The Effects of PCAOB Inspections on Auditor-Client Relationships
Andrew A. Acito
Eli Broad College of Business, Michigan State University
Chris E. Hogan
Eli Broad College of Business, Michigan State University
Richard D. Mergenthaler
Henry B. Tippie College of Business, The University of Iowa
Preliminary draft. Please do not circulate or cite without permission.
Revised: September 2013
Abstract
We investigate the effects of PCAOB inspections on the relationships between Big 4 auditors and their
clients using a new measure to capture the information garnered from the reports. Specifically, we
measure the relative importance of accounting standards to each client’s financial statements and
calculate the exposure to deficient auditing by relating their auditor’s inspection deficiencies to the
accounting standards. This measure of deficient auditing exposure is then adjusted by the exposure that
would occur for the average of the other Big 4 auditors. We find that our measure of relative exposure to
deficient auditing is positively related to auditor changes, but is not related to audit changes in audit fees.
These results suggest PCAOB inspections affect auditor-client relationships, but auditors do not have the
ability to increase fees to remediate deficient auditing, nor do they reduce fees to retain clients when they
have more deficiencies in areas important to the clients. Our findings have implications for understanding
and regulating the market for audit services.
We thank Jay Newquist and Kyle Peterson for their programming assistance and Courtney Shemka and June Sun for
their help with data collection. We also thank Brown Bag Participants at Michigan State University.
The Effects of PCAOB Inspections on Auditor-Client Relationships
1. Introduction
One of the most important aspects of the Sarbanes-Oxley Act of 2002 has been the creation of the
Public Company Accounting Oversight Board (PCAOB). The PCAOB is charged with establishing
standards for auditing public companies, inspecting auditors with publically held clients, and levying
penalties on auditors who do not comply with PCAOB standards. In 2005 testimony to the U.S. House of
Representatives, former PCAOB Chairman James McDonough emphasized the importance of auditor
inspections stating, “…the more significant, long-term effects of our work will be the product of our
oversight activities” (PCAOB 2005). Auditors, however, have sometimes been critical of what the
PCAOB identifies as deficiencies in its inspection reports. Deloitte, for example, responded to its 2007
PCAOB inspection report by saying, “reasonable judgments should be respected and not second-guessed"
(Johnson 2009). While the PCAOB’s inspection process has clearly led to significant changes in the way
audits are conducted, there has been little research that investigates the effects of reporting inspection
deficiencies.
We investigate whether PCAOB inspection reports affect auditor-client relationships. First, we
look for evidence of audit fee changes related to inspection issues. When the PCAOB identifies that an
audit is deficient in a particular area, the auditor is expected to take action to address the issue. The
auditor may work to improve auditing in the area of the deficiency, both at the client where the issue was
identified and at other clients, and this additional work may lead to higher fees. For example, the 2006
inspection report of PricewaterhouseCoopers states that at Issuer A, “The Firm failed to test the fair value
of warrants and stock-based compensation issued in two significant transactions during the year.”1 To
improve in this area, PricewaterhouseCoopers may revise its audit procedures not only for Issuer A, but
also for other issuers or clients that have high levels of warrants and stock-based transactions. Auditors
1
The PCAOB does not disclose the names of the clients where audit deficiencies are identified.
1
have a strong incentive to address their deficiencies because the PCAOB may see a failure to do so as a
lack of quality control. The PCAOB also evaluates auditors’ quality control systems, and can make these
deficiencies public if they are not resolved within one year (PCAOB 2012).2 Another possibility is that
an auditor will see the PCAOB’s attention on the area as an indication of higher audit risk and thus raise
fees on clients that have exposure to the area where the deficiency was identified. Conversely, audit fees
may decrease if auditors reduce fees in an attempt to retain the clients. An Auditor may agree to lower
fees to pacify a client that might otherwise dismiss the auditor because of a negative PCAOB inspection
report related to the client’s accounting (Abbot, Gunny, and Zhang 2013).
Second, we investigate whether PCAOB inspection reports affect the likelihood of Big 4 auditor
changes. A client may choose to dismiss an auditor that receives a negative PCAOB inspection report
(Abbot, Gunny, and Zhang 2013) and may be especially likely to dismiss their auditor or not engage a
new auditor if the auditor has deficiencies in an area that is important to the client’s accounting. The
PCAOB specifically suggests that audit committees ask auditors about deficiencies identified in their
audit as well as in audits of other clients that have similar accounting to the client (PCAOB 2012). It is
also possible that auditors want to avoid or discontinue relationships with clients that have high exposure
to certain accounting standards. Auditors may also be more likely to resign from or refuse to take on
marginal clients that use accounting standards that increase the auditors’ exposure to areas where
inspection deficiencies are likely.
An alternative to the hypothesized relations discussed above is that PCAOB inspection reports do
not affect auditor-client relationships for Big 4 auditors. There are reasons to believe that the PCAOB
reports are not informative to clients. For example, Lennox and Pittman (2010) find no evidence of a
relation between the number of issues identified in PCAOB inspections and changes in auditor market
2
As an example, in the 2008 and 2009 inspection reports for PricewaterhouseCoopers (PwC), the PCAOB
inspectors noted deficiencies related to auditing fair value measurements. Subsequently, Part II of those inspection
reports were released because the PCAOB felt PwC had not made sufficient progress in addressing quality controls
with respect to auditing fair value measurements, despite PwC taking actions which “included providing our audit
professionals with enhanced audit tools, training and additional technical guidance to promote more consistent audit
execution.” Of the Big 4 audit firms, only KPMG has yet to have a Part II of an annual inspection report released.
2
share. Gunny and Zhang (2013) also suggest inspection findings do not inform about audit quality for
larger audit firms, which the PCAOB inspects annually. Further, the PCAOB reports that auditors often
dismiss inspection findings in their communications with audit committees (PCAOB 2012).
To investigate the relation between PCAOB inspection results and auditor-client relationships, we
examine PCAOB inspection reports issued between 2005 and 2011 for Big 4 auditors. We search each
inspection report for a list of keywords developed by Folsom, Hribar, Mergenthaler, and Peterson (2012)
that relate to each accounting standard. If an inspection report contains a keyword related to a particular
accounting standard, this indicates the auditor had a deficiency related to auditing this standard. We also
determine how frequently the accounting standard keywords are used in each client’s financial statements
to understand the relative exposure each client has to the accounting standard. We then examine how
clients’ exposure to accounting standards related to areas where their auditor has PCAOB identified
deficiencies affects changes in audit fees and auditor changes.
We find that when a client-auditor pair has a high level of exposure to deficient auditing, based
on the client’s relative use of the accounting standards and the auditor’s deficiencies in these areas
relative to other Big 4 auditors, there is a higher likelihood of an auditor change occurring. There is also
lower deficient auditing exposure in the resulting new auditor-client pair. Deficient auditing exposure,
however, does not appear to be related to changes in audit fees for auditor-client relationships that are
maintained. Together, these results suggest that information in PCAOB inspection reports affect auditorclient relationships by helping to align clients’ accountings with Big 4 auditors’ expertise.
This paper makes a number of contributions to existing literature. First, the paper further
explores the effects of PCAOB inspections on auditor-client relationships. The literature to date has
found only limited evidence that PCAOB inspections affect these relationships. Lennox and Pittman
(2010) conclude clients do not find the inspection reports informative because auditors’ quality control
system deficiencies are not publically disclosed. Other studies provide evidence that triennially inspected
auditors (auditors with fewer than 100 clients) are more likely to be dismissed when they have GAAPdeficient PCAOB reports (Abbott, Gunny, and Zhang 2013) and that smaller, lower quality auditors exit
3
the audit market (Defond and Lennox 2011). Our study, however, focuses on the effects that PCAOB
inspection reports have on Big 4 auditors and their clients. Included in our study is an examination of the
reports’ effects on audit fees, an aspect of the auditor-client relationship that has received little attention
in prior studies on the effects of PCAOB inspection reports.
Second, our paper introduces a new approach to examining deficiencies identified in PCAOB
inspection reports. Specifically, we link PCAOB inspection deficiencies to firms’ financial statements
using an accounting standard keyword list developed by Folsom, Hribar, Mergenthaler, and Peterson
(2012). Prior work has relied on the number of issues identified in PCAOB inspection reports (e.g.,
Lennox and Pittman 2010; Abbott, Gunny, and Zhang 2013), which may be less relevant to clients’
auditor-engagement decisions than the types of deficiencies found. Our method is particularly
advantageous because the PCAOB releases little information about the clients where inspections issues
are identified.3 Because the client or the industry in which the client operates cannot usually be
identified, linking deficiencies related to accounting standards related with clients’ exposure to those
standards may be the best option for determining how specific types of PCAOB audit deficiencies affect
clients.
Our paper also has implications for auditing regulation and practice. Our results on PCAOB
inspection reports and audit fees provide information on whether the inspection reports improve audit
quality. We would expect fees to increase for clients with exposure to accounting standards related to
PCAOB audit deficiencies if auditors begin to do more testing to remediate these deficiencies. We would
expect lower fees if auditors reduce fees to pacify their clients after negative PCAOB reports. Lower fees
could hinder resolving the audit deficiencies and may lead to deficiencies in other areas if the auditor is
spread thin as a result of the audit fee cuts. Our results do not show evidence of either scenario, which
suggests that Big 4 audit firms with more deficiency exposure have little ability to negotiate higher fees,
but also do not appear to lower fees to retain clients. Further, our results provide evidence on PCAOB
3
Language in the SOX legislation actually prohibits the PCAOB from releasing the names of the clients where the inspection
issues are identified.
