact on shareholders. The
‘Evidence Collection’ category referred to questions about the use
of external evidence, general questions about what audit testing
was conducted, and more speci?c questions asking for more detail
on tests performed or reference to what further tests could/should
be done. The evidence collection questions asked to auditors of-
ten referred to audit evidence whereas questions to CFOs referred
to the data used by management. The third theme related to the
ACM gaining comfort for themselves and asking how the auditor,
CFO and internal auditor gained comfort. Often there was men-
tion of the term ‘comfort’, whether/how the auditor was satis?ed
66 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
or stronger questions asking the auditor/CFO to justify the large
change.
A different independent coder assessed the extent of prob-
ing of the questions asked, level of skepticism of the ACM, and
the evaluation of performance of the ACM from the perspective
of an audit committee Chairman.
6
The coder received a sepa-
rate page for each participant in random order showing a partic-
ipant number (i.e., no demographic data or experimental treat-
ment) and the questions asked by that participant directed to
the external auditor (EA), the CFO and the Head of Internal Au-
dit (IA) shown in three separate columns. The coder made four
ratings of probing – one for each of the three sets of questions
(EA, CFO and IA) and one overall assessment of the extent of
probing (see Table 4). These ratings were made on a 9-point Lik-
ert scale with end points of 1 (‘not at all probing’) and 9 (‘ex-
tremely probing’). Similar scales were used for the level of skep-
ticism shown by the ACM in their questioning and for when the
coder was asked to assume he was Chairman of the audit commit-
tee and rate the performance of the ACM on the task (‘bottom 10%
of ACMs’ to ‘top 10% of ACMs’). We refer to this coding as ‘level of
probing’.
The above coding refers to the package of questions asked by
each participant. Gri?th et al. (2015) suggest the importance of ap-
proaching evidence in a manner conducive to the critical thinking
aspect of skepticism and the importance of a targeted collection
of additional evidence (Hammersley, Johnstone, & Kadous, 2011).
This emphasis also appears bene?cial for ACMs. Therefore, ques-
tions were also coded by an author and a graduate research assis-
tant (who was blind to the research questions and experimental
treatments) on the basis of whether the question was (a) a general
request for additional evidence, or (b) was a targeted collection of
additional evidence by. The two coders achieved high reliability of
89.2 percent (Cohen et al., 2010).
7
We refer to this coding as ‘gen-
eral versus targeted evidence collection’.
Results
Manipulation checks
Manipulation checks for both independent variables indicate
that participants understood the manipulations. Speci?cally, 42 out
of 49 participants (85.6%) correctly identi?ed the information they
were provided about the audit environment in the case materials
related to an AJR.
8
Also, participants were asked to rate the in-
novativeness of the audit procedures on a 7-point scale (?3 = ‘ex-
tremely non-innovative’ and 3 = ‘extremely innovative’). The mean
response in the innovative treatment was 0.394, which was signif-
icantly greater than the mean response of ?1.715 in the standard
treatments (t
47
= 5.324; p < 0.001; one-tailed).
Hypothesis testing
Effect of an AJR given standard audit procedures (H1a & b)
H1a predicts that ACMs will feel greater accountability when
an AJR is present than when it is absent, given standard audit
procedures. Panel A of Table 2 summarizes the descriptive statis-
6
The independent coder is a present ACM, present Chairman of an audit com-
mittee and a senior member of the accounting profession (including being a former
audit partner from a Big 4 ?rm). The coder had previously participated in the study
and consequently he was familiar with the materials participants received and the
time it took to complete the study.
7
Disagreements on the application of a code to a question were resolved be-
tween the coders; where an agreement could not be reached, another author made
the ?nal decision, which was necessary for only six questions.
8
All seven participants who responded incorrectly to the AJR manipulation check
question were those in the AJR present treatments.
Table 2
Descriptive statistics of perceived accountability and overall comfort.
AJR
Absent Present
Panel A: Perceived accountability (standard error)
1
Procedures
Standard 1.90 2.68
(1.81) (0.37)
n = 17 n = 16
Innovative 2.53
(0.70)
n = 16
H1a: 2.68 versus 1.90 (t
31
= ?1.679; p = 0.052; one-tailed)
Panel B: Perceived overall comfort (standard error)
2
Procedures
Standard ?3.99 ?4.77
(2.87) (2.55)
n = 17 n = 16
Innovative ?3.17
(3.39)
n = 16
H2a: ?3.17 versus ?4.77 (t
30
= ?1.508; p = 0.071; one-tailed)
1
To elicit the participants’ accountability, we ask on a scale of ?3 (signi?cantly
not accountable) to +3 (signi?cantly accountable), “Given the audit environment
presented to you in the case material to what extent did you feel accountable to
ensure the reasonableness of the ?nancial statements?”.
2
The overall comfort scores were computed based on the factor loadings that
resulted from a factor analysis conducted on the responses provided by the partic-
ipants about their comfort level with respect to (1) the change in management’s
estimate of obsolete inventory, (2) the external auditor’s decision to allow man-
agement’s updated, smaller write-down of inventory, and (3) the difference in the
net pro?t that results from the different write-down amounts for obsolete inven-
tory. The responses were based on a scale of ?3 (completely uncomfortable) to +3
(completely comfortable). The responses to the three questions all loaded on one
factor that had an eigenvalue of 2.64.
tics regarding ACMs’ perceived accountability in ensuring the rea-
sonableness of the ?nancial statements. Consistent with our pre-
diction, given standard audit procedures, participants in the AJR
present treatment indicated that they felt signi?cantly greater
accountability (2.68 vs. 1.90) in ensuring the reasonableness of
the ?nancial statements than those in the AJR absent treatment
(t
31
= ?1.679; p = 0.052; one-tailed). Analysis of our additional
survey questions further suggest that this increase in accountabil-
ity is related to ACMs believing management’s estimates becom-
ing less conservative and auditors’ due diligence being negatively
impacted with the implementation of an AJR. Speci?cally, on a
7-point scale (?3 = ‘estimates will become extremely less conser-
vative’ and 3 = ‘estimates will become extremely more conserva-
tive’), the mean response of ?0.837 is signi?cantly less than 0
“not affected at all” (t
48
= ?8.452; p < 0.001; one-tailed). Partici-
pants were also asked to rate how their perception of the exter-
nal auditor’s due diligence would be affected on a 7-point scale
(?3 = ‘extremely negatively affected’ and 3 = ‘extremely positively
affected’). The mean response of ?0.306 is signi?cantly less than
0 ‘not affected at all’ (t
48
= ?2.399; p = 0.010; one-tailed). These
?ndings, together, suggest that ACMs feel greater accountability in
ensuring the reasonableness of the ?nancial statements when an
AJR is present due to the potential for management’s estimates
becoming more aggressive and auditors becoming less diligent in
auditing the estimates.
9
9
Analysis of the effect of an AJR given standard audit procedures on ACMs’ level
of perceived comfort (untabulated) is directionally consistent (but not statistically
signi?cant at p = 0.413; two-tailed) with our ?ndings for H1a where an AJR leads
to greater accountability. Speci?cally, implementation of an AJR appears to decrease
ACMs’ overall level of comfort from a mean of ?3.99 to ?4.77 (see Table 2, Panel
B).
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 67
Table 3
Quantity of questions developed by ACMs by experience and treatment.
Former Audit Partners Other Accountants Non-Accountants All ACMs
AJR
Absent/
Std
AJR
Present/
Std
AJR
Present/
Innov Total
AJR
Absent/
Std
AJR
Present/
Std
AJR
Present/
Innov Total
AJR
Absent/
Std
AJR
Present/
Std
AJR
Present/
Innov Total
AJR
Absent/
Std
AJR
Present/
Std
AJR
Present/
Innov
Grand
Total
n 5 5 5 15 6 6 8 20 6 5 3 14 17 16 16 49
To the EA
Total 22 29 24 75 34 25 30 89 24 14 11 49 80 68 65 213
Average 4.40 5.80 4.80 5.00 5.67 4.17 3.75 4.45 4.00 2.80 3.67 3.50 4.71 4.25 4.06 4.35
To the CFO
Total 20 26 20 66 30 23 39 92 24 20 9 53 74 69 68 211
Average 4.00 5.20 4.00 4.40 5.00 3.83 4.88 4.60 4.00 4.00 3.00 3.79 4.35 4.31 4.25 4.31
To the IA
Total 20 23 17 60 35 14 22 71 20 13 10 43 75 50 49 174
Average 4.00 4.60 3.40 4.00 5.83 2.33 2.75 3.55 3.33 2.60 3.33 3.07 4.41 3.13 3.06 3.55
All questions
Total 62 78 61 201 99 62 91 252 68 47 30 145 229 187 182 598
Average 12.40 15.60 12.20 13.40 16.50 10.33 11.38 12.60 11.33 9.40 10.00 10.36 13.47 11.69 11.38 12.20
H1b: 13.47 versus 11.69 (t
31
= 0.884; p = 0.384; two-tailed).
