Audit Committee Financial Expertise_ Corporate Governance and by sdfgsg234

VIEWS: 15 PAGES: 52

									   Audit Committee Financial Expertise, Corporate Governance and
             Accruals Quality: An Empirical Analysis




                                              Dan Dhaliwal
                                       Eller College of Management
                                           University of Arizona
                                         Tucson, AZ 85721 -0108
                                        dhaliwal@eller.arizona.edu

                                              Vic Naiker
                              The University of Auckland Business School
                                       University of Auckland
                                       Auckland, New Zealand
                                       v.naiker@auckland.ac.nz

                                              Farshid Navissi
                                  Department of Accounting and Finance
                                            Monash University
                                    Caulfield East, Vic 3145, Australia
                                 Farshid.Navissi@BusEco.monash.edu.au


                                        Current Version: April 2007




 Corresponding Author
Tel.: +1-520-621-2366
Fax: +1-520-621-3742
Email: dhaliwal@eller.arizona.edu

We thank participants at the 2006 American Accounting Association Annual Meeting and seminar participants at the
The University of Auckland, Monash University, Queensland University of Technology and LaTrobe University. We
are also indebted to Steven Cahan, Jerry Bowman, David Emanuel, Nicole Jenkins for their insightful suggestions.
   Audit Committee Financial Expertise, Corporate Governance and
             Accruals Quality: An Empirical Analysis



Abstract
The Sarbanes Oxley Act advocates the presence of financial experts on audit committees. However,
the requirement proved controversial, culminating with the stock exchanges adopting a broad
definition of financial expertise. We investigate the association between three types of audit
committee financial expertise (accounting, finance and supervisory expertise) and accruals quality,
in the presence of strong audit committee or board governance. Results indicate a positive relation
between accounting expertise and accruals quality, which is more pronounced in the presence of
strong audit committee governance. The findings suggest that future refinements to the financial
expertise definition must focus on accounting expertise.




Keywords: Audit committee; Accounting Expertise; Accruals quality
JEL Descriptors: M41, M43, M49, G34, G18, K22
Data Availability: Data are available from sources identified in the paper.
   Audit Committee Financial Expertise, Corporate Governance and
             Accruals Quality: An Empirical Analysis

1. Introduction

       The emergence of recent high profile fraudulent cases has motivated the U.S. Congress into

enacting the Sarbanes Oxley Act of 2002 (hereafter SOX), creating additional standards in the

manner in which the audit committee of the firm carries out its responsibility of preserving financial

reporting integrity. SOX has also introduced requirements relating to the composition of audit

committees, including one that requires the Securities Exchange Commission (hereafter SEC) to

adopt rules mandating audit committees of public firms to include at least one member who is a

financial expert or to disclose reasons for not adopting this requirement. While SOX proposed a

narrow definition of financial expertise to include individuals with experience in accounting or

auditing, the SEC and the U.S. stock exchanges controversially adopted a broader definition of

financial expertise, under which financial expertise could include accounting expertise, or any

experience in supervising employees with financial responsibilities and overseeing the performance

of companies.

       However, numerous prior studies have failed to provide strong evidence that financial

expertise (under the broad definition) positively influences audit committee effectiveness (Carcello

and Neal, 2003; Anderson et al., 2004; Lee et al., 2004; DeFond et al., 2005). The evidence,

therefore, suggests that the current definition of financial expertise may indeed be too encompassing

and lacks the effectiveness to ensure high financial reporting quality. A number of studies have

indirectly acknowledged the inefficiency of the broad definition of financial expertise by adopting

narrower versions of the definition that capture accounting and finance expertise. These studies

have produced more conclusive results documenting financial reporting advantages from the

presence of accounting or finance expertise on audit committees (Bédard et al., 2004; Archambeault


                                                  1
and DeZoort, 2001; Raghunandan et al., 2001; Raghunandan and Read, 2001; Krishnan, 2005).

However, these studies have not examined the influence of accounting and finance expertise

separately. This is important given that there is evidence to suggest that any expertise obtained as a

result of professional accounting and auditing certification or from any other accounting related

work experience contributes more significantly to audit committee effectiveness (DeFond et al.,

2005; Kalbers and Fogarty, 1993; McDaniel et al., 2002; DeZoort, 1997, 1998; Archambeault and

DeZoort, 2001; McMullen and Raghunandan, 1996; DeZoort and Salterio, 2001; Agrawal and

Chadha, 2005).

        This study decomposes the broad definition of financial expertise into three specific types of

financial expertise (accounting, finance, or supervisory expertise) and investigates which (if any) of

these three types of financial expertise in audit committees is most strongly associated with higher

quality financial reporting (proxied by accrual quality). More importantly, we contribute to the

extant literature on audit committee effectiveness, by investigating whether the effect of each type

of expertise on financial reporting quality is more pronounced, when complemented with strong

board or audit committee governance. Prior studies in this research stream have not considered

whether the three expertise types interact with other audit committee and board characteristics to

contribute to audit committee effectiveness. Hence, our study provides a better understanding of

which financial expertise and governance attributes are most likely to interact in promoting

financial reporting quality. Furthermore, previous studies utilize unique sample firms with

identifiable financial reporting problems, hence it is unclear whether their findings can be

generalized to firms that are not experiencing such problems.1 We use a sample of publicly traded

firms that, a priori, have no systematic discrepancies in financial reporting quality. Given that the

 1. The corporate events examined in these studies include the following: SEC enforcements and restatements
(McMullen and Raghunandan, 1996; Agrawal and Chadha, 2005; Farber, 2005), internal control problems (Krishnan,
2005), suspicious auditor switches and protection of auditors from dismissals (Archambeault and DeZoort, 2001;
Carcello and Neal, 2003). See Cohen et al. (2004) for a summary of dependent variables examined in studies relating to
board and audit committee characteristics. DeFond and Francis (2005) state that studies in this field that focus on
corporate irregularities have lower statistical power due to their smaller sample sizes.


                                                          2
SEC and U.S. stock exchanges still employ the broad definition of financial expertise, our results

can have significant implications for corporate governance policy setting by the SEC, by assessing

which of the three specific expertise types is likely to, either on its own or in conjunction with

strong corporate governance, offer the most significant contribution to audit committee

effectiveness.

        The results indicate that audit committee accounting expertise is positively associated with

accruals quality, suggesting that the specialized skills possessed by accounting experts make them

more effective in executing the audit committee’s primary responsibility of ensuring higher quality

financial reporting. Further, we find a significant positive interaction effect between audit

committee accounting expertise and strong audit committee governance, suggesting that the positive

association between audit committee accounting expertise and accruals quality is more pronounced

for firms with strong governance in audit committees.2 This finding suggests that, while accounting

experts in audit committees may ensure financial reporting quality, their ability to do so is

contingent upon strong governance in audit committees. Furthermore, the interaction between audit

committee accounting expertise and strong board governance is not statistically significant,

indicating that the effectiveness of accounting expertise is independent of board governance

strength. We find no evidence of an association between accruals quality and the presence of

finance or supervisory expertise in audit committees, even when complemented with strong audit

committee or board governance. Our results are generally insensitive to alternative measures of

accruals quality, alternative attributes of earnings quality, alternative measures of audit committee

governance strength, and controlling for self-selection bias.


 2. We measure the strength of audit committee governance using a summary measure that is based on three audit
committee characteristics that have been previously employed to substitute for audit committee effectiveness. These
characteristics include audit committee size, independence and number of meetings. We acknowledge that audit
committee financial expertise can also be employed as an additional characteristic in developing a summary measure of
audit committee governance strength. Hence, the results should be interpreted carefully. Specifically, our results show
how the positive association between audit committee financial expertise and accruals quality is stronger for firms
where other important audit committee characteristics collectively signal strong audit committee governance.


                                                          3
          The remainder of this study is organized as follows. Section 2 discusses background and

prior literature in this research stream. Section 3 describes sample selection procedure and data.

Section 4 explains the variable measurement, and Section 5 discusses the research design. Section 6

reports descriptive statistics, and the empirical results are discussed in Section 7. Section 8 reports

the results from additional tests, and Section 9 concludes the study.



2. Background and prior literature

         Audit committees have been long seen as a vital institution in assisting the board of directors

in enhancing the transparency and integrity of financial reporting (AICPA, 1967; SEC, 1972, 1974;

NYSE, 1973; Treadway Commission, 1987; COSO, 1992; POB, 1993; BRC, 1999; SOX, 2002).

Specifically, effective audit committees are expected to enhance financial reporting quality by

fulfilling its numerous responsibilities including, commenting on and approving accounting

policies, reviewing the financial statements, and maintaining and reviewing the adequacy of internal

controls. Moreover, audit committees are also expected to play an important role in enhancing the

effectiveness of external auditors over financial reporting quality by, assuming responsibilities for

the appointment and remuneration of external auditors, and discussing the scope of and reviewing

the auditors work.

         However, prior research indicates that the construct of audit committee effectiveness over

financial reporting is multidimensional and is affected by variety of audit committee characteristics

such as committee size ( Anderson et al., 2004; DeZoort and Salterio, 2001), independence (Klein,

2002; Bedard et al., 2004) and number of meetings (Menon and Williams, 1994; Beasley et al.,

2000).    Audit committee member financial expertise is another important dimension of audit

committee effectiveness that has gained the attention of regulators and academics (Treadway

Commission, 1987; GAO, 1991; POB, 1993; Kalbers and Fogarty, 1993; DeZoort, 1997, 1998;

BRC, 1999; SOX, 2002).         Advocates propose that the presence of financial experts in audit


                                                    4
committees will assist the committee in, critically analyzing accounting policies and financial

statements, identifying potential problems, and solving problems.

         Accordingly, Section 407 of SOX requires the SEC to adopt rules mandating that audit

committees of public firms must comprise at least one member who is a financial expert or

otherwise, and to disclose reasons for not adopting this requirement. However, the definition of

financial expertise adopted by SOX was very narrow and proved to be very controversial (Braswell

and Mauldin, 2004; DeFond et al., 2005; SEC, 2003). The definition classified individuals as

financial experts if they had obtained education and experience in accounting and auditing.3

However, critics believed that this definition was overly restrictive and severely limited the pool of

qualified financial experts. It was argued that even Alan Greenspan and Warren Buffett would not

have qualified as financial experts under the new definition (American Association of Bank

Directors, 2002; Bryan-Low, 2002).4 In response, the SEC adopted a more liberal definition of

financial expertise in its final rules, under which financial expertise could include accounting

expertise, or any experience in supervising employees with financial responsibilities and overseeing

the performance of companies.5 The wide-ranging definition of financial expertise was

subsequently adopted by the NASDAQ and AMEX, while the NYSE also implicitly adopted a

broad definition by delegating the task of interpreting financial expertise to the board of their



 3. These skills include (i) ability to understand financial statements and generally accepted accounting principles; (ii)
ability to assess the general application of such principles in connection with the accounting for estimates, accruals and
reserves; (iii) experience preparing, auditing, analyzing or evaluating financial statements that present a breadth and
level of complexity of accounting issues that are generally comparable to the breadth and complexity of issues that can
reasonably be expected to be raised by the registrant’s financial statements, or experience actively supervising one or
more persons engaged in such activities; (iv) ability to understand internal controls and the procedures for financial
reporting; and (v) ability to understand audit committee functions.
 4. Alan Greenspan is the Chairman of the Federal Reserve Board, and Warren Buffett is one of the most successful
investors in the United States.
 5. A similar controversy surrounded the 1999 Blue Ribbon Committee (BRC) recommendations on improving the
effectiveness of audit committees. For example, the BRC in its initial recommendations required audit committee
members to certify the conformity of the financial statements with Generally Accepted Auditing Principals (GAAP).
This requirement indicated that the BRC desired audit committee members to have expertise in the field of accounting.
The requirement proved to be very controversial, eventually leading to BRC removing the certification requirement and
replacing it with a requirement that required audit committee members to only state whether they believed that the
audited financial statements contained omissions or any untrue statements. See Scott (2001) for a complete discussion.


                                                            5
registrants. Under the broad definition, an audit committee member could be deemed a financial

expert if the member has work experience as a certified public accountant, auditor, chief financial

officer, financial comptroller, financial controller or accounting officer, or any work experience in

finance positions or as a chief executive officer (hereafter CEO) or company president. Hence,

financial expertise could comprise accounting or finance expertise, or any expertise entailed in

supervising the preparation of financial statements (supervisory expertise).

       Abbott et al. (2004), Abbott et al. (2003), and Farber (2005) employ the current broad

definition of financial expertise and document lower instances of earnings restatements, higher

demand for audit services and lower occurrence of financial fraud in firms with financial expertise

in audit committees. On the other hand, Carcello and Neal (2003), DeFond et al. (2005), Lee et al.