4
inspections improving audit quality by aligning clients with auditors that have more expertise in the
accounting standards used by the clients. This evidence has important implications for proposed rules
requiring auditor rotation. Auditor rotation could possibly have a negative effect on audit quality because
it could undo gains related to auditor-client realignment resulting from PCAOB inspections.
2. Background and Hypothesis Development
2.1
Background on PCAOB inspection process
The Public Company Accounting Oversight Board (PCAOB) was established by the Sarbanes-
Oxley Act of 2002 (SOX) as part of sweeping reforms aimed at restoring investor confidence in capital
markets after a number of high-profile accounting frauds. The PCAOB is charged with overseeing public
company auditors by establishing auditing standards, performing inspections, and assessing penalties for
failure to adhere to its standards. Prior to the establishment of the PCAOB, the audit profession was
largely self-regulated. The American Institute of Certified Public Accountants (AICPA) created auditing
standards and established a peer review process in which member firms selected another audit firm to
perform an annual review.4
The Sarbanes-Oxley Act requires audit firms that audit over 100 public clients to be inspected
annually and auditors with fewer public clients to be inspected every three years (PCAOB 2008). The
PCAOB inspections program uses a risk-based approach to selecting audit engagements, and audit areas
within those engagements, and thus the inspection findings are not necessarily representative of all audit
engagements (Gradison and Boster 2010). Inspectors may identify audit deficiencies as well as GAAP
departures, and these findings are made public in a report released typically several months following the
inspection (although the client is not identified in the public report). Quality control deficiencies
identified during the inspection process are only made public if they are not remediated to the Board’s
satisfaction within 12 months.
4
While this peer review process was self-regulated, there is some evidence that it provided audit clients with useful
information. Hilary and Lennox (2005) find that audit firms gained and lost clients depending on whether there were
positive or negative peer review opinions.
5
2.2 Impact of the PCAOB Inspection Process
To date, there is limited evidence on the impact of the PCAOB inspection process on audit
quality. In fact, the PCAOB and others have acknowledged it is difficult to measure “audit quality,”
which in turn makes it difficult to assess improvements in audit quality (e.g. Department of the Treasury,
2008 p. VIII:14-15). Carcello, Hollingsworth, and Mastrolia (2011) provide evidence related to audit
quality, as proxied by changes in abnormal accruals, for clients of the Big 4 audit firms following the first
two years of inspections. Carcello, Hollingsworth, and Mastrolia (2011) document a significant reduction
in abnormal accruals and conclude their results are consistent with an improvement in audit quality for the
Big 4 firms following PCAOB inspections. Lamoreaux (2013) compares audit quality, as proxied by the
propensity to issue going concern opinions as well as disclose material weaknesses in internal controls, in
foreign jurisdictions subject to PCAOB inspections versus those that have refused the PCAOB access to
inspect, and finds evidence of higher audit quality in those jurisdictions with “inspection exposure.” He
finds no evidence of such a difference prior to the PCAOB’s existence, and concludes exposure to the
PCAOB inspection process has resulted in an improvement in audit quality. Similarly, Gramling,
Krishnan and Zhang (2011) find an increased likelihood of issuing going concern opinions following
PCAOB inspections of triennially-inspected audit firms.
Evidence related to audit quality may also be inferred by the findings form DeFond and Lennox
(2011) who document that more than 600 small audit firms ceased auditing public clients after the
passage of SOX. DeFond and Lennox (2011) note that these were lower quality auditors since they either
avoided AICPA peer review or PCAOB inspection or were more likely to have deficiencies if they did
have a review or inspection (DeFond and Lennox 2011). The publicly-traded client firms would then be
forced to switch to an audit firm still registered with the PCAOB, presumably a higher quality audit firm.
The authors do in fact find support that audit quality, measured by the likelihood of receiving a going
concern opinion, increased for these client firms following the switch.
In this study, we seek to provide evidence on how the inspection process itself, and in particular
the PCAOB’s focus on high risk audit areas, may improve (or possibly reduce) audit quality. While prior
6
studies provide evidence of audit quality on average, as proxied by either the likelihood of issuing a going
concern opinion or abnormal accruals, we examine whether audit firms and clients respond to the
particular audit and/or GAAP deficiencies identified in the inspection process. As discussed earlier, the
PCAOB’s focus on particular audit areas will likely impact both audit firm behavior and client firms’
assessments of their current auditor. The inspection process sheds light on audit deficiencies and audit
firms may increase audit effort in these areas, or may even choose to avoid client firms with significant
exposure to particular high risk areas. In addition, the inspection process may provide evidence of an
audit firm’s quality which could result in client firms renegotiating audit fees or seeking a new auditor.
We discuss each of these potential impacts in greater detail below.
2.2.1 Audit Firm Responses to PCAOB Inspections
There tend to be common themes across PCAOB inspection reports in terms of the audit areas
reviewed and deficiencies identified. For example, in their summary of 2004-2007 inspection findings for
the annually inspected firms, the PCAOB notes that it continues to observe deficiencies in the significant
areas of revenue, fair value, management’s estimates, determination of materiality, and audit scope
(PCAOB 2008). The PCAOB also notes that it will continue to focus on these significant areas in future
inspections. In a study summarizing the PCAOB inspection findings for the annually inspected firms
over the period 2004-2009, Church and Shefchik (2012) also note the most common deficiencies involve
testing revenues, fair value measurements, other accounting estimates, and internal controls. Thus, the
engagements selected for review by the PCAOB are likely to be those with significant exposure to the
areas of concern. For example, the PCAOB paid particular attention to audits of companies in the
financial services sector following the financial crisis (PCAOB 2010).
There are various ways in which audit firms may respond following the identification and
disclosure of audit deficiencies. Audit firms may merely become more conservative in their reporting
decisions. For example, the findings of Lamoreaux (2013) may be evidence of changes in reporting
decisions in that audit firms increase the likelihood of issuing a going concern opinion and disclosing a
7
material weakness in internal controls. This does not necessarily imply they have changed their audit
approach to increase the amount of audit evidence gathered in these areas.
Alternatively, auditors may choose to increase audit effort in particular high risk areas they
believe the PCAOB will focus on during the inspection, or that were focus areas in previous inspections.
Carcello, Hollingsworth, and Mastrolia (2011) note that firms can adjust their audit approach, address
issues through their continuing education programs, and strengthen their internal work paper review
procedures in response to deficiencies identified in an inspection. For example, audit firms often note in
their written response to an inspection report that they are increasing training and documentation in
particular areas as a result of deficiencies identified by the PCAOB. We expect this would translate into
additional audit effort and fees for client firms with significant exposure to the audit area in which a
deficiency was identified (e.g. fair value estimates).
In some cases, an audit firm may choose to resign from clients with significant exposure in areas
where the audit firm performs poorly. When the PCAOB identifies a deficiency at an audit firm, the firm
may perceive clients with heavy exposure to the related accounting standard as higher risk because the
firm will have to make changes to its audit procedures for those clients. Further, auditors may perceive
higher reputation risk from auditing these clients because continuing to audit them may result in future
deficiencies identified by the PCAOB.
2.2.2 Client Firm Responses to PCAOB Inspection Reports
Since the passage and implementation of SOX, the audit committee has been responsible for the
hiring and firing the external auditor and for evaluating them on an annual basis. The PCAOB believes
their reports should be informative to audit committees making auditor acceptance and retention
decisions. PCAOB Chairman James Doty, for example, states, “Whether an audit committee's own
company audit is being reviewed as part of an inspection, or whether it's another company within the
same industry, PCAOB inspection reports provide insight into areas of risk and audit quality that are of
concern to all audit committees” (Whitehouse 2012).
8
In a discussion paper aimed at audit committee members, the PCAOB provides recommended
questions the audit committee should present to their audit firm in order to understand the implications of
inspection findings (PCAOB 2012). One of the issues they recommend audit committees discuss is
whether or not deficiencies identified on other audit engagements are related to audit areas of their
company. In addition, the PCAOB recommends the audit committee ask what the audit firm is doing to
remediate any identified deficiencies. This document suggests the PCAOB believes there is useful
information for audit committees in the inspection reports that should be considered in their annual
evaluation of the incumbent audit firm, even if their company’s audit was not selected for inspection.
Audit committees and management may react to this information in various ways. An optimistic
view is that they encourage their audit firm to invest in training and developing the appropriate audit
methodology to address the deficiencies, and state they support their auditor in increasing audit effort in
high risk areas. On the other hand, a pessimistic view is that the audit committee finds the inspection
reports uninformative, unrelated to their particular engagement, or dismisses the deficiencies as being
trivial and insignificant, and argues against any additional work that would increase fees.
Several studies discuss and examine the information content of PCAOB inspection reports, but
evidence on the usefulness of the reports is limited. One reason the reports may be viewed as
uninformative by audit committees and others is that the selection of audit engagements is risk-based and
therefore the findings are not generalizable or necessarily representative of audit quality for any particular
audit firm. Others suggest that the inspections focus on trivial and/or inconsequential audit deficiencies
(Glover, Prawitt, and Taylor 2009; DeFond 2010) and thus may not be informative to audit committees.
Despite these concerns, some evidence indicates that the reports are associated with audit quality or
perceptions of audit quality.
Gunny and Zhang (2012) examine whether PCAOB inspection reports provide information about
audit quality by associating the severity of deficiencies with proxies for audit quality. They find the
inspection reports for triennially-inspected firms provide information about audit quality, i.e. clients of
auditors with “seriously deficient” reports (reports that claim deficiencies lead to GAAP departures) have
9
significantly higher abnormal accruals and are more likely to restate their financial statements. However,
this report classification does not distinguish among the annually inspected audit firms, as there is less
variability in these reports in terms of the number and severity of deficiencies. In addition, Abbott,
Gunny, and Zhang (2013) find that clients of triennially inspected auditors with GAAP deficiencies
identified in their PCAOB inspection reports are more likely to dismiss their auditors, and this association
is stronger for companies with greater agency conflicts and more effective audit committees. The Abbott,
Gunny, and Zhang (2013) results suggest clients of triennially inspected audit firms use the inspection
reports as a signal of audit quality.