H2b: 11.69 versus 11.38 (t = 0.161; p = 0.873; two-tailed).
Observed background effect: Former audit partners ask more questions under an AJR (from 62 to 78) while other ACMs (other accountants and non-accountants) tend to
ask less questions when an AJR is implemented, decreasing from 167 (99 + 68) to 109 (62 + 47) (Fisher Exact Test; p = 0.002; two-tailed).
Former Audit Partners = previously held the position of an audit partner.
Other Accountants = other quali?ed accountants but not a former audit partner.
Non-Accountants = not quali?ed accountants.
All ACMs = Former Audit Partners plus Other Accountants plus Non-Accountants.
AJR Absent/Std = AJR absent/standard procedures treatment.
AJR Present/Std = AJR present/standard procedures treatment.
AJR Present/Innov = AJR present/innovative procedures treatment.
EA = External Auditor.
CFO = Chief Financial O?cer.
IA = Internal Auditor.
Total = total number of questions asked by participants in each treatment.
Average = total divided number of participants in the treatment group
.H1b predicts that the greater accountability in the presence of
an AJR, given standard audit procedures, will lead ACMs to be more
skeptical and ask more questions. Table 3 summarizes the total and
average number of questions directed to the three different par-
ties by the three experimental treatments and background.
10
We
?nd no signi?cant effect of an AJR (absent vs. present) given stan-
dard procedures on the total number of questions asked by ACMs
(13.47 vs. 11.69; t
31
= 0.884; p = 0.384; two-tailed, Table 3).
11
Fur-
ther, there was also no signi?cant effect of an AJR given standard
procedures in terms of the number of questions directed to the
external auditor (4.71 vs. 4.25; t
31
= 0.638; p = 0.528; two-tailed),
the CFO (4.35 vs. 4.31; t
31
= 0.058; p = 0.954; two-tailed) and the
internal auditor (4.41 vs. 3.13; t
31
= 1.517; p = 0.139; two-tailed).
Interestingly, we ?nd a signi?cant background effect, where for-
mer audit partners react signi?cantly differently compared to oth-
ers. Speci?cally, former audit partners become more skeptical and
ask more questions under an AJR (from 62 to 78) while other ACMs
tend to become less skeptical and ask less questions when an AJR
is implemented, decreasing from 167 (99 + 68) to 109 (62 + 47)
(Fisher Exact Test; p = 0.002; two-tailed; see Table 3).
10
We also used the independent coder’s rating of how probing the total package
of questions were to test H1b, but ?nd no signi?cant results (t
31
= 0.013; p = 0.990;
two-tailed). See Table 4 for the independent coder’s ratings.
11
Untabulated results show that ACMs with external audit experience asked sig-
ni?cantly more questions in total than those without external audit experience
(p = 0.091). Other demographic factors, such as experience as an ACM of a pub-
lic ?rm, being a CPA or CA, or having been a former audit partner did not have
signi?cant effects on the total number of questions asked (p > 0.10).
Effect of innovativeness of procedures given an AJR (H2a & b)
H2a predicts that given an AJR, ACMs will have more com-
fort in the ?nancial statements when the auditor uses innovative
rather than standard procedures. Panel B of Table 2 summarizes
the descriptive statistics regarding ACMs’ perceived overall level of
comfort.
12
Consistent with our predictions, given an AJR, we ?nd
that ACMs in the innovative treatment felt a higher level of com-
fort regarding the issues surrounding the accounting estimate than
those in the standard treatments (?3.17 vs. ?4.77; t
30
= ?1.508;
p = 0.071; one-tailed) providing marginally signi?cant support for
H2a.
13
Analysis of our additional survey questions suggest that the
higher level of comfort associated with the conduct of innovative
audit procedures rather than standard audit procedures is likely
due to ACMs believing that a move from standard to innovative
audit procedures will lead to higher audit quality. Speci?cally, par-
ticipants were asked to indicate the extent to which they believe
audit quality would improve if auditors move from standard audit
procedures to more innovative audit procedures (?3 = ‘extremely
12
The overall level of comfort was computed based on the factor loadings that
resulted from a factor analysis conducted on the responses provided by the partic-
ipants about their comfort level with respect to (1) the change in management’s
estimate of obsolete inventory, (2) the external auditor’s decision to allow manage-
ment’s updated, smaller write-down of inventory, and (3) the difference in the net
pro?t that results from the different write-down amounts for obsolete inventory.
The responses to the three questions all loaded on one factor that had an eigen-
value of 2.64. The factor loadings were 0.911, 0.960, and 0.943, respectfully. Analysis
results based on the three separate measures are consistent with the analysis based
on the overall level of comfort.
13
Analysis of the effect of a move towards more innovative audit procedures given
an AJR on ACMs’ perceived accountability (untabulated) is not statistically signi?-
cant (2.68–2.53; p = 0.457; two-tailed) (see Table 2, Panel A).
68 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
low’ and 3 = ‘extremely high’). The mean response of 1.078 was
signi?cantly higher than 0 ‘neutral’ (t
48
= 7.584; p < 0.001; one-
tailed).
H2b tests whether given an AJR, the greater comfort level that
follows from the external auditors’ use of innovative procedures
rather than standard procedures will lead ACMs to question the ac-
counting estimate to a lesser/greater extent when the auditor uses
innovative procedures. We ?nd no signi?cant effect of the inno-
vativeness of the procedures, given an AJR, on the total number of
questions asked by the ACMs (standard: 11.69 vs. innovative: 11.38;
t
30
= 0.161; p = 0.873; two-tailed).
14
Hence, we do not reject the
null hypothesis for H2b. Such results are likely due to the fact that
although ACMs feel more comfort when the audit partner adopts
more innovative audit procedures, their unfamiliarity with the pro-
cedures lead them to be at least as questioning of the estimate as
when standard audit procedures are used.
15
Analysis of the demographic factors, however, again suggest a
potential ACM background effect. Speci?cally, for the two treat-
ments with AJR present, former audit partners as a whole decrease
the number of questions they ask from 78 to 61 when innova-
tive audit procedures are used rather than standard procedures
(see Table 3). Those without an audit partner background, how-
ever, show a slight increase in their questioning behavior from 109
(62 + 47) to 121 (91 + 30) (see Table 3) when innovative proce-
dures are used. The difference between the two groups is signif-
icant at p = 0.108 (Fisher Exact Test; two-tailed). The directional
difference in our results suggests that while the use of innovative
procedures increase ACMs’ overall perception of comfort regarding
the accounting estimate, former audit partners who are more fa-
miliar with the innovative procedures may exhibit a lower level of
skeptical behavior when innovative procedures are used, and the
unfamiliarity with innovative procedures lead those without an au-
dit partner background to keep asking probing questions. We sug-
gest the need for future research addressing this potential expla-
nation.
The overall pattern of results for H1 and H2 can be seen in
Fig. 1. From Fig. 1, Panel A it can be seen that adoption of an AJR
would increase ACM accountability and this would occur whether
the change to innovative procedures also occurred. We conjecture
that if the move increases ACM comfort, as found in H2a, this
would attenuate, or work against, an increase in accountability
from the introduction of an AJR. Overall, the combined results sug-
gest that the increase in ACM accountability from a move to an AJR
is likely to be robust to a change of the type of audit procedures
from standard to innovative. In addition, Fig. 1, Panel B suggests
that if an AJR is implemented, then innovative audit procedures
are needed if ACM comfort is considered an important priority.
Supplementary analysis
Does ACM background affect the level of probing and the types of
questions asked (RQ1)
In this section we divide participants into three background
groups: former audit partners, other quali?ed accountants and
14
Consistent with the results of H2b, the effect of the move towards more in-
novative procedures given AJR did not have a signi?cant effect on the participants’
response to the question that asked on a scale of ?3 (signi?cantly less than av-
erage amount of questioning would be warranted) to +3 (signi?cantly more than
average amount of questioning would be warranted), to what extent they would
like to question auditors and/or management about the appropriateness/su?ciency
of the write-down (Standard: 2.406 vs. Innovative: 2.244; t
30
= ?0.68; p = 0.501;
two-tailed).
15
We also test H2b using the independent coder’s rating of how probing the total
package of questions were but ?nd no signi?cant results (t
30
= ?0.222; p = 0.825;
two-tailed) (see Table 4, Panel B).
non-accountants. We then consider the differences in the level
of probing of the questions of the different background groups
(Table 4, Panel A) and the types of questions they ask (Table 5).
We outline below that while the level of probing did not vary with
ACM background, the quantity of questions asked and the type of
questions asked did vary with ACM background.
