For an sec restricted client, which activity would be considered categorically prohibited?

Section 201(g) of the Sarbanes-Oxley Act (the Act) makes it unlawful for a registered public accounting firm to provide certain, but not all, non-audit services to audit clients that are subject to the Securities and Exchange Act of 1934. Included in the prohibited list are bookkeeping, financial information system design and valuation services. Under section 202(h), an audit firm may provide to audit clients any non-audit service that is not specifically proscribed by section 202(g) subject to pre-approval by the audit committee. Further, section 202(g)(9) allows the Public Company Accounting Oversight Board (PCAOB) to ban the contemporaneous provision of any other non-audit services, if the Board deems the provision of such service to be impermissible. Although these prohibitions represent significant interventions in the operations of the non-audit service market, they have not been supported by systematic research. For instance, what is the basis for banning some, but not other, non-audit services and how is the PCAOB going to determine which non-audit services to ban in the future?

This study revisits the debate over whether accounting firms, in their zeal to provide highly profitable non-audit services, might (1) accept riskier clients and (2) fail to adequately respond to this increased risk thereby compromising audit quality. Although SEC concerns over auditor independence (SEC, 2001) combined with the Sarbanes-Oxley legislation of 2002 have dramatically curtailed the types of non-audit services that auditors provide to their audit clients, the provision of non-audit services remains a significant source of revenue for the firms (Bryan-Low, 2002). In fact, for fiscal years ending in 2003, over 40% of the fees that clients paid to their auditors were for non-audit services (Plitch and Rapoport, 2004).

There appears to be a trend to expand the scope of prohibited services. In 2001 the profession, under pressure from the SEC agreed to limit the types of permissible services. The list of prohibited services was further expanded by the Sarbanes-Oxley Act. More recently, tax consulting has been questioned. Large international accounting firms are being scrutinized for their roles in providing clients with aggressive tax advice (WSJ, 2004) raising the possibility that additional restrictions maybe placed on auditors’ ability to provide non-audit services. Daniel Goelzer of the PCAOB questioned whether certain tax services should be disallowed (2003). Further, in July 2004 the PCAOB held a roundtable to discuss the value of allowing auditors to continue to provide tax services.

Because these prohibitions constitute an intrusion into the operations of a market and potentially impose unnecessary costs on market participants, it is important that they are guided by research.1 For example, while it has been argued that non-audit services may impair auditor independence, there is reason to believe the provision of non-audit services can actually improve audit quality by allowing the auditor to better understand the client’s business. Empirical evidence examining the relationship between non-audit fees and independence-in-fact is equivocal (Frankel et al., 2002, Ashbaugh et al., 2003).

Further, work by Kinney et al. (2004) (1) was unable to document a significant association between restatements and the provision of some non-audit services that have been banned and (2) reports evidence of a negative relationship between restatements and the fees paid for tax services. Such results suggest that perhaps some of the prohibitions against the provision of non-audit services may not necessarily be in the public interest. Along those lines, the California Employees’ Retirement System (Calpers) is rethinking its policy related to auditors providing non-audit services. In 2003 the Calpers board unanimously adopted a policy of withholding support for audit committee members that authorize auditors to perform non-audit services. However, this policy may have been a knee-jerk reaction as some board members are now reported to be reconsidering the policy (Sidel, 2004).

Although the Sarbanes-Oxley Act banned some services, it is nevertheless important to examine these issues to determine whether the justification for the ban can be found in auditors’ opportunistic decision making. If so, then the ban can be said to be in the public interest. If not, an alternative rationale for the ban must be found. To address these issues, we analyze data collected in 1999 and 2000, prior to both the demise of Enron and the November 2001 adoption of the SEC’s more restrictive independence rules that limited the types of services that audit firms could provide to their audit clients.