In contrast, Lennox and Pittman (2010) conclude that less information about audit quality is
conveyed with the inception of the PCAOB inspection process, as compared to what was conveyed under
the prior peer review regime, and that audit clients do not find the PCOAB inspection reports useful in
signaling audit quality. Their conclusion is based on a lack of association between the number (or
unexpected number) of deficiencies disclosed in PCAOB inspection reports and changes in audit firm
market shares. Lennox and Pittman (2010) also note there is criticism over the initial confidentiality of
quality control concerns (Part II of the PCAOB reports), and suggest this lack of transparency contributes
to the lack of informativeness of the reports. However, as noted by DeFond (2010), even if the reports are
“uninformative” to client firms, the inspection process may still result in improvements in audit quality.
In a more recent study examining the public disclosure of Part II findings for triennially inspected firms
that do not remediate quality control deficiencies to the satisfaction of the PCAOB within 12 months,
Ragothaman (2012) finds the PCAOB reports disclose a greater number of quality control deficiencies
relative to peer review reports, suggesting the PCAOB is more harsh than the peer reviewers and may
provide an ex ante incentive to improve audit quality.
2.2.3 Types of Deficiencies and the Audit Market
It is important to understand how audit firms and their clients react to the deficiencies as this has
implications for audit market concentration and audit quality. In the final report from The Department of
the Treasury Advisory Committee on the Auditing Profession, the committee notes “Auditing firms,
10
public companies, market participants, academics, investors and others reasoned that large public
companies with operations in multiple countries need auditing firms with global resources and technical
and industry expertise to deal with an increasingly complex business and financial reporting
environment” (Department of the Treasury 2008, p. VIII:2). The committee concludes that these needs
restrict the choice of auditor for the large multinational companies to just the global auditing firms. As
accounting standards increase in complexity and the PCAOB focuses on the auditing of these most
complex standards, more clients may be driven to seek the expertise of a global auditing firm, and even
limit their choices among the Big 4 accounting firms.
At the same time, regulators express concern that too much concentration in the audit market
decreases quality-based competition and may result in reduced audit quality (Department of the Treasury
2008, p. VIII:2-3). To the extent we observe auditor-switching activity, rather than increasing audit fees
related to improved audit procedures in response to deficiencies identified in PCAOB inspection reports,
reduced audit market competition could be an unintended consequence of the inspection process.
2.3 Hypotheses
There is little prior research on how auditors and clients react to PCAOB inspection reports and
how these reports affect auditor-client relationships. While there is some evidence that clients are more
likely to dismiss smaller, triennially-inspected audit firms with inspection deficiencies, there is no such
evidence for clients of Big 4 auditors. An important question is whether audit deficiencies affect auditorclient relationships in terms of audit fees charged and the engagement decisions of both clients and
auditors, especially for Big 4 auditors who audit the majority of large publically traded companies.
As noted above, audit firms may respond to PCAOB inspection findings by increasing training
and audit effort in the specifically-identified areas of concern, and may face more pressure to do so if the
deficiencies are specific to the auditor as opposed to deficiencies identified for all auditors. We expect
this would result in higher audit effort and higher audit fees for clients with significant exposure to the
accounting standards related to the identified audit deficiency, leading to our first hypothesis. It is also
possible, however, that audit firms may decrease their fees in order to pacify clients considering
11
dismissing auditors that have deficiencies identified in areas important to those clients (Abbot, Gunny,
and Zhang 2013). Such a fee reduction, is also more likely to occur if the deficiencies are specific to the
auditor.
Hypothesis 1: Audit fees are increased (reduced) when a Big 4 auditor’s inspection report
indicates the exposure to deficient auditing, given the client’s use of accounting standards, is
higher than it would be for other Big 4 auditors.
As an alternative to increased audit effort, audit firms may selectively resign from audit
engagements or price themselves out of the market to avoid significant exposure to particular high risk
audit engagements. Client firms may also choose to dismiss their audit firm if they believe they do not
have sufficient expertise in auditing certain high risk areas. In both cases, we expect an increase in
switching activity away from an audit firm without expertise to an audit firm perceived to have greater
expertise. These predictions lead to the following two hypotheses.
Hypothesis 2a: An auditor-client engagement is less likely to continue when a Big 4 auditor’s
inspection report indicates the exposure to deficient auditing, given the client’s use of accounting
standards, is higher than it would be for other Big 4 auditors.
Hypothesis 2b: On average, clients switching Big 4 auditors is will engage an auditor that results
in lower exposure to deficient auditing than existed with the prior auditor.
3. Research Design
3.1 Data
We begin our analysis with PCAOB inspection reports for the Big 4 auditing firms between 2005
and 2011.5 The PCAOB conducts annual inspections of these of these auditing firms, resulting in 28
inspection reports. The reports issued in a given year are usually based on inspections that occur in the
prior year and are assumed to be related to audits of the clients’ fiscal years ending the year prior to the
inspection work.
5
All PCOAB inspection reports are available at PCAOB.org. Full PCAOB inspections on Big 4 auditors are first
issued in 2005. In 2004, the PCAOB issued limited inspection reports on these firms, reviewing 16 audits at each
firm, whereas recent inspections typically review more than 50 audits. To the extent audit fee changes or auditor
changes occurred due to these limited reports, we expect not including the limited reports to bias against identifying
a relation between PCAOB reports and audit-client relationships.
12
Part I.A of the inspection reports, “Review of Audit Engagements,” includes descriptions of the
audit deficiencies identified by the PCAOB during their inspections. For this section in each inspection
report, we use a keyword list for U.S. GAAP standards developed by Folsom, Hribar, Mergenthaler, and
Peterson (2012) to identify accounting standards related to audit deficiencies detailed by the PCAOB.6
An auditor is considered to be “deficient” in auditing accounting related to a standard if a keyword related
to the accounting standard is identified in the inspection report. Whether a deficiency related to a certain
accounting standard is identified is a function not only of the auditing proficiency, but also a function of
the extent to which its clients use the accounting standard and the PCAOB’s focus on the area. However,
we believe that given Big 4 firms have broad client bases, and given that the PCAOB uses a risk-based
approach to selecting the audits it reviews and its focus on certain accounting issues, the likelihood of the
PCAOB reviewing auditing related to the accounting standards is relatively consistent across firms.
Table 1 displays the number of Big 4-firm inspection reports with deficiencies related to each
accounting standard that were identified using the keyword list.7 We find that standards related to
intangible assets (SFAS 142), business combinations (SFAS 141), and income taxes (SFAS 109) are most
commonly identified in the reports. These findings are aligned with the PCAOB’s 2008 report, which
stated that inspection issues were frequently identified relating to accounting estimates, fair value
estimates, and income taxes. Inspections also frequently identified issues related to revenue, which
corresponds to standards SOP 97-2, SAB 101, and SFAS 48 in the key word list (PCAOB 2008). In
addition, the PCAOB identifies inventory, fair value of financial instruments, and valuation of pension
plan assets as areas where audit firms frequently failed to test controls (PCAOB 2010). Keywords for
6
Appendix A provides a partial list of the keywords used and the related standards. See Folsom, Hribar,
Mergenthaler, and Peterson (2012) for complete details on the development of the keyword list. Folsom, Hribar,
Mergenthaler, and Peterson (2012) validate the list’s ability to identify accounting standard usage in 10-K filings by
examining the word counts across industries, correlating the word counts with dollar amounts, and having the list
reviewed by the national office of a Big 4 accounting firm. Appendix B provides the details of selected deficiencies
identified by the PCAOB.
7
Results for certain standards that contain overlapping terms are removed to avoid double counting. For example,
we include the results for search terms related to SFAS 123R, but not related to SFAS 123 and ABP 25.
13
SFAS 107, SFAS 115, SFAS 133, and SFAS 87 are related to the fair value of financial instruments and
pensions, however, our keyword list does not identify any deficiencies related to inventory.
We use the same keyword list to identify the relative importance of each accounting standard to
each audit client. Specifically, we match the keyword list to Big 4 auditor clients’ machine-readable 10-K
filings on EDGAR and count the number of times keywords related to each standard are identified in each
10-K filing. These counts are then standardized, by subtracting the mean and dividing by the standard
deviation of the counts for each standard-year, to create a measure of the relative impact (RELIMPACT)
that each standard has for a client (Folsom, Hribar, Mergenthaler and Peterson 2012). The minimum
standardized count value for each accounting standard is added back to the variable to keep all values
positive.
We match relative impact numbers from the 10-K filings to the auditor deficiencies identified in
the PCAOB inspection reports using the fiscal year of the reviewed audits and the companies’ auditors
listed on Compustat. The combination of this data allows us to calculate the amount of exposure there is
to auditing considered deficient by the PCAOB for each client-auditor pair. Specifically, we measure
deficient auditing exposure (DEFEXP) as
∑
where
RELIMPACT = client’s standardized keyword count for an accounting standard;
DEFICIENCY = indicator variable equal to one if the PCAOB identifies a client’s auditor as
having a deficiency related to the accounting standard, and zero otherwise.
To measure deficient auditing exposure with the current auditor relative to the exposure to
deficiencies they would face with other auditors, we calculate an adjusted auditor deficiency exposure
(ADJEXP) as
where
14
ALTEXP = the mean of deficient auditing exposures calculated as if the client was instead
audited by the other three Big 4 auditors.