(2004), and Anderson et al. (2004) use a similar definition but fail to document any strong financial

reporting advantages arising from audit committee financial expertise. These results cast doubts on

the ability of some types of financial expertise to significantly influence audit committee

effectiveness. Indeed, anecdotal evidence suggests that supervisory responsibility (as in the case of

a CEO or President) does not ensure an adequate understanding of accounting matters for an audit

committee member (Livingston, 2003). Plitch and Ceron (2003) state that the majority of CEOs

would not refer to themselves as financial experts. These views question the ability of CEOs and

Presidents to significantly contribute to the monitoring role of audit committees. A number of

studies have indirectly acknowledged the inefficiency of the broad definition of financial expertise

by adopting narrower versions of the definition that capture accounting or finance expertise. These

studies have collectively produced more consistent results documenting numerous financial

reporting advantages from the presence of financial expertise in audit committees (Bédard et al.,

2004; Archambeault and DeZoort, 2001; Raghunandan et al., 2001; Raghunandan and Read, 2001;

Krishnan, 2005). However, since these studies have not examined the influence of accounting or




                                                  6
finance expertise separately, it is possible that the results of these studies are being driven by one of

these expertise types (e.g., accounting expertise).

         Accounting expertise may be more important for audit committee members than any other

expertise, since “best practices” suggest that audit committee members are responsible for tasks that

require high degrees of accounting sophistication. Prior studies argue that financial reporting issues

involve the highest level of technical detail among audit committee effective areas (Kalbers and

Fogarty, 1993; Green, 1994), and ideal audit committee members should have knowledge of

accounting concepts and the auditing process to enhance their understanding of the financial

reporting process, recognize problems, ask probing questions of the management and auditor and

make leadership contributions to audit committees (McDaniel et al., 2002; Libby and Luft, 1993;

Bull and Sharp, 1989; Lipman, 2004; Scarpati, 2003).6 Archival evidence suggests that audit

committee accounting expertise is negatively associated with SEC enforcements and restatements

(McMullen and Raghunandan, 1996; Agrawal and Chadha, 2005)7 and suspicious auditor switches

(Archambeault and DeZoort, 2001), and positively associated with firm credit ratings (Ashbaugh-

Skaife et al., 2006) and the likelihood of supporting auditors in financial reporting disputes with

management (DeZoort and Salterio, 2001). DeFond et al. (2005) document positive market

reactions to the appointment of new audit committee members with accounting expertise, but no

reactions to the appointment of audit committee members with non-accounting expertise. It is

therefore likely that accounting expertise, relative to other expertise, can contribute more to the

effectiveness of audit committees.

         The ability of financial experts to ensure financial reporting quality may, however, be

contingent upon the presence of a strong board and audit committee governance environment. The

 6. Audit partners and audit committee members themselves believe that members skilled in accounting are more
likely to collaborate with other corporate participants, such as internal auditors, and serve as effective guardians of the
financial reporting process (Cohen et al., 2002; DeZoort, 1997, 1998).
 7. Braswell and Mauldin (2004) point out that the results of Abbott et al. (2004) are qualitatively weaker than those of
Agrawal and Chadha (2005), which could signal that the results of Abbott et al. (2004) using the broad definition of
financial expertise were driven by accounting expertise that was not separately tested.


                                                            7
report of the Blue Ribbon Committee (1999) states that “audit committee performance relies on the

practices and attitudes of the entire board.” Prior research also suggests that firms with strong board

governance attributes (e.g., director independence and share ownership) are more likely to support

and empower audit committees (Menon and Williams, 1994; Collier and Gregory, 1999; Cohen et

al., 2002; Bédard et al., 2004). Beasley and Salterio (2001) find that firms with strong board

governance attributes are more likely to voluntarily form audit committees composed of members

with relevant financial reporting and audit committee knowledge and experience. Studies on audit

committee governance attributes (e.g., audit committee size, independence, and number of

meetings) suggest that such attributes actively promote more effective utilization of any available

specialist knowledge and expertise. For instance, Abbott and Parker (2000) find that independent

audit committees and audit committees that meet more often are more likely to perceive industry

specialization as an important skill in external auditors and accordingly appoint industry specialist

auditors. DeZoort and Salterio (2001) examine disputes between auditors and corporate

management, and find that independent audit committee members are more likely to understand the

external auditors’ position, particularly when the position is based on economic substance. Further

support is provided by Archambeault and DeZoort (2001), who find that audit committee

independence and size are negatively related to the probability of firms dismissing auditors

subsequent to the disclosure of a reportable event (e.g., receiving a modified audit opinion). While

these studies focus on the effective utilization of auditor expertise, some recent studies argue that

strong audit committee governance attributes may even lead to more effective use of any expertise

possessed by individual audit committee members (Cohen et al., 2004).

       DeFond et al. (2005) argue that firms with strong corporate governance are more likely to

value accounting experts on audit committees because such expertise complements the

effectiveness of a strong corporate governance setting. Their findings indicate that the positive

market reaction to the appointment of accounting experts to audit committees is largely


                                                  8
concentrated among firms with relatively strong corporate governance. This result suggests that

investors perceive the benefits of an accounting expert to be contextual and contingent upon firms

possessing strong governance. However, whether the market’s perception results in enhanced

financial reporting quality is an unanswered question. A concurrent working paper, Carcello et al.

(2006), fails to document any evidence of a pronounced association between audit committee

accounting expertise and financial reporting quality (discretionary accruals) in the presence of

strong overall governance, constructed from a broad range of corporate governance mechanisms.

However, this insignificant finding could be due to the effect of audit committee accounting

expertise being more pronounced when complemented with specific corporate governance

mechanisms. Indeed, DeFond and Francis (2005) state that it is important for researchers to evaluate

whether certain governance attributes are more important than others in complementing financial

expertise. In this study, we attempt to address this issue by examining whether the relationship of

the three different types of financial expertise (accounting, finance and supervisory) with accruals

quality is more profound when complemented with a strong audit committee or board governance

environment.8,9



3. Sample selection and data

         Panel A of Table 1 outlines the sample selection procedure. The original sample is based on

firms with coverage in the Investor Responsibility Research Center board practices database

(IRRC).


 8. Controlling for total population of corporate governance mechanisms is onerous as it is time consuming and costly
to obtain data for the various dimensions of governance. Furthermore, it is difficult to outline an appropriate structural
model than links the various governance components together.
9. The study by Carcello et al. (2006) came to our attention following the presentation of our paper at the 2006
American Accounting Association annual meeting. There are at least two other notable differences between our study
and Carcello et al. (2006). First, Carcello et al. (2006) consider the financial expertise type of individuals from 283
firms who are designated the status of financial experts under the SOX requirements, while we consider the financial
expertise of all audit committee members from a considerably larger sample (1,114 firms). Second, our study measures
the financial reporting quality of each sample firm over a longer period and considers additional sources of biographical
information to determine the financial expertise of audit committee members.


                                                            9
                                 <<< INSERT TABLE 1 ABOUT HERE >>>

         The IRRC database contains corporate governance information on company directors, such

as director independence, committee memberships, work experience, and other related data. Using

this database, we identify an initial sample of directors who are audit committee members within

each firm covered by IRRC during the period 1995 through 1998.10,11 The initial sample includes

6,021 individuals who served as audit committee members in 2,034 firms. We require our sample

firms to have coverage for financial data on the Compustat Quarterly Files from the period 1995 to

1998.12 This results in 19,436 usable Compustat observations for 1,749 firms that have 5,394

individuals in total serving as audit committee members.13 Further, the sample firms must have at

least eight quarters of data for computing changes in working capital accruals, cash flow from

operations and changes in sales. This reduces the sample size to 18,909 usable quarterly Compustat

observations for 1,418 firms with 3,991 individuals serving as audit committee members. Similar to

previous studies, we exclude banks, financial institutions, utilities, and firms in regulated industries,

which reduces the sample size to 15,269 observations for 1,114 firms with 3,176 individuals serving




10. We choose our sample period as 1995–1998 after considering several factors. First, we restrict our study to a period
prior to 1999, in order to examine firms that voluntarily adopted the best governance practices related to audit
committee expertise, size, independence and meetings before the release of the report and recommendations of the Blue
Ribbon Committee on improving the effectiveness of corporate audit committees in 1999. DeFond and Francis (2005)
and Bédard et al. (2004) also suggest that it is optimal for future researchers to employ such time periods to investigate
issues such as those examined in this study. Second, the measure of accruals quality is based on a modified version (see
McNichols 2002) of the model employed by Dechow and Dichev (2002) and is constructed using data from a number of
interim periods. In order to maximize the number of interim observations used in computing our accruals quality
measure, we select 1995 as the starting point of our sample period, as the IRRC database began its coverage around
1995. To maintain consistency, we developed measures of independent variables over the same time period used to
measure the dependent variable.
11. Since proxy statements form the initial source of the information provided in IRRC, we used 1996 disclosures to
extract audit committee details relating to the 1995 year, 1997 disclosures to extract details for the 1996 year and so on.
12. Similar to Dechow and Dichev (2002), we require the sample firms to have at least eight periods of financial
information in order to assess the quality of accruals. We use quarterly data from the statement of cash flow as using
annual data will result in each firm consisting of a maximum of only four observations in our sample period 1995–1998.
We control for seasonality issues that are inevitably induced by the use of quarterly data.
13. Useable Compustat observations are quarters with complete data on cash flow from operations, change in sales, and
data used to calculate working capital accruals.


                                                           10
as audit committee members.14 Finally, we extract information on any missing audit committee

members by examining the proxy statements of each firm. This results in a final sample of 3,510

individuals serving as audit committee members in 1,114 firms with 15,269 quarters of usable data

available on Compustat.15 Panel B of Table 1 describes the industry membership of firms in the

final sample, assigned on the basis of two-digit SIC codes. Fifty-three industries are represented by

sample firms, with 30 industries represented by at least 10 firms. Further, approximately 36 percent

of the sample firms are collectively represented by firms operating in the four industries: chemicals

and allied products (9.34 percent), industrial machinery and equipment (7.90 percent), electronic

and other electric equipment (8.53 percent), and business services (9.87 percent ).



4. Measurement of variables

         We employ all available data from Compustat, IRRC, proxy statements, and Hoovers during

the sample period 1995 through 1998 to construct the measures of the independent variables and the

dependent variable (accruals quality).

4.1 Accrual quality

         The measure of accruals quality is derived from a modified version of the Dechow and

Dichev (2002) model, suggested by McNichols (2002), and is based on the standard deviation of

residuals (STDRES) from firm-specific regressions of changes in working capital accruals on

lagged, current, and future cash flows from operations, change in sales, and quarterly dummy

variables, estimated over the sample period. (See Appendix).16,17 The residuals would intuitively



14. These firms often have incentives to report earnings levels linked to regulatory oversight or linked economic
benefits flowing to regulators. To focus on cases without these complications, we exclude firms with SIC codes
between 4400 and 5000 and between 6000 and 6500 from the sample.
15. The final sample is considerably larger than the sample sizes employed in recent archival audit committee studies
such as Abbott et al. 2004 (176 firms), Lee et al. 2004 (190 firms), Bédard et al. 2004 (300 firms), Anderson et al. 2004
(252 firms), Ragunandan and Rama 2003 (199 firms), Geiger and Rama 2003 (132 firms), Carcello and Neal 2003 (374
firms), Abbott et al. 2003 (492 firms), Klein 2002 (692 firms), and Beasley and Salterio 2001 (627 firms).
16. We chose to examine working capital accruals rather than total accruals as prior studies find a larger presence of
managerial discretion in relation to working capital accruals (Dechow et al., 1996).


                                                          11
represent abnormal accruals. Under this measure of accruals quality, larger standard deviation of the

residuals (higher volatility in abnormal accruals) would imply lower accruals quality.

        As with any other proxy of financial reporting quality, our measure of accruals quality is

subject to limitations. For example, it is possible that a negative contemporaneous association

between accruals and cashflows undermines the ability of the Dechow and Dichev (2002) model to

capture financial reporting quality (Wysocki, 2006). Hence, to corroborate our findings, we report

results from sensitivity tests, later in the paper, that replicate our analyses using alternative

measures of accruals and earnings quality including earnings persistence, predictability and

smoothness.

4.2 Financial expertise

        We decompose SEC-defined financial expertise into three specific types of financial

expertise, namely, accounting, finance, and supervisory expertise. We assign accounting expertise

to audit committee members who currently have (or have previously had) work experience as

certified public accountants, chief financial officers, vice presidents of finance, financial controllers,

or any other major accounting positions. Finance expertise is assigned to audit committee members

who currently have (or have previously had) work experience as investment bankers, financial

analysts, or any other financial management roles. This is because regulators also consider those

individuals with experience in analyzing or evaluating financial statements as financial experts.

Finally, we assign supervisory expertise to audit committee members who currently have (or

previously have had) work experience as chief executive officers or company presidents.18 On the

basis of this classification scheme, we construct three dichotomous independent variables. The first


17. Our study employs data from the statement of cash flows and a modified version of the methodology employed in
Dechow and Dichev (2002) to measure accruals quality. Collins and Hribar (2002) argue that accrual measures derived
from the statement of cash flows are less likely to contaminate computations of abnormalities in accruals, while Francis
et al. (2005) state that uncertainty in accruals is best captured by the accruals quality measure developed by Dechow
and Dichev (2002).
18. Audit committee members who have more than one type of expertise are assigned to multiple categories to control
for the different types of expertise they possess.


                                                          12
variable, ACCTG, is coded 1 if the audit committee includes at least one member with accounting

expertise in each year during the sample period, and 0 otherwise. Similarly, FINANCE (SUPER) is

coded 1 if the audit committee includes at least one member with finance (supervisory) expertise in

each year during the sample period, and 0 otherwise.

