Audit Committee Characteristics and Financial Misstatement:

A Study of the Efficacy of Certain

Blue Ribbon Committee Recommendations

Lawrence J. Abbott
Assistant Professor
University of Memphis
(corresponding author)
ljabbott@memphis.edu

Susan Parker
Assistant Professor
Santa Clara University
sparker@scu.edu

Gary F. Peters
Assistant Professor
University of Georgia
Gpeters@terry.uga.edu

March, 2002

We wish to express our sincere thanks to two anonymous reviewers and to the editor for
numerous helpful comments and suggestions. We also wish to thank Dana Hermanson
and participants of the 2001 American Accounting Association Auditing Section
Meeting.

Audit Committee Characteristics and Financial Misstatement:

A Study of the Efficacy of Certain
Blue Ribbon Committee Recommendations

SUMMARY

This study addresses the impact of certain audit committee characteristics

identified by the Blue Ribbon Committee on Improving the Effectiveness of Corporate
Audit Committees (BRC) on the likelihood of financial misstatement. We examine 41
firms which issued fraudulent reports and 88 firms which restated annual results (without
allegations of fraud) in the period 1991-1999, together with matched pairs control groups
of similar size, exchange listing, industry and auditor type.
We find that the independence of the audit committee and whether the committee
meets at least four times per year exhibit a significant and negative association with the
occurrence of financial reporting restatements. We also document a significant positive
association between an audit committee that lacks a member with financial expertise and
the occurrence of financial reporting restatements. However, only audit committee
independence and the lack of financial expertise exhibit a negative (positive) association
with financial reporting fraud. Our results underscore the importance of the BRCs first
two recommendations, both of which concern audit committee independence. Our results
also highlight the need for financial expertise as a means of strengthening the monitoring
and oversight role that the audit committee plays in the financial reporting process.
Keywords: Blue Ribbon Committee; Audit Committee; Corporate Governance; Fraud;
Financial Restatements.
Data Availability: All data are available from public sources.

INTRODUCTION
In response to apparent increases in financial reporting misstatements and at the
behest of the Securities and Exchange Commission (SEC), the New York Stock
Exchange (NYSE) and the National Association of Securities Dealers (NASD) formed
the Blue Ribbon Committee on Improving the Effectiveness of Corporate Audit
Committees (BRC).1 The BRCs purpose was to develop recommendations aimed at
improving financial reporting by strengthening the audit committees role as a financial
monitor. In October 1999, the BRC issued a summary statement of its recommendations
(BRC 1999). The BRCs recommendations were championed by the National
Association of Corporate Directors (NACD 1999) and by then SEC Chairman Arthur
Levitt (Levitt 1998), who stated Qualified, committed independent and tough-minded
audit committees represent the most reliable guardians of the public interest.
The BRCs recommendations address audit committee member independence,
financial literacy and expertise, and committee size, as well as increased audit committee
autonomy over external auditor engagement. The BRC also made several
recommendations related to audit committee charters and communications with the
external auditor and shareholders. The SEC, major stock exchanges and the AICPA
subsequently adopted certain BRC recommendations regarding audit committees.2 The

The number of class action lawsuits involving financial reporting misstatements has grown rapidly since
the passage of the 1995 Private Security Litigation Act. Furthermore, approximately 60% of post-1995
suits allege accounting irregularities. The numbers of companies named in suits of this type were 114 in
1996, 181 in 1997, 244 in 1998, 216 in 1999, and 117 in the first seven months of 2000. These figures are
taken from the Stanford Securities Class Action Clearinghouse (http:// securities.Stanford.edu).
2
Our focus in this paper is on testing the efficacy of the regulatory system recommended by the BRC.
However, we include an appendix which details the differences between the BRC recommendations and the
regulations proscribed by the NYSE, Nasdaq and SEC.
1

SEC rules became effective December 16, 2000.3 The purpose of this study is to
examine the efficacy of the BRC recommendations (to the extent possible using archival
data) in the context of financial reporting misstatement.
In examining our research question, we investigate whether firms that voluntarily
adopted the recommendations of the BRC were less likely to experience financial
reporting misstatements. Specifically, we examine 41 firms which issued fraudulent
reports and 88 firms which restated annual results (without an allegation of fraud by the
SEC) in the period 1991 to 1999 i.e. the period before the effective implementation date
of the BRC recommendations. Following Beasley (1996) we construct an exchange-,
industry-, size-, and time-matched control set of firms for both samples. We then
combine these into an overall sample of 129 misstatement firms and 129 control firms.
We then study the association between certain BRC recommendations (audit committee
independence, financial expertise, minimum audit committee size, and external auditor
engagement rights) and the likelihood of financial reporting misstatement.4 Because the
BRC recommends audit committee oversight of the external auditors quarterly financial
statement review, we also test the impact of a threshold level of four audit committee
meetings. Because other BRC recommendations, such as those related to communications
with external auditors and other audit committee duties disclosed in charters, cannot be

These rules replace older exchange-listing rules which required the disclosure of certain audit committee
member affiliations and number of meetings. These rules are effective for proxy statements filed after
December 15, 2000, and are contained in 17 CFR Parts 210, 228, 229 and 240, Release No. 34-42266; File
No. S7-22-99 (http://www.sec.gov/rules.html).
4
We also test the fraud and restatement samples separately.
2

determined through public disclosures during the sample period, they are not tested in
this study.5
We find that the audit committee independence and activity (whether the
committee meets at least four times per year) exhibit a significant and negative
association with the occurrence of financial reporting restatements. We also find a
significant and positive association between the occurrence of financial reporting
restatement and the lack of financial expertise (the presence of a financial expert) on the
audit committee. Only audit committee independence and the lack of financial expertise
exhibit a significant negative (positive) association with financial reporting fraud.
Empirically, our results extend the univariate tests of Beasley, Carcello, Hermanson and
Lapides (BCHL 2000) into a large sample, multivariate environment to determine which
particular audit committee characteristics determine audit committee effectiveness. We
also extend BCHL by examining audit committee size, financial expertise and external
auditor engagement rights. Our study supports the contentions of the BRC regarding the
importance of independent, knowledgeable and diligent audit committee members.
The remainder of this paper is organized as follows: The next section reviews
prior literature and develops hypotheses, followed by discussions of sample selection,
methodology, results and limitations. The final section details our policy implications and
conclusions.

Information on audit committee charters is typically obtained through survey. However, because our
study covers a period of almost a decade, and because many misstatement firms are no longer in existence,
we believe that surveying firms on this historical information is not feasible.
3

HYPOTHESIS DEVELOPMENT
Prior Literature
While the presence of an audit committee has been required of firms trading on
the three major exchanges since 1992, it is unclear whether the mere presence of an audit
committee ensures effective discharge of its oversight duties.6 Prior studies have not
extensively addressed the audit committee characteristics recommended by the BRC. For
example, Beasley (1996) finds that only characteristics of the full board are associated
with the likelihood of fraudulent reporting, while the presence of an audit committee or
the proportion of outsiders serving on the committee are not significant. However,
McMullen (1996) and Dechow et al. (1996) find evidence that suggests the presence of
an audit committee is a factor in an overall monitoring environment associated with less
aggressive reporting.
Abbott et al. (2000) attempt to address audit committee effectiveness by replacing
the dichotomous audit committee presence variable with a single variable encompassing
audit committee independence and activity. They argue that Beasleys (1996) failure to
find audit committee existence as a deterrent to fraud is one of form (audit committee
presence) versus substance (audit committee characteristics). Beasley, Carcello,
Hermanson and Lapides (BCHL 2000) further this line of research by examining audit
committee composition and activity separately. BCHL limit their study to three industries
(technology, healthcare and financial). In two of the industries (technology and
healthcare), the authors find univariate differences in audit committee presence between
fraud firms and a no-fraud industry benchmark. In all three industries, BCHL find

Audit committees have been required for NYSE-listed companies since 1978, for NASDAQ firms since
1989 and for AMEX firms since 1992.
4

significant univariate differences in audit committees comprised purely of outsiders and

more independent boards between fraud firms and the no-fraud benchmark.
It should be noted, however, that BCHL recognize an important limitation to their
research design. Namely, their analysis does not employ a discriminant model with oneto-one matching of fraud firms with non-fraud firms. Furthermore, their univariate
analysis cannot examine the impact of various audit committee characteristics in coexistence. Thus, it remains an empirical question which particular audit committee
characteristics are most likely to be associated with a reduction in financial reporting
misstatement.

BRC Recommendations and Hypotheses

The BRCs recommendations were aimed at decreasing the likelihood of both
fraud and financial reporting restatements resulting from abusive earnings management
(i.e. earnings management which obscures the true financial performance of the
company). The BRC report notes that the committee did not focus on fraud per se, but
that many recommendations may reduce the possibility of fraud (BRC 1999). The
committees focus is on the large grey area where discretion and subjective judgements
bear on the quality of financial reporting. The report also notes in the discussion of the
costs and benefits of implementing the recommendations, that a net benefit would result
to the extent that instances of outright fraud, as well as other practices that result in
lower quality financial reporting are reduced with improved oversight. In this spirit, we
follow the BRCs recommendations by investigating both financial reporting

restatements and financial reporting fraud when developing our sample selection criteria.7
In the succeeding sections, we develop from the BRC recommendations a roster of
empirically testable audit committee characteristics.