This adjustment to the deficient auditing exposure ensures the variable captures the auditing deficiency
exposure that is specific to the auditor-client pair as opposed to exposure to auditing related to accounting
standards that is particularly high risk or a particular focus of the PCAOB. The adjusted deficiency
exposure (ADJEXP) can be thought of as the misalignment between a client’s accounting standard usage
and the auditor’s ability to audit those accounting standards.
Figure 1 provides time trends for the adjusted deficiency exposure (ADJEXP) and its components
over the fiscal years of the audits inspected by the PCAOB. In general, both adjusted deficiency
exposure and the raw values of deficiency exposure (DEFEXP) are decreasing over time. These trends
indicate that number of deficiencies is decreasing and that alignment between clients’ accounting and
auditor expertise is improving. The increase in adjusted deficiency exposure and its components in 2007
is primarily driven by one auditor’s inspection, which contained new deficiencies related to taxes and
intangible assets.
Other data for the models is obtained from Audit Analytics and Compustat. Data on audit fees
and auditor changes are from Audit Analytics. Client data for control variables in the models is obtained
from Compustat. We require sample observations to have all necessary data items for the models and
exclude financial firms (SIC 60-69) because these clients are likely to have different audit fee structures
from other clients (Fields, Fraser, and Wilkins 2004).
3.2 Models
3.2.1 Audit Fee Changes
We investigate Hypothesis 1 on whether changes in deficient auditing exposure affects audit fees
using a model of audit fee changes. Specifically, we use an ordinary least squares (OLS) regression of
audit fee changes on changes in deficient auditing exposure and control variables. The variables in the
model are based primarily on work by Ettredge, Li, and Scholz (2007) and Hogan and Wilkins (2008).
The form of the model is
15
where
ΔFEEit = change in the natural logarithm of audit fees for client i from fiscal year t-1 to fiscal
year t;
ΔADJEXPit = change in adjusted auditing exposure between the current PCAOB report issued
during fiscal year t and the last PCAOB report prior to start of fiscal year t for client i’s auditor;
ΔSIZEit = change in the natural logarithm of total assets for client i from fiscal year t-1 to fiscal
year t;
ΔROAit = change in return on assets for client i from fiscal year t-1 to fiscal year t;
ΔABSAACCit = change in the absolute value of abnormal accruals for client i from fiscal year t-1
to fiscal year t;
ΔNSOPit = change in an indicator variable for a non-standard audit opinion for client i from fiscal
year t-1 to fiscal year t (1 indicates change from standard to non-standard, -1 indicates change
from non-standard to standard, 0 indicates no change);
ΔMWit = change in an indicator variable for the existence of an internal control material
weakness for client i from fiscal year t-1 to fiscal year t (1 indicates change from no weakness to
weakness, -1 indicates change from weakness to no weakness, 0 indicates no change);
ΔLEVit = change in leverage for client i from fiscal year t-1 to fiscal year t;
ΔINVit = change in inventory scaled by total assets for client i from fiscal year t-1 to fiscal year t;
ΔRECit = change in receivables scaled by total assets for client i from fiscal year t-1 to fiscal year
t;
ΔM&Ait = change in an indicator variable for mergers and acquisition activity for client i from
fiscal year t-1 to fiscal year t (1 indicates change from no M&A activity to M&A activity, -1
indicates change from M&A activity to no M&A activity, 0 indicates no change);
ΔSEGSit = change in the number of reported business segments for client i from fiscal year t-1 to
fiscal year t;
ΔSPITEMit = change in an indicator variable for the existence of special items for client i from
fiscal year t-1 to fiscal year t (1 indicates change from no special items to special items, -1
indicates change from special items to no special items, 0 indicates no change);
ΔRESTRUCTit = change in an indicator variable for restructuring charges for client i from fiscal
year t-1 to fiscal year t (1 indicates change from no restructuring charges to restructuring charges,
-1 indicates change from restructuring charges to no restructuring charges, 0 indicates no change);
YEAR = indicator variables equal to one if the observation for the corresponding fiscal year, and
zero otherwise.
Complete variable definitions are available in Appendix C.
The variable of interest, ΔADJEXP, measures the change in deficient auditing exposure for the
client-auditor pair, relative to the average exposure that would occur if the client used other Big 4
auditors. A positive coefficient on ΔADJEXP would indicate that audit firms increase their fees, either
due to increased risk or increased audit effort, when the auditor-client relationship experiences an increase
in exposure to deficient auditing. A negative coefficient would be consistent with auditors attempting to
16
pacify clients that might otherwise dismiss the auditor and replace them with an auditor with fewer audit
deficiencies in areas that are important to the client.
We control for a number of other variables we expect to be related to changes in audit fees from
one year to the next. We expect that increases in firm size (ΔSIZE), inventory (ΔINV), receivables
(ΔREC), and business segments (ΔSEGS) will be positively related to changes in audit fees because these
variables are related to audit effort. Similarly, additional audit effort is likely to increase audit fees when
clients engage in more special transactions such as restructuring (ΔRESTRUCT), mergers and acquisitions
(ΔM&A), or special items (ΔSPITEM) or when clients have new auditing issues that lead to non-standard
audit opinions (ΔNSOP) or the disclosure of material weaknesses in internal control (ΔMW) (Ettredge, Li,
and Scholz 2007; Hogan and Wilkins 2008). We also control for variables related to audit risk that will
increase audit effort or the premium the auditor charges for bearing the risk. Changes in audit fees are
expected to be negatively related to changes in return on assets (ΔROA) and positively related to changes
in the absolute value of abnormal accruals (ΔABSAACC) and changes in leverage (ΔLEV). Finally, we
control for the client fiscal YEAR to the capture time trend in audit fee changes.
3.2.2 Auditor Changes
We use a logistic regression model to investigate Hypothesis 2 on whether the adjusted auditing
deficiency exposure affects the likelihood of auditor changes occurring. The unit of analysis is a clientinspection observation. For each annual PCAOB inspection report (r), there is one client-inspection
observation for each publically traded client of the audit firm that has the necessary data. The dependent
variable (AUDCH) is equal to one if an auditor change occurs between the inspection report release date
and the lesser of one year or the auditor’s next inspection report release, and zero otherwise. The control
variables in the model are based on Ettredge, Li, and Scholz (2007). The form of the model is
where
17
AUDCHir = indicator equal to one if client i changes auditors between the time its auditor’s
inspection report is released and the lesser of one year or the auditor’s next inspection report
release, and zero otherwise;
ADJEXPir = adjusted auditing exposure calculated for client i using inspection report r;
SIZEir = natural logarithm of total assets for client i for the fiscal year ending prior to the release
of inspection report r;
ABFEEir = abnormal audit fees for client i for the fiscal year ending prior to the release of
inspection report r;
NSOPir = indicator variable equal to one if client i has a non-standard audit opinion for the fiscal
year ending prior to the release of inspection report r, and zero otherwise;
MWir = indicator variable equal to one if client i has an internal control material weakness for the
fiscal year ending prior to the release of inspection report r, and zero otherwise;
LEVir = leverage for client i for the fiscal year ending prior to the release of inspection report r;
ABSAACCir = absolute value of abnormal accruals for client i for the fiscal year ending prior to
the release of inspection report r;
AUDINDSHRir = percent of the square root of client assets audited by client i’s auditor in client
i’s two-digit standard industry code for the fiscal year ending prior to the release of inspection
report r;
INDUSTRY = indicator variables equal to one if firm i is a member of the two-digit Standard
Industry Classification (SIC) code, and zero otherwise;
YEAR = indicator variables equal to one if the observation for the corresponding fiscal year, and
zero otherwise.
Complete variable definitions are available in Appendix C.
In this model, the variable of interest, ADJEXP, is the auditing deficiency exposure for the clientauditor pair less the average exposure that would have occurred if the client used other Big 4 auditors.
We expect a positive relation between ADJEXP and auditor changes, which would indicate that auditor
changes are more likely to occur when auditors have deficiencies related to areas of accounting important
to the client. The client may be more likely to dismiss the auditor because other auditors would have
fewer deficiencies or the auditor may be more likely to resign because the client is seen as being higher
risk to the auditor.
Other variables likely to affect auditor-client engagement decisions are included in the model as
controls. We expect SIZE to be negatively related to auditor changes (AUDCH), because large clients are
likely to be important to the audit firms. We also expect the auditor’s market share of the industry
(AUDINDSHR) to be negatively related to auditor changes because both the auditor and the client benefit
from the auditor’s industry expertise. Positive coefficients are expected on leverage (LEV) and abnormal
accruals (ABSAACC) because higher values of these variables are associated with more risk, which will
18
increase the likelihood of auditor resignations. Abnormal fees (ABFEE) are expected to have a positive
relation with auditor changes because higher abnormal fees may be a sign that the auditor considers the
client to be high risk and because clients are more likely to leave auditors charging high fees (Hribar,
Kravet, and Wilson 2010; Ettredge, Li, and Sholz 2007). Further, we expect non-standard audit opinions
(NSOP) and material weakness disclosures (MW) to be positively related to auditor changes because these
variables indicate higher risk and auditor-client tension. Finally, we include indicator variables to control
for INDUSTRY and fiscal YEAR effects on auditor changes.
4. Results
4.1 Audit Fee Changes
Table 2 provides descriptive statistics for the audit fee changes test. The sample is comprised of
year-to-year audit fee changes for Big 4 auditor clients that maintain the same auditor between the years.
The sample is further limited to observations where a PCAOB inspection report is issued for the client’s
auditor between the years. These data restrictions and data requirements to the model yield a sample of
6,769 audit fee change observations. The “average” client firm is charged about $2.6 million in audit fees
and the overall trend in fees (ΔFEE) is flat over the sample period. Further, the average firm in the
sample is increasing in size (ΔSIZE), decreasing in return on assets (ΔROA), and is also decreasing in
adjusted auditing exposure (ΔADJEXP), and the likelihood of having a non-standard audit opinion
(ΔNSOP) or a material weakness in internal control (ΔMW).