        Prior research predominantly uses a firm’s proxy statement to determine the financial

expertise of the firm’s audit committee members. Singular proxy statements typically contain

information about the work experience of audit committee members for recent years only. For

example, some proxy statements simply state that an audit committee member has been a director of

the company for the last five years. In some cases it is difficult to accurately determine the type of

financial expertise possessed by an audit committee member from a singular proxy statement

(DeZoort et al., 2002; DeFond and Francis, 2005). We consider two additional sources of

biographical information when constructing the measures of financial expertise: (1) proxy

statements of other companies in which the member has held directorships during the sample

period, and the Hoovers database.19

4.3 Corporate governance

4.3.1 Audit committee governance

          We capture the governance strength of each sample firm’s audit committee by

constructing a summary measure (AC GOV SCORE) that captures of the overall strength of the

audit committee. Specifically, AC GOV SCORE is derived from three commonly used audit

committee characteristics: audit committee size (AC SIZE), audit committee independence (AC

IND), and audit committee meetings (AC MEET). To develop the summary measure, we create

dichotomous measures of the three audit committee governance characteristics for each sample




19. Our conclusions remain unchanged when we replicate our analyses after measuring financial expertise using
singular proxy statements only.


                                                     13
firm, with a value of 1 representing strong governance and a value of 0 representing weak

governance:

1) Audit committee size (AC SIZE)—Firms with larger audit committees devote more resources to
   oversee the financial reporting and internal control systems (Anderson et al., 2004) and facilitate
   quality discussions among audit committee members (DeZoort and Salterio, 2001).20 Empirical
   evidence indicates that firms with larger audit committees are less likely to make suspicious
   auditor switches (Archambeault and DeZoort, 2001) and more likely to have lower cost of debt
   (Anderson et. al., 2004). Since stock exchanges now require their registrants to have at least
   three directors on the registrants’ audit committees, we code sample firms 1 if they have at least
   three members on their audit committee in each year during the sample period, and 0 otherwise.

2) Audit committee independence (AC IND)—There is considerable evidence depicting a positive
   relationship between audit committee independence and financial reporting integrity (Klein,
   2002; Bédard et al., 2004). Anderson et al. (2004) provide evidence that fully independent audit
   committees are associated with a significantly lower cost of debt. We code sample firms 1 if
   their audit committees are entirely composed of independent members in each year during the
   sample period, and 0 otherwise.21

3) Audit committee meetings (AC MEET)—Menon and Williams (1994) argue that audit
   committees that do not meet or meet only once are unlikely to be effective monitors while audit
   committees that meet several times exert more serious efforts in monitoring management. Other
   supporting evidence indicates that firms whose audit committees meet less often are more likely
   to engage in fraudulent behavior (Beasley et al., 2000), face reporting problems (McMullen and
   Raghunandan, 1996), and make suspicious auditor switches (Archambeault and DeZoort, 2001).
   We code sample firms 1 if the audit committee met at least four times in each year during the
   sample period, and 0 otherwise.22




20. Increasing the number of individuals involved in an activity substantially decreases the opportunity for wrongdoing
by reducing the likelihood of collusion among the individuals (Kiger and Scheiner, 1997).
21. Independent audit committee members include members who are not employees, not related to an executive officer
of the firm, not receiving excess compensation from the firm (except for the directors’ fees), not shareholders or
executive officers in other firms that have significant economic ties with the firm, and not members who hold
compensation committee interlocking directorships in other firms.
22. Our threshold is based on the recommendations of the Blue Ribbon Committee (1999) and other governance
institutions (e.g., National Association of Corporate Directors, 1999). Prior studies that have employed a threshold of
four audit committee meetings include Abbott et al. (2003, 2004) and Carcello et al. (2002).


                                                         14
       The three dichotomous variables are then summed up to obtain the summary governance

measure (AC GOV SCORE). Such a measure is likely to better capture the strength of a firm’s

overall governance environment than individual measures (DeFond et al., 2005; Bushman et al.,

2004). An overall measure of audit committee strength (ACGOV) is then constructed by coding it 1

if a sample firm’s AC GOV SCORE is greater than or equal to two (strong audit committee

governance), and 0 otherwise.

4.3.2 Board governance

       Using an approach similar to that discussed above, we construct a summary measure of the

strength of a firm’s board (BD GOV SCORE) employing four commonly used board characteristics:

board size (BOARD SIZE), board independence (BOARD IND), director shareholding (SHARE

OWN), and CEO-Chair duality (DUALITY). To construct the summary measure, we create the

following dichotomous variables for the four board governance characteristics:


1) Board size (BOARD SIZE)—Jensen (1993) and Lipton and Lorsch (1992) suggest that large
   boards are less effective monitors and are easier for CEOs to influence. Yermack (1996),
   Eisenberg et al. (1998), and Loderer and Peyer (2002) find a significant negative relationship
   between board size and firm value. We code sample firms 1 if the average board size during the
   sample period is less than the sample median, and 0 otherwise.

2) Board independence (BOARD IND)—Outside directors are argued to have incentives to carry
   out their monitoring tasks more effectively, not collude with top managers in expropriating
   shareholder wealth (Fama and Jensen, 1983), and objectively question and evaluate
   management performance (Carcello and Neal, 2003). Existing empirical evidence indicates that
   independent directors are associated with stronger corporate governance and financial reporting
   integrity (Rosenstein and Wyatt, 1990; Dechow et al., 1996; Beasley, 1996; Core et al., 1999).
   Accordingly, we code sample firms 1 if the average proportion of independent directors in the
   firm during the sample period is greater than 60 percent , and 0 otherwise.

3) Share ownership (SHARE OWN)—Higher equity ownership on the part of the directors is likely
   to motivate them to question managerial policies (Patton and Baker, 1987), because directors’
   decisions impact their wealth (Minow and Bingham, 1995). Prior studies have documented that
   larger stock ownership by directors is positively related to financial reporting quality (Gerety
   and Lehn, 1997; Beasley, 1996; Shivdasani, 1993). We code sample firms 1 if the average
   cumulative percentage of stock held by outside directors during the sample period is larger than
   the sample median, and 0 otherwise.




                                                15
4) CEO-Chair Duality (DUALITY) —Jensen (1993) posits that CEOs who also hold the position of
   board chairman (Duality) exert undue influence on the board, compromising the strength of the
   board’s governance. Existing empirical evidence suggests that Duality is associated with higher
   instances of SEC accounting enforcement actions for alleged violations of GAAP (Dechow et
   al., 1996; Farber, 2005), higher control of risk and higher audit effort resulting in higher audit
   fees (Tsui et al., 2001), lower sensitivity of CEO turnover to firm performance (Goyal and Park,
   2002), and lower instances of voluntary corporate disclosure (Gul and Leung, 2004). We code
   sample firms 1 if the CEO and board chairman positions are not held by the same individual
   during the sample period, and 0 otherwise.


        The four dichotomous board governance variables are then added up to obtain the BD GOV

SCORE. An overall measure of board strength (BDGOV) is then constructed by coding it 1 if a

sample firm’s BD GOV SCORE is greater than or equal to three (strong board governance), and 0

otherwise.

4.4 Control variables

        Dechow and Dichev (2002) identify several innate factors that affect accruals quality. They

document that accruals quality is negatively related to standard deviation of cash flow from

operations (σ(CFO)), frequency of reporting negative earnings (NEGEARN), standard deviation of

sales revenue (σ(SALES)), and operating cycle (OPCYCLE) while accruals quality is positively

related to the log of total assets (ASSETS). Therefore, we control for these variables. To control for

seasonality in sales, and cash flow from operations, we subtract from these variables, their mean

values calculated over corresponding quarters during a five-year prior to the sample period (1990

through 1994), prior to computing the variables’ standard deviations.23 In addition, we control for

three additional innate factors that Francis et al. (2004) find to be related to Dechow and Dichev’s

(2002) measure of accruals quality. Specifically, they show that accruals quality is negatively

related to the ratio of total intangibles to total sales (INTINT), while accruals quality is positively




23. Levi (2005) and Wild and Seber (2000) also employ this approach of controlling for seasonality.


                                                         16
related to the absence of reported intangibles (INTDUM), and the ratio of property, plant, and

equipment to total assets (CAPINT).24

        We also control for other variables that could be related to the quality of financial reporting.

McVay (2006) argues that firm managers wishing to manage core earnings upwards shift expenses

that should be classified as core expenses to special items. Myers and Skinner (2002) also show that

firms are more likely to report positive special items when the firms’ earnings would otherwise be

unusually low. Other studies documenting the increased reporting of special items overtime and the

use of special items to manage markets’ perceptions of firm performance include Elliott and Hanna

(1996), Bradshaw and Sloan (2002), and Kinney and Trezevant (1997). Accordingly, we employ

the average absolute value of special items scaled by assets during the sample period, to control for

reporting of special items (SPITEM). Loebbecke et al. (1989) and Bell et al. (1993) argue that the

managements of high-growth firms are more likely to misstate financial statements during a

downturn to give an appearance of stable growth, and Abbott et al. (2004) argue that a firm’s rate of

growth can negatively impact the strength of the firm’s internal control and accounting system.

Firm growth (GROWTH) is controlled by the average book to market ratio of the sample firms

during the sample period. It has also been documented that issuing firms report income-increasing

accruals around the time of seasoned equity offerings (Teoh et al., 1998; Rangan, 1998;

Shivakumar, 2000; Zhou and Elder, 2004) in order to inflate stock prices. We control for additional

offerings by employing a dichotomous variable (ISSUE) that is coded 1 if a sample firm issues

additional stock during the sample period, and 0 otherwise. Beasley (1996) and Bell et al. (1993)

state that poor financial performance often causes management to place undue emphasis on




24. In addition to the innate factors identified by Dechow and Dichev (2002) and Francis et al. (2004), Hribar and
Nichols (2006) provide evidence on the correlation between unsigned measures of accruals quality and other firm
characteristics such as market value of equity, sales growth, leverage, and book-to-market ratios. Our conclusions
remain robust when we include these firm characteristics as independent variables.


                                                       17
earnings, which will increase the likelihood of financial statement fraud.25 Frost (1997) and Koch

(2002) find that disclosures of financially sound firms are of higher credibility than the disclosures

of financially distressed firms. Further, Mercer (2004) argues that the lower credibility of

disclosures from financially distressed firms arises from the greater incentives that managers have

for misrepresentation in those circumstances. To control for financial distress, we use a

dichotomous variable (DISTRESS) that is coded 1 if the sample firm experienced financial distress

during the sample period, and 0 otherwise.26 Finally, given that some industries are more highly

represented by the sample firms, we also control for the impact of industry effects on accruals

quality by employing industry dummies (INDUSTRY) representing the two-digit SIC codes.



5. Research design

         We employ a cross-sectional regression model to formally test the relative contribution of

the three types of audit committee financial expertise to accruals quality. The model employs the

standard deviation of residuals (STDRES) from firm-specific regressions of changes in working

capital accruals on lagged, current, and future cash flows from operations, change in sales, and

quarterly dummy variables as a proxy for accruals quality (dependent variable). Under this measure

of accruals quality, a larger standard deviation of the residuals (higher volatility in abnormal

accruals) implies lower accruals quality. The independent variables include the three financial

expertise, two governance, and control variables, as well as variables that capture the interaction

effects between the financial expertise and governance variables. The regression model used for our

analyses is as follows:



25. Mutchler et al. (1997) and Louwers (1998) find a positive relation between financial distress and audit going-
concern uncertainty disclosures.
26. We use the Altman (1968) Z-Score to measure financial distress, where an average Z-Score of below 1.81 signals
financial distress for a firm. Given that the Altman Z-Score uses financial ratios reflecting firm liquidity, profitability,
leverage, solvency, and activity to predict whether financial distress (Altman, 1968), we do not separately include the
financial ratio variables as additional control variables.