Independence
BRC recommendations 1 and 2 address the definition of director independence
and requirements for independent audit committees. These recommendations are based
upon the BRCs contention that independent directors are better able to objectively
evaluate the propriety of managements accounting, internal control and reporting
practices. In recommendation 1, the BRC defines independence as the exclusion of
current and former employees, relatives of management, and persons receiving
compensation from the company (except directors fees). The BRC also recommends
excluding directors who are a partner in, a controlling shareholder or an executive officer
of, any for-profit business organization to which the corporation made or from which the
corporation received significant payments in the last 5 years. Compensation committee
interlocking directorships are also recommended for exclusion. In recommendation 2, the
BRC states that all listed companies with market capitalizations of greater than $200
million should have audit committees composed entirely of independent directors. 8
These recommendations are consistent with prior literature which finds that
greater director independence is associated with greater monitoring of management in a
variety of settings (Weisbach 1988; Baysinger and Butler 1985; Beasley 1996; Carcello
7

We provide results for the fraud-only subsample in later sections.

The most important departure from the BRC recommendations regarding independence is that the
exchanges did not adopt exemptions for smaller firms, as recommended by the BRC, but apply the new

and Neal 2000a and others). Improved monitoring by independent directors is expected to
result from incentives for such directors to build and maintain reputations as decision
control experts (Fama and Jensen 1983). Although audit committee service may increase
a directors reputation as a monitor, it also exacerbates the potential reputational damage
should a financial misstatement occur while the director serves on the audit committee.
Consequently, the preservation of reputational capital serves as one motivation for higher
quality monitoring on the part of outside, independent directors.9
A second, complementary motivation for more effective monitoring on the part of
independent audit committee members is director liability. Outside non-audit committee
directors may be able to demonstrate fulfillment of their fiduciary duties by asserting
reliance upon the audit committee for issues such as the adequacy of the firms financial
reporting, internal control structure and relationship with its external auditor (Reinstein et
al. 1984). This is because the audit committee is specifically charged with overseeing the
financial reporting process (Berton 1995; Lublin and McDonald 1998; Lublin and Nelson
1998; Bailey 2000). In cases of shareholder lawsuits alleging financial statement fraud,
this strategy can potentially subrogate outside director liability to audit committee
members. As such, even if the audit committee director is effectively shielded from
personal financial liability by insurance or indemnification, he or she faces costs, given
the time involved in mounting a defense (Sahlman 1990). Hence, litigation concerns
provide a secondary motivation for increased audit committee diligence on the part of
outside, independent audit committee directors.

rules to all firms. For a detailed explanation of the differences between the BRCs recommendations and
the adoptions per the NYSE and Nasdaq/AMEX exchanges, please see the appendix.

We expect that audit committees comprised entirely of independent directors will

be more willing and able to confront management about financial reporting matters and
thus will exhibit fewer incidents of financial reporting misstatements. This generates our
first hypothesis (stated in alternative form):
H1: The presence of an audit committee comprised entirely of independent
directors - per the BRCs independence recommendations - is associated with a
lower incidence of financial reporting misstatement.

Expertise and Audit Committee Size

The BRCs third recommendation is a response to increasingly complex
accounting and auditing issues facing audit committees. This recommendation requires
(except for small firms) a minimum of three audit committee directors, each of whom is
financially literate and at least 1 of whom has accounting or related financial
management expertise. Braiotta (2000) states that, in general, the audit committee should
be large enough to have members with a good mix of business judgement and experience,
but not so large as to be unwieldy. This recommendation is also likely rooted in a desire
to elevate the status of the audit committee. The BRC notes that the complexity of the
accounting and financial matters reviewed merit the commitment of considerable director
resources to the audit committee. We expect firms that commit more directorial resources
to their audit committees (in the form of three or more directors on the audit committee)
to be less prone to financial misstatement. This recommendation generates our second
hypothesis (stated in alternative form):
H2: The presence of an audit committee with at least three directors is associated
with a lower incidence of financial reporting misstatement.
9

Reputational capital can often result in substantial financial gain. In 1999, average compensation for
outside directors at the largest 200 industrial and service companies was $133,672. Outside director
compensation grew 70% between 1995 and 1999 (Pearl Meyers & Partners, quoted in Schellhardt (1999)).
8

The BRCs recommendation with respect to expertise is consistent with the POBs (1993)
position that the effectiveness of the audit committee is affected, first and foremost, by
the expertise of members of audit committees in the areas of accounting and financial
reporting, internal controls and auditing. The BRC defines financial literacy as the
ability to read and understand fundamental financial statements. The BRC also states that
accounting or financial management expertise may be demonstrated by employment
experience in finance or accounting, a CPA certification or comparable experience, or a
position as a CEO or other senior officer with financial oversight responsibilities.
The BRCs assertion that effective audit committees must be comprised of
members who possess ample financial knowledge is consistent with previous research on
audit knowledge. Audit committee members could come from a wide variety of
backgrounds and may not have the experience or technical knowledge needed for
effective accounting and auditing oversight (Kalbers and Fogarty 1993). Likewise, the
audit committees oversight role may be discounted by the external auditor if the auditor
believes the audit committee does not possesses the knowledge necessary to understand
technical auditing and financial reporting matters (Knapp 1987). Furthermore,
knowledgeable audit committees are better equipped to understand auditor judgements
and discern the substance of disagreements between management and the external auditor
(DeZoort 1998; DeZoort and Salterio 2001).
These results suggest that audit committee members who possess financial
literacy/expertise provide additional support for external auditors when discussing
accounting issues and disagreements with management. These results lead us to expect
that the outcome of greater audit committee financial knowledge will be less financial

misstatement, in that disputes resulting in SEC action will be more likely to be resolved
in the auditors favor prior to financial statement release. This generates our third
hypothesis (stated in alternative form):
H3: The presence of an audit committee without at least one director meeting the
BRCs recommendation for accounting or related financial management expertise
is associated with a higher incidence of financial reporting misstatement.
It should be noted that because of the difficulty of developing a measure of
literacy (which is defined in terms of skills, rather than professional background) from
public information, we confine this hypothesis to addressing the effect of the presence of
financial experts on the audit committee.10

Oversight Responsibility for Outside Auditors Engagement

The remaining BRC recommendations address the application of these audit
committee oversight duties. In particular, recommendation 6 details the relation between
the audit committee and the external auditor. The BRC report notes that it is essential to
the integrity of the audit process that all parties recognize that the audit committee and
the board are the entities to which auditors are accountable. As a result, the BRC
recommends that the audit committee charter specify that the outside auditor is ultimately
accountable to the board of directors and the audit committee. The charter should also
specify that the audit committee has the ultimate authority to select, evaluate, and where
appropriate, replace the outside auditor (or to nominate the outside auditor to be proposed
for shareholder approval in any proxy statement).

10

In addition, our measure of expertise is similar to that adopted by the Nasdaq, but not to that of the
NYSE, which allows the board substantial discretion. Please see the appendix for further details.
10

Research underscores the importance of the audit committee in the selection and
replacement of the external auditor. Carcello and Neal (2000b) find that financially
distressed firms with independent audit committees are more likely to receive a going
concern qualification, consistent with increased audit committee support for the auditor in
negotiations with management. These findings suggest that explicitly conferring auditor
engagement rights on the audit committee may further increase the auditors ability to
withstand management pressures, potentially resulting in less aggressive accounting
and/or more thorough audit procedures. This leads to our fourth hypothesis (stated in
alternative form):
H4: The presence of an audit committee possessing control over external auditor
selection and/or termination decision rights (as disclosed in the proxy statement)
is associated with a lower incidence of financial reporting misstatement.

Other BRC Recommendations

BRC recommendations 4 and 5 require the audit committee to have a formal
charter and to disclose the status of the charter and whether the audit committee fulfilled
its designated responsibilities during the proxy statement period. Recommendations 7-10
deal (either directly or indirectly) with the responsibilities and activities of the audit
committee, requiring the committee to evaluate auditor independence (number 7), to
discuss accounting quality with the auditor (number 8), and to hold similar discussions
with management and the auditor (number 9). Recommendation 10 requires interim
reviews, and also discussion of these review-related issues with the auditor.
BRC recommendation 8, requiring outside auditors to hold specific discussions
with audit committees was implemented through the issuance of Statement on Auditing
Standards No. 90, Audit Committee Communications (SAS 90), superceding Statement

11

No. 61 (SAS 61), Communication with Audit Committees. BRC recommendation nine
was implemented to require the audit committee to state in the proxy statement whether,
based upon review and discussion with the management and auditors, the committee
recommends that the financial statements be included in the Form 10-K or 10-KSB. BRC
recommendation 10, requires timely quarterly reviews and discussion with the audit
committee of matters consistent with Statement on Auditing Standards No. 71 (SAS 71).
It is most likely that the preferred method of examining the efficacy of these
remaining recommendations is through the use of survey instruments. We believe our
sample time period (1991-1999), and the liquidation of many sample firms following
SEC sanction, precludes obtaining survey information on audit committee charters and
reviews (not required disclosures during the sample period). Thus, we do not attempt to
test these recommendations directly. However, we do indirectly investigate the diligence
of the audit committee in carrying out their assigned duties by observing the impact of
audit committee meeting frequency. The BRC (as discussed above) and the Treadway
Commission (1987), recommend a frequency of at least four audit committee meetings
per year (consistent with reviews of quarterly financial statements). Therefore meeting
frequency may indirectly provide information on the value of audit committee monitoring
of quarterly statements and their diligence in carrying out their responsibilities.
Correspondingly, Menon and Williams (1994) posit that meeting frequency is a
signal of audit committee director liability concern and audit committee diligence. Menon
and Williams (1994) find that more meetings are positively associated with the presence
of outside directors. Abbott et al. (2000) utilize the number of audit committee meetings
in constructing a composite audit committee effectiveness variable. This variable, which

12

is coded 1 when the audit committee is comprised entirely of outsiders and meets at
least twice annually, is negatively related to the incidence of financial misstatement.11 In
this study, we use four meetings as our threshold. While the BRC did not make a specific
recommendation for meeting frequency, it implied that audit committees meet at least
quarterly. Specifically, the BRC recommends that all publicly held firms have their
interim quarterly financial statements reviewed by the external auditor and that the audit
committee discuss the results of the quarterly reviews with the external auditors, resulting
in four meetings. If quarterly meetings are associated with greater audit committee
diligence in their monitoring duties, then we expect quarterly meetings to be associated
with a lower level of misstatement. This leads to our fifth hypothesis (stated in alternative
terms):
H5: An audit committee that meets at least four times annually is negatively
associated with the probability of financial reporting misstatement.