In Table 2 Panel B, we separate the sample into observations with and without increases in audit
fees. Audit fees increase in 3,429 observations or approximately 51 percent of the sample. There is only
a weakly significant (Z-Stat = 1.68) difference in the median change in adjusted auditing exposure
(ΔADJEXP) between client firms with and without audit fee increases, and no significant difference in
mean values. Client firms experiencing increases in audit fees tend to be increasing in size (ΔSIZE), and
have increases in receivables (ΔREC) and inventory (ΔINV). Clients with increasing audit fees also tend
to have increasing leverage (ΔLEV), decreasing ROA (ΔROA), increasing absolute value of abnormal
19
accruals (ΔABSAACC), and new non-standard audit opinions (ΔNSOP) and material weaknesses (ΔMW).
Further, these client firms also tend to have new mergers and acquisition activity (ΔM&A) and new
special items (ΔSPITEM). Surprisingly, however, increasing audit fees appear to be related to a decline in
restructuring rather than an increase in it (ΔRESTRUCT).
Results for the audit fee changes regression are presented in Table 3. The coefficient on the
change in adjusted deficient auditing exposure (ΔADJEXP) is insignificant, indicating a lack of evidence
for changes in audit fees related to exposure to deficient auditing for clients that do not change auditors.
This result suggests that auditors with deficiencies have difficulty increasing audit fees to cover the cost
of additional audit procedures needed to remediate their deficient auditing, but also suggests auditors do
not cut fees in order to retain clients where there is high exposure to deficient auditing.
Other significant variables in the model have the predicted signs. We find audit fee increases
(ΔFEE) are positively related to increases in size (ΔSIZE), receivables (ΔREC), inventory (ΔINV),
leverage (ΔLEV), and the absolute value of abnormal accruals (ΔABSAACC), and negatively related to
return on assets (ΔROA) consistent with these variables increasing audit risk and effort. Further, increase
in audit fees are related to new mergers and acquisitions activity (ΔM&A), new special items (ΔSPITEM),
and new material weaknesses in internal control (ΔMW), which indicates these items also increase audit
risk and fees.
4.2 Auditor Changes
Descriptive statistics for the auditor changes test are provided in Table 4. Limiting the sample to
Big 4 clients and data requirements yields a sample of 11,359 client-inspection report observations. The
inspection reports lead to auditor changes within the lesser of one-year and the next inspection report in
456 observations or four percent of the sample. On average, the client firms in the sample have total
assets (SIZE) of approximately $700 million and long-term-debt-to-asset ratios (LEV) of slightly less than
0.2.
Table 4 Panel B separates the sample into observations with and without auditor changes. The
two groups are significantly different in mean and median for all variables in the auditor change model
20
except for the likelihood of non-standard audit opinions (NSOP) and the auditor’s market share of the
audit clients in the industry (AUDINDSHR). Importantly, the auditor switching observations have
significantly higher values of adjusted audit deficiency exposure (ADJEXP) than observations where the
auditor-client relationship is maintained. The clients switching auditors also are smaller companies
(SIZE) with higher leverage (LEV), higher abnormal audit fees (ABFEE) more material weaknesses in
internal control (MW), and a larger absolute value of abnormal accruals (ABSAACC).
Table 5 displays the results of a logistic regression where the dependent variable equal to one if
there is an auditor switch, and zero otherwise. The coefficient on adjusted deficient auditing exposure
(ADJEXP), the variable of interest, is positive and significant at p < 0.05. This result indicates that
PCAOB reports affect auditor-client relationships by leading to more auditor changes among auditorclient pairs where exposure to PCAOB-identified auditing deficiencies is high. This finding is in contrast
to the conclusions of Lennox and Pitman (2010) who find that the overall number of issues identified in
PCAOB inspection reports is not related to changes in auditor market share. Together, the results suggest
that while PCAOB inspections do not change overall auditor market share, they do play a role in aligning
clients’ accounting with auditors’ abilities.
A number of the control variables also have significant coefficients. Consistent with expectations
there is a significantly positive relation between the likelihood of auditor changes and both non-standard
audit opinions (NSOP) and the disclosure of material weaknesses in internal control (MW). These
relations are consistent with auditors resigning from higher risk clients and clients dismissing auditors in
search of more favorable audit reports. We also find significantly negative relations between auditor
changes and SIZE and the auditor’s market share of the client’s industry (AUDINDSHR). These results
are consistent with auditors wanting to retain their larger clients and with both auditors and clients
wanting to maintain relationships where the auditor has more specialization in the industry.
Table 6 examines how deficient auditing exposure (DEFEXP) changes for clients that switch
from one Big 4 auditor to another. Within the Big 4, auditor changes occur in 157 cases or about 34% of
21
our auditor change observations (with the remaining changes in our sample being from a Big 4 firm to a
non-Big 4 firm). In these cases, the average deficient auditing exposure (DEFEXP) significantly changes
from 6.44 before the switch to 5.83 after the switch (t-stat = -2.63). This result indicates that the new
auditor-client pairings established from an auditor change result in lower deficient auditing exposure,
which is one manner in which deficient auditing exposure is reduced overall.
Overall, the results show that information in PCAOB inspection reports does affect auditor-client
relationships in the Big 4 audit market. Specifically, auditor changes are more likely when an auditorclient pairing creates more exposure to deficient auditing than the average pairing of the client with other
Big 4 auditors. Further, when a new auditor-client pair is established, it has lower deficient auditing
exposure than the prior pairing.
5. Conclusion
PCAOB inspection reports are highly visible critiques of auditors’ work, but many have
questioned whether the reports contain useful information. Our study investigates whether inspections
affect auditor-client relationships by examining the relation between inspection findings and audit fee
changes and auditor changes for Big 4 auditors and their clients. We examine this relation using a new
measure of the information in the inspection reports that accounts for both clients’ use of accounting
standards and their auditors’ deficiencies related to these standards compared to other Big 4 firms’
deficiencies.
We find that our measure of relative exposure to deficient auditing is positively related to auditor
changes, but is not related to changes in audit fees. These results suggest PCAOB inspections affect
auditor changes, but auditors do not have the ability to increase fees to remediate deficient auditing. It
also does not appear, however, that auditors reduce fees to retain clients when they have more
deficiencies in areas important to the clients. Our findings are in contrast to Lennox and Pittman (2010),
who find no relation between the number of deficiencies identified and changes in auditor market share.
22
Our study has import implications for auditing regulation and practice. Our results on PCAOB
inspection reports and audit fees suggest it may be difficult for auditors remediate their deficiencies since
it appears they are not able to pass higher audit costs on to clients. Our findings with respect to auditor
changes indicate that PCAOB inspections do yield useful information and that this information is used to
improve the alignment between auditor expertise and clients’ use of accounting standards. While this
trend may be beneficial in the short term, narrowing audit expertise could have a negative effect on audit
market competition in the future. Further, our study has implications for proposed rules on auditor
rotation. Such a requirement, if implemented, could have a negative effect on gains in audit quality
resulting from auditor-client realignment.
23
References
Abbott, L., K. Gunny, and T. Zhang. 2013. When the PCAOB talks, who listens? Evidence from
stakeholder reaction to GAAP-deficient PCAOB inspection reports of small auditors. Auditing: A
Journal of Practice & Theory 32(2): 1-31.
Carcello, J.V., C. Hollingsworth, and S.A. Mastrolia. 2011. The effect of PCAOB inspections on Big 4
audit quality. Research in Accounting Regulation 23 (2): 85-96.
Church, B.K. and L.B. Shefchik. 2012. PCAOB inspections and large accounting firms. Accounting
Horizons 26 (1): 43-63.
DeFond, M.L., 2010. How should the auditors be audited? Comparing the PCAOB Inspections with the
AICPA Peer Reviews. Journal of Accounting & Economics 49 (1/2): 104-108.
DeFond, M.L. and C. Lennox. 2011. The effect of SOX on small auditor exits and audit quality. Journal
of Accounting & Economics 52 (1): 21-40.
Department of the Treasury. 2008. Advisory Committee on the Auditing Profession. (October 6).
Ettredge, M.L., C. Li, and S. Scholz. 2007. Audit Fees and Auditor Dismissals in the Sarbanes-Oxley Era.
Accounting Horizons 21 (4): 371-386.
Farrell, J., and H. Shabad. 2005. The focus of future PCAOB auditor inspections. CPA Journal 75 (6), 9.
Fields, L.P., D. Fraser, and M.S. Wilkins. 2004. An investigation of the pricing of audit services for
financial institutions. Journal of Accounting and Public Policy 23(1): 53-77.
Folsom, D., P. Hribar, R. Mergenthaler, and K. Peterson. 2012. Principles-Based Standards and Earnings
Attributes. Working Paper, University of Iowa.
Glover, S.M., D.F. Prawitt, and M. H. Taylor. 2009. Auditor standard setting and inspection for U.S.
public companies: A critical assessment and recommendations for fundamental
change. Accounting Horizons 23 (2): 221-237.
Gradison, B. and R. Boster. 2010. The PCAOB’s first seven years: A retrospection. Current Issues in
Auditing 4(1): A9-A20.
Gramling, A., J. Krishnan, and Y. Zhang. 2011. Are PCAOB identified audit deficiencies associated with
a change in reporting decisions of triennially inspected audit firms? Auditing: A Journal of
Practice and Theory 30 (3)59-79.
Gunny, K.A. and T.C. Zhang. 2013. PCAOB inspection reports and audit quality. Journal of Accounting
and Public Policy 32: 136-160.