                                                            18
STDRES i = α 0 + α 1 ACCTG i + α 2 FINANCE i + α 3 SUPER i + α 4 ACGOV i + α 5 BDGOV i + α 6 ACCTG i ∗ ACGOV i
           + α 7 ACCTG i ∗ BDGOV i + α 8 FINANCE i ∗ ACGOV i + α 9 FINANCE i ∗ BDGOV i + α 10 SUPER i ∗ ACGOV i
           + α 11 SUPER i ∗ BDGOV i + α 12 ASSETS i + α 13σ (CFO ) i + α 14 NEGEARN i + α 15σ ( SALES ) i + α 16 OPCYCLE i
           + α 17 INTINT i + α 18 INTDUM i + α 19 CAPINT i + α 20 SPITEM i + α 21GROWTH i + α 22 ISSUE i + α 23 DISTRESS     i
              n
           + ∑ γ j INDUSTRY j + ε i                                                                                      (1)
              j =1




Where

STDRESi                   = Standard deviation of the residuals of firm i during the period 1995-1998
                         from firm-specific regressions of changes in working capital accruals on
                         lagged, current, and future cash flows from operations, change in sales, and
                         quarterly dummy variables;
ACCTGi                   = 1 if the audit committee of firm i includes at least one member with
                         accounting expertise in each year during the period 1995-1998, and 0
                         otherwise;
FINANCEi                 =1 if the audit committee of firm i includes at least one member with finance
                         expertise in each year during the period 1995-1998, and 0 otherwise;
SUPERi                   =1 if the audit committee of firm i includes at least one member with
                         supervisory expertise in each year during the period 1995-1998, and 0
                         otherwise;
ACGOVi                   =1 if the AC GOV SCORE of firm i is greater than or equal to two, and 0
                         otherwise;
BDGOVi                   =1 if the BD GOV SCORE of firm i is greater than or equal to three, and 0
                         otherwise;
ASSETSi                  = Log of average total assets of firm i during the period 1995-1998;
σ(CFO)i                  = Standard deviation of seasonally adjusted cash flow from operations of firm i
                         during the period 1995-1998;
NEGEARNi                 = Proportion of fiscal quarters during the period 1995-1998 where firm i is
                         reporting negative earnings;
σ(SALES)i                = Standard deviation of seasonally adjusted sales revenue of firm i during the
                         period 1995-1998;
OPCYCLEi                 = Log of average operating cycle of firm i during the period 1995-1998;
INTINTi                  = Average ratio of reported R&D and advertising expense to total sales
                         revenue of firm i during the period 1995-1998;
INTDUMi                  = 1 if INTINTi = 0, and 0 otherwise;
CAPINTi                  = Average ratio of net book value of property, plant, and equipment to total
                         assets of firm i during the period 1995-1998;
SPITEMi                  = Average absolute value of special items scaled by total assets of firm i during
                         the period 1995-1998;
GROWTHi                  =Average book to market equity ratio for firm i during the period 1995-1998;
ISSUEi                   =1 if firm i issues securities during the period 1995-1998, and 0 otherwise;27
DISTRESSi                =1 if the average Altman Z-Score of firm i over the period 1995-1998 signals
                         financial distress, and 0 otherwise; and28

27. Firms are deemed to have issued securities if they experienced at least a 10 percent increase in number of
outstanding shares in any consecutive fiscal quarters.


                                                               19
INDUSTRYi               = Industry indicators representing the industry membership of firm i over the
                        period 1995-1998.


A significant negative coefficient for a financial expertise variable (e.g., α1 for ACCTG), when the

regression is run without the interaction terms, would indicate that firms with that type of audit

committee financial expertise have lower accruals volatility (higher accruals quality). When the

regression is run after including the interaction terms, the financial expertise variables capture the

association between the type of financial expertise and accruals quality for firms with weak

governance. A significant negative coefficient for a variable that captures the interaction between a

financial expertise and governance variable (e.g., α6 for ACCTG * ACGOV) would indicate whether

firms with that type of audit committee financial expertise and strength in that governance attribute

have lower accruals volatility (higher accruals quality) than firms that have that type of audit

committee financial expertise but weak governance.



6. Descriptive statistics

6.1 Firm characteristics

        Panel A of Table 2 reports summary statistics used to construct the financial expertise and

governance variables for the 1,114 sample firms, over the sample period. Descriptive statistics on

continuous data are based on average firm-level values during the sample period.

                                <<< INSERT TABLE 2 ABOUT HERE >>>

        Using the broad definition of financial expertise, the mean number of financial experts on

audit committees is 1.145, indicating that even before the introduction of SOX, a large proportion of

the firms in the sample had satisfied the requirement of having at least one financial expert on the



28. An average Altman Z-Score of below 1.81 during the sample period signals financial distress for a firm. Annual
Altman Z-Scores are determined using the following formula: 1.2 x Working Capital + 1.4 x Retained Earnings + 3.3 x
EBIT + 0.6 x Market Capitalization + 1 x Sales, where working capital, retained earnings, EBIT, and sales are scaled by
total assets.


                                                         20
audit committee. A breakdown of financial expertise into its three components of accounting

expertise, finance expertise, and supervisory expertise indicates that supervisory expertise is the

largest contributor of financial expertise in audit committees (mean = 0.548), while accounting

expertise is the smallest contributor (mean = 0.266). The Blue Ribbon Committee (1999)

recommended that firms maintain audit committees consisting of at least three members who were

all fully independent. The upper quartile audit committee size of 3 members indicates that only a

small proportion of the sample firms maintained audit committees consisting of at least three

members. The mean (median) proportion of independent members in audit committees is 0.800

(0.698), suggesting that many firms had a considerable number of non-independent members in

their audit committees. The average number of audit committee meetings is 2.726 (median is

2.500), which is lower than the minimum threshold of four meetings subsequently recommended by

the Blue Ribbon Committee (1999). Summary statistics on board characteristics indicate that the

average (median) number of directors on the boards of the sample firms is 8.920 (8.667), and the

mean (median) proportion of independent directors on the boards is 0.577 (0.591). The average

cumulative proportion of shares held by the independent directors is 0.054 (median is 0.008), and

the CEO holds the position of board chairman in approximately 58 percent (649 firms) of the

sample firms.

       Panel B of Table 2 reports descriptive statistics for the control variables. The mean log of

total assets is 6.553 and only slightly larger than its median (6.467), indicating the absence of strong

skewness in the size of the sample firms. The mean (median) standard deviation of seasonally

adjusted cash flow from operations is 0.026 (0.018), while the sample firms reported negative

earnings in about 15.6 percent of their fiscal quarters during the sample period. The mean (median)

standard deviation of seasonally adjusted sales is 0.069 (0.052). These descriptive statistics are

similar to those reported by Levi (2005), who also employs the Dechow and Dichev (2002) model

using quarterly financial data adjusted for seasonality. We find that the mean (median) log of the


                                                  21
sample firms’ operating cycle is 4.878 (5.015). Summary statistics on the reporting of intangibles

indicate that about 35 percent of the sample firms do not report any R&D and advertising expenses,

while the average (median) ratio of total R&D and advertising to total sales is 0.074 (0.016). The

mean (median) ratio of property, plant, and equipment to total assets is 0.315 (0.276). The summary

statistics on intangibles and capital expenditures are similar to those reported by Francis et al.

(2004). Turning to the remaining summary statistics, the mean (median) special items amounts to

0.6 percent (0.3 percent) of total assets, and statistics on firm growth indicate that, on average, the

book value of equity amounts to 44.7 percent (median is 40.3 percent) of the market value of equity.

Finally, we find that 271 firms (24 percent) issued further securities, and 268 firms (24 percent)

experienced financial stress during the sample period.

6.2 Accrual quality

       Table 3 reports descriptive statistics on the STDRES of the sample firms, where larger

STDRES implies lower accruals quality.

                            <<< INSERT TABLE 3 ABOUT HERE >>>

       Panel A of Table 3 reports summary statistics on the STDRES of firms with and without

accounting expertise in their audit committees. Across the full sample, firms with audit committee

accounting expertise (ACCTG = 1) have a lower mean and median STDRES in comparison to firms

without audit committee accounting expertise (ACCTG = 0). A Wilcoxon two-sample test indicates

a significant difference between the STDRES of firms with and without audit committee accounting

expertise at the 1 percent level. The remaining results in panel A of Table 3 report the STDRES for

firms with and without accounting expertise in their audit committee, given the strength of the audit

committee and board governance environment. The results show that firms with strong audit

committee governance (ACGOV = 1) have significantly lower mean and median STDRES when

they have audit committee accounting expertise (ACCTG = 1) relative to when they do not have

audit committee accounting expertise (ACCTG = 0) at 1 percent level. There is no significant


                                                  22
difference between the mean and median STDRES of firms with and without accounting expertise in

their audit committees when they have weak audit committee governance (ACGOV = 0). These

results suggest that audit committee accounting expertise is only likely to positively contribute

toward accruals quality when there is strong audit committee governance in place. However, these

results should be interpreted cautiously given that they are obtained from a univariate test that does

not control for the effects of other variables. Turning to the results relating to board governance

strength, the STDRES is significantly lower for firms with audit committee accounting expertise

regardless of whether firms have strong or weak board governance. These results suggest that the

marginal contribution of audit committee accounting expertise toward accruals quality is unaffected

by board governance strength.

       Panel B of Table 3 reports similar summary statistics on the STDRES of all firms and for

firms with and without finance expertise in their audit committees. In all instances reported, the

results show that firms with audit committee finance expertise (FINANCE = 1) do not have

significantly lower STDRES in comparison with firms without audit committee finance expertise

(FINANCE = 1). These findings suggest that finance expertise in audit committees is unlikely to,

either on its own, or in conjunction with strong corporate governance, be related to higher accruals

quality. Similar to the results reported in panel B, the results from panel C of Table 3 indicate that

supervisory expertise is unlikely to be related to higher accruals quality with one exception: firms

with strong audit committee governance (ACGOV = 1) have significantly higher STDRES when

they have audit committee supervisory expertise (SUPER = 1) relative to when they do not have

audit committee accounting expertise (SUPER = 0). This result is surprising as it suggests that the

presence of supervisory expertise on strong audit committees is likely to be negatively associated

with accruals quality.




                                                 23
7. Empirical results


         Table 4 reports the correlation matrix for the dependent and independent variables.

                                 <<< INSERT TABLE 4 ABOUT HERE >>>


         The correlation coefficients for the industry dummies are not reported. On the whole,

although there are a few large correlation coefficients, the coefficients are not large enough to

prohibit the use of a multivariate regression analysis.29

         Table 5 presents the results from the cross-sectional regression of STDRES on the three

measures of audit committee financial expertise (ACCTG, FINANCE, and SUPER), the governance

variables (ACGOV and BDGOV), variables that capture the interaction between the expertise and

governance variables, and the 13 control variables. Coefficients for industry dummies are not

reported.

                                 <<< INSERT TABLE 5 ABOUT HERE >>>

         The second column of Table 5 reports the results from Model 1, which examines the

association of the three types of financial expertise with STDRES, without considering any

interaction effects between the expertise and governance variables. We find that STDRES is

negatively and significantly related to ACCTG (at the 1 percent level). Given that lower STDRES

signals higher accruals quality, this finding suggests that audit committee accounting expertise is

positively related to accruals quality. On the other hand, the results show that accruals quality is not

significantly related to FINANCE, while surprisingly, the results indicate that accruals quality is

negatively related to SUPER (at the 5 percent level). These results do not support the stance of the

SEC and stock exchanges on allowing the inclusion of supervisory and finance expertise in the

29. Because we employ a large number of independent variables in the regression analysis, there could still be a
possibility of each independent variable being collectively correlated with the other independent variables. The matrix
of bivariate correlation in Table 4 does not help detect such multiple correlations. In such instances, the use of variance
inflation factors helps determine the multiple correlations between an independent variable and the rest of the
independent variables (Berry and Feldman, 1985). We compute and examine the variance inflation factors for the
independent variables but find no signs of any such correlation problems.


                                                           24
definition of financial expertise. The results also indicate that the governance variables, ACGOV

and BDGOV, are both positively associated with accruals quality at the 1 percent level. This is

consistent with the findings of prior research that a strong governance environment is associated

with improved monitoring of the financial reporting process. Consistent with Dechow and Dichev

(2002), we find that the variables σ(CFO), NEGEARN, σ(SALES), and OPCYCLE are negatively

related to accruals quality (at the 1 percent level), while accruals quality is positively related to

ASSETS (at the 1 percent level). The results also indicate that INTINT is negatively related to

accruals quality (at the 5 percent level), while INTDUM and CAPINT are both positively related to

accruals quality at the 1 percent level and 10 percent level, respectively. The results for INTINT,

INTDUM, and CAPINT are consistent with those observed by Francis et al. (2004). Turning to the

results for the remaining control variables, we find that SPITEM, ISSUE, and DISTRESS are all

negatively related to accruals quality at the 1 percent, 5 percent, and 5 percent levels respectively.

These results are consistent with prior studies, signaling lower accruals quality in firms reporting

more special items, firms that issue additional stock, and firms that are under financial distress.

       Model 2 includes as independent variables the measure of accounting expertise (ACCTG),

the governance and control variables, and two variables that capture the interaction between

ACCTG and the governance variables (ACCTG * ACGOV and ACCTG * BDGOV). The results from

this regression, reported in the third column of Table 5, show the interaction term ACCTG             *


ACGOV is significantly negative at the 1 percent level, implying that the positive association

between accounting expertise and accruals quality is more prominent in firms that have strong audit

committee governance. In other words, audit committee accounting expertise has a stronger effect

when other characteristics of the audit committee promote strong governance. The interaction term

ACCTG    *   BDGOV is not statistically significant, indicating that the effectiveness of accounting

expertise is independent of board governance strength. Results for the governance and control

variables are qualitatively similar to those observed for Model 1. Models 3 and 4 include the


                                                  25
variable FINANCE (SUPER) instead of ACCTG and variables that capture their interactions with

the variables ACGOV and BDGOV. The results of these regressions are reported in columns four

and five of Table 5. The results show that FINANCE and the interactions of FINANCE with the two

governance variables are insignificant, indicating that finance expertise is unlikely to influence

accruals quality even in the presence of strong audit committee and board governance. The

comparative results for SUPER indicate that SUPER and the interaction of SUPER with BDGOV

are unrelated to accruals quality. The coefficient of the interaction between SUPER and ACGOV is

positive and weakly significant (at the 10 percent level). Again, this result is puzzling. The results

for the governance and control variables are qualitatively similar to those reported earlier.

       Model 5 includes variables capturing all three types of financial expertise (ACCTG,

FINANCE, and SUPER), governance variables (ACGOV and BDGOV), six interaction terms that

interact each of the three financial expertise variables with the two governance variables, and the 13

control variables. The results reported in the last column of Table 5 are consistent with those

reported earlier with one exception: the interaction effect between SUPER and ACGOV is no longer

statistically significant. Overall, our results strongly suggest that accounting expertise in audit

committees is the only expertise that is likely to contribute to audit committee effectiveness and this

contribution is even more pronounced in the presence of audit committees with strong attributes.