Table 4, Panel A summarizes the mean scores related to
the level of probing of the independent coder for each evalu-
ation category split across the three background groups (on a
range of 9-point scales, where higher numbers indicate higher
performance). Analysis of the scores in Table 4 Panel A show
ACMs asked probing questions of the EA (mean = 6.00), CFO
(mean = 6.08), IA (mean = 5.51), and in their total package of ques-
tions (mean = 5.92). Mean scores for all categories were signi?-
cantly above the midpoint of the scale (p a7d? 0.046; two-tailed).
Their overall level of skepticism (mean = 7.00) and performance
level (mean = 6.06) were also signi?cantly higher than the mid-
point of the scale (p a7d? 0.002; two-tailed).
16
Next we consider the types of questions asked by ACMs and the
effects of demographic background. Table 5 categorizes all ques-
tions asked by ACMs of the EAs, CFOs and IAs split by demographic
background. We incorporate columns for former audit partners
(n = 15), other accountants (n = 20) and non-accountants (n = 14).
The table classi?es questions based on our coding of questions
about reporting fairly, evidence collection and gaining comfort.
Our analysis discusses average questions asked per participant
and provides evidence that ACM background affects the type and
quantity of questions ACMs ask in their oversight process, as seen
from the following key insights from Table 5. First, all three ACM
background groups asked the most questions to EAs, next most to
CFOs and the least to IAs (former audit partners 6.5, 5.5 and 5.1
respectively; other accountants 6.0, 5.4 and 4.5 respectively; non-
accountants 4.9, 4.4 and 3.6 respectively – see average questions
per participant in Table 5, row 4C). This trend holds for questions
related to evidence collection (see Table 5, row 2B where the to-
tal for Column A of 2.5 is greater than the totals for Columns B
and C, of 1.9 and 1.6 respectively) and questions related to gain-
ing comfort (see Table 5, row 3B where the total of Column A
of 2.0 is greater than the totals of Columns B and C, 1.3 and 1.1
respectively). However, for the reports fairly category, the average
questions per participant to EAs is less (1.1, see Table 5, Column A,
row 1B) than for either CFOs or IAs (1.6 and 1.3 respectively as per
Table 5, Columns B and C, row 1B). As the evidence collection and
comfort questions
17
mainly relate to processes, while the reports
fairly questions mainly relate to outcomes, this suggests that ACMs
generally preferred to rely on EAs to obtain information about pro-
cesses that underlie the reported accounting estimate. On the other
hand, the questions directed to CFOs and IAs (see Table 5, row 1A)
by ACMs in general, and former audit partners in particular, were
more focused on outcomes. These questions, particularly those un-
der rows 1.1 and 1.2 of Table 5, related to the accuracy of the esti-
mates developed by management.
Second, the main differences between the types of questions
asked were between former audit partners (n = 15) and non-
accountants (n = 14) with former audit partners asking more than
double the number of questions on average per participant in the
‘reports fairly’ category to EAs (1.5 vs. 0.6; Column A, row 1B),
16
As noted in the testing of H2a and H2b, analysis of the scores of how prob-
ing the total package of questions were in Table 4 Panel B shows no signi?cant
difference based on the introduction of an AJR or the type of procedures used by
the auditor. The largest difference in the mean level of probing of questions asked
to external auditors by former audit partners (mean = 6.53) compared to all other
ACMs (means of 5.75 and 5.79) is signi?cant at p = 0.127, two tailed.
17
While many gaining comfort questions related to the processes used to gain
comfort, some questions speci?cally asked why the revised estimate was preferable
and thus related directly to an outcome.
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 69
Fig. 1. Observed results of perceived accountability and overall comfort.
1
To elicit the participants’ accountability, we ask on a scale of ?3 (signi?cantly not accountable) to +3 (signi?cantly accountable), “Given the audit environment presented
to you in the case material to what extent did you feel accountable to ensure the reasonableness of the ?nancial statements?”.
2
The overall comfort scores were computed based on the factor loadings that resulted from a factor analysis conducted on the responses provided by the participants about
their comfort level with respect to (1) the change in management’s estimate of obsolete inventory, (2) the external auditor’s decision to allow management’s updated, smaller
write-down of inventory, and (3) the difference in the net pro?t that results from the different write-down amounts for obsolete inventory. The responses were based on a
scale of ?3 (completely uncomfortable) to +3 (completely comfortable). The responses to the three questions all loaded on one factor that had an eigenvalue of 2.64.
to CFOs (2.5 vs. 1.1), and to IAs (1.7 vs. 0.6) (Table 5, Columns B
and C, row 1B). The difference between the former audit partners
and non-accountants was particularly evident in rows 1.1 and 1.2,
which were questions related to ?nancial statement accuracy.
Third, the emphasis on questions on evidence collection (rows
2.1–2.4 in Table 5) varies with background group. For questions to
EAs (see Table 5, Column A, row 2C), over 40% of the questions
were in the evidence collection category for all three ACM back-
ground groups (45.4%, 41.2% and 40.6% respectively, Table 5, Col-
umn A, row 2C). However, this pattern changes for questions asked
to CFOs (Table 5, Column B, row 2C); of the three ACM background
groups, former audit partners asked the least number of questions
to CFOs in the evidence collection category (20.7% of their ques-
tions) compared to other accountants (51.9% of their questions) and
non-accountants (36.1% of their questions). These differences are
particularly evident in category 2.4 (business impact and manage-
ment’s turnaround strategy).
Differences in the extent and type of questions asked by former audit
partner ACMs and other ACMs split by treatments (RQ2)
The above analysis considers differences in the three ACM back-
ground groups for the three experimental treatments combined.
In Table 6 we provide additional qualitative analysis to build on
our hypothesis testing. Given the small cell sizes when we split
our 49 participants across three experimental treatments and three
background groups from Table 5, in Table 6 we combine other ac-
countant ACMs and non-accountant ACMs into one group. Below
we consider differences in the questioning behavior of former au-
dit partners (n = 15) and other ACMs (n = 34) depending on the
70 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
Table 4
Mean evaluation scores of the extent to which questions are probing.
Mean scores
n Questions to EA Questions to CFO Questions to IA Question package Skepticism Performance
Panel A: By participant background
Former Audit Partners 15 6.53 6.07 5.93 6.13 7.07 6.33
Other Accountants 20 5.75 6.15 5.50 5.95 7.00 6.05
Non-Accountants 14 5.79 6.00 5.21 5.64 6.93 5.79
Overall 49 6.00 6.08 5.51 5.92 7.00 6.06
Panel B: By experimental treatment
AJR Absent/Standard Procedures Treatment
Former Audit Partners 5 6.40 5.60 6.00 5.60 7.00 6.00
Other Accountants 6 6.17 6.33 6.50 6.33 7.33 6.17
Non-Accountants 6 6.17 5.50 4.67 5.67 6.83 5.50
Total 17 6.24 5.82 5.71 5.88 7.06 5.88
AJR Present/Standard Procedures Treatment
Former Audit Partners 5 7.00 6.80 6.00 7.00 7.40 7.00
Other Accountants 6 5.50 5.67 5.00 5.50 6.83 5.67
Non-Accountants 5 5.20 6.20 5.00 5.20 6.80 5.80
Total 16 5.88 6.19 5.31 5.88 7.00 6.13
AJR Present/Innovative Procedures Treatment
Former Audit Partners 5 6.20 5.80 5.80 5.80 6.80 6.00
Other Accountants 8 5.63 6.38 5.13 6.00 6.88 6.25
Non-Accountants 3 6.00 6.67 6.67 6.33 7.33 6.33
Total 16 5.88 6.25 5.63 6.00 6.94 6.19
All ratings in this table are done by an independent coder with ACM experience who was blind to demographic details using 9-point Likert scales.
Questions to EA = questions directed to the External Auditor.
Questions to CFO = questions directed to the CFO.
Questions to IA = questions directed to the Head of Internal Audit.
Question Package = all questions to EAs, CFOs and IAs.
Skepticism= level of skepticism ACMs showed in their questioning behavior.
Performance = the performance of the ACM on this task from the perspective of the Chair of the Audit Committee.
Former Audit Partners = previously held the position of an audit partner.
Other Accountants = other quali?ed accountants but not a former audit partner.
Non-Accountants = not quali?ed accountants.
presence/absence of an AJR and the use of innovative versus stan-
dard procedures. Overall, these additional analyses show signi?cant
differences between former audit partners and other ACMs in the
number of questions asked, the types of questions asked, the ex-
tent to which the questions show higher levels of skepticism and
how probing the questions are. These differences are outlined be-
low and suggest the need for future research to further examine
the underlying differences in questioning behavior of ACMs of dif-
ferent backgrounds.