Hillary, G. and C. Lennox. 2005. The credibility of self-regulation: Evidence from the accounting
profession’s peer review program. Journal of Accounting and Economics 40: 211-229.
24
Hogan, C.E and M.S. Wilkins. 2008. Evidence on the Audit Risk Model: Do Auditors Increase Audit
Fees in the Presence of Internal Control Deficiencies? Contemporary Accounting Research 25(1):
219-242.
Hribar, P., T. Kravet, and R. Wilson. 2010. A New Measure of Accounting Quality. Working Paper, The
University of Iowa.
Lamoreaux, P.T. 2013. Does PCAOB Inspection Exposure Affect Auditor Reporting Decisions? Working
Paper, Arizona State University.
Lennox, C., and J. Pittman. 2010. Auditing the auditors: Evidence on the recent reforms to the external
monitoring of audit firms. Journal of Accounting & Economics 49 (1/2): 84-103.
Public Company Accounting Oversight Board (PCAOB). 2008. Report on the PCAOB’s 2004, 2005,
2006, and 2007 Inspections of Domestic Annually Inspected Firms. PCAOB Release No. 200808. Washington DC.: PCAOB.
Public Company Accounting Oversight Board (PCAOB). 2010. Report on Observations of PCAOB
Inspectors Related to Audit Risk Areas Affected by the Economic Crisis. PCAOB Release No.
2010-006. Washington DC.: PCAOB.
Public Company Accounting Oversight Board (PCAOB). 2012. Information for Audit Committees about
the PCAOB Inspection Process. PCAOB Release No. 2012-003. Washington DC.: PCAOB.
Ragothaman, S. 2012. Does Watching the Watchdogs: An Examination of the PCAOB Quality Control
Inspection Reports on Triennially Inspected Audit Firms and the AICPA Peer Review Reports.
Working Paper, The University of South Dakota.
Whitehouse, T. 2012. PCAOB Offers Audit Committees Hints on How to Use Inspection Findings.
Compliance Week. August 1.
25
Appendix A. Keyword Search Terms by Standard
This appendix contains a partial listing of keyword search terms from Folsom, Hribar, Mergenthaler and Peterson (2012) used to
create DEFEXP, ALTEXP, and ADJEXP. In our textual analysis, we also include additional variations of these keywords and the
name of each standard. A complete list is available from the authors. Any keywords with “/#/” signifies the number of adjacent
words we search for relevant terms. For example, “Warrant /5/ Debt” signifies we search for keyword “warrant” within five words
(forwards or backwards) of the keyword “debt.” The “|” reflects an ‘OR’ for the associated keywords within parentheses. For example,
“(stock|share)-based compensation” would match either “stock-based compensation” or “share-based compensation”.
Standard
APB 2
APB 4
Keyword #1
investment credit* /10/
tax*
investment credit* /10/
tax*
Keyword #2
tax /10/ deferral method
Keyword #3
Keyword #4
allowable investment
credit
tax* /10/ flow-through
method
APB 9
extraordinary items
extraordinary gain
extraordinary loss
APB 14
warrant /5/ debt
convertible /5/ debt
stock purchase warrant*
conversion option /5/
debt
APB 16
business combination
merge* /5/ pool*
acqui* /5/ pool
merge* /5/ purchase
APB 17
goodwill
intangible asset*
goodwill /5/ amortiz*
APB 18
equity method
significant influence
share of earnings
share of loss(es)
change in accounting
principle
non[-]interest bearing /2/
note
invest* /5/ permanent /5/
foreign /5/ tax
stock-based
compensation
early /5/ extinguish* /5/
debt
non[-]monetary
transaction
change in accounting
estimate
non[-]interest bearing /2/
receivable
undistributed earnings /5/
subsidiar*
change in reporting entity
error /5/ previously
issued financial statement
impute* /2/ interest
note /5/ discount
accounting for income
taxes /3/ special areas
unremitted earnings /5/
subsidiar*
restricted stock /5/ grant
option(s) /5/ issue
non[-]monetary exchange
non[-]reciprocal transfer
APB 30
discontinued operations
extraordinary items
disposal /5/ segment
ARB 43 Ch. 3a
current (asset|liabilit*) /5/
classif* /5/ year
current (asset|liabilit*) /5/
classif* /5/ operating
cycle
working capital
ARB 43 Ch. 3b
right (of|to) setoff
right (of|to) offset
ARB 43 Ch. 4
lower of cost or market
/1/ impairment
inventory /1/ impairment
ARB 43 Ch. 7a
quasi-reorganization
corporate readjustment
ARB 43 Ch. 7b
stock dividends
stock split*
APB 20
APB 21
APB 23
APB 25
APB 26
APB 29
option(s) /5/ grant
early /5/ extinguish* /5/
liabilit*
inventory pricing
split-ups
unusual /5/ infrequent*
firm purchase
commitment
Appendix A. Keyword Search Terms by Standard (Cont.)
Standard
Keyword #1
Keyword #2
Keyword #3
Keyword #4
ARB 43 Ch. 9a
depreciation
ARB 43 Ch. 9b
Depreciation /5/
Appreciat* Asset*
depreciation /5/
appreciat*
depreciation /15/ quasireorganization
ARB 43 Ch. 10a
real estate taxes
property taxes
real estate /5/ tax*
property /5/ tax*
war /5/ contract /5/
terminat*
defense /5/ contract /5/
terminat*
war and defense contract
/5/ terminat*
ARB 43 Ch. 11a
ARB 43 Ch. 11b
ARB 43 Ch. 11c
ARB 43 Ch. 12
cost plus /5/ fee* /5/
contract*
government contract /5/
renegotiat*
fixed fee /5/ war
contract* /5/ terminat*
foreign earnings /5/
presentation
ARB 45
percentage of completion
long term construction
construction /2/ progress
cost /2/ excess /2/ billings
ARB 51
consolidat* /5/ financial
statement
intercompany /5/
eliminat*
controlling financial
interest
minority interest
Concepts 5 & 6
earned /5/ revenue
earned /5/ sales
realizable future benefit
probable future benefit
SFAS 2
research and develop*
research /5/ cost*
SFAS 5
conting* liab*
conting* gain
conting* /5/ loss
SFAS 7
develop* stage /10/
enterpris*
develop* stage /10/ corp*
develop* stage /10/
company
SFAS 13
lease
bargain purchase option
bargain renewal option
SFAS 15
trouble* debt restruc*
debt /5/ restruc*
debt restruct /5/ settle*
debt /5/ modifi*
SFAS 16
prior period adjustment
SFAS 19
exploration
mineral rights
proved reserves
unproved reserves
SFAS 34
interest /3/ capitaliz*
self-constructed asset /5/
debt
self-constructed asset /5/
interest
SFAS 35
defined benefit /5/
pension
defined benefit /5/ plan
SFAS 43
compensat* absence*
vacation Accru*
sick accru*
SFAS 45
franchise fee
franchise /5/ sales
franchise /5/ revenue
27
conting* /5/ reasonably
possible
planned principal
operations /5/
commenced
transfer* ownership /5/
lesee
illness accru*
Appendix A. Keyword Search Terms by Standard (Cont.)
Standard
Keyword #1
Keyword #2
Keyword #3
SFAS 47
purchase commitment
purchase obligation
SFAS 48
revenue /5/ right of return
sales /5/ right of return
SFAS 49
sale of inventory /5/
financing arrange*
SFAS 50
record* industry
music industry
SFAS 51
cable television
cable /5/ hookup
SFAS 52
reporting currency
foreign currency
SFAS 53
motion picture
license /5/ film
SFAS 57
related part*
SFAS 60
insurance contract
SFAS 61
title plant
SFAS 63
broadcasting industr*
SFAS 65
mortgage loans
SFAS 66
sale* /5/ real estate
retail land sale*
SFAS 67
capitaliz* /5/ real estate
/5/ acquisition
SFAS 68
fund* /5/ research and
development
SFAS 71
cost-based rates /3/
regulat*
SFAS 77
Keyword #4
long term commitment
long term obligation
music /5/ advance royalty
record /5/ advance
royalty
functional currency
translation adjustment
insurance /3/ shortduration /2/ contract*
insurance /3/ longduration /2/ contract*
insurance /2/ claim* cost
network affiliation
agreement*
mortgage-backed
securities
exhibition rights /5/
license agreement*
loan fees
commitment fees
capitaliz* /5/ real estate
/5/ develop*
research and
development
arrangement*
accounting for the effects
of certain types of
regulation
capitaliz* /5/ real estate
/5/ construction
research and
development /5/
obligation*
capitaliz* /5/ real estate
/5/ sale
regulat* asset*
regulat* liabilit*
receivable /5/ recourse
transfer /5/ receivable
transferor /5/ receivable
SFAS 80
futures contract
futures /5/ hedge
SFAS 86
technological feasibility
/5/ software
SFAS 87
Pension
SFAS 88
settlement /5/ defined
benefit (pension|plan)
internal* /5/ develop* /5/
software
projected benefit
obligation
curtailment /5/ defined
benefit (pension|plan)
28
cost /5/ software /5/ sold
accumulated benefit
obligation
termination benefit* /10/
employ*
capitalized software cost
funding of plan assets
Appendix A. Keyword Search Terms by Standard (Cont.)
Standard
SFAS 97
SFAS 101
SFAS 105
SFAS 106
SFAS 107
Keyword #1
universal life type
discontinuation of
application of FASB
statement no. 71
disclos* /5/ financial
instrument*
post[-]retirement benefits
other than pensions
disclos* /5/ financial*
instrument* /5/ fair value
Keyword #2
Keyword #3
Keyword #4
long-duration contract
/10/ mortality
cease* to meet the
criteria /5/ SFAS [No.]