8. Additional tests

8.1 Alternative measures of accruals quality

       Prior studies have employed other approaches in computing accruals quality. To check

whether our results are robust to alternative measures of accruals quality, we repeat the analyses in

Table 5 by considering three other proxies of accruals quality. The first additional proxy is the

average absolute value of residuals from firm-specific time series estimations of the modified

Dechow and Dichev (2002) model. This measure of accruals quality focuses on the magnitude


                                                  26
rather than the volatility of abnormal accruals and is based on the intuition that larger residuals from

the time series estimations of the modified Dechow and Dichev (2002) model represent lower

accruals quality. Our second additional measure of accruals quality is based on the residuals from

the modified Dechow and Dichev (2002) cross-sectional model estimated quarterly for each two-

digit SIC industry group with at least eight observations. Accruals quality is represented by the

standard deviation a firm’s quarterly residuals from these regressions over the sample period. The

third additional measure of accruals quality is represented by a firm’s average absolute value of

residuals from industry-specific quarterly estimations of the modified Dechow and Dichev (2002)

model. After employing these alternative measures of accruals quality, the results (not reported) are

qualitatively similar to those reported earlier with the following exceptions: (1) under all three

additional measures of accruals quality, ACCTG is no longer significant in the results for Models 2

and 5 in Table 5, and (2) the results from the second additional accruals quality proxy indicates that

the interaction between ACCTG and ACGOV is significant at the 5 percent level.

8.2 Alternative attributes of earnings quality

       Accruals quality is merely one attribute of earnings quality. In addition to considering

Dechow and Dichev’s (2002) measure of accruals quality, Francis et al. (2004) consider three

further accounting-based attributes of earnings quality: earnings persistence, predictability and

smoothness. Under these three alternative measures of earnings quality, firms are perceived to have

higher earnings quality if they have higher earnings persistence, higher earnings predictability, and

lower earnings smoothness. To test whether our results are robust to alternative measures of

earnings quality, we repeat the analyses in Table 5 by considering earnings persistence,

predictability, and smoothness as alternative proxies of earnings quality during the sample period.

Consistent with prior research, we measure firm i’s earnings persistence as the slope coefficient

estimate from its first-order autoregressive model for earnings per share after controlling for any

seasonality effects. Following Francis et al. (2004), we derive firm i’s earnings predictability from


                                                  27
the square root of the estimated error variance from the earnings persistence equation described

above. Finally, firm i’s earnings smoothness is defined as the ratio of the standard deviation of

scaled, seasonally-adjusted net income before extraordinary items to the standard deviation of

scaled, seasonally-adjusted cash flows from operations. After employing these alternative measures

of earnings quality, the results (not reported) are qualitatively similar to those reported earlier with

the following exceptions: (1) under all three additional measures of earnings quality, ACCTG is no

longer significant in the results for Models 2 and 5 in Table 5, (2) the results from using the

earnings predictability measure indicates that ACCTG is also no longer significant in the results for

Model 1, and (3) under the earnings persistence measure, the interaction between ACCTG and

ACGOV is significant at the 5 percent level.

8.3 Alternative definitions for audit committee governance strength

       The earlier analysis classifies governance in audit committees (ACGOV) as being strong if

an audit committee has an AC GOV SCORE of at least two, whereby AC GOV SCORE is the sum of

three dichotomous variables. These dichotomous variables are coded 1 if an audit committee has a

minimum audit committee size (AC SIZE) of three members, fully independent members (AC IND),

and met (AC MEETING) on an average of four times per year during the sample period,

respectively. To test the sensitivity of our results, we repeat the analysis in Table 5 after employing

a continuous measure of governance strength by redefining ACGOV as the AC GOV SCORE. The

results from this analysis (not reported) indicate that while ACCTG is no longer statistically

significant for Models 2 and 5 in Table 5, the variable that captures the interaction between ACGOV

and ACCTG remains strongly significant at the 1 percent level. The comparative results for

FINANCE and SUPER remain statistically insignificant. We also repeat the analysis in Table 5 after

replacing ACGOV with the three dichotomous audit committee variables themselves and nine

variables that capture the interaction of these dichotomous variables with ACCTG, FINANCE, and

SUPER. This analysis may add further insights that have been masked by the use of summary


                                                  28
measure of corporate governance. The results (not reported) indicate that ACCTG is positively

associated with accruals quality at the 5 percent level, while the three variables capturing the

interaction of ACCTG with the variables AC SIZE, AC IND, and AC MEETING are positively

associated with accruals quality at the 5 percent, 10 percent, and 10 percent levels, respectively. The

results suggest that, of the audit committee attributes examined in the study, audit committee size is

the most important audit committee attribute that synergizes the impact of audit committee

accounting expertise on accruals quality. The interaction effects of the three audit committee

variables with FINANCE and SUPER are not statistically significant.            Consistent with prior

research, we find that AC SIZE and AC IND are positively related to accruals quality at the 5

percent and 1 percent levels respectively. We further examine the sensitivity of the results by

repeating the analysis in Table 5 after redefining ACGOV several times based on the numerous

permutations using the three dichotomous audit committee variables. Such an exercise could reveal

the critical combinations of audit committee attributes that best promote the effective use of audit

committee accounting expertise in increasing accruals quality. However, the results (not reported)

do not indicate that any particular combinations of audit committee attributes are more successful

than others in promoting the effectiveness of audit committee accounting expertise. In all instances

examined, the interaction between ACCTG and ACGOV remains significant at conventional levels.

Again, the interaction effects concerning FINANCE and SUPER remain insignificant.

8.4 Potential role of board characteristics

       The earlier results indicate that strong boards do not significantly promote the impact of

audit committee accounting expertise on accruals quality. This finding is somewhat unexpected

given that strong boards are expected to support and empower audit committees, implying more

effective utilization of any expertise within the audit committees. We further investigate the role

that board attributes may play by repeating the analysis in Table 5 after replacing BDGOV with the

four dichotomous board variables used in constructing BDGOV: BOARD SIZE (size of board),


                                                  29
BOARD IND (degree of independence in board), SHARE OWN (share ownership of independent

directors), and DUALITY (whether the CEO was also the board chairman). We also introduce 12

independent variables that capture the interaction of these four dichotomous variables with ACCTG,

FINANCE, and SUPER. The results (not reported) indicate that all four board characteristics are

positively associated with accruals quality.30 However, while ACCTG remains significant at the 5

percent level, none of the 12 variables that capture the interaction of the individual board

characteristics with ACCTG, FINANCE, and SUPER are statistically significant. We further

investigate the potential role of board governance in synergizing the effects of the three types of

financial expertise, by developing alternative measures of board governance based on a continuous

measure or different combinations of the four board characteristics. We also introduce a three-way

interaction variable using ACCTG, ACGOV, and BDGOV to evaluate whether positive interaction

between audit committee accounting expertise and strong audit committees is more pronounced in

firms that also have strong board governance. We also replicate this three-way interaction analysis

using FINANCE and SUPER. The results (not reported) from these additional analyses do not

produce any evidence to suggest that board characteristics help promote the effective use of any

type of financial expertise in ensuring the quality of financial reporting.

8.5 Controlling for self-selection bias

        Our analysis thus far, assumes that a firm’s choice of accounting expertise is exogenously

given. However, it is possible that firms do not randomly appoint accounting experts but rather

self-select accounting experts based on certain firm characteristics. This suggests that accrual

quality and the presence of audit committee accounting experts are endogenous.                          From an

econometric perspective, the presence of any self-selection would introduce a bias in our earlier

regression analysis, yielding inconsistent parameter estimates (Maddala, 1983). To address this



30. Specifically, BOARD SIZE and DUALITY are both related to accruals quality at the 1 percent level, while BOARD
IND and SHARE OWN are both related to accruals quality at the 5 percent level.


                                                       30
potential self-selection issue, we adopt the two stage “treatment effects” procedure of Heckman

(1979) and Lee (1979). In the first stage, we estimate a self-selection probit regression that

evaluates whether the presence of an accounting expert is related to certain firm characteristics.

The parameter estimates are then used to compute inverse Mills ratios (IMR). In the second stage,

we replicate our analyses in Table 5 after including the computed IMR from the first stage as an

additional independent variable, to account for potential self-selection bias. Our self-selection

model is based on the model employed by Agrawal and Chadha (2005) to examine firm

characteristics that drive the presence of accounting experts on boards:31

    ACCTG i = β 0 + β 1 LSALES i + β 2 OPERF i + β 3 SLSGROWTH   i   + β 4 LEVERAGE i + β 5 CAPINT i + β 6 STOCKVOL   i

                + β 7 AGE i + β 8 BOARDSIZE i + ∑ k =1 δ k STOCKEXCHG
                                                  n
                                                                         k   +ν i                              (2 )

Where

LSALESi                  = Log of average total sales of firm i during the sample period;
OPERFi                   = Average ratio of operating performance to total assets of firm i during the
                        sample period;
SLSGROWTHi               = Average sales growth of firm i during the sample period;
LEVERAGEi                = Average ratio of long-term debt to firm value of firm i during the sample
                        period;
CAPINTi                  = Average ratio of total assets to number of employees of firm i during the
                        sample period;
STOCKVOLi                = Standard deviation of stock returns of firm i during the sample period;
AGEi                     = Age of firm i at the start of the sample period;
BOARDSIZEi               = Average board size of firm i during the sample period; and
STOCKEXCHGi              = Stock exchange indicators representing the exchange firm i traded on during
                        the sample period.


         The results from our first stage self-selection model (not reported), indicate that CAPINT

and STOCKVOL are both positively associated with the presence of audit committee accounting

experts at the 1 percent level, while OPERF is also positively related to the presence of accounting

experts but at the 5 percent level. We also find that AGE has a significant negative relationship




31. We also control for the control for the exchange the firm traded on, as Deli and Gillan (2000) show that the
probability of firms having independent and active audit committee is related to trading venue.


                                                          31
with the presence of accounting experts (at the 5 percent level). These results are consistent with

the expectations of Agrawal and Chadha (2005).

       The results from our second stage analysis (not reported) indicate that the parameter

estimate of IMR is positive and significant at the 1 percent level. This suggests that it is important

to explicitly control for self-selection bias. However, the results for ACCTG and the interaction of

ACCTG with ACGOV are qualitatively similar to those reported earlier. Hence, our findings remain

robust after the correction for the endogeneity between accrual quality and presence of audit

committee accounting experts.



9. Conclusion

       Recent developments in corporate governance suggest that financial expertise in audit

committees is integral to increasing the effectiveness of audit committees in monitoring the

financial reporting process. However, there has been an ongoing controversy regarding the current

definition of financial expertise adopted by the SEC and stock exchanges. Numerous studies have

failed to provide any significant evidence on the ability of financial experts (as currently defined) to

influence the financial reporting process, strongly suggesting that the current definition of financial

expertise is too broad and encompasses skills that do not contribute to audit committee

effectiveness. This study decomposes the current definition of financial expertise into three specific

types of financial expertise (accounting, finance, and supervisory expertise) and then investigates

the association of the three specific types of audit committee financial expertise to accruals quality.

Furthermore, we examine whether a particular type of audit committee financial expertise has a

stronger effect on accruals quality in the presence of strong governance.

       Multivariate tests indicate a significant positive relation between accounting expertise in

audit committees and accruals quality but no significant association between accruals quality and

the presence of finance or supervisory expertise in audit committees. This finding supports the


                                                  32
notion that the current definition of financial expertise is too broad and supports any future

refinements to the financial expertise definition to include only those skills that result in accounting

expertise. Further, we find that the positive association between audit committees with accounting

expertise and accruals quality is more pronounced in the presence of strong audit committee

governance, but not in the presence of strong board governance. Consistent with prior research, we

also find that both strong audit committee governance and board governance result in accruals

quality.




                                                  33
References

Abbott, L. J., Parker, S., 2000.     Fraudulent financial reporting: Auditor selection and audit

       committee characteristics. Auditing: A Journal of Practice & Theory 19, 47-66.

Abbott, L. J., Parker, S., Peters, G. F., Raghunandan, K., 2003. The association between audit

       committee characteristics and audit fees. Auditing: A Journal of Practice & Theory 22, 17-

       32.

Abbott, L. J., Parker, S., Peters, G. F., 2004. Audit committee characteristics and restatements.

       Auditing: A Journal of Practice & Theory 23, 69-87.

Agrawal, A., Chadha, S., 2005. Corporate governance and accounting scandals. Journal of Law and

       Economics 48, 371-406.

Altman, E., 1968. Financial ratios, discriminant analysis and the prediction of corporate bankruptcy.

       Journal of Finance 23, 589-609.

American Association of Bank Directors, 2002. Letter to the SEC – RE: The Sarbanes Oxley Act

       of 2002.

American Institute of Certified Public Accountants (AICPA), 1967. Executive committee statement

       on audit committees of board of directors. Journal of Accountancy 124, 10.

Anderson, R. C., Mansi, S. A., Reeb, D. M., 2004. Board characteristics, accounting report

       integrity, and the cost of debt. Journal of Accounting and Economics 37, 315-342.