First, we examine some speci?c differences between former au-
dit partners and other ACMs across the AJR absent and the AJR
present treatments (for both these treatments standard procedures
are used).
18
The former audit partners in the AJR present treatment
ask more questions compared to those in the AJR absent treatment
while the opposite pattern holds for the other ACMs. The direction
of these results holds for questions asked to EAs (Table 6, Panel A),
CFOs (Table 6, Panel B) and IAs (Table 6, Panel C). The former au-
dit partners ask EAs 8.8 questions per participant on average in
the AJR present treatment versus 5.0 per participant in the AJR
absent treatment while other ACMs ask 6.3 questions per partic-
ipant in the AJR absent treatment versus 4.9 questions per partici-
pant in the AJR present treatment (see Table 6, row 5). This differ-
ence between the two groups is statistically signi?cant at p = 0.004
(Fisher Exact Test; two-tailed) showing that former audit partners
in the AJR present treatment ask more questions than those in the
AJR absent treatment but for other ACMs the reverse holds. The
18
Table 6 provides both total numbers of questions asked and average number of
questions asked. As the results do not change depending on this choice (i.e., con-
stant cell sizes for former audit partners and similar cell sizes for other ACMs), for
clarity we report per participant numbers (i.e. mean questions asked).
same pattern holds for questions directed to IAs (Table 6, Panel
C). Former audit partners in the AJR present treatment ask more
questions to IAs (6.2 per participant) than those in the AJR absent
treatment (4.8 per participant) while other ACMs in the AJR ab-
sent treatment ask more questions (5.2 per participant) than those
in the AJR present treatment (3.1 per participant; see Table 6, row
15). These differences are signi?cant (Fisher Exact Test; p = 0.017;
two-tailed). We note that the above results for former audit part-
ners for both questions to EAs and IAs are consistent with H1b,
whereas the results for other ACMs are in the opposite direction to
the predictions of H1b. However, caution is required in interpret-
ing these results given the small number of former audit partners
in each treatment group. The direction of these ?ndings is con-
sistent across the three types of questions. We illustrate this with
data in Panel A, which shows that for the former audit partners,
the average number of questions per participant is higher for the
AJR present treatment participants than the AJR absent treatment
participants across all three types of questions: reports fairly (1.8
vs. 1.2), evidence collection (4.4 vs. 2.0) and gaining comfort (2.2
vs. 1.8) (Table 6, Panel A, Column A, rows 1–3). We ?nd the oppo-
site for the three types of questions for the other ACMs with the
AJR absent treatment being larger than the AJR present treatment
in each case (1.3 vs. 0.4, 2.5 vs. 2.2 and 2.2 vs. 2.0) (Table 6, Panel
A, Column B, rows 1–3).
Second, there are both similarities and differences in the types
of questions directed to CFOs compared to those directed to
EAs and IAs discussed above. The questions directed to the CFO
(Table 6, row 10) follow a similar pattern to questions to EAs and
IAs with the former audit partners in the AJR present treatment
asking more questions (6.4 per participant) compared to those in
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 71
Table 5
Types of questions directed to the External Auditor, CFO and Head of Internal Audit.
Column A: Questions
Directed to EA
Column B: Questions
Directed to CFO
Column C: Questions
Directed to IA
Former
Audit
Partners
Other
Accts
Non-
Accts Total
Former
Audit
Partners
Other
Accts
Non-
Accts Total
Former
Audit
Partners
Other
Accts
Non-
Accts Total
Rown 15 20 14 49 15 20 14 49 15 20 14 49
Reports Fairly
1.1 Conservatism, integrity, bias and pressure exerted by
management
8 8 3 19 11 5 4 20 8 7 2 17
1.2 Loan covenants, incentives, pressure to perform, effect on
remuneration and shareholders
5 7 1 13 17 9 5 31 2 7 1 10
1.3 Management decision to use non-sales staff in the
assessment
5 0 3 8 4 3 3 10 5 6 2 13
1.4 The control environment for inventory; surprises 4 6 2 12 5 8 4 17 10 9 3 22
1A Total Questions: Reports Fairly 22 21 9 52 37 25 16 78 25 29 8 62
1B Reports Fairly Questions per Participant 1.5 1.1 0.6 1.1 2.5 1.3 1.1 1.6 1.7 1.5 0.6 1.3
1C % of Grand Total Questions 22.7 17.6 13.0 18.2 45.1 23.1 26.2 31.1 32.5 32.6 16.0 28.7
Evidence Collection
2.1 Use of external evidence 7 10 6 23 1 9 2 12 2 5 0 7
2.2 What testing was done (general question) 1 6 1 8 1 1 1 3 4 3 2 9
2.3 Asking for further details of audit testing done, reference
to further tests that could be done, and the extent of
reliance on the internal auditor
31 26 20 77 5 12 6 23 14 13 15 42
2.4 Business impact including management’s turnaround
strategy
5 7 1 13 10 34 13 57 8 6 7 21
2A Total Questions: Evidence Collection 44 49 28 121 17 56 22 95 28 27 24 79
2B Evidence Collection Questions per Participant 2.9 2.5 2.0 2.5 1.1 2.8 1.6 1.9 1.9 1.4 1.7 1.6
2C % of Grand Total Questions 45.4 41.2 40.6 42.5 20.7 51.9 36.1 37.8 36.4 30.3 48.0 36.6
Gaining Comfort
3.1 Terminology related to comfort/comfortable/con?dent 3 9 3 15 4 5 7 16 5 9 2 16
3.2 Question as to whether/how the auditor/CFO was satis?ed 9 8 6 23 4 2 0 6 3 5 1 9
3.3 Questions asking for justi?cation by the auditor/CFO
including why is the revised estimate preferable
8 13 9 30 15 12 10 37 9 10 8 27
3.4 Whether the accounting treatment and audit procedures
met accounting/audit standards, were consistent
industry/other clients and internally reviewed
9 11 11 31 1 1 2 4 0 1 2 3
3A Total Questions: Gaining Comfort 29 41 29 99 24 20 19 63 17 25 13 55
3B Gaining Comfort Questions per Participant 1.9 2.1 2.1 2.0 1.6 1.0 1.4 1.3 1.1 1.3 0.9 1.1
3C % of Grand Total Questions 29.9 34.5 42.0 34.7 29.3 18.5 31.1 25.1 22.1 28.1 26.0 25.5
4A Other Questions 2 8 3 13 4 7 4 15 7 8 5 20
4B Grand Total of Questions 97 119 69 285 82 108 61 251 77 89 50 216
4C Average Questions per Participant 6.5 6.0 4.9 5.8 5.5 5.4 4.4 5.1 5.1 4.5 3.6 4.4
The total number of classi?cations for EAs of 285 is greater than the total number of questions asked (213 questions) shown in Table 3 as some questions ?t under more
than one category. Similarly, classi?cations of 251 for CFOs and 216 for IAs is greater than the number of questions asked (211 and 174 questions respectively) shown in
Table 3.
Reports Fairly = questions related to whether the reported earnings is fairly presented and not overstated.
Evidence Collection = questions related to the nature and extent of evidence collected by the auditors or management.
Gaining Comfort = questions related to the ACM gaining comfort about the ?nancial statements.
Questions Directed to EA = questions asked by ACMs to the external auditor.
Questions Directed to CFO = questions asked by ACMs to the CFO.
Questions Directed to IA = questions asked by ACMs to the Head of Internal Audit.
Former Audit Partners = previously held the position of an audit partner.
Other Accountants = other quali?ed accountants but not a former audit partner.
Non-Accountants = not quali?ed accountants.
the AJR absent treatment (4.8 per participant). In comparison, the
other ACMs asked less questions in the AJR present treatment (4.6
per participant) compared to the AJR absent treatment (5.0 per
participant). While these results are in the same direction as the
results for questions to EAs and IAs, they are not statistically signif-
icant (Fisher Exact Test; p = 0.192; two-tailed). However, one sig-
ni?cant difference to the above trends relates to questions about
evidence collection asked to CFOs compared to EAs (Table 6, rows
2 and 7). While former audit partners on average asked 2.0 and
4.4 evidence collection questions to EAs (for AJR absent and AJR
present respectively) (Table 6, Column A, row 2), they asked con-
siderably fewer evidence collection questions to CFOs across these
treatments (1.0 and 1.2 questions) (Table 6, Column A, row 7). Also,
while the above trends show that other ACMs ask more ques-
tions to CFOs overall in the AJR absent treatment versus the AJR
present treatment (5.0 vs. 4.6, row 10), this was not the case
for the evidence collection questions to CFOs (AJR absent = 1.8;
AJR present = 2.3; Table 6, Column B, row 7). In fact, this is the
only case where other ACMs asked less questions in the AJR ab-
sent treatment compared to AJR present treatment. It appears that
while former audit partners did not see evidence collection ques-
tions to CFOs to be very important (i.e., they relied on EAs and
IAs), this was not the case for other ACMs who increased evidence
collection questions to CFOs when an AJR was present.
