Audit committee accounting expertise, CEO power, and audit pricing^*

* Accepted by Suresh Radhakrishnan upon recommendation of Inder Khurana.

Abstract

The Sarbanes–Oxley Act of 2002 (SOX) mandates that all listed firms disclose whether they have a financial expert on the audit committee, highlighting the committee’s expertise. However, some argue that non-accounting financial experts, compared to accounting financial experts, are not sufficient to ensure audit committee effectiveness because the former lack accounting knowledge. Accounting experts on audit committees may require higher audit efforts, while auditors may assess audit committees with accounting financial experts as effective, decreasing audit efforts. This paper first inspects the effect of audit committee accounting expertise on audit fees as a proxy for audit efforts, and then investigates whether the effect is moderated by powerful CEOs. Using post-SOX period data, our results show that, on average, firms with accounting experts on audit committees are more likely to pay higher audit fees, and the effect is less pronounced when a powerful CEO manages a firm.

Keywords:

• Accounting experts
• audit committee
• audit fees
• CEO power

JEL Classification:

• G30
• G34
• G38
• M12
• M40
• M41
• M42
• H83

Acknowledgements

We appreciate the helpful comments from Jong-Hag Choi, Jong Eun Lee, and the workshop participants at the 2013 Korean Accounting Association and 2014 American Accounting Association annual meetings.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

* Accepted by Suresh Radhakrishnan upon recommendation of Inder Khurana.

1. In detail, SOX Section 301 mandates that the audit committee be composed entirely of independent outside directors. SOX Section 407 also requires that all listed firms disclose whether they have a financial expert on the audit committee or, if not, to explain why they do not. In line with SOX, NYSE, AMEX, and NASDAQ have proposed even stricter requirements, insisting that listed firms have at least one financial expert on the audit committee.

2. There is an exemption in the independence rule (SOX 301(C) EXEMPTION AUTHORITY). Thus, we observe that a small number of firms still do not have fully independent audit committees.

3. For details, refer to Defond, Hann, and Hu (2005), Krishnan and Visvanathan (2008), and Dhaliwal, Naiker, and Navissi (2010).

4. According to Cohen, Krishnamoorthy, and Wright (2010), the adoption of SOX has led audit committees to be more active and diligent. Especially, this phenomenon is more pronounced for accounting experts because they have greater litigation and reputation risks relative to other audit committee members (Cost and Miller 2005; Srinivasan 2005). We discuss this result in Section 6.5 later.

5. Gotti et al. (2012) also examine the relationship between CEO ownership and audit fees. However, they do not simultaneously consider the characteristics of the audit committee, the most responsible party in the audit fee decision.

6. Singhvi, Raghunandan, and Mishra (2013) revisit this research question. They show that findings by Davidson, Xie, and Xu (2004) and Defond, Hann, and Hu (2005) are not valid post-SOX.

7. They define a financial expert as a member with experience in the role of either CPA, CFO, controller, treasurer, vice president for finance, investment banker, or venture capitalist. We think that their definition is closer to the conventional accounting expert definition.

8. Carcello et al. (2002) and Abbott et al. (2003) use the broad definition of financial expert, not the narrow one.

9. Findings by Carcello et al. (2002) on the association between audit committee characteristics and audit fees are not robust. The significant results disappear when they concurrently control for board and audit committee characteristics.

10. Throughout the paper, we examine the difference between independent accounting experts and independent non-accounting financial experts. Major stock exchanges such as NYSE/AMEX and NASDAQ require all listed firms to have at least one financial expert on fully independent audit committees. Using hand-collected data in 2003, Williams (2005) confirms that about 99% of large firms have a financial expert. On the other hand, Krishnan and Visvanathan (2009) compare accounting experts and non-accounting experts, including non-accounting financial experts and non-experts, regardless of their independence.

11. Petersen (2009) remarks that, if the number of firm clusters is much greater than the number of year clusters, it is enough to use one-way clustered standard errors in the firm dimension.

12. Following Reichelt and Wang (2010), we define a binary choice variable MSALead with a value of 1 if the audit market leader’s market share exceeds that of the second-largest market share auditor by 10%, within a two-digit SIC category in a particular year, and in a particular Metropolitan Statistical Area (MSA).

13. One might doubt whether multicollinearity problems would result from closely related control variables. In an untabulated correlation table, we check that our correlation matrix is similar to one in previous literature (Collier and Gregory 1996; Carcello et al. 2002; Choi et al. 2010).

14. When we extend the sample period to commence from 2000, our results are not changed.

15. Petersen (2009) remarks that Fama–Macbeth statistics are not designed to address time-series correlation (i.e. firm effect). Thus, if there are severe firm effects as well as time effects, Fama–Macbeth statistics are not better than t statistics based on cluster-adjusted standard errors. Thus, we admit that Fama–Macbeth statistics may not be “more robust” than the statistics shown in Table 3 and Table 4.

16. Our analysis is based on samples from years 2000 to 2002 because companies must comply with SOX section 407 in their financial statements starting with fiscal year ending 15 July 2003.