Archambeault, D., DeZoort, F. T., 2001. Auditor opinion shopping and the audit committee: An

       analysis of suspicious auditor switches. International Journal of Auditing 5, 33-52.

Ashbaugh-Skaife, H., Collins, D. W., LaFond, R., 2006. The effects of corporate governance on

       firms’ credit ratings. Journal of Accounting and Economics 46, 203-243.

Beasley, M. S., 1996.     An empirical analysis of the relation between the board of director

       composition and financial fraud. The Accounting Review 71, 443-465.




                                                 34
Beasley, M. S., Carcello, J. V., Hermanson, D. R., Lapides, P. D., 2000. Fraudulent financial

       reporting: Consideration of industry traits and corporate governance mechanisms.

       Accounting Horizons 14, 441-454.

Beasley, M. S., Salterio, S. E., 2001. The relationship between board characteristics and voluntary

       improvements in audit committee composition and expertise. Contemporary Accounting

       Research 18, 539-570.

Bédard, J., Chtourou, S. M., Courteau, L., 2004. The effect of audit committee expertise,

       independence, and activity on aggressive earnings management. Auditing: A Journal of

       Practice & Theory 23, 13-35.

Bell, T. B., Szykowny, S., Willingham, J. J., 1993. Assessing the likelihood of fraudulent financial

       reporting: A cascaded logit approach. Working paper, KPMG Peat Marwick.

Berry, W., Feldman, S., 1985. Multiple Regression in Practice. Sage Publications, Beverly Hills,

       CA.

Blue Ribbon Committee on Improving the Effectiveness of Corporate Audit Committees (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 The National Association of Securities Dealers.

Bradshaw, M., Sloan, R., 2002. GAAP versus the Street: An empirical assessment of two

       alternative definitions of earnings. Journal of Accounting Research 40, 41–66.

Braswell, M., Mauldin, E., 2004. Changes in audit committee financial expertise. Working paper,

       University of Missouri.

Bryan-Low, C., 2002. Defining moment for the SEC: Who’s a financial expert? Wall Street

       Journal December 9, 2002: C1.

Bull, I., Sharp, F. C., 1989. Advising clients on Treadway audit committee recommendations.

       Journal of Accountancy 167, 46-52.


                                                35
Bushman, R., Chen, Q., Engel, E., Smith, A., 2004.              Financial accounting information,

       organizational complexity and corporate governance systems. Journal of Accounting and

       Economics 37, 167–201.

Carcello, J., Hermanson, D., Neal, T. L., 2002. Disclosure in audit committee charters and reports.

       Accounting Horizons 16, 291-304.

Carcello, J. V., Neal, T. L., 2003. Audit committee characteristics and auditor dismissals following

       “new” going-concern reports. The Accounting Review 78, 95-117.

Carcello, J. V., Hollingsworth, C. W., Klein, A., Neal, T. L., 2006. Audit committee financial

       expertise, competing governance mechanisms, and earnings management. Working paper,

       University of Tennessee.

Cohen, J., Krishnamoorthy, G., Wright, A. M., 2002. Corporate governance and the audit process.

       Contemporary Accounting Research 9, 573-594.

Cohen, J., Krishnamoorthy, G., Wright, A. M., 2004.         The corporate governance mosaic and

       financial reporting quality. Journal of Accounting Literature 23, 87-152.

Collier, P., Gregory, A., 1999. Audit committee activity and agency costs. Journal of Accounting

       and Public Policy 18, 311-332.

Collins, D. W., Hribar, P., 2002. Errors in estimating accruals: Implications for empirical research.

       Journal of Accounting Research 40, 105-135.

Committee of Sponsoring Organizations of the Treadway Commission (COSO), 1992. Internal

       control integrated framework. COSO, New York, NY.

Core, J. E., Holthausen, R. W., Larcker, D. F., 1999. Corporate governance, chief executive officer

       compensation, and firm performance. Journal of Financial Economics 51, 371–406.

Dechow, P. M., Sloan, R. G., Sweeney, A. P., 1996. Causes and consequences of earnings

       manipulation: An analysis of firms subject to enforcement actions by the SEC.

       Contemporary Accounting Research 13, 1-36.


                                                 36
Dechow, P. M., Dichev, I. D., 2002. The quality of accruals and earnings: The role of accrual

       estimation errors. The Accounting Review 77, 35-59.

DeFond, M. L., Hann, R. N., Hu, X., 2005. Does the market value financial expertise on audit

       committees of board of directors? Journal of Accounting Research 43, 153-193.

DeFond, M. L., Francis, J. R., 2005. Audit research after Sarbanes-Oxley. Auditing: A Journal of

       Practice & Theory 24, 5-30.

Deli, D. N., Gillan, S. L., 2000. On the demand for independent and active audit committees.

       Journal of Corporate Finance 6, 427-445.

DeZoort, F. T., 1997. An investigation of audit committees’ oversight responsibilities. Abacus 33,

       208-27.

DeZoort, F. T., 1998. An analysis of experience effects on audit committee members’ oversight

       judgments. Accounting, Organizations and Society 23, 1–21.

DeZoort, F. T., Salterio, S., 2001. The effects of corporate governance experience and financial

       reporting and audit knowledge on audit committee members’ judgments. Auditing: A

       Journal of Practice & Theory 20, 31-47.

DeZoort, F. T., Hermanson, D. R., Archambeault, D. S., Reed, S. A., 2002. Toward a theory of

       audit committee effectiveness: A synthesis of the empirical audit committee literature.

       Journal of Accounting Literature 21, 38-75.

Eisenberg, T., Sundgren, S., Wells, M., 1998. Larger board size and decreasing firm value in small

       firms. Journal of Financial Economics 48, 35-54.

Elliott, J. A., Hanna, J. D., 1996. Repeated accounting write-offs and the information content of

       earning. Journal of Accounting Research 34, 135–155.

Fama, E. F., Jensen, M. C., 1983. Separation of ownership and control. Journal of Law and

       Economics 26, 301-325.




                                                  37
Farber, D., 2005. Restoring trust after fraud: Does corporate governance matter? The Accounting

       Review 80, 539-561.

Francis, J., LaFond, R., Olsson, P., Schipper, K., 2004. Costs of equity and earnings attributes. The

       Accounting Review 79, 967-1010.

Francis, J., LaFond, R., Olsson, P., Schipper, K., 2005. The market pricing of accruals quality.

       Journal of Accounting and Economics 39, 295-327.

Frost, C., 1997. Disclosure policy choices of UK firms receiving modified audit reports. Journal of

       Accounting and Economics 23, 163-187.

Geiger, M. A., Rama, D. V., 2003. Audit fees, non-audit fees, and auditor reporting on stressed

       companies. Auditing: A Journal of Practice & Theory 22, 53-69.

General Accounting Office (GAO), 1991. Audit committees: Legislation needed to strengthen bank

       oversight. United States General Accounting Office, Washington, DC.

Gerety, M., Lehn, K., 1997. The causes and consequences of accounting fraud. Managerial and

       Decision Economics 18, 587-599.

Goyal, V. K., Park, C. W., 2002. Board leadership structure and CEO turnover. Journal of

       Corporate Finance 8, 49-66.

Green, D. L., 1994. Canadian audit committees and their contribution to corporate governance.

       Journal of International Accounting, Auditing and Taxation 3, 135-151.

Gul, F. A., Leung, S., 2004. Board leadership, outside directors' expertise and voluntary corporate

       disclosures. Journal of Accounting and Public Policy 23, 351-379.

Heckman, J., 1979. Sample selection bias as a specification error. Econometrica 47, 153-162.

Hribar, P., Nichols, D. C., 2006. The use of unsigned earnings quality measures in tests of earnings

       management. Working paper, Cornell University.

Jensen, M., 1993. The modern industrial revolution, exit, and failure of internal control systems.

       Journal of Finance 48, 831-880.


                                                 38
Jones, J. J., 1991. Earnings management during import relief investigations. Journal of Accounting

       Research 29, 193-228.

Kalbers, L. P., Fogarty, T. J., 1993. Audit committee effectiveness: An empirical investigation of

       the contribution of power. Auditing: A Journal of Practice & Theory 12, 24-49.

Kiger, J., Scheiner, J., 1997. Auditing. Second edition. Houghton Mifflin Company, Boston, MA.

Kinney, M., Trezevant, R., 1997. The use of special items to manage earnings and perceptions.

       Journal of Financial Statement Analysis 3, 45–53.

Klein, A., 2002. Audit committee, board of director characteristics, and earnings management.

       Journal of Accounting and Economics 33, 375–400.

Koch, A., 2002. Financial distress and the credibility of management earnings forecasts. Working

       paper, Carnegie Mellon University.

Krishnan, J., 2005. Audit committee quality and internal control: An empirical analysis. The

       Accounting Review 80, 649-675.

Lee, H. Y., Mande, V., Ortman, R., 2004. The effect of audit committee and board of director

       independence on auditor resignation. Auditing: A Journal of Practice & Theory 23, 131-

       146.

Lee, L-F., 1979. Identification and estimation in binary choice models with limited (censored)

       dependent variables. Econometrica 47, 977-996.

Levi, S., 2005. Voluntary disclosure of accruals in preliminary-earnings announcements and the

       pricing of accruals. Working paper, New York University.

Libby, R., Luft, J., 1993. Determinants of judgment performance in accounting settings: Ability,

       knowledge, motivation, and environment. Accounting, Organizations and Society 18, 425–

       450.

Lipman, F. D., 2004. Six common mistakes of audit committees. Directors and Boards 28, 30.




                                               39
Lipton, M., Lorsch, J. W., 1992. A modest proposal for improved corporate governance. Business

       Lawyer 48, 59-77.

Livingston, P., 2003. Financial experts on audit committees. Financial Executive 19, 6-7.

Loderer, C., Peyer, U., 2002. Board overlap, seat accumulation and share prices. European

       Financial Management 8, 165-192.

Loebbecke, J. K., Eining, M. M., Willingham, J. J., 1989. Auditors’ experience with material

       irregularities: Frequency, nature and delectability. Auditing: A Journal of Practice & Theory

       9, 1-28.

Louwers, T., 1998. The relation between going-concern opinions and the auditor’s loss function.

       Journal of Accounting Research 36, 143-156.

Lovell, M., 1963. Seasonal adjustment of economic time series and multiple regression analysis.

       Journal of the American Statistical Association 58, 993-1010.

Maddala, G. S., 1983. Limited-dependent and qualitative variables in econometrics. Cambridge

       University Press, New York, NY.

McDaniel, L., Martin, R. D., Maines, L. A., 2002. Evaluating financial reporting quality: The

       effects of financial expertise vs. financial literacy. The Accounting Review 77, 139-167.

McMullen, D. A., Raghunandan, K., 1996. Enhancing audit committee effectiveness. Journal of

       Accountancy 182, 79-81.

McNichols, M., 2002. Discussion of the quality of accruals and earnings: The role of accrual

       estimation errors. The Accounting Review 77, 61-69.

McVay, S. E., 2005. Earnings management using classification shifting: An examination of core

       earnings and special items. The Accounting Review 81, 501-531.

Menon, K., Williams, J. D., 1994. The use of audit committees for monitoring. Journal of

       Accounting and Public Policy 13, 121-139.




                                                40
Mercer, M., 2004. How do investors assess the credibility of management disclosures? Accounting

       Horizons. 18, 185-196.

Minow, N., Bingham, K., 1995. “The ideal board” in corporate governance. Blackwell Publishers,

       Cambridge, MA.

Mutchler, J. F., Hopwood, W., McKeown, J. C., 1997. The influence of contrary information and

       mitigating factors on audit opinion decisions on bankrupt companies. Journal of Accounting

       Research 35, 295-310.

Myers, L., Skinner, D., 2002. Earnings momentum and earnings management. Working paper,

       University of Michigan.

National Association of Corporate Directors (NACD), 1999. Report of the blue ribbon commission

       on audit committees. NACD, Washington, DC.

National Commission on Fraudulent Financial Reporting (Treadway Commission), 1987. Report of

       the National Commission on fraudulent financial reporting. United States Government

       Printing Office, Washington, DC.

New York Stock Exchange (NYSE), 1973. Response to the white paper questionnaire concerning

       recommendations and comments on financial reporting to shareholders and related matters.

       NYSE, New York, NY.

Patton, A., Baker, J., 1987. Why do not directors rock the boat? Harvard Business Review 65, 10-

       12.

Plitch, P., Ceron, C. F., 2003. Finance expertise differs by regulator. Wall Street Journal April 16:

       B4D.

Public Oversight Board (POB), 1993. In the public interest: Issues confronting the accounting

       profession. AICPA, Stamford, CT.

Raghunandan. K., Read, W. J., Rama, D. V., 2001. Audit committee composition, "gray directors,"

       and interaction with internal auditing. Accounting Horizons 15, 105-118.


                                                 41
Raghunandan. K., Read, W. J., 2001. The state of audit committees. Journal of Accountancy 191,

       57-60.

Raghunandan. K., Rama, D. V., 2003. Audit committee composition and shareholder actions:

       Evidence from voting on auditor ratification. Auditing: A Journal of Practice & Theory 22,

       253-263.

Rajgopal, S., Venkatachalam, M., Kotha, S., 2003. The value-relevance of network advantages: The

       case of e-commerce firms. Journal of Accounting Research 41, 135-162.