Third, comparison is also be made between the innovative au-
dit procedures treatment and the standard procedures treatment
(for both these treatments AJR is present). The trend for former
audit partners was to ask less questions in the innovative proce-
dures treatment compared to the standard procedures treatment to
72 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
Table 6
Total questions (average per participant) from all ACMs split by treatments.
Column A: Former Audit Partner ACMs Column B: Other ACMs
AJR Absent/
Standard
AJR Present/
Standard
AJR Present/
Innovative Total
AJR Absent/
Standard
AJR Present/
Standard
AJR Present/
Innovative Total
Grand
Total
Row n 5 5 5 15 12 11 11 34 49
Panel A: Questions directed to the External Auditor (EA)
1 Reports Fairly: Total Questions
(Average)
6(1.2) 9(1.8) 7(1.4) 22(1.5) 15(1.3) 4(0.4) 11(1.0) 30(0.9) 52(1.1)
2 Evidence Collection: Total Questions
(Average)
10(2.0) 22(4.4) 12(2.4) 44(2.9) 30(2.5) 24(2.2) 23(2.1) 77(2.3) 121(2.5)
3 Gaining Comfort: Total Questions
(Average)
9(1.8) 11(2.2) 9(1.8) 29(1.9) 26(2.2) 22(2.0) 22(2.0) 70(2.1) 99(2.0)
4 Other 0 2 0 2 4 4 3 11 13
5 Grand Total per Treatment 25(5.0) 44(8.8) 28(5.6) 97(6.5) 75(6.3) 54(4.9) 59(5.4)
188(5.5)
285(5.8)
Panel B: Questions directed to the CFO
6 Reports Fairly: Total Questions
(Average)
12(2.4) 15(3.0) 10(2.0) 37(2.5) 18(1.5) 12(1.1) 11(1.0) 41(1.2) 78(1.6)
7 Evidence Collection: Total Questions
(Average)
5(1.0) 6(1.2) 6(1.2) 17(1.1) 21(1.8) 25(2.3) 32(2.9) 78(2.3) 95(1.9)
8 Gaining Comfort: Total Questions
(Average)
6(1.2) 9(1.8) 9(1.8) 24(1.6) 16(1.3) 10(0.9) 13(1.2) 39(1.1) 63(1.1)
9 Other 1 2 1 4 5 4 2 11 15
10 Grand Total per Treatment 24(4.8) 32(6.4) 26(5.2) 82(5.5) 60(5.0) 51(4.6) 58(5.3)
169(5.0)
251(5.1)
Panel C: Questions directed to the Internal Auditor (IA)
11 Reports Fairly: Total Questions
(Average)
8(1.6) 11(2.2) 6(1.2) 25(1.7) 15(1.3) 9(0.8) 13(1.2) 37(1.1) 62(1.3)
12 Evidence Collection: Total Questions
(Average)
9(1.8) 11(2.2) 8(1.6) 28(1.9) 26(2.2) 12(1.1) 13(1.2) 51(1.5) 79(1.6)
13 Gaining Comfort: Total Questions
(Average)
5(1.0) 7(1.4) 5(1.0) 17(1.1) 17(1.4) 7(0.6) 14(1.3) 38(1.1) 55(1.1)
14 Other 2 2 3 7 4 6 3 13 20
15 Grand Total per Treatment 24(4.8) 31(6.2) 22(4.4) 77(5.1) 62(5.2) 34(3.1) 43(3.9) 139(4.1) 216(4.4)
Former Audit Partner ACMs = previously held the position of an audit partner.
Other ACMs = other quali?ed accountants who are not former audit partners (20) and not quali?ed accountants (14).
AJR Absent/Std = AJR absent treatment/standard procedures.
AJR Present/Std = AJR present treatment/standard procedures.
AJR Present/Innov = AJR present treatment/innovative procedures.
Reports fairly = questions related to whether the reported earnings is fairly presented and not overstated.
Evidence collection = questions related to the nature and extent of evidence collected by the auditors or management.
Gaining comfort = questions related to the ACM gaining comfort about the ?nancial statements.
EAs (5.6 vs. 8.8, Table 6, Panel A, row 5), CFOs (5.2 vs. 6.4; Table 6,
Panel B, row 10) and IAs (4.4 vs. 6.2, Table 6, Panel C, row 15), sug-
gesting former audit partners are more comfortable with the use
of innovative procedures. The comparisons for other ACMs show
the opposite trend where the mean number of questions asked
per participant was greater in the innovative procedures treatment
than the standard procedures treatment (5.4 vs. 4.9, 5.3 vs. 4.6, 3.9
vs. 3.1 for questions to EAs, CFOs and IAs respectively; Table 6, Col-
umn B, rows 5, 10 and 15 respectively). These differences between
former audit partner ACMs and other ACMs would have the oppo-
site effect on the results for H2b with former audit partners ask-
ing less questions when innovative audit procedures are used, and
other ACMs asking more questions when innovative procedures are
used. One explanation for these effects is that the other ACMs are
likely to be less familiar with innovative procedures and, therefore,
ask more questions.
Fourth, all of the above analyses have related to the number of
questions asked to EAs, CFOs and IAs. Here we return to Table 4,
Panel B where the independent coder’s evaluation of the level of
probing is categorized by both participant background and experi-
mental treatment. Across all columns in Table 4, Panel B the scores
in row AJR present/standard procedures for the former audit part-
ners (7.00, 6.80, 6.00, 7.00, 7.40 and 7.00) were higher than the
equivalent scores for the other two treatments. Speci?cally, we
observe the scores for former audit partners were higher in the
AJR present/standard procedures while the opposite pattern held
for ‘other accountant’ ACMs (5.50, 5.67, 5.00, 5.50, 6.83 and 5.67).
We also note that when comparing the two treatments with AJR
present, the scores for former audit partners decreased with a
move towards more innovative procedures, while we observe the
opposite trend for other ACMs. These results are consistent with
our earlier ?ndings of higher skepticism by former audit partners
in the AJR present/standard procedures treatment.
Fifth, as described in the research methods, we categorized
questions on the basis of general versus targeted evidence collec-
tion, with the latter being more conducive to higher levels of skep-
ticism (Gri?th et al., 2015). Comparing the two treatments with
standard procedures we found that the former audit partners as a
group showed an increase for the need for more targeted evidence
given an AJR (AJR absent: 4.40 vs. AJR present: 5.80) while other
ACMs showed a decrease (4.83–3.55) in the need for more targeted
evidence (p = 0.059; Fisher Exact Test; two-tailed; untabulated).
The impact of a change to an AJR and more innovative audit
procedures on ACMs’ perceptions (RQ3)?
Here we consider the responses of our participants to a number
of post-experiment questions. Our results show that ACMs believe
that an AJR will lead management estimates to become somewhat
less conservative. The mean response of ?0.837 is signi?cantly less
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 73
than 0 ‘not affected at all’ (p < 0.001; two-tailed).
19
This provides
a potential explanation for our results for H1a where we ?nd the
introduction of an AJR increases ACMs’ perceived accountability to-
wards ensuring the reasonableness of the ?nancial statements. We
suggest that the belief that management estimates will become
somewhat less conservative with an AJR is likely to increase ACMs’
perceived accountability about ensuring the reasonableness of the
accounting estimate.
There have been suggestions that external auditor due diligence
could be adversely affected by an AJR (Pozen, 2008). Consistent
with such argument, we ?nd that ACMs who participated in our
study believe an AJR will lead auditor due diligence to be some-
what negatively affected. The mean response of ?0.306 is signi?-
cantly lower than 0 ‘not affected at all’ (p = 0.020; two-tailed). The
different demographic factors do not have a signi?cant effect on
the results (p > 0.100).
We also ?nd that ACMs believe that a change to more innova-
tive procedures by auditors would increase audit quality. The mean
response, 1.078, is signi?cantly greater than 0 ‘neutral’ (p < 0.001;
two-tailed). This provides a potential explanation for our results
for H2a where we ?nd auditors’ use of more innovative proce-
dures under an AJR framework increases ACMs’ perceived comfort
regarding the accounting estimate. The different demographic fac-
tors do not have a signi?cant effect on how ACMs perceive this is-
sue (p a7e? 0.100; two-tailed). Also, ACMs believe that the change
to more innovative procedures will not have a signi?cant effect on
inspectors. The mean response of -0.071 is not signi?cantly differ-
ent from 0 ‘neutral’ (p = 0.688; two-tailed).