71
disclos* /5/ off balance
sheet risk
post[-]retirement health
care benefit
retrospective deposit
method
post[-]retirement benefit
plan
post[-]retirement /5/
health care
tax asset
tax /5/ temporary
difference
limited payment contract
regulated enterprise
disclos* /5/ credit risk
SFAS 109
income tax*
tax liability
SFAS 113
reinsurance
retrocession
SFAS 115
available-for-sale /5/
securit*
trading /5/ securit
held-to-maturity /5/
securit*
other than temporary
impairment /10/
investment
SFAS 116
accounting for
contributions received
and contributions made
nonreciprocal transfer
donor imposed restriction
donor imposed condition
SFAS 119
disclos* /5/ derivative
disclos* /5/ futures
contract
disclos* /5/ forward
contract
disclos* /5/ swap
contract
SFAS 121
impair* /5/ long-lived
dispos* /5/ long-lived
(stock|share)-based
compensation
(stock|share)-based
compensation
option(s) /5/
(grant*|issue*|award*)
option(s) /5/
(grant*|issue*|award*)
restricted stock /5/
(grant*|issue*|award*)
restricted stock /5/
(grant*|issue*|award*)
SFAS 125
transfer /5/ financ* asset*
servic* /5/ financ* asset*
extinguish* /2/ liabilit*
transfer /5/ receivable
SFAS 130
comprehensive income
SFAS 133
derivativ*
hedg*
underlying /5/ notional
amount*
put /2/ option
SFAS 140
transfer* /5/ financ*
asset*
servic* /5/ financ* asset*
extinguish* /2/ liabilit*
extinguish* /2/ debt
SFAS 141
business combination
merge* /5/ purchase
acqui* /5/ purchase
merge* /5/ contingent
consideration
SFAS 142
Goodwill
intangible asset*
implied fair value /5/
carrying amount
impair* /5/ goodwill
SFAS 143
asset* retirement
obligation
SFAS 144
impair* /5/ long-lived
SFAS 123
SFAS 123r
dispos* /5/ long-lived
29
grant date
grant date
Appendix A. Keyword Search Terms by Standard (Cont.)
Standard
SFAS 146
SFAS 150
SFAS 154
EITF 94-03
EITF 00-21
SOP 97-2
SAB 101
Keyword #1
restruct*
(exp*|charg*|activit*|rese
rv*)
instruments with
characteristics of both
liabilities and equity
change in accounting
principle
Keyword #2
Keyword #3
exit or disposal activit*
Keyword #4
exit /5/ disposal activity
/5/ termination benefits
freestanding financial
instrument
change in accounting
estimate
change in reporting entity
error /5/ previously
issued financial statement
restruct* exp*
restruct* charg*
restruct* activit*
restruct* reserv*
revenue /10/ multiple
deliverables
multiple element /5/
software
persuasive evidence /5/
arrangement
multiple deliverable
arrangement*
objective evidence /2/
element* /5/ fair value
persuasive evidence /5/
agreement
direct cost /10/ multiple
deliverable
vendor specific objective
evidence
unit /5/ value /5/ stand
alone basis
software /10/ revenue
recognition
fee /2/ fixed /2/
determinable
30
delivery /5/ occur*
Appendix B. Selected PCAOB Identified Audit Deficiencies
SFAS 131 – Ernst and Young 2004
In this audit, the Firm failed to identify a departure from GAAP that it should have identified and
addressed before issuing its audit report. The issuer disclosed in the notes to its financial statements two
reportable segments despite the presence of information that indicated the issuer was organized in more
than two reportable segments as defined by SFAS No. 131, Disclosures about Segments of an Enterprise
and Related Information. The improper aggregation of reporting segments resulted in the offsetting of
operating profits at one segment with losses at another.
SFAS 57 – KPMG 2005
In addition, the issuer failed to disclose in its financial statements (1) certain lease arrangements; (2)
information concerning an impaired asset and the issuer's method for determining the fair value of the
asset; and (3) information concerning related party transactions between the issuer and entities in which
its chairman and chief executive officer had an ownership interest. Omitting these disclosures was
inconsistent with, respectively, SFAS No. 13, Accounting for Leases; SFAS No. 144, Accounting for the
Impairment or Disposal of Long-Lived Assets; and SFAS No. 57, Related Party Disclosures.
SFAS 123R – PriceWaterhouseCoopers 2006
The Firm failed to test the fair value of warrants and stock-based compensation issued in two significant
transactions during the year.
SFAS 142 – Deloitte & Touche LLP 2007
The issuer performed its annual impairment test of goodwill as of an interim date and, during 2005,
elected to carry forward the fair values of all but one of its reporting units that it had used in its 2004
impairment test, as permitted under certain conditions by Statement of Financial Accounting Standards
("SFAS") No. 142, Goodwill and Other Intangible Assets ("SFAS No. 142"). The issuer asserted that the
conditions for doing so were satisfied, including the condition that the likelihood that the fair values of its
business units had declined below their book values was remote. The Firm failed to test that assertion
sufficiently (limiting its procedures to obtaining memoranda from management and making inquiries of
management), despite the following factors: (1) a decline in the issuer's market capitalization by over 25
percent between the annual goodwill impairment tests; (2) the shut down of certain operating assets and
the resulting impairment charges at one business unit; (3) the sales of certain long-lived assets during
2005; and (4) the reduction in the issuer's debt rating during 2005. In addition, despite the fact that the
issuer was exploring options to sell or spin off certain business units, the Firm failed to evaluate
sufficiently (limiting its procedures to obtaining memoranda from management and making inquiries of
management) whether the issuer's determination of the fair values of those business units appropriately
took into account whether all or any portion of those business units was more likely than not to be sold or
disposed of.
31
Appendix C. Variable Definitions
Variable
FEE
Definition
Change in variable from fiscal year t-1 to fiscal year t.
Abnormal audit fees calculated as the residual from an OLS regression of the natural logarithm
of audit fees on size, sales, inventory and receivables, debt, number of business segments,
existence of a loss, existence of a going concern issue, timing of the fiscal year end and audit
work, industry, and year determinants. Data are from Audit Analytics and COMPUSTAT.
Absolute value of performance-adjusted abnormal accruals. Abnormal accruals are calculated
using a cross-sectional modified Jones model, estimated by two-digit SIC code. Performance
adjustment subtracts median abnormal accruals from companies in the same decile of prioryear return on assets, two-digit SIC code, and year. Data are from COMPUSTAT.
DEFEXP minus ALTEXP.
The average of DEFEXP calculated for each of the other Big 4 accounting firms as if they were
auditing the client in the corresponding fiscal year.
Indicator variable equal to one if the company announces a change in auditor in the fiscal year,
and zero otherwise.. Data are from Audit Analytics.
Percent of the square root of client assets audited by the firm in the two-digit standard industry
code. Data are from COMPUSTAT.
The sum of the products of the relative impact of each accounting standard for a company and
an indicator for whether the PCAOB finds a related issue for the company's auditor. The
relative impact of an accounting standard for a company is calculated by standardizing the
count of keywords related to the standard in the company's 10-K filing. Whether the PCAOB
finds a related issue is determined by the presence of keywords in Section I.A of the PCAOB
inspection report for the fiscal year corresponding to the 10-K filing.
Natural logarithm of audit fees. Data are from Audit Analytics.
INV
Total inventory scaled by total assets (INVT/AT). Data are from COMPUSTAT.
LEV
M&A
Total long-term debt (DLTT) divided by total assets (AT). Data are from COMPUSTAT.
Indicator variable equal to one if the company has merger and acquisition activity, as indicated
by the sales footnote (SALE_FN), and zero otherwise. Data are from COMPUSTAT.
Indicator variable equal to one if the company has a material weakness in internal control
disclosed as required by SOX 404, and zero otherwise. Data are from COMPUSTAT.
Indicator variable equal to one if the company has a non-standard audit opinion, and zero
otherwise. Data are from COMPUSTAT.
Total receivables scaled by total assets (RECT/AT). Data are from COMPUSTAT.
Indicator variable equal to one if the company is undergoing restructuring, as indicated by the
disclosure of restructuring costs (RCA, RCP, RCEPS, RCD), and zero otherwise. Data are
from COMPUSTAT.
Return on assets (IB/AT) for the fiscal year. Data are from COMPUSTAT.
Natural logarithm of one plus the number of business segments disclosed (BUSSEG). Data are
from COMPUSTAT.
Natural logarithm of total assets (AT). Data are from COMPUSTAT.
Indicator variable equal to one if the company has a special item (SPI) in the fiscal year, and
zero otherwise. Data are from COMPUSTAT.
Δ
ABFEE
ABSAACC
ADJEXP
ALTEXP
AUDCH
AUDINDSHR
DEFEXP
MW
NSOP
REC
RESTRUCT
ROA
SEGS
SIZE
SPITEM
32
Figure 1. Deficient Auditing Exposure Time Trend
This figure displays the trend of deficient auditing exposure (DEFEXP), the exposure calculated if alternative Big 4 auditors
were used (ALTEXP), and the adjusted audit exposure (ADJEXP) over the client-fiscal years for which inspections were
performed. The figure uses the means the variables in each fiscal year. The figure is based on A 11,359 client-inspections for
the Big 4 audit firms.
7.00
6.00
5.00
4.00
DEFEXP
3.00
ALTEXP
ADJEXP
2.00
1.00
0.00
-1.00
2003
2004
2005
2006
33
2007
2008
Table 1. Big 4-Firm Auditing Deficiencies
This table reports on the number of Big 4 firms with deficiencies related to each accounting standard for each fiscal year on
which the inspections took place. Only accounting standards related to at least one PCAOB-identified audit deficiency are
displayed in the figure.