Rangan, S., 1998. Earnings management and the performance of seasoned equity offerings. Journal

       of Financial Economics 50, 101–122.

Rosenstein, S., Wyatt, J. G., 1990. Outside directors, board independence, and shareholder wealth.

       Journal of Financial Economics 26, 175–91.

Sarbanes-Oxley Act of 2002 (SOX), 2002. Public Law No. 107-204, 116 Stat. 745, sec. 1-1107.

       GPO, Washington, DC.

Scarpati, S. A., 2003. CPAs as audit committee members. Journal of Accountancy 196, 32-35.

Scott, H. S., 2001. The SEC, the audit committee rules, and the marketplaces: Corporate

       governance and the future. Washington University Law Quarterly 79, 549-571.

Securities and Exchange Commission (SEC), 1972. Standing audit committees composed of outside

       directors: Accounting Series Release No. 123. SEC, Washington, DC.

Securities and Exchange Commission (SEC), 1974. Notice of amendments to require increased

       disclosure of relationships between registrants and their independent public accountants:

       Accounting Series Release No. 165. SEC, Washington, DC.

Securities and Exchange Commission (SEC), 2003. Final rule: Disclosure required by sections 406

       and 407 of the Sarbanes-Oxley Act of 2002. SEC, Washington, DC.

Shivakumar, L., 2000. Do firms mislead investors by overstating earnings before seasoned equity

       offerings? Journal of Accounting and Economics 29, 339–370.


                                               42
Shivdasani, A., 1993. Board composition, ownership structure, and hostile takeovers. Journal of

       Accounting and Economics 16, 167-198.

Teoh, S. H., Welch, I., Wong, T. J., 1998. Earnings management and post-issue underperformance

       of seasoned equity offerings. Journal of Financial Economics 50, 63-99.

Tsui, J., Jaggi, B., Gul, F., 2001. CEO determination, growth opportunities, and their impact on

       audit fees. Journal of Accounting, Auditing, and Finance 16, 189-208.

Wild, C. J., Seber, G. A. F., 2000. Chance encounters: A first course in data analysis and inference.

       Wiley, New York, NY.

Wysocki. P., 2006. Assessing earnings and accruals quality: U.S. and international evidence.

       Working paper, MIT.

Yermack, D., 1996. Higher market valuation of companies with a small board of directors. Journal

       of Financial Economics 40, 185–211.

Zhou, J., Elder, R., 2004. Audit quality and earnings management by seasoned equity offering

firms. Asia-Pacific Journal of Accounting and Economics 11, 95-120.




                                                 43
Appendix

Accruals quality

We derive our measure of accruals quality using a modified version of the Dechow and Dichev (2002) model that
relates working capital accruals32 to lagged, current and future cash flows from operations. The model developed by
Dechow and Dichev (2002) embodies the intuition that current accruals are estimates of future cash flow realizations,
and that accruals quality is an inverse function of the precision of these estimates. Their measure of accruals quality is
based on the residuals from firm-specific regressions of changes in working capital accruals on lagged, current and
future cash flows from operations. We employ two modifications to the model. First, following McNichols (2002), we
include change in sales33 (employed in the Jones 1991 model) as an additional independent variable. McNichols (2002)
demonstrates the benefits of considering the implications of both the Jones (1991) and Dechow and Dichev (2002)
models, in developing more powerful approaches to estimating accruals quality in the presence of management
discretion. Second, following Lovell (1963) and Rajgopal et al. (2003) we control for potential seasonality effects. The
model is as follows:

                                                                                              3
∆WCi ,t = φ 0,i + φ1,i CFOi ,t −1 + φ 2,i CFOi ,t + φ 3,i CFOi ,t +1 + φ 4,i ∆SALES i ,t + ∑ φ 5 n ,i QTRi ,n + ε i ,t
                                                                                             n =1



where ∆WC i ,t is the change in working capital accruals of firm i in quarter t, measured using data from the statement of
cash flows34 scaled by average assets (Compustat item 44), CFOi ,t is the cash flow from operations of firm i in quarter t
(Compustat item 108) scaled by average assets, ∆SALES i ,t is the change in sales of firm i in quarter t (Compustat item
2) scaled by average assets, QTRi ,n is a dummy variable that takes a value of 1 when the observation is from fiscal
quarter n and 0 otherwise, and ε i,t is the residual of firm i in quarter t.

For each sample firm, we estimate regression (1) using the available time series quarterly data for that firm during
1995–1998. These regressions result in quarterly residuals ( ε i,t ) for each firm. These quarterly residuals would

intuitively represent abnormal changes in working capital accruals. Similar to Dechow and Dichev (2002), we measure
accruals quality by calculating the standard deviation of firm i’s estimated residuals from regression (1). This measure
is based on the time-series mechanics of accruals and on the intuition that any large positive abnormal accruals will be
offset by future negative abnormal accruals. Under this measure of accruals quality, higher volatility in abnormal
accruals (larger standard deviation of the residuals) would imply lower accruals quality.




32. We chose to examine working capital accruals rather than total accruals as previous research finds a larger presence
of managerial discretion in relation to working capital accruals (Dechow et al., 1996).
33. We considered but chose not to include property, plant, and equipment (PPE) as an additional independent variable,
as PPE is normally related to long term-accruals, while our dependent variable (working capital accruals) is a measure
of short-term accruals.
34. More specifically, change in working capital accruals is measured as the increase in accounts receivable
(Compustat item 103) plus the increase in inventory (Compustat item 104) plus the decrease in accounts payable and
accrued liabilities (Compustat item 105) plus decreases in taxes accrued (Compustat item 106) plus the increase
(decrease) in other assets (liabilities) (Compustat item 107). Quarterly data on all of these items reflect a year-to-date
figure and consequently had to be adjusted to reflect quarterly changes.


                                                                           44
Table 1
Sample derivation and industry membership

Panel A: Sample derivation
                                                                                        Individuals
                                                                                      Serving as Audit
                                                                                        Committee        Usable Compustat
                                                                                         Members           Observations
                                                                             Firms

Original sample obtained from IRRC                                           2,034        6,021
Less firms with unavailable data on quarterly Compustat files                 235          627
                                                                             1,749        5,394              19,436
Less firms with less than eight observations                                  331         1,403               527
                                                                             1,418        3,991              18,909
Less banks and financial institutions                                          89          288               1,123
                                                                             1,329        3,703              17,789
Less firms in regulated industries                                            215          527               2,517
                                                                             1,114        3,176              15,269
Add missing information obtained from proxy statements                          .          334                  .
Final Test Sample                                                            1,114        3,510              15,269

Panel B: Industry membership of sample firms
Two-Digit                                                                            Number                % of
SIC Codes                     Industry Name                                          of Firms            Sample
    10                        Metal mining                                                 11              0.99
    13                        Oil and gas extraction                                       47              4.22
    20                        Food and kindred products                                    36              3.23
    22                        Textile mill products                                        17              1.53
    23                        Apparel and other textile products                           15              1.35
    24                        Lumber and wood products                                     11              0.99
    26                        Paper and allied products                                    25              2.24
    27                        Printing and publishing                                      34              3.05
    28                        Chemicals and allied products                              104               9.34
    29                        Petroleum and coal products                                  18              1.62
    30                        Rubber and misc. plastics products                           19              1.71
    32                        Stone, clay and glass products                               13              1.17
    33                        Primary metal industries                                     33              2.96
    34                        Fabricated metal products                                    27              2.42
    35                        Industrial machinery and equipment                           88              7.90
    36                        Electronic and other electric equipment                      95              8.53
    37                        Transportation equipment                                     44              3.95
    38                        Instruments and related products                             54              4.85
    39                        Misc. manufacturing industries                               12              1.08
    50                        Wholesale trade-durable goods                                38              3.41
    51                        Wholesale trade-nondurable goods                             18              1.62
    53                        General merchandise stores                                   14              1.26
    54                        Food stores                                                  11              0.99
    56                        Apparel and accessory stores                                 24              2.15
    57                        Furniture and home furnishings stores                        11              0.99
    58                        Eating and drinking places                                   18              1.62
    59                        Miscellaneous retail                                         29              2.60
    73                        Business services                                          110               9.87
    80                        Health services                                              16              1.44
    87                        Engineering and management services                          10              0.90
    Total                     30 INDUSTRIES                                            1,002              89.95
    Others                    23 INDUSTRIES                                              112              10.05
    TOTAL SAMPLE              53 INDUSTRIES                                            1,114             100.00




                                                                        45
Table 2
Descriptive statistics over the period of 1995-1998. Sample consists of 1,114 firms.
Variable                                                                       Mean            Q1     Median            Q3   Std.
                                                                                                                             Dev.
Panel A: Data used to construct test variables
Number of AC Members with Financial Expertise*                                 1.145      0.333      1.000      2.000     0.919
Number of AC Members with Accounting Expertise                                 0.266      0.000      0.000      0.333     0.473
Number of AC Members with Finance Expertise                                    0.317      0.000      0.000      0.667     0.537
Number of AC Members with Supervisory Expertise                                0.548      0.000      0.000      1.000     0.709
Number of AC Members                                                           2.737      2.000      2.667      3.000     1.096
Proportion of Independent AC Members                                           0.698      0.500      0.800      1.000     0.324
Number of AC Meetings                                                          2.726      2.000      2.500      3.500     1.324
Number of Board Members                                                        8.920      7.000      8.667     10.500     2.604
Proportion of Independent Board Members                                        0.577      0.444      0.591      0.714     0.181
Ownership of Independent Board Members                                         0.054      0.001      0.008      0.031     0.235
CEO is the Board Chairman                                                      649 out of 1,114 firms (58%) had the same person
                                                                               perform the roles of CEO and board chairman during
                                                                               the period 1995 to 1998.
Panel B: Data used to construct control variables
Log (Total Assets)                                                             6.536       5.587      6.467       7.398     1.456
Standard Deviation of Cash Flows from Operations                               0.026       0.011      0.018       0.029     0.039
Proportion of Quarters Reporting Negative Earnings                             0.156       0.000      0.063        0.188    0.208
Standard Deviation of Sales                                                    0.069       0.029      0.052        0.085    0.062
Log (Operating Cycle)                                                          4.878       4.560      5.015        5.166    0.604
Reporting of R&D and Advertising                                               386 out of 1,114 firms (35%) did not report any
                                                                               R&D and Advertising during the period 1995 to
                                                                               1998.
Total R&D and Advertising as a Proportion of Sales                             0.074       0.000      0.016        0.073    0.321
Book value of PP&E to Total Assets Ratio                                       0.315       0.158      0.276        0.422    0.209
Absolute Value of Special Items to Assets                                      0.006       0.000      0.003        0.008    0.010
Book to Market Ratio                                                           0.447       0.258      0.403       0.571     0.604
Issue of Securities                                                            271 out of 1,114 firms (24%) issued securities during
                                                                               the period 1995 to 1998.
Financial Distress                                                             268 out of 1,114 firms (24%) were under financial
                                                                               stress during the period 1995 to 1998.
Descriptive statistics on continuous data are based on average firm-level values during the period 1995 through 1998.
*
    = Definition of financial expertise based on the broad definition of financial expertise




                                                                             46
Table 3
Descriptive statistics on accruals quality of sample firms. Sample consists of 1,114 firms.
The computation of accruals quality (STDRES) is based on the Dechow and Dichev (2002) methodology and is based on the standard deviation of firm i’s estimated residuals
                                                                                                                                   3
( ε i,t ) from firm-level regression: ∆WCi ,t = φ0,i + φ1,i CFOi ,t −1 + φ 2,i CFOi ,t + φ3,i CFOi ,t +1 + φ 4,i ∆SALES i ,t + ∑ φ5 n ,i QTRi ,n + ε i ,t
                                                                                                                                  n =1


Panel A: Accruals quality based on presence/absence of accounting expertise in audit committees
                                   FULL SAMPLE                              ACGOV = 1                               ACGOV = 0                                BDGOV = 1                              BDGOV = 0
                           ACCTG =1           ACCTG = 0           ACCTG =1             ACCTG = 0            ACCTG =1         ACCTG = 0              ACCTG =1          ACCTG = 0            ACCTG =1          ACCTG = 0
Mean STDRES                  0.0102             0.0130              0.0042               0.0098              0.0140            0.0143                 0.0091            0.0121               0.0104            0.0132
Median STDRES                0.0064             0.0088              0.0036               0.0074              0.0101            0.0098                 0.0068            0.0094               0.0063            0.0087
No. of Observations            220                894                  87                  643                 133               643                     46               210                  174               684
Difference P-value                     0.0000                                  0.0000                                 0.6733                                   0.0480                                 0.0000

Panel B: Accruals quality based on presence/absence of finance expertise in audit committees
                                   FULL SAMPLE                               ACGOV = 1                               ACGOV = 0                               BDGOV = 1                              BDGOV = 0
                          FINANCE = 1        FINANCE = 0         FINANCE = 1         FINANCE = 0           FINANCE = 1       FINANCE = 0           FINANCE = 1       FINANCE = 0          FINANCE = 1       FINANCE = 0
Mean STDRES                  0.0119             0.0126               0.0079              0.0086               0.0144            0.0142                0.0122            0.0113               0.0118            0.0130
Median STDRES                0.0080             0.0086               0.0056              0.0064               0.0097            0.0099                0.0088            0.0087               0.0078            0.0086
No. of Observations             283               831                  108                 230                  175               601                    76               180                  207               651
Difference P-value                     0.3159                                  0.5053                                  0.9372                                  0.4521                                 0.1297