Discussion and conclusions
Changes to the audit environment have led to suggestions for
the need to change the regulatory framework for auditors includ-
ing the introduction of an AJR and relatedly the use of more in-
novative audit procedures (Peecher et al., 2013). However, such
changes are not without their critics with some stakeholders sug-
gesting that they could negatively affect the judgments of man-
agement and auditors as well as the oversight process of the au-
dit committee, resulting in decreased ?nancial statement quality
(Treasury, 2008). Here we focus on the effect of an AJR on the
oversight process of ACMs given the increased role audit commit-
tees play in monitoring the ?nancial reporting and audit process,
especially those that relate to complex accounting estimates (CAQ,
2014; FRC, 2013a).
Our key ?ndings show that introducing an AJR increases ACMs’
perceived accountability in ensuring the reasonableness of the ?-
nancial statements. Further analysis of our survey questions shows
that this result is related to ACMs believing accounting estimates
are likely to become less conservative and auditors’ due diligence
to be negatively impacted with the introduction of an AJR. To gain
a greater understanding of the impact of an AJR we consider the
effects of its introduction both with standard and innovative au-
dit procedures. The latter is important because an AJR is likely to
encourage the adoption of innovative procedures. We ?nd that a
move towards more innovative procedures under an AJR frame-
work increases ACMs’ perceived overall comfort regarding the ac-
counting issue at hand. Results from our survey questions suggest
19
This analysis is conducted for all participants in the three conditions together
as the questions speci?cally ask the participants to assume an AJR rule exists
or a move towards innovative procedures. Thus, we did not expect any differ-
ence between treatments. For all questions examined under RQ3, t-tests were
conducted between our three treatments and no signi?cant results were found
(p a7e? 0.165; two-tailed). We also obtain participants’ perceptions of the internal
audit quality and ?nd no signi?cant difference between our experimental condi-
tions (p a7e? 0.269; two-tailed).
that this result stems from ACMs believing innovative procedures
lead to higher audit quality than standard procedures. However,
despite the greater accountability and comfort associated with the
introduction of an AJR and a move towards more innovative proce-
dures, our ?ndings show that this does not necessarily lead ACMs
to be more skeptical and ask more probing questions when an AJR
is introduced or ask less probing questions when innovative proce-
dures are used by the auditor.
We note that while an AJR did not lead to more questions be-
ing asked than without an AJR, it also did not lead ACMs to ask less
questions either. Despite the concern that introducing some form
of an AJR will lead to a potential decrease in ?nancial statement
quality due to less conservative ?nancial statement preparation or
lower audit quality due to lower auditor diligence or less audit
committee scrutiny, we did not ?nd any evidence of the latter. In
fact, our supplementary analysis, using an independent coder, on
the level of probing of ACM questions shows ACMs under all treat-
ments exercised a high level of skepticism and asked very probing
questions. We ?nd no evidence that the extent to which ACMs are
probing is adversely affected by the introduction of an AJR either
with standard audit procedures or innovative audit procedures.
Hence, despite the concern for reduced ?nancial reporting quality
with the introduction of an AJR (Pozen, 2008; Treasury, 2008), our
results suggest ACMs will continue to ask probing questions with a
change to an AJR. Moreover, given the stated advantages of an AJR
combined with innovative audit procedures (Peecher et al., 2013),
we suggest further consideration to be given to such changes un-
less future research shows negative consequences on other parts of
the ?nancial reporting chain such as preparers or auditors.
The large number of questions asked and the high level of prob-
ing by ACMs noted above is consistent with the results of inter-
views by Cohen et al. (2010) that in the post-SOX period, ACMs are
asking probing and di?cult questions. This is in contrast to earlier
literature (e.g., Cohen, Krishnamoorthy, & Wright, 2002; Turley &
Zaman, 2007) suggesting ACMs were not effective questioners. Re-
cent professional literature suggests the importance of audit com-
mittees seeking to understand whether or not auditors have exer-
cised an appropriate degree of skepticism on issues where there
is a disagreement between auditors and management as well as
issues where the two parties have agreed (FRC, 2012). The FRC
(2012) further suggests that ACMs should require an explanation
of the auditor’s rationale for particular conclusions, the alternative
that was considered and the reasons for speci?c judgments be-
ing considered the most appropriate. The participants in our study
asked numerous such questions as illustrated in Table 5. These
questions related to such factors as the accuracy of the ?nancial
statements, the processes used by the external auditors, CFOs and
internal auditors in collecting evidence and how each of these
groups gained comfort in the accounting numbers.
Our paper also provides insights on the likely effects of an AJR
on managers and auditors from the viewpoint of ACMs. Looking at
the views of these ACMs is important as they work closely with
both managers and auditors, and will have informed opinions of
the likely effects of an AJR. While there was no evidence of an AJR
negatively affecting the performance of ACMs, ACMs in general be-
lieve that an AJR may lead to management estimates that are less
conservative and that the introduction of an AJR may lead to less
auditor diligence. If these perceptions of ACMs are accurate, this
could potentially lead to overall lower ?nancial statement qual-
ity. However, in a multi-period setting, as management and audi-
tors become aware that ACMs are likely to continue to ask probing
questions, there are incentives for auditors to further improve au-
dit quality with consequent incentives for management to become
more conservative in their estimates.
Supplementary analysis of the questions developed by our ACM
participants show that ACM background (particularly, background
74 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
as a former audit partner) leads to signi?cant differences in how
ACMs are affected by our speci?c manipulations of an AJR and the
nature of audit procedures. In particular, former audit partners be-
came more skeptical and asked more questions when an AJR is
present versus absent, while other ACMs decrease their question-
ing behavior. These results were consistent across the three types
of question categories (reports fairly, evidence collection and gain-
ing comfort). We suggest that a likely explanation for this effect
is that former audit partners better understand the relationship
between auditors and inspectors, and the reasons for suggestions
of second-guessing in the inspection process. Other differences be-
tween former audit partners and other ACMs is that the former
audit partners place a greater reliance on asking ‘evidence collec-
tion’ questions to EAs rather than CFOs, they appear more comfort-
able with innovative audit procedures, they looked for more tar-
geted evidence and their questions were most probing when they
were part of the AJR present/standard audit procedures treatment.
The differences in questioning behavior between ACMs of differ-
ent backgrounds have the potential for cognitive stimulation (Chen,
Trotman, & Zhou, 2015) when the audit committee works as an
interacting group. Future research should examine how and when
these bene?ts will occur. The above ?ndings also support the em-
phasis placed on the importance of ?nancial expertise (e.g., Glover
& Prawitt, 2013; IAASB, 2014) and on diversity among members of
the audit committee (Beasley et al., 2009; Cohen et al., 2010; Ernst,
2013).
Sometimes there are unintended consequences of new controls
and regulations (see Bell et al., 2005) and as a result the intended
consequences are not realized. For example, one potential bene?t
of an AJR is that it is likely to encourage auditors to use less de-
fensive/more innovative audit procedures. A negative consequence
would be if this leads ACMs to ask less probing questions when
more innovative procedures were used. Our results suggest that
this is not likely to be a problem as we ?nd no differences in the
number of questions asked or the level of probing of those ques-
tions when innovative rather than standard procedures were used.
If an AJR is implemented, this may imply that auditors should con-
sider engaging in more innovative procedures as our study ?nds
that these procedures provide ACMs with greater comfort. Also,
as ACMs prefer innovative procedures, and see them as providing
higher audit quality, regulators could consider the potential for en-
couragement and greater acceptance of such procedures.
The introduction of an AJR together with innovative procedures
presents a cost-bene?t tradeoff for ACMs. Previous research has
suggested that an AJR will result in greater use of innovative proce-
dures (Peecher et al., 2013), and if this occurs, there are bene?ts for
ACMs as we ?nd that they are more comfortable with innovative
procedures and see them as providing higher audit quality. How-
ever, for ACMs there is also a cost as they need to be concerned
that the introduction of an AJR may result in potentially greater
reporting aggressiveness by management and potentially less dili-
gence by auditors.