METHODOLOGY
Sample Selection
We began our sample selection by reading Accounting and Auditing Enforcement
Releases (AAERs) filed between January 1, 1991 and June 30, 2000. These were
obtained from the Commerce Clearinghouse Internet Research database, and from the
SECs EDGAR web site. We included firms in our preliminary sample if the following
conditions were met:

11

Note that Abbott et al. (2000) did not investigate whether audit committee independence or audit
committee activity as stand-alone variables reduce the incidence of financial misstatement. Further, APP
did not examine the relation between other audit committee characteristics such as financial expertise,
external auditor termination rights or audit committee directorships and the incidence of financial
misstatement. Finally, APP tested only AAER alleged fraud firms from the period 1980 1994 and did not
include restatement firms in their sample.
13

1)
2)
3)
4)
5)

The allegation of fraud or financial misstatement contained in the AAER

applied to the period between January 1, 1991 and December 31, 1999;
The subject firm was listed on the AMEX or NYSE, or quoted on the
Nasdaq;
The allegation applied to annual earnings;
The AAER referred to a misstatement of financial statements other than
those contained in a registration statement for an initial public offering;
The alleged financial misstatement was not exclusively the result of a
defalcation (theft of corporate assets for personal use).

Our primary reason for imposing our first two restrictions was to increase the
likelihood that the subject firm had an audit committee. Firms listed on the three major
exchanges were required to have audit committees during the time frame considered. 12
In terms of our third restriction, we limited our sample to those firms with
misstatements in the annual financial statements to better control for variation that might
exist in the degree of external audit scrutiny received by the sample and control firms. In
particular, during the sample period, SEC registrants were not required to have a timely
review of interim statements by their external auditors. Consequently, it would be
difficult to make predictions as to the potential effectiveness of the audit committee in
monitoring the interim, quarterly financial reporting process when there may have not
been any external auditor involvement in the reporting of these results. This enables us to
focus more narrowly on the role of the audit committee as overseer in the financial
reporting process.
With respect to our fourth restriction, very few companies appear to form audit
committees and/or hold audit committee meetings in the year immediately following their
initial public offering (Parker 1997). Thus, they are unlikely to have audit committees at
the time the registration documents are filed. In addition, the monitoring environment for

14

firms during the IPO period is likely to be systematically different from that of post-IPO
firms because of the increased scrutiny by underwriters, auditors and legal representatives
during this period, presenting difficulties in selecting a matching control firm. Our final
restriction was imposed since control weaknesses leading to a defalcation might not be
similar to those related to a pure financial misstatement.
After examining all AAERs from the period January 1, 1991 to June 30, 2000
and imposing the above five restrictions, we arrived at a preliminary sample of 113 firms.
This is summarized in Table 1 (Panels A and B). Table 1, Panel A, provides a
breakdown of the initial 113 firms identified during the AAER search. For 56 firms we
could not locate proxy statement data on the composition of the audit committee and the
number of committee meetings held in the first year of the misstatement. We lose an
additional ten firms due to lack of Compustat data for the last year preceding the
misstatement. We lose four firms due to the inability to find an adequate matching firm.
This yields an AAER sample of 41 fraud firms and two restatement firms. The two
restatement firms are segregated and added to the restatement sample described below
(see Table 1, Panel A).
We then supplemented our sample of firms from the AAER search with firms
disclosing restated earnings. We searched the Dow Jones Interactive Database for the
term restatement appearing in the lead paragraph or headline of articles from The Wall
Street Journal, Barrons, the Dow Jones Corporate Filings Alert, Dow Jones Business
News, Dow Jones News Service and PR Newswire between January 1, 1991 and
December 31, 1999. This search resulted in 3033 articles found. Of these, 1425 were

12

The AMEX regulations had not yet taken effect in 1991, however, our sample from that year consists
only of Nasdaq and NYSE firms.
15

multiple references to restatements, thus resulting in 1608 unique financial restatements.

Of these 1608, 358 were not related to financial statement restatements, 153 referred to
foreign firms, 553 were restatements of quarterly statements only, 46 announcements
related to fiscal years ending before January 1, 1991, 236 referred to restatements caused
by mergers and acquisitions, divestitures, stock- or tax-related changes, changes in
accounting principles or related to restatements of financials included in registration
statements or unaudited annual reports. This results in a preliminary sample of 262
annual restatements.13
Of the 262 annual restatements, 22 firms were banks, 72 were not on Compustat
for the year prior to the first year of restatement, and we could not obtain proxy
information covering the restated year for 18. Twenty-seven were included in the AAER
sanction sample, and 19 were omitted because they were not listed on the NYSE, Nasdaq
or AMEX (16), or because they were not suitable for inclusion for another reason (3). We
could not find a suitable match for 18 firms, resulting in a final sample from this source
of 86. Including the two restatement firms from the AAER search (described above)
brings our total restatement sample to 88. Adding the 41 fraud firms brings our total
misstatement (restatements and fraud) sample to 129. This is summarized in Table 1,
Panel B.

Control Firms Selection

Following Beasley (1996), we match each sample firm with a control firm on the

13

Of these sample firms, 184 occurred during the 1995-1999 period. This compares reasonably to the
annual restatement sample of 204 for a similar period in Palmrose and Scholz (2000), although Palmrose
and Scholz use a more exhaustive search methodology.
16

same exchange with a market value of equity within 30% of the sample firm in the year
preceding the fraud or misstatement.14 We also match on the basis of Big 5 (Big 6) or non
Big 5 (Big 6) auditor. We select the control firm from among those meeting the above
criteria in the same 4-digit SIC code whenever possible. In cases in which we are unable
to locate a control firm from the corresponding 4-digit SIC code, we select from firms
matching at the 3-digit or 2-digit level.15 Table 1, Panel C shows 53 (24) of our
restatement (fraud) firms were matched with a corresponding control firm at the 4-digit
level, 20 (11) restatement (fraud) firms were matched at the three-digit level, and the
remaining 15 (6) restatement (fraud) firms were matched at the 2-digit level. Table 1,
panel D shows that approximately 50% of our observations in both the fraud and
restatement samples are from four high-tech industries: 35xx (computers), 36xx
(electronic components) and 38xx (instruments) and 73xx (software and computer
programming). Table 1, Panel E shows the sample observations, by year. The majority of
the restatements occurred in the later years of the sample, while the majority of frauds
identified occurred in the earlier years. We return to this issue in the discussion of our
results.

Research Design
Test Variables
Table 2 summarizes our test and control variables. We obtain information on the
board and audit committee from the firms proxy statement for the period covering the

14

None of the control firms were identified as having financial misstatements in the period from 1991 to
September 30, 2001.
15
Four sample firms were matched on a basis of total assets because no equity-value matching firm could
be located in the same 2-digit industry.
17

first year of financial misstatement. The critical proxy sections were those which include
the directors biographical information, executive salary information, board meetings and
committees, employment contracts, related party transactions, certain transactions and
compensation committee interlocks. For each director serving on the audit committee
during the target year, we determine first whether the director was a current or former
employee of the firm (acceptable under Nasdaq and AMEX regulations during the sample
period) or had any other disclosed affiliation with the company or its management, other
than board service. These grey directors may have financial incentives to support
management which offset the potential costs of less effective financial oversight. INDEP
is coded 1 if the audit committee consists entirely of independent auditors (per the
BRCs independence definition) and 0 otherwise. In terms of the BRCs audit
committee size recommendation, MINSIZE is coded 1 if the audit committee consists
of at least three directors and 0 otherwise.
We believe the BRCs requirement that every member possess basic financial
literacy is impossible to evaluate from the currently available public information, since
such literacy may be associated with a wide variety of backgrounds and qualifications.
However, the BRC requirement that at least one member possess financial expertise is
more precisely stated and testable using archival data. We attempt to operationalize the
BRCs recommendation that at least one member possess past employment experience
in finance or accounting, requisite professional certification in accounting, or any other
comparable experience or background which results in the individuals financial
sophistication, including being or having been a CEO or other senior officer with

18

financial oversight responsibilities.16 We code NOEXPERT as 1 if the audit

committee does not include at least one director who is (or has been) a CPA, investment
banker or venture capitalist, served as CFO or controller, or has held a senior
management position (CEO, President, EVP, SVP, VP) with financial responsibilities.17
AUDCHOSE is coded 1 if the annual proxy statement discloses that the audit
committee possesses external auditor retention/nomination/dismissal rights, and 0
otherwise. This variable is likely to be measured with substantial noise. During the
sample period, the content of proxy statement disclosures of audit committee activities
was not specified, and thus to some extent we may be measuring propensity to disclose
instead of the specific audit committee activity.
In terms of our fifth hypothesis, we also obtain the number of meetings held
during the first misstatement year. MINMEET is dichotomously coded 1 if the audit
committee met at least four times during the first misstatement year; and 0 otherwise.
Control Variables
Our sample selection process (discussed above) results in the inclusion of firms
alleged by the SEC to have fraudulently misstated their financial statements, firms which
were alleged by the SEC to misstate, but without fraudulent intent, and those which
restated financial statements without disclosed SEC action. It is not clear how the SEC
makes a determination regarding which firms to investigate and what allegations are
appropriate. Feroz et al. (1991) note that the SEC has more candidates for investigation
than it can pursue and ranks targets based on probability of success and the potential

16

We thank an anonymous reviewer for assistance in appropriately operationalizing this variable.