Standard
SFAS 142
SFAS 141
SFAS 109
SFAS 13
SFAS 133
SOP9 7-2
SFAS 5
SFAS 65
SFAS 144
APB 18
SFAS 50
ARB 43 Ch. 9a
SFAS 2
SFAS 107
SFAS 115
SFAS 131
SFAS 87
SFAS 123R
SFAS 52
SFAS 146
SAB 101
APB 14
APB 20
APB 23
SFAS 19
SFAS 48
SFAS 57
SFAS 86
SFAS 130
SFAS 143
2003
3
3
2
3
2
0
0
0
1
0
0
2
0
0
0
2
0
0
1
1
0
0
1
1
1
0
1
0
1
0
2004
4
2
2
2
0
1
0
0
0
1
0
1
1
0
0
0
0
1
1
0
0
1
0
0
0
0
0
0
0
1
Inspection Fiscal Year
2005
2006
4
3
2
2
2
1
1
1
1
1
0
2
0
1
0
2
1
1
0
1
1
1
0
0
1
0
0
1
0
1
1
0
0
0
1
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
34
2007
4
2
2
0
1
2
2
1
0
0
1
0
1
1
1
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
2008
2
1
2
0
2
0
1
1
1
1
0
0
0
1
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
Total
20
12
11
7
7
5
4
4
4
3
3
3
3
3
3
3
2
2
2
2
1
1
1
1
1
1
1
1
1
1
Table 2. Descriptive Statistics for Audit Fee Changes
This table reports descriptive statistics for client firm-year observations used in the audit fee changes test. In Panel B, the mean tstatistics are from two-sample t-tests and median Z-statistics are from a Wilcoxon two-sample test. Significance at the 10%, 5%,
and 1% levels are denoted *, **, and ***. All continuous variables are Winsorized at the 1% and 99% level. Variable definitions
appear in Appendix C.
Panel A: Descriptive Statistics
N
Mean
StDev
1Q
Median
3Q
ΔFEE
6,769
0.006
0.217
-0.097
0.003
0.094
ΔADJEXP
6,769
-0.085
3.889
-2.506
-0.094
2.529
ΔSIZE
6,769
0.032
0.228
-0.057
0.035
0.125
ΔROA
6,769
-0.005
0.405
-0.038
-0.001
0.029
ΔABSAACC
6,769
0.000
0.094
-0.032
0.000
0.031
ΔSEGS
6,769
-0.015
0.806
0.000
0.000
0.000
ΔNSOP
6,769
-0.053
0.546
0.000
0.000
0.000
ΔICD
6,769
-0.013
0.224
0.000
0.000
0.000
ΔLEV
6,769
0.002
0.102
-0.024
0.000
0.017
ΔREC
ΔINV
ΔM&A
ΔSPITEM
ΔRESTRUCT
6,769
6,769
6,769
6,769
6,769
0.007
0.004
-0.010
0.021
0.000
0.051
0.035
0.478
0.498
0.032
-0.011
-0.003
0.000
0.000
0.000
0.002
0.000
0.000
0.000
0.000
0.021
0.011
0.000
0.000
0.000
Panel B: Comparison of Positive and Non-Positive Audit Fee Changes Observations
ΔFEE
ΔADJEXP
ΔSIZE
ΔROA
ΔABSAACC
ΔSEGS
ΔNSOP
ΔICD
ΔLEV
ΔREC
ΔINV
ΔM&A
ΔSPITEM
ΔRESTRUCT
Positive Fee Change
(N=3,429)
Mean
Median
0.152
0.093
-0.021
0.051
0.075
0.063
-0.018
-0.003
0.006
0.000
-0.026
0.000
-0.009
0.000
-0.002
0.000
0.009
0.000
0.014
0.006
0.009
0.000
0.044
0.000
0.040
0.000
-0.001
0.000
Non-Positive Fee Change
(N=3,340)
Mean
Median
-0.145
-0.098
-0.151
-0.235
-0.013
0.011
0.008
0.001
-0.005
-0.001
-0.004
0.000
-0.099
0.000
-0.024
0.000
-0.006
0.000
-0.001
0.000
-0.001
0.000
-0.065
0.000
0.001
0.000
0.001
0.000
35
Between Sample Tests
Mean
Median
77.04 *** 71.24 ***
1.37
1.68 *
16.12 *** 16.64 ***
-2.68 *** -5.26 ***
5.12 ***
4.22 ***
-1.11
-0.35
6.74 ***
6.61 ***
4.08 ***
4.06 ***
5.82 ***
6.25 ***
11.84 *** 13.77 ***
12.23 *** 12.40 ***
9.46 ***
9.35 ***
3.23 ***
3.25 ***
-1.90 *
-1.88 *
Table 3. Model of Audit Fee Changes
This table reports the results of a client firm-year ordinary-least-squares (OLS) regression with audit fee change (ΔFEE) as the
dependent variable. The sample is comprised of 6,769 client-year observations. Significance at the 10%, 5%, and 1% levels are
denoted *, **, and ***. All continuous variables are Winsorized at the 1% and 99% level. Variable definitions appear in
Appendix C.
Parameter
Prediction
Coefficient
?
+/+
0.070
0.001
0.183
ΔROA (B3)
ΔABSAACC (B4)
ΔSEGS (B5)
ΔNSOP (B6)
ΔICD (B7)
ΔLEV (B8)
ΔREC (B9)
ΔINV (B10)
ΔM&A (B11)
ΔSPITEM (B12)
+
+
+
+
+
+
+
+
+
-0.021
0.117
0.005
0.008
0.071
0.103
0.261
0.457
0.031
0.025
ΔRESTRUCT (B13)
+
-0.104
Intercept (B0)
ΔADJEXP (B1)
ΔSIZE (B2)
Year Fixed Effects
t-Statistic
7.47 ***
1.08
11.09 ***
-2.48 **
3.57 ***
1.49
1.55
4.04
3.29
3.74
5.34
5.14
5.06
-1.36
Included
N
Clusters
6,769
2,115
R2
0.20
36
***
***
***
***
***
***
Table 4. Descriptive Statistics for Auditor Changes
This table reports descriptive statistics for client firm-inspection report observations used in the audit fee changes test. In Panel
B, the mean t-statistics are from two-sample t-tests and median Z-statistics are from a Wilcoxon two-sample test. Significance at
the 10%, 5%, and 1% levels are denoted *, **, and ***. All continuous variables are Winsorized at the 1% and 99% level.
Variable definitions appear in Appendix C.
Panel A: Descriptive Statistics
N
Mean
StDev
1Q
Median
3Q
AUDCH
11,359
0.040
0.196
0.000
0.000
0.000
ADJEXP
11,359
-0.120
3.212
-2.120
-0.147
1.643
SIZE
11,359
6.533
1.810
5.297
6.498
7.706
ABFEE
11,359
0.081
0.503
-0.255
0.082
0.408
NSOP
11,359
0.608
0.488
0.000
1.000
1.000
MW
11,359
0.061
0.239
0.000
0.000
0.000
LEV
11,359
0.187
0.211
0.001
0.136
0.290
ABSAACC
11,359
0.059
0.078
0.012
0.033
0.073
AUDINDSHR
11,359
0.247
0.085
0.190
0.234
0.299
Panel B: Comparison of Auditor Switching and Non-Auditor Switching Observations
ADJEXP
SIZE
ABFEE
NSOP
MW
LEV
ABSAACC
AUDINDSHR
Auditor Switch
(N=456)
Mean
Median
0.288
0.138
5.554
5.495
0.131
0.153
0.621
1.000
0.147
0.000
0.171
0.106
0.079
0.041
0.240
0.228
No Auditor Switch
(N=10,903)
Mean
Median
-0.137
-0.159
6.573
6.542
0.079
0.080
0.607
1.000
0.057
0.000
0.188
0.137
0.058
0.033
0.247
0.235
37
Between Sample Tests
Mean
Median
2.77 ***
2.86 ***
-11.86 *** -11.33 ***
1.84 *
2.26 **
0.58
0.58
5.35 ***
7.82 ***
1.70 *
2.17 **
4.31 ***
3.67 ***
-1.61
-1.44
Table 5. Model of Auditor Changes
This table reports the results of tests of a client-firm clustered logistic regression with auditor change (AUDCH) as the dependent
variable. The sample is comprised of 11,359 client firm-inspection report observations, 456 of which switch auditors
(AUDCH=1) and 10,903 of which do not switch auditors (AUDCH = 0). Significance at the 10%, 5%, and 1% levels are denoted
*, **, and ***. All continuous variables are Winsorized at the 1% and 99% level. Variable definitions appear in Appendix C.
Parameter
Prediction
Coefficient
Intercept (B0)
ADJEXP (B1)
?
+
1.715
0.037
SIZE (B2)
ABFEE (B3)
+
-0.441
0.113
NSOP (B4)
MW (B5)
LEV (B6)
ABSAACC (B7)
AUDINDSHR (B8)
+
+
+
+
-
0.373
0.794
0.182
0.760
-1.084
Industry Fixed Effects
Year Fixed Effects
Chi-Square
56.08 ***
4.24 **
285.09 ***
1.20
11.16 ***
25.09 ***
0.68
1.98
3.89 **
Included
Included
N
Clusters
Psuedo-R2
11,359
3,053
0.66
38
Table 6. Auditing Exposure Change from Auditor Changes
This table reports the results of a paired t-test of differences in deficient auditing exposure before and after auditor changes. The
sample is comprised of 157 client-firms that change from one Big 4 auditor to another. Significance at the 10%, 5%, and 1%
levels are denoted *, **, and ***. Variable definitions appear in Appendix C.
Auditor-Client Pairing
New Pairing
Prior Pairing
Difference
Paired t-Test
DEFEXP
5.38
6.44
-1.06
-2.63 **
N
157
39