A number of limitations to our study should be noted. First,
we examined the individual judgments of ACMs. However, audit
committees are teams usually consisting of three or more mem-
bers and there can be both process gains and losses from group
interaction. Future research could begin to examine process gains
and losses in audit committees by considering the outputs of both
nominal and interacting teams. Second, we selected a case where
the preparer had decreased the amount of an inventory obsoles-
cence account from their original estimate, and the auditors had
accepted the reduced estimate of the write-down. Using a case
where the auditor had disagreed with the client may affect the
questioning behavior of ACMs and should be considered in future
research. Third, our manipulation of standard versus innovative
procedures was based on previous literature but there are other
alternative descriptions that could have been used and differences
in the nature of the innovative procedures could be considered in
future research. Fourth, as discussed earlier, we did not include
an AJR absent/innovative procedures condition in our design. How-
ever, we note that existing audit standards, both in the US and in-
ternationally, do not preclude auditors from using innovative pro-
cedures and, therefore, future research could examine the effects of
innovative procedures when an AJR rule is absent but other incen-
tives for the use of innovative procedures exist. Fifth, our qualita-
tive analysis of questions asked by ACMs suggests potential inter-
action effects between participant background and the treatment
received. Deeper analysis of issues related to the background of
ACMs will require selection of participants from two diverse back-
grounds (e.g., former audit partners versus non-accountant ACMs)
with larger cell sizes. Sixth, we examined the effect of an AJR on
ACMs’ judgments in a one-period experimental setting. Future re-
search could consider examining the effect in a multi-period set-
ting to see if ACMs adjust their accountability levels after they
see how auditors actually behave in terms of being more or less
diligent with the implementation of an AJR and how auditors and
management react to the level of questioning by ACMs.
Acknowledgements
We thank the participants at the 5th Audit Quality Conference
in Venice, the 2014 Accounting, Organizations and Society Confer-
ence on Accounting Estimates (October 2014) and the Northeast-
ern University Research Workshop. We also thank Sandra Vera-
Muñoz (our commentator), Lisa Koonce (editor), two anonymous
reviewers and the helpful comments of Michael Coleman, Brian
Fitzgerald, Gary Monroe, Mark Nelson, Mark Peecher, Steve Salte-
rio, Roger Simnett, Ira Solomon and Arnie Wright. We acknowledge
a research grant from Chartered Accountants Australia and New
Zealand and the ?nancial support of an ARC Australian Professo-
rial Fellowship to Ken Trotman. The ?nancial support of Deloitte
& Touche LLP for their support of the AOS Estimates Conference is
gratefully acknowledged.
Appendix A
AJR Absent
Current regulation of auditors in Australia
In the current audit environment, auditors’ judgments, even
those that are complex and involve di?cult future estimates, can
be second guessed as courts can rule that alternative judgments
should have been reached. In addition, irrespective of how reason-
able the auditors’ judgment process was, regulatory inspectors (e.g.
ASIC in Australia; PCAOB in the USA) currently have wide authority
to determine that the judgments of auditors are ‘de?cient’.
AJR Present
New regulation of auditors in Australia
Please assume that new legislation in Australia has intro-
duced an Audit Judgment Rule whereby courts and regulatory
inspectors (e.g. ASIC in Australia; PCAOB in the USA) will not
second guess auditor judgments provided the auditor meets
certain key tests. Under an Audit Judgment Rule, a reasonableness
test is applied and even if an inspector develops an estimate that
differs from an auditor’s estimate, uses a different method to reach
an estimate than used by the auditor, or believes that his/her own
estimation process is superior to the auditor’s process, the inspec-
tor still can conclude that the auditor’s estimate is not unreason-
able. The key test is whether the auditor has used ‘reasoned eval-
uation made in good faith and in a rigorous, thoughtful and delib-
erate manner’. In summary, unless the auditor’s judgment process
is unreasonable, the auditor’s judgments cannot be said to be ‘de-
?cient’.
Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76 75
References
American Institute of Certi?ed Public Accountants (AICPA) (2007). Consideration of
fraud in a ?nancial statement audit. AU Section 316. New York, NY: AICPA.
Australian Securities & Investment Commission (ASIC) (2012). Report 317: Au-
dit inspection program report for 2011-12. Accessed May 2014.
Beasley, M. S., Carcello, J. V., Hermanson, D. R., & Neal, T. L. (2009). The audit com-
mittee oversight process. Contemporary Accounting Research, 26(1), 65–122.
Bell, T. B., Peecher, M. E., & Solomon, I. (2005). The 21st century public company
audit: Conceptual elements of KPMG’s global audit methodology. Montvale, New
Jersey: KPMG.
Bonner, S. E. (2008). Judgment and decision making in accounting. Upper Saddle River,
NJ: Pearson Prentice Hall.
Bratten, B., Gaynor, L. M., McDaniel, L., Montague, N. R., & Sierra, G. E. (2013). The
audit of fair values and other estimates: The effects of underlying environmen-
tal, task, and auditor-speci?c factors. Auditing: A Journal of Practice & Theory,
32(Supplement 1), 7–44.
Centre for Audit Quality (CAQ) (2014). CAQ approach to audit quality indicators.
Accessed May 2014.
Chen, X., Trotman, K. T., & Zhou, F. (2015). Nominal versus interacting electronic
fraud brainstorming in hierarchical audit teams? The Accounting Review, 90(1),
175–198.
Christensen, B. E., Glover, S. M., & Wood, D. A. (2012). Extreme estimation uncer-
tainty in fair value estimates: Implications for audit assurance. Auditing: A Jour-
nal of Practice & Theory, 31(1), 127–146.
Christensen, B. E., Glover, S. M., & Wood, D. A. (2013). Extreme estimation uncer-
tainty and audit assurance. Current Issues in Auditing, 7(1), 36–42.
Church, B. K., & Shefchik, L. B. (2012). PCAOB inspections and large accounting ?rms.
Accounting Horizons, 26(1), 43–63.
Cohen, J., Gaynor, L. M., Krishnamoorthy, G., & Wright, A. M. (2007). Auditor com-
munications with the audit committee and the board of directors: Policy rec-
ommendations and opportunities for future research. Accounting Horizons, 21(2),
165–187.
Cohen, J., Krishnamoorthy, G., & Wright, A. M. (2002). Corporate governance and the
audit process. Contemporary Accounting Research, 19(4), 573–594.
Cohen, J., Krishnamoorthy, G., & Wright, A. M. (2004). The corporate governance
mosaic and ?nancial reporting quality. Journal of Accounting Literature, 23, 87–
152.
Cohen, J., Krishnamoorthy, G., & Wright, A. M. (2010). Corporate governance in the
post-Sarbanes-Oxley era: Auditors’ experiences. Contemporary Accounting Re-
search, 27, 751–786.
Contessotto, C., & Moroney, R. (2013). The association between audit committee ef-
fectiveness and audit risk. Accounting & Finance, 54(2), 393–418.
Deloitte, L. L. P. (2014). Audit committee brief: A New Era in audit committee reporting.
Accessed May 2014.
DeZoort, F. T. (1998). An analysis of experience effects on audit committee members’
oversight judgment. Accounting, Organizations and Society, 23(1), 1–21.
DeZoort, F. T., Harrison, P., & Taylor, M. (2006). Accountability and auditors’ mate-
riality judgments: The effects of differential pressure strength on conservatism,
variability, and effort. Accounting, Organizations and Society, 31(4–5), 373–390.
DeZoort, F. T., Houston, R. W., & Hermanson, D. R. (2003). Audit committee mem-
ber support for proposed audit adjustments: A source credibility perspective.
Auditing: A Journal of Practice & Theory, 22(2), 189–205.
DeZoort, F. T., & Salterio, S. E. (2001). The effects of corporate governance experi-
ence and ?nancial-reporting and audit knowledge on audit committee mem-
bers’ judgments. Auditing: A Journal of Practice & Theory, 20(2), 31–47.
Dhaliwal, D., Naiker, V., & Navissi, F. (2010). The association between accruals qual-
ity and the characteristics of accounting experts and mix of expertise on Audit
Committees. Contemporary Accounting Research, 27(3), 787–827.
Ernst & Young (2013). Insights for European Audit Committees: Greater Business
Challenges Call for Stronger Audit Committees. Issue 18. Accessed May 2014.
European Commission (2010). Green paper audit policy: Lessons from the Crisis.
Federation of European Accountants (FEE) (2014). Opening a discussion: The fu-
ture of audit and assurance. Accessed May 2014.
Financial Reporting Council (FRC) (2012). Auditing practices board: Professional scep-
ticism – establishing a common understanding and rea?rming its central role in
delivering audit quality. London: Financial Reporting Council.
Financial Reporting Council (FRC) (2013a). Audit quality inspections an-
nual report 2012/2013.
Accessed May 2014.
Financial Reporting Council (FRC) (2013b). Audit quality thematic review: Ma-
teriality. Accessed May 2014.
Gendron, Y., & Bédard, J. (2006). On the constitution of audit committee effective-
ness. Accounting, Organizations and Society, 31(3), 211–239.
Gendron, Y., Bédard, J., & Gosselin, M. (2004). Getting inside the black box: A ?eld
study of practices in “effective” audit committees. Auditing: A Journal of Practice
& Theory, 23(1), 153–171.
Gibbins, M., & Newton, J. (1994). An empirical exploration of complex accountability
in public accounting. Journal of Accounting Research, 32, 165–186.
Glover, S. M., & Prawitt, D. F. (2013). Enhancing auditor professional skepticism.