Our data are limited in that SEC regulations require biographical information in the proxy for only the
preceding five year period. We extend the evaluation of expertise to test whether certain categories of
expertise are significantly related to misstatement events (discussed in the sensitivity analysis section).
17

19

message value. If the choice of whether to bring a fraud allegation is dependent on these
institutional factors, our categorization between fraud and restatement firms may be
noisy. Thus, in developing the control variables we draw from both the literature on
financial misstatements resulting from error and the fraud-related literature, and include
the same control variables in testing both samples.
In order to arrive at a manageable set of control variables we synthesize the fraud
and financial reporting error literature across three dimensions: Internal Control
Environment, Pressure / Incentives, and Management Capacity. Prior literature has
generally shown that these three factors are determinants of financial misstatements
(Loebbecke and Willingham, 1988).
Internal Control Environment. In studies involving auditor identified errors,
Hylas and Ashton (1982), Kreutzfeldt and Wallace (1986) and Wallace and Kreutzfeldt
(1995) find that errors are associated with internal control environment factors such as
weak controls, poor communications, personnel structure and lack of financial
knowledge. Likewise, Loebbecke and Willingham (1988) find that weak controls and
internal decentralization are positively associated with the risk of financial misstatements.
Several studies investigate the role of internal and external monitoring
mechanisms in mitigating the risk of financial misstatement (DeFond and Jiambalvo
1991; Beasley 1996; Bell and Carcello 2000). For example, Beasley (1996) finds that the
percentage of outside directors on the full board is negatively related to the incidence of
fraud, consistent with the incentives of outside directors to enhance their human capital
by demonstrating decision control expertise. Borrowing from this previous literature, we
control for the impact of the internal control environment. BLOCK is the proportion of

20

stock controlled by unaffiliated 5% owners. BOARDSIZE is the number of directors on

the full board, %OUTSIDER is the proportion of unaffiliated directors serving on the full
board and OUTOWN is the proportion of voting control held by outside board members.
We expect BLOCK, %OUTSIDER, OUTOWN to be negatively associated with
misstatement risk, consistent with greater monitoring. We expect BOARDSIZE to be
positively associated with misstatement likelihood, due to less effective monitoring
associated with more unwieldy boards. Finally, AGEPUB is the number of years the
company has been publicly traded. Older firms are expected to have more developed
internal control systems and processes and thus a lower likelihood of financial
misstatement.
Other characteristics, such as the competence of accounting personnel and the
quality of management controls are difficult to assess from archival data. In later tests, we
include an indicator variable (discussed in the sensitivity analysis section) for the
presence or absence of an internal audit function. We use the indicator variable as an
additional proxy for the quality of the internal control environment.
Pressure / Incentives. In studies involving financial errors and fraud risk,
Kreutzfeldt and Wallace (1986), Wallace and Kreutzfeldt (1995), DeFond and Jiambalvo
(1991), Loebbecke and Willingham (1988), and Bell, et al. (1991) find a positive
association between performance pressures and the risk of financial misstatements. For
example, Kreutzfeldt and Wallace (1986) find that firms with liquidity and profitability
problems were more likely to make more errors. Likewise, firms subject to capital market
pressures were at a greater risk of financial misstatements.

21

Borrowing from this previous literature, we control for the impact of the
pressures/incentives by including the following variables. TROUBLE measures the
financial distress of the firm. We base the trouble variable on the Altman Z score in the
year prior to misstatement (Altman 1968).18 FINANCE captures the need for additional
financing used by Dechow et al. (1996). The need for financing is calculated as:

(Operating cashflowt Average Capital Expenditures t-3-t-1)/Current Assets t-1

Firms with scores less than 0.5 are expected to need external financing within 2
years, which provides an incentive to manipulate earnings and assets in an attempt to
qualify for lower cost financing (Dechow et al. 1996). FINANCE is coded 1 for these
firms and 0 otherwise19. Finally, GROWTH is the average growth rate of total assets in
the two years preceding the misstatement. The firms rate of growth can affect the
internal control environment, or the need to sustain high growth can provide an incentive
for misstatement. We predict a positive relation with the incidence of financial
misstatement and our TROUBLE, FINANCE and GROWTH variables.

Management Capacity. Finally, management capacity to engage in financial

misstatements is considered. Loebbecke and Willingham (1988) posit that primary
characteristics related to increased fraud risk also include domination of operational and
financial decision-making by a single person (among others mentioned above). In follow18

For the 20 firms for which an Altman Z cannot be constructed from Compustat data, we use an indicator
variable with the value 1 if the firm has experienced losses in 3 of the preceding 6 years. Our results are
not sensitive to the exclusion of these 20 firms.
19
For 30 firms for which we are unable to obtain enough information to calculate the external financing
need, we examine whether the firm registered a securities offering (based on notice in the Investment
Dealers Digest) or received additional debt or equity financing in the first year of misstatement or the
succeeding year.
22

up studies using the same data set, but different methodologies (Loebbecke et al. 1989;
Bell and Carcello 2000), the existence of a dominant management, aggressive attitude to
financial reporting, auditor disputes and undue pressure placed on auditors (together with
other factors mentioned above) were again validated as fraud indicators.
Because of the limits of archival data, we cannot control for management attitude.
However, in addition to controls identified earlier, we attempt to control for indicators of
management capacity or CEO dominance by using the following variables. CEOCHAIR
is an indicator variable coded 1 if the CEO chairs the board. MGROWN is the
proportion of voting control held by management in the year of misstatement. High
management ownership can indicate dominance of decision-making by an individual or
small group, demonstrate alignment of manager and shareholder interests or provide
managers with incentives to misstate financial results. Therefore we do not make a
directional prediction for this variable. FOUNDER is coded 1 if the CEO is also the
company founder. We predict a positive association with financial restatement risk for
both CEOCHAIR and FOUNDER.

Regression model
Our research design is similar to that used in Beasley (1996) and Dechow et al.
(1996). We employ logit regressions on our size-, exchange-, industry- and Big 6/nonBig 6 auditor-matched sample to test whether firms which voluntarily adopted audit
committee characteristics consistent with certain BRC recommendations (those which
can be objectively evaluated from public information) have a lower likelihood of
financial misstatement. Our regression model is:

23

MISSTATEMENT = + 1INDEP + 2MINSIZE + 3NOEXPERT + 4AUDCHOSE

+ 5MINMEET + 6BLOCK + 7BOARDSIZE + 8%OUTSIDER + 9OUTOWN
+ 10AGEPUB + 11TROUBLE + 12FINANCE + 13GROWTH + 14CEOCHAIR
+ 15MGROWN + 16FOUNDER +

where the variables are as described previously.

RESULTS
Descriptive Statistics and Univariate Analysis
Table 3 compares the misstatement and control firms. In Panel A, we present the
mean, median and standard deviations for the restatement firms and their matched control
firms. The restatement and control firms differ significantly (p-values of .05 or less) in
the independence of the audit committee, absence of an expert, audit committee
responsibility for auditor selection and whether the committee meets at least 4 times per
year. The mean number of restatement (control) firms with fully independent audit
committees is 50% (72%). Twenty-four percent of the restatement firms lack a financial
expert on the audit committee, while only 8% of the control firms lack an expert. Fiftyone percent (65%) of the restatement (control) firms disclose auditor selection
responsibilities for the audit committee and 15% (38%) of the restatement (control) firms
meet at least four times during the year. With respect to the control variables (proxying
for alternate monitoring mechanisms, incentives for misstatement and management
dominance), the restatement and control samples differ only in the levels of management
ownership restatement firms have significantly (p-value < .10) higher levels of
ownership (26%) than control firms (14%). Restatement firms have mean (median) total
assets of $989.3 million ($129.2 million). The mean (median) control firm size was

24

$1067.5 million ($126.2 million). The size differences were not significant, indicating
that the matching process was successful with respect to size.
Panel B shows the results for the fraud firms and the matched control firms.
Differences in test variables between these firms are significant (p-value < .01) only with
respect to audit committee independence and the absence of a financial expert on the
audit committee. The proportion of sample (control) firms with an entirely independent
audit committee is 22% (73%). The proportion of fraud firms lacking a financial expert
on the audit committee (36%) is significantly lower than the proportion of control firms
(9.8%). However, the sample and control firms do not differ significantly in whether the
committee includes at least three members, whether committees hold at least 4 meetings,
and the disclosure of auditor selection rights.20
Several differences are noted between the fraud firms and control firms with
respect to the control variables. The fraud firms and control firms differ in the presence of
a strong central figure, an element of the internal control environment. Sample firms have
higher levels of managerial ownership (p<0.05), are more likely to have a CEO who
serves as chair (p<0.01), and to have a CEO who is also the founder (p<0.01). In
addition, the fraud firms experienced significantly higher average growth prior to the
fraud period (122% vs. 28%). We did not find differences in the alternate monitoring
mechanisms (BLOCK, OUTOWN), or in other control variables. Fraud firms vary in
assets from $3 million to $6.416 billion, with mean (median) total assets of $726 million
($56 million). The mean (median) control firm size was $612 million ($53.12 milllion).