Panel C: Accruals quality based on presence/absence of supervisory expertise in audit committees
                                  FULL SAMPLE                               ACGOV = 1                                ACGOV = 0                                BDGOV = 1                             BDGOV = 0
                           SUPER = 1          SUPER = 0           SUPER =1             SUPER = 0            SUPER = 1         SUPER = 0              SUPER =1          SUPER = 0           SUPER = 1         SUPER = 0
Mean STDRES                  0.0124             0.0125              0.0089               0.0074               0.0151            0.0138                 0.0118            0.0113              0.0126            0.0128
Median STDRES                0.0083             0.0084              0.0068               0.0055               0.0101            0.0095                 0.0082            0.0088              0.0083            0.0084
No. of Observations             473               641                 210                  128                  263               513                    108               148                 365               493
Difference P-value                     0.7485                                  0.0373                                  0.4803                                   0.9891                                0.7710
Variable definitions:
ACCTG                       = 1 if the audit committee of the firm i included at least one member with accounting expertise during the period 1995-1998, and 0 otherwise
FINANCE                     = 1 if the audit committee of the firm i included at least one member with finance expertise during the period 1995-1998, and 0 otherwise
SUPER                       = 1 if the audit committee of the firm i included at least one member with supervisory skills during the period 1995-1998, and 0 otherwise
ACGOV                       = 1 if the AC GOV SCORE of firm i is greater than or equal to two, and 0 otherwise
BDGOV                       = 1 if the BD GOV SCORE of firm i is greater than or equal to 3, and 0 otherwise
AC GOV SCORE                = A summary measure which captures the overall strength of firm i's audit committee and is equal to the sum of the following three dichotomous audit committee variables:
                              AC SIZE            = 1 if the size of audit committee of firm i was greater or equal to three members in each year during the period 1995-1998, and 0 otherwise
                              AC IND             = 1 if the audit committee of firm i was composed entirely of independent audit committee members in each year during the period 1995-1998, and 0 otherwise
                              AC MEET            = 1 if the audit committee of firm i met more than three times in each year during the period 1995-1998, and 0 otherwise
BD GOV SCORE                = A summary measure which captures the overall strength of firm i's board and is equal to the sum of the following four dichotomous audit committee variables:
                              BOARD SIZE         = 1 if the average size of board of firm i during the period 1995-1998 was less than the sample median, and 0 otherwise
                              BOARD IND          = 1 if the average proportion of independent directors in firm i’s is greater than 60% during the period 1995-1998, and 0 otherwise
                              SHARE OWN          = 1 if the average proportion of firm i's shares held by independent directors during the period 1995-1998 is greater than or equal to the sample median, and 0 otherwise
                              DUALITY            = 1 if the CEO of firm i was also its board chairman during the period 1995-1998, and 0 otherwise




                                                                                                              47
Table 4
Correlation matrix of the dependent and independent variables. Sample consists of 1,114 firms.

              STDRES      ACCTG      FINANCE      SUPER      ACGOV       BDGOV      ASSETS      σ(CFO)     NEGEARN       Σ(SALES)     OPCYCLE       INTINT     INTDUM       CAPINT      SPITEM      GROWTH       ISSUE     DISTRESS

STDRES        1.000       -0.083     -0.024       -0.004     -0.199      -0.035     -0.351      0.532      0.378          0.391        0.142         0.175     -0.116       -0.208      0.268       -0.028        0.090    0.108
ACCTG                     1.000      0.063         0.085      0.099      -0.024      0.006      0.020      0.062          0.046        0.017        -0.023      0.018        0.023      0.004       0.020         0.081    -0.050
FINANCE                              1.000         0.041      0.099      0.054      -0.039      0.012      -0.022         0.029        0.054        -0.008     -0.065       -0.000      0.003       -0.027        0.092    0.007
SUPER                                              1.000      0.263      -0.003      0.061      -0.008     -0.030         0.009        0.010         0.033      0.008       -0.020      -0.001      -0.002        0.046    0.024
ACGOV                                                         1.000      0.025       0.017      -0.099     -0.004        -0.083       -0.024        -0.021     -0.033        0.033      0.006       0.013        -0.028    -0.048
BDGOV                                                                    1.000      -0.242      0.053      0.085          0.058        0.000         0.072     -0.066       -0.048      0.077       -0.005        0.068    0.025
ASSETS                                                                               1.000      -0.356     -0.272        -0.257       -0.116        -0.149     -0.016        0.208      -0.129      -0.016       -0.064    -0.017
σ(CFO)                                                                                          1.000      0.338          0.459        0.039         0.052      0.008       -0.117      0.178       -0.023        0.037    0.081
NEGEARN                                                                                                    1.000          0.133        0.017         0.211      0.009       -0.025      0.391       0.043         0.046    -0.187
σ(SALES)                                                                                                                  1.000        0.026         0.013      0.016       -0.233      0.104       -0.074        0.161    0.169
OPCYCLE                                                                                                                                1.000         0.101     -0.206       -0.411      -0.019      -0.029        0.047    0.055
INTINT                                                                                                                                               1.000     -0.169       -0.130      0.098       -0.052        0.081    0.025
INTDUM                                                                                                                                                          1.000        0.248      -0.082      0.060         0.097    -0.068
CAPINT                                                                                                                                                                       1.000      -0.122      0.049         0.013    -0.108
SPITEM                                                                                                                                                                                  1.000       0.018         0.032    -0.053
GROWTH                                                                                                                                                                                              1.000        -0.054    -0.105
ISSUE                                                                                                                                                                                                             1.000    0.012
DISTRESS                                                                                                                                                                                                                     1.000
Pearson correlations significant at the 5% level are in bold figures
Variable definitions:
STDRESi = Quality of accruals in firm i measured as the standard deviation of the residuals from firm-specific regression during the sample period (1995-1998); ACCTGi = 1 if the audit committee of the firm i included at least one
member with accounting expertise in each year during the sample period, and 0 otherwise; FINANCEi = 1 if the audit committee of the firm i included at least one member with finance expertise in each year during the sample period,
and 0 otherwise; SUPERi = 1 if the audit committee of the firm i included at least one member with supervisory skills in each year during the sample period, and 0 otherwise; ACGOVi = 1 if the AC GOV SCORE of firm i is greater than
or equal to two, and 0 otherwise; BDGOVi = 1 if the BD GOV SCORE of firm i is greater than or equal to 3, and 0 otherwise; ASSETSi = log of average total assets of firm i during the sample period; σ(CFO)i = standard deviation of
seasonally adjusted working capital accruals of firm i during the sample period; NEGEARNi = Proportion of fiscal quarters during the sample period where firm i reporting negative earnings; σ(SALES)i = standard deviation of
seasonally sales revenue of firm i during the sample period; OPCYCLEi = Log of average operating cycle of firm i during the sample period; INTINTi = Average ratio of reported R&D and advertising expense to total sales revenue of
firm i during the sample period; INTDUMi = 1 if INTINTi = 0, and 0 otherwise; CAPINTi = Average ratio of net book value of property, plant and equipment to total assets of firm i during the sample period; SPITEMi = Average
absolute value of special items scaled by total assets of firm i during the sample period; GROWTHi = Average book to market equity ratio for firm i during the sample period; ISSUEi = 1 if firm i issued securities during the sample
period and 0 otherwise; and DISTRESSi = 1 if the average Altman Z Score of firm i over the sample period signaled financial distress for the firm and 0 otherwise




                                                                                                                 48
Table 5
Cross-sectional regression test of accruals quality on expertise, governance, interaction and
control variables. Sample consists of 1,114 firms.
STDRES i = β 0 + β1 ACCTG i + β 2 FINANCE i + β 3 SUPERi + β 4 ACGOVi + β 5 BDGOVi + β 6 ACCTG i ∗ ACGOVi

              + β 7 ACCTG i ∗ BDGOVi + β 8 FINANCE i ∗ ACGOVi + β 9 FINANCE i ∗ BDGOVi + β10 SUPERi ∗ ACGOVi

              + β11 SUPERi ∗ BDGOVi + β12 ASSETS i + β13σ (CFO ) i + β14 NEGEARN i + β15σ ( SALES ) i + β16 OPCYCLE i

              + β17 INTINTi + β18 INTDUM i + β19 CAPINTi + β 20 SPITEM i + β 21GROWTH i + β 22 ISSUE i + β 23 DISTRESS i
                 n
              + ∑ γ j INDUSTRY j + ε i                                                                                 (2)
                 j =1


                                      Model 1                    Model 2                   Model 3       Model 4         Model 5
                                      0.0090                    -0.0108                    -0.0088       -0.0085        -0.0097
INTERCEPT                             0.1094                    -0.0687 *                  -0.1187       -0.1248        -0.0931 *
                                     -0.0032                    -0.0015                                                 -0.0016
ACCTG                                -0.0000 ***                -0.0851 *                                               -0.0755 *
                                     -0.0001                                               -0.0009                      -0.0008
FINANCE                               0.4625                                               -0.1760                      -0.2047
                                     -0.0012                                                             -0.0005        -0.0008
SUPER                                -0.0232 **                                                          -0.2549        -0.1877
                                     -0.0036                    -0.0022                    -0.0032       -0.0050        -0.0031
ACGOV                                -0.0000 ***                -0.0021 ***                -0.0000 ***   -0.0054 ***    -0.0054 ***
                                     -0.0042                    -0.0048                    -0.0044       -0.0038        -0.0044
BDGOV                                -0.0000 ***                -0.0000 ***                -0.0000 ***   -0.0000 ***    -0.0000 ***
                                                                -0.0048                                                 -0.0046
ACCTG * ACGOV                                                   -0.0013 ***                                             -0.0021 ***
                                                                -0.0006                                                 -0.0006
ACCTG * BDGOV                                                   -0.3839                                                 -0.3700
                                                                                           -0.0012                      -0.0010
FINANCE * ACGOV                                                                            -0.2195                      -0.2451
                                                                                           -0.0013                      -0.0013
FINANCE * BDGOV                                                                            -0.2089                      -0.2233
                                                                                                         -0.0021        -0.0017
SUPER * ACGOV                                                                                            -0.0687 *      -0.1224
                                                                                                         -0.0005        -0.0008
SUPER * BDGOV                                                                                            -0.3690        -0.2962
                                     -0.0011                    -0.0011                    -0.0011       -0.0011        -0.0011
ASSETS                               -0.0000 ***                -0.0000 ***                -0.0000 ***   -0.0000 ***    -0.0000 ***
                                     -0.1145                    -0.1144                    -0.1163       -0.1154        -0.1146
σ(CFO)                               -0.0000 ***                -0.0000 ***                -0.0000 ***   -0.0000 ***    -0.0000 ***
                                     -0.0121                    -0.0116                    -0.0114       -0.0120        -0.0117
NEGEARN                              -0.0000 ***                -0.0000 ***                -0.0000 ***   -0.0000 ***    -0.0000 ***
                                     -0.0363                    -0.0358                    -0.0358       -0.0355        -0.0360
σ(SALES)                             -0.0000 ***                -0.0000 ***                -0.0000 ***   -0.0000 ***    -0.0000 ***
                                     -0.0018                    -0.0018                    -0.0018       -0.0018        -0.0019
OPCYCLE                              -0.0043 ***                -0.0053 ***                -0.0052 ***   -0.0062 ***    -0.0039 ***
                                     -0.0023                    -0.0025                    -0.0025       -0.0025        -0.0025
INTINT                               -0.0117 **                 -0.0074 ***                -0.0073 ***   -0.0079 ***    -0.0086 ***
                                     -0.0028                    -0.0028                    -0.0027       -0.0026        -0.0028
INTDUM                               -0.0003 ***                -0.0003 ***                -0.0004 ***   -0.0008 ***    -0.0004 ***
                                     -0.0035                    -0.0036                    -0.0040       -0.0042        -0.0037
CAPINT                               -0.0726 *                  -0.0644 *                  -0.0482 **    -0.0396 **     -0.0604 *
                                     -0.1350                    -0.1453                    -0.1398       -0.1377        -0.1434
SPITEM                               -0.0000 ***                -0.0000 ***                -0.0000 ***   -0.0000 ***    -0.0000 ***
                                     -0.0005                    -0.0005                    -0.0004       -0.0005        -0.0005
GROWTH                               -0.1694                    -0.1727                    -0.1904       -0.1819        -0.1774
                                     -0.0017                    -0.0016                    -0.0015       -0.0015        -0.0016
ISSUE                                -0.0121 **                 -0.0172 **                 -0.0269 **    -0.0253 **     -0.0143 **
                                     -0.0014                    -0.0015                    -0.0015       -0.0015        -0.0015
DISTRESS                             -0.0217 **                 -0.0166 **                 -0.0163 **    -0.0146 **     -0.0162 **
Adjusted R2                          -0.4552 ***                -0.4572 ***                -0.4457 ***   -0.4468 ***    -0.4579 ***
See Table 4 for Variable definitions
(***), (**) and (*) denote significance at the 0.01, 0.05, and 0.10 levels, respectively



                                                                                49
50

								
To top