Accessed October 2014.
Grant Thornton (2009). Response Letter in Release No. 104-2009-117: Report account-
ing oversight board.
Gri?th, E. E., Hammersley, J. S., & Kadous, K. (2014). Audits of complex estimates as
veri?cation of management numbers: How institutional pressures shape prac-
tice. Contemporary Accounting Research, in press.
Gri?th, E. E., Hammersley, J. S., Kadous, K., & Young, D. (2015). Auditor mindsets
and audits of complex estimates. Journal of Accounting Research, 53(1), 49–77.
Hammersley, J. S., Johnstone, K. M., & Kadous, K. (2011). How do audit seniors re-
spond to heightened fraud risk? Auditing: A Journal of Practice & Theory, 30(3),
81–101.
Hoffman, V., & Patton, J. (1997). Accountability, the dilution effect, and conservatism
in auditor’s fraud judgments. Journal of Accounting Research, 35(2), 227–237.
Hurtt, R. K. (2010). Development of a scale to measure professional skepticism. Au-
diting: A Journal of Practice & Theory, 29(1), 149–171.
Hurtt, R. K., Brown-Liburd, H., Earley, C. E., & Krishnamoorthy, G. (2013). Research
on auditor professional skepticism: Literature synthesis and opportunities for
future research. Auditing: A Journal of Practice & Theory, 32(Supplement 1), 45–
97.
International Auditing and Assurance Standards Board (IAASB) (2014). A framework
for audit quality: Key elements that create an environment for audit quality.
Accessed May 2014.
International Forum of Independent Audit Regulators (IFIAR) (2014).
2013 annual report & accounts. Accessed May 2014.
International Federation of Accountants (IFAC) (2011). Audit quality: An IAASB per-
spective. New York: IFAC.
International Standards of Auditing (ISA) 240. The Auditor’s responsibilities relating to
fraud in an audit of ?nancial statements. 2013 Handbook of International Qual-
ity Control, Auditing, Review, Other Assurance, and Related Services Pronounce-
ments, IFAC.
Kang, Y. J. (2014). Are audit committees more challenging given a sophisticated investor
base? Does the answer change given anticipation of additional mandatory audit
report Disclosure? University of Massachusetts-Amherst, Working Paper.
Koonce, L., Anderson, U., & Marchant, G. (1995). Justi?cation of decisions in auditing.
Journal of Accounting Research, 33(2), 369–384.
KPMG (2011). Elevating professional judgment in auditing. KPMG LLP.
Krishnan, G. V., & Visvanathan, G. (2008). Does the SOX de?nition of an accounting
expert matter? The association between audit committee directors’ account-
ing expertise and accounting conservatism. Contemporary Accounting Research,
25(3), 827–857.
Lerner, J. S., & Tetlock, P. E. (1994). Accountability and social cognition. In V. S. Ra-
machandran (Series Ed.), Encyclopedia of human behavior: Vol. 1 (pp. 3098–3121).
San Diego, CA: Academic Press.
Lerner, J. S., & Tetlock, P. E. (1999). Accounting for the effects of accountability. Psy-
chological Bulletin, 125(2), 255–275.
Libby, R., & Luft, J. (1993). Determinants of judgment performance in accounting
settings: Ability, knowledge, motivation, and environment. Accounting, Organi-
zations and Society, 18(5), 425–450.
Martin, R. D., Rich, J. S., & Wilks, T. J. (2006). Auditing fair value measurements: A
synthesis of relevant research. Accounting Horizons, 20(3), 287–303.
McDaniel, L., Martin, R. D., & Maines, L. A. (2002). Evaluating ?nancial reporting
quality: The effects of ?nancial expertise vs. ?nancial literacy. The Accounting
Review, 77(s-1), 139–167.
Monks, R., & Minow, N. (2004). Corporate governance. Malden, MA: Blackwell Pub-
lishing Ltd..
National Association of Corporate Directors (NACD) (2013). Honing skepticism. NACD
Directorship. . January/February 2013, 29-37.
Nelson, M. W. (2009). A model and literature review of professional skepticism in
auditing. Auditing: Journal of Practice & Theory, 28(2), 1–34.
Nelson, M. W., Elliott, J. A., & TarpIey, R. L. (2002). Evidence from auditors about
managers’ and auditors’ earnings management decisions. The Accounting Review,
77(4), 175–202.
Nelson, M., & Tan, H.-T. (2005). Judgment and decision making research in auditing:
A task, person, and interpersonal interaction perspective. Auditing: A Journal of
Practice & Theory, 24, 41–71.
O’Connell, J., & Boutros, A. S. (2002). Treating medical malpractice claims under a
variant of the business judgment rule. Notre Dame Law Review, 77, 373–436.
Palmrose, Z. (2006). Maintaining the value and viability of auditors as gatekeepers
under SOX: An auditing master proposal. In U. Fuchita, & R. Litan (Eds.), Finan-
cial gatekeepers: Can they protect investors (pp. 103–135). Baltimore, MD: The
Brookings Institute.
Public Company Accounting Oversight Board (PCAOB) (2010). Report on observations
of PCAOB inspectors related to audit risk areas affected by the economic crisis. Re-
lease No. 2010-006. Washington, D.C.: PCAOB.
76 Y.J. Kang et al. / Accounting, Organizations and Society 46 (2015) 59–76
Public Company Accounting Oversight Board (PCAOB) (2011). Concept release on au-
ditor independence and audit ?rm rotation. Release No. 2011-006. Washington,
D.C.: PCAOB.
Public Company Accounting Oversight Board (PCAOB) (2012). Auditing standard no.
16 communications with audit committees (AS 16). Washington, D.C.: PCAOB.
Public Company Accounting Oversight Board (PCAOB) (2013). Report on 2007–2010
inspections of domestic ?rms that audit 100 or fewer public companies. Accessed
May 2014.
Peecher, M. E. (1996). The in?uence of auditor’s justi?cation processes on their de-
cisions: A cognitive model and experimental evidence. Journal of Accounting Re-
search, 34(1), 125–140.
Peecher, M. E., Schwartz, R., & Solomon, I. (2007). It’s all about audit quality: Per-
spectives on strategic-systems auditing. Accounting, Organizations and Society,
32(4–5), 463–485.
Peecher, M. E., Solomon, I., & Trotman, K. T. (2013). An accountability framework for
?nancial statement auditors and related research questions. Accounting, Organi-
zations and Society, 38(8), 596–620.
Pomeroy, B. (2010). Audit committee investigation of signi?cant accounting deci-
sions. Auditing: A Journal of Practice & Theory, 29(1), 173–205.
Pozen, R. (2008). Final report of the advisory committee on improvements to ?nan-
cial reporting to the United States securities and exchange commission. Washington
DC: Securities and Exchange Commission Advisory Committee.
Rowe, S. P. (2014). Auditor Comfort Under Uncertainty. Tulane University Working
Paper.
Securities and Exchange Commission (SEC) 2011. Measurement uncertainty in ?nan-
cial reporting: How much to recognize and how best to communicate it. Brie?ng
Paper for the Financial Reporting Series Roundtable Discussion. Accessed Oc-
tober 2014.
Siegel-Jacobs, K., & Yates, J. F. (1996). Effects of procedural and outcome accountabil-
ity on judgment quality. Organizational Behavior and Human Decision Processes,
65(1), 1–17.
Srinivasan, S. (2005). Consequences of ?nancial reporting failure for outside direc-
tors: Evidence from accounting restatements and audit committee members.
Journal of Accounting Research, 43(2), 291–334.
Tetlock, P. E. (1983). Accountability and complexity of thought. Journal of Personality
and Social Psychology, 45(1), 74–83.
Tetlock, P. E. (1992). The impact of accountability on judgment and choice: Toward
a social contingency model. Advances in Experimental Social Psychology, 25, 331–
376.
Tetlock, P. E., & Boettger, R. (1989). Accountability: A social magni?er of the dilution
effect. Journal of Personality and Social Psychology, 57(3), 388–398.
Tetlock, P. E. (1985). Accountability: The neglected social context of judgment and
choice. In B. M. Staw, & L. Cummings (Eds.), Research in organizational behavior,
7 (pp. 297–332). Greenwich, Connecticut: JAI Press.
The Department of the Treasury (Treasury) (2008). Advisory committee on the audit-
ing profession. Final Report, October 6, 2008.
Turley, S., & Zaman, M. (2007). Audit committee effectiveness: Informal processes
and behavioural effects. Accounting, Auditing & Accountability Journal, 20(5),
765–788.
Vera-Muñoz, S. C. (2005). Corporate governance reforms: Rede?ned expectations
of audit committee responsibilities and effectiveness. Journal of Business Ethics,
62(2), 115–127.
doc_117372679.pdf