20

In order to determine whether the independence, activity and expertise levels have increased we split the
sample into those firms with misstatements in 1991-1995 and those with misstatements in 1996-1999. We
find that the average variable levels have not changed significantly over time for either the control or
sample firms, despite the more intense scrutiny afforded audit committees in the later period.
25

The difference in size is not significant, indicating that the matching process was
successful with respect to size.
In other univariate results (not reported) we examine how the fraud and
restatement samples differ from each other. Only three significant differences emerged.
The fraud sample was significantly smaller than the restatement sample, significantly less
likely to have an independent audit committee and were from an earlier sample year21.
Table 4 shows the correlations between the regression variables. We find
numerous significant correlations between the corporate governance-related variables.
These variables appear primarily to function as complements. The presence of an
independent audit committee is positively related to the presence of an expert, meetings
greater than four, and a higher proportion of outsiders on the overall board. Holding four
or more meetings is related positively to the audit committee size of at least three, the
presence of an expert and the assignment of auditor selection duties to the committee.
While these correlations are significant, none is greater than 0.22. We also note negative
correlations between audit committee independence and growth, and audit committee
independence and the presence of a founder as CEO. Board size is related positively to
the minimum audit committee size and number of meetings, auditor selection duties for
the committee, the presence of an expert and the proportion of outsiders on the board.
The highest of these correlations is 0.41 (board size and minimum committee size)22.

21

This last result is not surprising given the length of time it takes for the SEC to commence enforcement
action.
22
Regression diagnostics on the multivariate results reported in the succeeding section do not indicate the
presence of multicollinearity among the explanatory variables.
26

Multivariate Results
Table 5 shows the regression results. Our results for the full sample (Column A)
are consistent with hypotheses 1, 3 and 5, but do not support hypotheses 2 and 4. The
presence of a completely independent audit committee and higher levels of committee
activity (measured by holding a minimum of 4 meetings) are significantly related to a
lower incidence of misstatement (p < 0.01). In addition, the absence of a director meeting
the BRC guidelines of a financial expert (NOEXPERT) is significantly related to an
increased probability of financial misstatement (p < 0.01). A committee of at least three
members and the disclosure of auditor engagement rights are not significantly related to
the incidence of misstatement. Consistent with prior research, the likelihood of financial
misstatement is also higher for firms with larger boards and those that have a CEO
serving as chair.
Table 5, Columns B and C show separate regressions for each subsample. The
results for the restatement sample (Column B) are similar to those for the overall sample,
supporting hypotheses 1,3 and 5. Among the control variables, there are significant
negative associations between the presence of blockholders (BLOCK) and the number of
years the firm has been publicly traded (AGEPUB) and the incidence of restatement.
Consistent with the conjectures of prior literature, board size is positively related to the
incidence of restatement.
The results in the fraud sample (Column C) support only hypotheses 1 and 3. The
presence of an independent audit committee is associated with a lower likelihood of
fraud, while the absence of an expert is associated with a greater likelihood of fraud.

27

Among the control variables, fraud firms experienced higher growth in the years
preceding the fraud, and also have larger boards.
In contrast to Beasley (1996), the presence of outsiders on the board is not
significantly associated with a lower incidence of fraud. In further tests (not reported)
these variables remain insignificant even in the absence of the audit committee-related
variables. We believe that the differences in our results may be related to differences in
sample firm selection procedures. Our sample firms are much larger than the sample used
by Beasley. The mean (median) total assets for Beasleys sample is $103 million ($11
million), and 83% of Beasleys firms are traded on the Nasdaq. The mean (median) firm
size in our overall sample is $596 million ($76 million). The mean (median) firm size in
our fraud sub-sample is $726.30 million ($55.8 million). We believe size will affect the
corporate governance results for two reasons. First, in an extremely small firm the board
may be able to execute its activities, including financial monitoring, without recourse to a
committee structure. Second, in small firms the addition of one or more independent
directors may be a strong signal of managements commitment to reliable reporting.
There are also some differences in addition to firm size that may be important.
First, we exclude misstatements contained in registration statements, while Beasley does
not. Recently public firms (prior to the new audit committee guidelines, which specify
how soon after IPO firms on the three primary exchanges must form audit committees)
are unlikely to have audit committees (Parker 1997), and therefore any board monitoring
will take place in the absence of an audit committee structure. Also, the presence of other
monitors around the IPO, such as underwriters and heightened auditor scrutiny may have
an unknown effect (either as a substitute or complement) on corporate governance

28

structures. A second difference is that Beasley includes an unknown number of

misstatements in quarterly financial statements, while we exclude these, for reasons
discussed earlier. We also restrict our sample to firms that traded on one of the three
major exchanges to facilitate matching on this characteristic. Lastly, during our sample
period the degree of regulatory and public attention on corporate governance mechanisms
and the audit committee, in particular, was heightened. The cumulative effect of these
differences between our sample and Beasleys is unknown and therefore we do not assert
that our results are directly comparable to Beasleys.

Sensitivity Analyses
We test for the possibility that undiscovered (later revealed) misstatements exist
in control firms matched with our misstatement firms. When we omit misstatement firms
after 1998 from our tests we obtain results similar to those for the full sample. Because
the BRC recommendations exempt small firms (those with market capitalizations under
$200 million) from certain requirements for audit committee independence, we examine
the results with small companies omitted. The results are substantially similar, except that
the auditor engagement rights variable (AUDCHOSE) is marginally associated with a
lower incidence of fraud in the over $200 million sample (p < 0.10). In order to test
whether high-profile fraud cases were driving our results, we omitted the four most
publicized fraud cases in our sample: Sunbeam, Rite-Aid, Cendant and McKesson
HBOC. The results remained unchanged for both the fraud and combined samples.
With regard to potentially correlated omitted variables, we also test other boardrelated characteristics including average tenure of outside directors and the number of
directorships held by outside directors. None of these variables are significant at

29

conventional levels, nor do they impact the results reported in Table 5. We also add a
variable coded 1 if the firms proxy statement discloses the presence of an internal audit
function. The internal audit variable is not significant.23 Finally, none of the individual
expert types (i.e. CPA, venture capitalist) was associated with financial misstatements.24

LIMITATIONS
This study has at least five important limitations. First, because we wish to
provide timely information that may be of interest to policymakers, we test our
hypotheses in a historical sample of firms. This precludes the more complete test of BRC
recommendations that would be possible using a survey instrument to obtain non-public
information. This is especially true for the area of financial literacy (DeZoort and Salterio
2001; Maines, Martin and McDaniel 2001). As we noted previously, we address only the
expertise measure, and only to the extent allowed by archival data. It is not certain to
what extent literacy and expertise may be substitutes, or what effect the proportion of
directors with literacy or expertise may have. Second, these same limitations also force us
to adopt a piecemeal approach to evaluating the efficacy of certain recommendations,
even though the BRC report stresses that its recommendations should be adopted in their
entirety. Third, our meeting frequency threshold is only one measure of audit committee
activity and may be relatively crude. Fourth, our lack of results for meeting frequency for
the fraud firms does not necessarily suggest that it is not an important element of audit
23

This may be due to our inability to accurately identify all firms with internal audit functions, or to assess
the effectiveness of those functions.
24
The breakdown of experts was as follows: 83 venture capitalists/investment bankers, 27 CFO/controllers,
134 CEO/COO/Presidents, 23 senior financial vice presidents, 25 vice presidents with financial
responsibilities, 4 audit firm partners, 7 commercial bankers, 14 expert CPAs. These numbers are not
mutually exclusive due to categorization overlap. For example, several individuals had both CPA and CFO
experience in the prior five years per the annual proxy statement description.
30

committee effectiveness. Although its coefficient is in the prediction direction, the

relative lack of observations may have diminished the power of our test for this element.
Therefore, we believe it remains an empirical question if meeting frequency is
systematically unrelated to the incidence of misstatement. Lastly, we cannot rule out the
possibility that both audit committee characteristics and the lower incidence of financial
misstatements are related to an undiscovered correlated, omitted variable.

POLICY IMPLICATIONS AND CONCLUSIONS

The Blue Ribbon Committee developed ten recommendations intended to
improve financial monitoring by improving the independence and effectiveness of the
audit committee. We find that audit committees comprised entirely of independent
directors, that meet at least 4 times a year or that lack a member with financial expertise
are significantly associated with financial misstatement. Committees which do not meet
minimum size requirements or which do not disclose auditor engagement rights are not
associated with a higher misstatement incidence. We also find that characteristics of the
full board are unrelated to the incidence of misstatement. This is consistent with the
recent regulatory emphasis on the audit committees role in financial reporting, and the
delegation of significant responsibilities to an independent subset of the board. When we
evaluate our fraud and restatement subgroups separately, we find that only audit
committee independence and the absence of a financial expert are related to fraud
incidence. Results for restatement firms are similar to those in the full sample. Our
results, which are derived from a sample period predating the implementation of the BRC

31

recommendations, suggest that their adoption may result in more effective financial
monitoring.
Although are results are derived from the pre-BRC implementation period, four
factors suggests that our evidence remains timely and relevant. First, current research
suggests that corporate audit committees have yet to attain the standards suggested by the
BRC. In particular, in a study from the 2001 proxy statement season, Abbott et al. (2002)
find that of a sample of over 500 firms, only 74% of audit committees were comprised
entirely of outsiders, only 57% met at least 4 times annually and 79% possessed a
member with financial expertise. An even smaller proportion of firms have adopted all
the BRC recommendations: only 38% of those audit committees were simultaneously
independent, met at least 4 times annually and possessed a member with financial
expertise. This suggests considerable room for improvement in corporate audit
committees. Second, the exchange-listing rules on independence and expertise as
adopted differ from the BRC recommendations and allow for substantial board discretion
in determining audit committee member independence and financial expertise (see
appendix for details). Third, it is possible that enforcement by the exchanges will not be
rigorous, given that the exchanges are motivated to maintain their clientele of listed
companies. Finally, highly publicized accounting restatement cases such as Enron (where
three of the six audit committee directors had financial ties to management) and others
suggest that financial misstatement and the adequacy of financial monitoring continue to
be very important issues facing regulators, professionals and academics.
Although our study should be interpreted in the light of the limitations discussed
above, we note that our results confirm the importance attached by the BRC to audit

32

committee independence and expertise in the fraud sample, and also to the number of
meetings (the importance of meetings is implied by the BRC emphasis on quarterly
financial statement reviews) for restatement firms. Our research is consistent with the
implementation of certain BRC recommendations leading to a decrease in financial
misstatement. We believe our study contributes to an area that has become increasingly
important in recent years, as investor losses from high-profile financial misstatements
have continued to mount. We hope that other researchers will obtain more detailed data
on audit committee members, their activities and interactions with external auditors (see
DeZoort, Hermanson and Houston, 2000a; 2000b) and audit committee charters in order
to further our knowledge about audit committee characteristics and the financial reporting
process.

33

TABLE 1
Sample Selection
PANEL A: Fraud Sample Selection

AAERs filed January 1991 December 1999, alleging fraudulent or

negligent misstatement of annual financial reports between 1991 1999.
Less: Firms for which proxies were unavailable for the year of the sanction.
Less: Firms for which financial data was not available on Compustat.
Less: Firms for which we could not locate an appropriate control firm.
Less: Remaining AAER firms which do not allege fraud
Final sample of fraud firms

113
(56)
(10)
(4)
(2)
41

PANEL B: Restatement Sample Selection

Annual restatements located from article search

Less: Financial Institutions
Less: Compustat information not available
Less: Proxy statement not available
Less: Included in the SEC sanction fraud sample
Less: Not on the NYSE, AMEX or Nasdaq, or other reason for non-inclusion
Less: No control firm located
Firms located from this source
Add: Restatement firms located via AAERs
Restatement firms, total
Total observations (fraud and restatement)
Total sample, including control firms

262
(22)
(72)
(18)
(27)
(19)
(18)
86
2
88
129
258

PANEL C: Matches by(Restatement/Fraud):

4-digit SIC code

53/24

3-digit SIC code

20/11

2-digit SIC code

15/6

PANEL D: Sample observations by 2-digit SIC code

28xx Chemicals and Allied Products, Pharmaceuticals

35xx Industrial Machinery, Computers
36xx Electronic Components & Other Electrical Equipment
38xx Instruments
49xx Electric, Gas, Sanitary Services
50xx Durable Goods
59xx Miscellaneous Retail
73xx Business Services, Computer Programming & Software
80xx Health Services
2-digit SIC codes with less than 3 observations
Total

Restatement
3
9
5
9
5
3
4
19
5
26
88

Fraud
3
5
5
3
1
0
1
12
0
11
41

PANEL E: Sample observations by year (Restatement/Fraud):

3/12
4/5
1991
1994
2/6
10/1
1992
1995
3/8
18/2
1993
1996

1997
1998
1999

22/6
23/1
3/0

34

TABLE 2
Description of Variables
Variable Name

Description

Expected
Sign

Prior Studies

BRC Recommendation Measures

INDEP

MINSIZE
NOEXPERT

AUDCHOSE

MINMEET

Indicator variable with the value 1 if all audit

committee members are independent by BRC
definition.
Indicator variable with the value 1 if the audit
committee consists of at least three members.
Indicator variable with the value 1 if audit
committee includes no directors with financial
expertise per the BRCs definition.
Indicator variable with the value 1 if annual proxy
statement discloses that audit committee has
external auditor engagement/termination rights.
Indicator variable with the value 1 if the audit
committee meets at least four times annually
during the year of the fraud; 0 otherwise.

Beasley et al. (2000);

BRC (1999)

BRC (1999)

BRC (1999)

BRC (1999)

Menon and Williams

(1994); BRC (1999)

Beasley (1996);
Dechow et al. (1996)

+
-

Beasley (1996)
Beasley (1996)

Beasley (1996)

Beasley (1996)

Loebbecke et a.
(1989);
Bell et al. (1991)
Dechow et al. (1996)

Internal Control Environment

BLOCK

BOARDSIZE
%OUTSIDER
OUTOWN
AGEPUB

The cumulative percentage of outstanding

common stock shares held by 5%+ blockholders
not affiliated with management.
The number of directors on the board.
Percentage of the board members who are nonemployee directors.
The cumulative percentage of stock held by
outside directors.
The number of years the company has been
publicly traded.

Pressure/Incentives
TROUBLE

FINANCE

GROWTH

Indicator variable equal to one if the firm is

financially distressed, based on the Altman Zscore in the year preceding misstatement.
(Operating cashflowt Average Capital
Expenditures t-3-t-1)/Current Assets t-1
Firms with scores less than 0.5 are expected to
need external financing within 2 years, and are
coded 1.
The average percentage change in total assets for
the two years ending before the year of the
misstatement.

Loebbecke et al.
(1989);
Bell et al. (1991)

Loebbecke et al.
(1989);

Beasley (1996)

Dechow et al. (1996)

Management Capacity
CEOCHAIR

MGROWN
FOUNDER

Indicator variable with a value 1 if the chairperson

of the board holds the managerial positions of
CEO or president and a value of zero otherwise.
The cumulative ownership % in the firm held by
insiders (e.g. managers) who serve on the board.
Indicator variable with the value 1 if the firms
founder retains the post of either CEO or Chair.

35

TABLE 3 Panel A
Univariate Results Full Sample
Restatement Firms
(n = 129)

Control Firms
(n = 129)
Diff. in MannMedian Std Dev means Whitney
1.000
0.453 -0.216 8.95***

Variable Name
INDEP

Mean Median Std Dev

0.500
0.500
0.503

Mean
0.716

MINSIZE

0.614

1.000

0.489

0.727

1.000

0.448

-0.113 2.58

NOEXPERT

0.239

0.000

0.429

0.079

0.000

0.272

0.160 8.64***

AUDCHOSE

0.511

1.000

0.503

0.647

1.000

0.480

-0.136 3.38**

MINMEET

0.147

0.000

0.357

0.375

0.000

0.486

-0.210 12.47***

BLOCK

0.128

0.085

0.137

0.162

0.139

0.164

-0.034 2.23

BOARDSIZE

6.988

7.000

2.619

7.375

7.000

2.488

-0.387 1.01

%OUTSIDER

0.609

0.600

0.158

0.629

0.633

0.152

-0.024 0.77

OUTOWN

0.042

0.005

0.092

0.029

0.009

0.054

0.013 1.37

AGEPUB

6.591

7.000

3.307

7.473

10.000

3.028

-0.882 1.44

TROUBLE

0.342

0.000

0.477

0.278

0.000

0.451

0.064 0.73

FINANCE

0.215

0.000

0.892

0.093

0.000

0.355

0.122 1.44

GROWTH

1.111

0.401

2.695

0.892

0.152

3.068

0.219 0.25

CEOCHAIR

0.784

1.000

0.414

0.705

1.000

0.458

0.079 1.46

MGROWN

0.258

0.124

0.694

0.144

0.084

0.159

0.114 2.72*

FOUNDER

0.318

0.000

0.468

0.238

0.000

0.429

0.080 1.38

TOTAL ASSETS

989.3

129.2

1788.2

1067.5

126.2

1972.4

78.2

0.56

Note: Variables are defined in table 2, with the exception of TOTAL ASSETS, which
equals total firm assets in millions.
Significance levels:
*, **, ***

p-value < .10, .05, .01, respectively.

36

TABLE 3 Panel B
Univariate Results Fraud Subsample
Fraud Firms (n = 41)

Control Firms (n = 41)

Diff. in MannMedian Std Dev means Whitney
1.000
0.449 -0.513 28.5***

Variable Name
INDEP

Mean Median Std Dev

0.219
0.000
0.419

Mean
0.732

MINSIZE

0.708

1.000

0.461

0.634

1.000

0.488

0.074 0.48

NOEXPERT

0.365

0.000

0.487

0.098

0.000

0.300

0.267 8.96***

AUDCHOSE

0.536

1.000

0.504

0.585

1.000

0.491

-0.049 0.194

MINMEET

0.146

0.000

0.358

0.292

0.000

0.461

-0.146 2.58

BLOCK

0.095

0.071

0.105

0.119

0.089

0.134

0.006 0.82

BOARDSIZE

7.950

7.000

4.630

7.320

7.000

2.733

0.630 0.57

%OUTSIDER

0.579

0.575

0.193

0.618

0.625

0.175

-0.039 0.85

OUTOWN

0.032

0.007

0.193

0.061

0.010

0.132

-0.029 1.64

AGEPUB

8.912

8.000

3.544

7.888

8.000

3.451

0.024 0.05

TROUBLE

0.171

0.000

0.381

0.268

0.000

0.449

-0.093 1.13

FINANCE

0.293

0.000

0.461

0.219

0.000

0.419

0.074 0.57

GROWTH

1.220

0.036

2.728

0.278

0.152

0.383

0.942 4.80**

CEOCHAIR

0.878

1.000

0.331

0.634

1.000

0.487

0.244 7.62***

MGROWN

0.190

0.163

0.173

0.107

0.027

0.169

0.083 4.84**

FOUNDER

0.488

0.000

0.506

0.171

0.000

0.381

0.317 10.27***

TOTAL ASSETS

726.30

55.80

1477.02

612.37

53.12

930.66 113.93 0.18

Note: Variables are defined in table 2, with the exception of TOTAL ASSETS, which
equals total firm assets in millions.
Significance levels:
*, **, ***

p-value < .10, .05, .01, respectively.

37

TABLE 4

Correlation Matrix
Variable
Name
Indep
Minsize
Noexpert
Audchose
Minmeet

Indep

1. 00
0. 00

Minsize

Noexpert

-0. 098 -0. 19

0. 12
0. 00
1. 00 -0. 22
0. 00
0. 00
1. 00
0. 00

Audchose

Minmeet

0. 08 0. 16
0. 18 0. 01
0. 13 0. 17
0. 04 0. 01
-0. 18 -0. 13
0. 00 0. 03
1. 00 0. 15
0. 00 0. 02
1. 00
0. 00

Block

Block

0. 08
0. 20
0. 01
0. 83
-0. 04
0. 56
-0. 09
0. 16
-0. 02
0. 78
1. 00
0. 00

Boardsize
%Outsiders
Outown
Agepub
Trouble
Finance
Growth
CEOchair
Mgrown

Boardsize %Outsiders Outown

0. 04
0. 55
0. 41
0. 00
-0. 18
0. 00
0. 14
0. 03
0. 27
0. 00
-0. 07
0. 29
1. 00
0. 00

0. 34
0. 00
0. 18
0. 01
-0. 12
0. 05
0. 14
0. 02
0. 21
0. 00
0. 21
0. 00
0. 19
0. 00
1. 00
0. 00

0. 10
0. 12
-0. 12
0. 05
0. 11
0. 17
-0. 17
0. 01
-0. 04
0. 56
0. 12
0. 06
-0. 09
0. 13
0. 12
0. 06
1. 00
0. 00

Agepub

0. 07
0. 30
0. 10
0. 10
0. 04
0. 53
0. 10
0. 13
-0. 02
0. 74
0. 00
0. 97
0. 24
0. 00
0. 21
0. 00
-0. 03
0. 65
1. 00
0. 00

Trouble

-0. 05
0. 40
-0. 15
0. 02
0. 07
0. 31
-0. 05
0. 40
-0. 01
0. 84
-0. 09
0. 14
-0. 18
0. 00
-0. 07
0. 31
0. 26
0. 00
-0. 11
0. 09
1. 00
0. 00

Finance

-0. 08
0. 18
-0. 08
0. 18
0. 12
0. 06
-0. 09
0. 15
-0. 13
0. 04
0. 05
0. 42
-0. 10
0. 10
-0. 07
0. 27
0. 02
0. 69
-0. 06
0. 33
-0. 01
0. 86
1. 00
0. 00

Growth

-0. 15
0. 02
-0. 00
0. 99
0. 02
0. 80
-0. 00
0. 97
-0. 02
0. 77
-0. 04
0. 57
-0. 16
0. 01
-0. 06
0. 35
0. 01
0. 92
-0. 07
0. 26
0. 01
0. 85
0. 11
0. 09
1. 00
0. 00

CEOChair Mgrown Founder

-0. 08
0. 22
0. 01
0. 86
-0. 03
0. 67
0. 01
0. 93
-0. 12
0. 05
-0. 08
0. 20
-0. 03
0. 64
-0. 01
0. 86
-0. 19
0. 00
0. 24
0. 00
0. 06
0. 30
0. 01
0. 88
0. 07
0. 24
1. 00
0. 00

-0. 03
0. 65
0. 00
0. 94
0. 04
0. 48
0. 00
0. 99
-0. 11
0. 08
-0. 09
0. 13
-0. 06
0. 31
-0. 04
0. 50
-0. 06
0. 31
-0. 07
0. 23
-0. 11
0. 09
0. 04
0. 57
0. 01
0. 85
0. 06
0. 30
1. 00
0. 00

-0. 22
0. 00
-0. 04
0. 57
0. 09
0. 14
0. 04
0. 51
-0. 12
0. 06
-0. 09
0. 13
-0. 09
0. 14
-0. 28
0. 00
-0. 15
0. 01
-0. 09
0. 17
-0. 12
0. 07
0. 11
0. 08
0. 15
0. 02
0. 30
0. 00
0. 12
0. 05

Note: Independent variables are defined in Table 2.

(p-values listed below correlation coefficients).

38

TABLE 5
Logistic Regression of Financial Misstatements on
Audit Committee Characteristics and Control Variables
Column A

Column B

Column C

Expected
Sign

Full
Sample
Coef.
Wald
Est.
2
-0.775
1.044

Restatement
Sub-sample
Coef.
Wald
Est.
2
0.198
0.034

Fraud
Sub-sample
Coef.
Wald
Est.
2
-1.720
1.212

INDEP

-1.174*** 13.894

MINSIZE

-0.364

NOEXPERT

1.231***

8.206

AUDCHOSE

-0.305

1.020

MINMEET

-1.028***

BLOCK

-1.356

BOARDSIZE

0.124**

3.444

0.143*

2.549

%OUTSIDER

1.085

1.123

1.635

1.624

-0.458

0.049

OUTOWN

0.288

0.027

1.980

0.704

-0.238

0.003

AGEPUB

-0.011

0.274

-0.107*

2.629

0.017

0.413

TROUBLE

-0.030

0.009

-0.064

0.025

-0.198

0.078

FINANCE

0.038

0.025

-0.033

0.017

0.050

0.004

GROWTH

0.033

0.356

-0.026

0.174

1.479** 3.268

CEOCHAIR

0.502*

2.016

0.143

0.117

0.995

1.519

MGROWN

None

0.579

0.886

0.466

0.811

-1.417

0.386

Independent
Variable
Intercept

1.075

-1.042*** 7.177
-0.529

1.501

1.097** 4.083
-0.449

-2.234*** 9.182
-0.551

0.464

1.668** 2.989

1.561

0.190

0.075

7.701

-1.365*** 9.370

-0.285

0.105

1.530

-2.131** 2.751

-0.444

0.018

0.212*

FOUNDER
+
0.328
0.871
0.152
0.126
0.826
Observations
258
176
82
Adjusted R2
0.2756
0.2623
0.4823
59.74***
55.76***
64.82***
Model 2
Note: Dependent Variable = 1 if financial misstatement is present, 0 otherwise.
Independent variables are defined in table 2.

2.021

1.093

*, **, *** = p-value < .10, .05, .01, respectively .

39

APPENDIX

This appendix contains a brief explanation of the Blue Ribbon Committee (BRC)
recommendations, and consequent implementation by regulatory bodies.

Recommendations 1 and 2: Recommends that audit committee members have no

relationship to the firm which might interfere with exercise of independence from
corporation and management such as: a) employment by the corporation or affiliates, for
the current year or the past five years, b) compensation from the corporation or affiliates
other than receipt of retirement benefits or director compensation, c) immediate family
members of current or former employees, d) serving as a partner, controlling shareholder
or executive of a for-profit business to which the corporation made or from which the
corporation received significant payments. However, directors with one or more of
these relationships might still be appointed to the AC in limited and exceptional
circumstances if the board determines that service is in the best interests of the firms and
its shareholders and explains the exception in the proxy (BRC, 10-11). The BRC
suggested a threshold of 5% of gross revenues or $200,000, which ever is greater, to be
considered significant. Recommendation 2 suggests that all firms with market
capitalizations over $200 million adopt the preceding independence rules for all audit
committee directors (BRC, 11-12).
Implementation: Both the NASD/AMEX and NYSE disqualify current employees and
relatives of current management. Former employees and their relatives are barred from

40

service for a period of three years after severance (not the five years recommended by the
BRC). Compensation committee interlocks are also forbidden. The NASD/AMEX
disqualifies directors whose personal compensation (excluding board fees) is more than
$60,000 per year, and directors who are partners or executives of for-profit businesses
receiving revenues in excess of $200,000 or 5% of total revenues from the company. A
director who is not independent (as long as he/she is not a current employee or relative of
management) may serve if the board determines that audit committee service is in the
best interests of the firm. The NYSE has no compensation-based guidelines, and
continues to allow the full board substantial discretion to assess whether an existing
business relationship impairs independence, and to allow one former employee or relative
of such within the three year severance period to serve. However, both the exchanges
and the SEC have adopted rules requiring that when the Board appoints a nonindependent member, the nature of the relationship that makes an AC member nonindependent and the reasons for the appointment be disclosed.
The most important departure from the BRC recommendations regarding
independence is that the exchanges did not adopt exemptions for smaller firms, as
recommended by the BRC, but apply the new rules to all firms.25 The SEC disclosure
rules also require that public companies which do not trade on major exchanges state
whether, if they have an audit committee, the members are independent, and under which
definition of independence they are judged.

25

This outcome is consistent with the SECs position with respect to the application of disclosure rules
(discussed below) to small registrants. The SEC notes in section 2 of the final rules on audit committee
disclosure that they have seen instances of financial fraud at both small and large companies, and that
improvement in the financial reporting process at all companies is important for promoting investor
conference. One exception is that the NASD/AMEX does not apply its audit committee membership rules
to small business filers.
41

Recommendation 3: Establishes a target audit committee size of at least three members,

each with the ability to read and understand financial statements. Further, one member
should possess financial expertise, demonstrated by past employment experience in
finance or accounting, requisite professional certification in accounting, or any other
comparable experience or background which results in the individuals financial
sophistication, including being or having been a CEO or other senior officer with
financial oversight responsibilities (BRC, 12).
Implementation: Guidelines on literacy and expertise were operationalized differently by
the exchanges.26 The NASD/AMEX substantially adopted the BRC definitions of
expertise and literacy (Rule 4350 (d) (2a)). However, the NYSE elected to leave
substantial discretion in the Boards hands to set expertise requirements: the NYSE rules
call for at least one member of the audit committee to have accounting or related
financial management expertise, as the board interprets in its business judgement (Section
303.01 (B) (2c)).

Recommendations 4, 5: Require the audit committee to have a formal charter and to

disclose the status of the charter and whether the audit committee fulfilled its designated
responsibilities during the proxy statement period (BRC, 12).
Implementation: NASD/NYSE exchange listing rules also incorporate these
recommendations. The SEC has adopted rules requiring firms to include a report from
their audit committee in their proxies (including certain stipulated statements), provide

42

disclosures regarding the independence of their audit committee members (7), and
include in the proxy a copy of the audit committee charter (4).27

Recommendations 6-10: Specify that the outside auditor is ultimately responsible to the
Board and audit committee, and that these shareholder representatives be responsible for
selecting or replacing the outside auditor (6), requiring the committee to evaluate auditor
independence (7), requiring the committee to discuss accounting quality with the auditor
(8), and to hold similar discussions with management and the auditor, and to place a letter
from the audit committee in the annual report and 10-K (or 10-KSB) stating that the
committee has held such discussions and based upon discussions with management and
the auditors, believe the financial statements are fairly presented in accordance with
GAAP (9) (BRC, 14-16).. Recommendation 10 requires timely reviews of quarterly
financial results and discussion of review-related issues with the auditor (BRC, 16).
Implementation: The SECs charter-related regulations require a statement that the
auditor is responsible to the Board and the committee (6), and that the committee must
ensure that it receives and evaluates information on the auditors independence and the
effects that disclosed relationships and services may have on the auditors objectivity (7).
Recommendation 8 was implemented through the issuance of Statement on Auditing
Standards No. 90, Audit Committee Communications (SAS 90), amending Statement on
Auditing Standards No. 61, Communications With Audit Committees (SAS 61).

26

The NASD/AMEX rules are included in section 4350 of the NASD Manual Marketplace Rules and can
be viewed in their entireity at http://nasd.com. The NYSE rules are included in section 303 of the NYSE
Listed Company Manual, which may be viewed at http://www.nyse.com/listed/listed.htm.
27
These rules are effective for proxy statements filed after December 15, 2000, and are contained in 17
CFR Parts 210, 228, 229 and 240, Release No. 34-42266; File No. S7-22-99 and may be viewed at the SEC
web site http://www.sec.gov/rules.
43

Recommendation 9 was implemented through an SEC requirement that the audit

committee to state in the proxy statement whether, based upon review and discussion
with the management and auditors, the committee recommends that the financial
statements be included in the Form 10-K or 10-KSB. Recommendation 10 on timely
quarterly reviews was implemented through an amendment to Rule 10-01(d) of SEC
Regulation S-X and Item 310(b) of Regulation S-B.

44

REFERENCES

Abbott, L. J., Y. Park, and S. Parker. 2000. The effects of audit committee activity and
independence on corporate fraud. Managerial Finance, Vol. 26, No. 26: 55-67.
Abbott, L.J., S. Parker, G. Peters and K. Raghunandan. 2002. Audit committees, audit
fees and non-audit fees. Working paper. Santa Clara University.
Altman, E. I. 1968. Financial ratios, discriminant analysis and the prediction of corporate
bankruptcy. Journal of Finance (September): 589-609.
American Institute of Certified Public Accountants(AICPA). 1988. Communication With
Audit Committees . Statement on Auditing Standards No. 61. New York, NY:
AICPA.
---. 1992. Interim Financial Information. Statement on Auditing Standards No. 71. New
York, NY: AICPA.
----1998. Audit Committee Communications. Statement on Auditing Standards No. 90.
New York, NY: AICPA.
Bailey. J. 2000. Second helping: Star rescuers take on Waste Management and end up
tarnished. The Wall Street Journal (February 29).
Baysinger, B. D., and H. N. Butler. 1985. Corporate governance and the board of
directors: Performance effects of changes in board composition. Journal of Law,
Economics and Organization 1 (Fall): 101-124.
Beasley, M. 1996. An empirical analysis of the relation between the board of director
composition and financial statement fraud. The Accounting Review, Vol. 71, No.
4: 443-465.
Beasley, M., Carcello, J. V., Hermanson, D. and P. Lapides. 2000. Fraudulent financial
reporting: Consideration of industry traits and corporate governance mechanisms.
Accounting Horizons, Vol. 14, No. 4 (December): 441 454.
Bell, T. B. and J. V. Carcello. 2000. A decision aid for assessing the likelihood of
fraudulent financial reporting. Auditing: A Journal of Practice & Theory 19(1):
169-184.
Bell, T. B., S. Szykowny, and J. J. Willingham. 1991. Assessing the likelihood of
fraudulent financial reporting: A cascaded logit approach. Working paper,
KPMG, Peat Marwick, Montvale, NJ.

45

Berton, L. 1995. Corporate woes put board audit panels in the spotlight. The Wall Street
Journal (April 7).
Blue Ribbon Committee (BRC). 1999. Report and Recommendations of the Blue Ribbon
Committee on Improving the Effectiveness of Corporate Audit Committees. New
York, NY: New York Stock Exchange and National Association of Securities
Dealers. 1999. Stamford, CT.
Braiotta, L. 2000. The Audit Committee Handbook Second Edition. New York: John
Wiley & Sons, Inc.
Carcello, J. and T. Neal. 2000a. Audit committee characteristics and auditor reporting.
The Accounting Review, Volume 75, No. 4(October): 453 - 467.
Carcello, J. and T. Neal. 2000b. Audit committee composition and auditor changes.
Working paper. University of Tennessee.
Dechow, P. M., R. G. Sloan, and A. P. Sweeney. 1996. Causes and consequences of
earnings manipulation: An analysis of firms subject to enforcement actions by the
SEC. Contemporary Accounting Research 13: 1-36.
DeFond, M.L and J. Jiambalvo. 1991. Incidence and circumstances of accounting errors.
The Accounting Review 66 (July): 643-55.
DeZoort, F.T. 1998. An analysis of experience effects on audit committee members'
oversight judgments. Accounting, Organizations and Society (January): 1-22.
DeZoort, F.T., Hermanson, D. and R. Houston. 2002a. Audit committee member
support for proposed audit adjustments: Effects of financial report timing,
proximity to earnings target and external auditor consistency. Working paper.
University of Alabama.
Dezoort, F.T.,
,
. 2002b. Audit committee support for auditors: The effects
of materiality, justification and accounting precision. Working paper. University
of Alabama.
DeZoort, F.T. and S. Salterio. 2001. The effects of corporate governance experience and
financial reporting and audit knowledge on audit committee members
judgements. Forthcoming. Auditing: A Journal of Theory and Practice.
Fama, E., and M. Jensen. 1983. Agency problems and residual claims. Journal of Law
and Economics 26 (2): 327-349.
Feroz, E.H., K. Park and V.S. Pastena. 1991. The financial and market effects of the SECs
accounting and auditing enforcement releases. Journal of Accounting Research 29
(Supplement): 107-142.

46

Hylas, R.E. and R.H. Ashton. 1982. Audit detection of financial statement errors. The
Accounting Review 57 (October): 751-65.
Kalbers, L. P., T. J. Fogarty. 1993. Audit committee effectiveness: An empirical
investigation of the contribution of power. Auditing: A Journal of Practice &
Theory (Spring): 24-49.
Knapp, M. C. 1987. An empirical study of audit committee support for auditors involved
in technical disputes with client management. The Accounting Review (July): 578588.
Kreutzfeldt, R.W. and W.A Wallace. 1986. Error characteristics in audit populations:
Their profile and relationship to environmental factors. Auditing: A Journal of
Practice and Theory 6 (Fall): 20-43.
Levitt, A. 1998. Corporate Governance: Integrity in the Information Age Speech
presented at Tulane University, March 12.
Loebbecke, J. K. and J. J. Willingham. 1988. Review of SEC accounting and auditing
enforcement releases. Working paper. University of Utah.
Loebbecke, J. K., M. M. Eining, and J. J. Willingham. 1989. Auditors experience with
material irregularities: Frequency, nature, and detectability. Auditing: A Journal
of Practice and Theory 9 (Fall): 1-28.
Lublin, J. S. and E. MacDonald. 1998. Scandals signal laxity of audit panels. The Wall
Street Journal (July 17): B1, B5.
______, and E. Nelson. 1998. Cendant audit panels ties raise questions. The Wall Street
Journal (July 24): C1, C4.
Maines, L., Martin, R. and L. McDaniel. 2002. The effects of financial expertise versus
financial literacy. Forthcoming. The Accounting Review.
McMullen, D. A. 1996. Audit committee performance: An investigation of the
consequences associated with audit committees. Auditing: A Journal of Practice
and Theory 15, (Spring): 87- 103.
Menon, K., and J. D. Williams. 1994. The use of audit committees for monitoring.
Journal of Accounting and Public Policy 13 (Spring): 121-139.
National Association of Corporate Directors (NACD). 1999. Report of the NACD Blue
Ribbon Commission on Audit Committees. Washington, DC: NACD.
Palmrose, Z.V. and S. Scholz. 2000. Auditor independence and non-audit services.
What do the restatements suggest? Working paper. University of Kansas.

47

Parker, S. 1997. The effect of audit committees on financial reporting. Unpublished

dissertation, University of Oregon.
Public Oversight Board. 1993. A Special Report by the Public Oversight Board of the
SEC Practice Section, AICPA.
Reinstein, A., J. Callaghan, and L. Braiotta Jr. 1984. Corporate audit committees:
Reducing directors' legal liabilities. Journal of Urban Law 61: 375-389.
Sahlman, W. A. 1990. Why sane people shouldnt serve on public boards. Harvard
Business Review (May-June): 28-35.
Schellhardt, T. D. 1999. More directors are raking in six-figure pay. The Wall Street
Journal (October 29).
Treadway Commission. 1987. The report of the national commission on fraudulent
financial reporting, National Commission on Fraudulent Financial Reporting,
Washington D.C.
Wallace, W.A and R.W. Kreutzfeldt. 1995. The relation of inherent and control risks to
audit adjustments. Journal of Accounting, Auditing and Finance (Summer): 45984.
Weisbach, M. S. 1988. Outside directors and CEO turnover. Journal of Financial
Economics 20: 431-460.

48