In this study, we use a behavioral experiment to add to the existing archival-based literature on the potential effect of providing non-audit services on auditor independence. In general, we study whether an auditor’s judgments and decision making are potentially affected by the provision of non-audit services. Specifically, our study examines the joint effects of the provision of non-audit services and management integrity (i.e., a factor related to client risk profile, Beaulieu, 2001, AICPA, 1994, Asare et al., 1994, Huss and Jacobs, 1991) on auditors’ client acceptance and staffing decisions.2

Johnstone (2000) modeled and tested a two stage risk-based client acceptance process, and found that experienced audit partners viewed engagement profitability as an essential component of an auditor’s business risk. Further, Johnstone found that the partners did not significantly engage in risk adaptation strategies to make riskier clients more attractive. Thus, building on Johnstone’s (2000) work, we examine whether the potential to earn non-audit fees, and thus increase a potential engagement’s profitability, will entice auditors to accept high risk clients. Secondly, we build on the work of Johnstone and Bedard (2001) to experimentally investigate whether in cases where higher risk clients are accepted, auditors will make staffing decisions to compensate for such additional risk in an effort to maintain audit quality.

Thus, in this analysis, we use a controlled experiment to examine the effect that the potential to sell non-audit services to a prospective audit client has on auditor judgments. To further examine the government’s concerns, we manipulate and analyze the effect of the client’s risk profile on auditor judgments. Prior research has shown that the client’s risk profile is an important factor that negatively affects the client acceptance decision (e.g., Farmer et al., 1987). However, if auditors accept risky clients that offer the potential for additional non-audit fees, they should also modify their audit approach (potentially resulting in a more costly audit) to accommodate the heightened risk.3 To do otherwise would support regulators’ claims that provision of non-audit services compromise audit quality. Johnstone and Bedard (2001) found that for prospective clients that were bid for, auditors planned to use more industry and high risk specialists where error risk factors were identified. Further, they found that audit engagement teams where non-audit services were provided were more likely to utilize industry experts than were engagement teams where non-audit services were not provided. Building on Johnstone and Bedard’s (2001) archival-based results, we not only examine the auditors’ risk assessments and client acceptance decisions, we also examine whether auditors modified their audit approaches in light of heightened risk.

The data analyzed in this paper was collected from seventy-three audit partners who were asked to evaluate a prospective hypothetical client. They were randomly assigned to one of the four experimental conditions created by fully crossing the potential to provide non-audit services and management integrity. The partners were asked to evaluate the audit firm’s business risk, make an acceptance decision, and propose a staffing plan for the potential engagement.

The data analysis yields two primary findings. First, we find that management integrity affected the acceptance decision, but there was no evidence that it interacted with the potential to provide non-audit services. Second, we also find evidence that auditors reallocated professional staff time in response to management integrity issues. In addition, we found that management integrity, but not the potential to provide non-audit services, is associated with the level of perceived business risk. Overall, our results in this experimental setting do not appear to lend support for regulators’ concerns that public accounting firms were engaged in an excessively aggressive pursuit of clients who could be provided non-audit services. Further, auditors staffing strategies appear to vary with the level of client risks. Thus, at least in this experimental setting, the evidence does not support claims that auditor independence-in-fact was necessarily compromised by the potential to provide non-audit services.

This study adds to the literature as follows: we use an experiment to test a two-stage, risk-based model of client acceptance (Johnstone, 2000) within the context of the potential to provide or not provide non-audit services. The results of those tests appear to be consistent with audit firms’ assertions that the potential to provide non-audit services does not necessarily drive decision-making in the client acceptance process. We also build on Johnstone and Bedard’s archival study (2001) to experimentally investigate the extent to which staff allocations are modified in light of the client risk profile. These findings indicate that more experienced staff play a greater role in the audits of riskier clients.

Perspectives on regulations

Regulations, such as the prohibition of non-audit services for audit clients, can be analyzed from three perspectives: (1) Public Interest, (2) Public Administration, (3) Public Choice (Harrison et al., 2004). The Public Interest analysis examines when and whether a regulatory intervention is likely to improve the operation of a market economy. It answers the question of whether the public is made better off by a regulatory scheme. For instance, economic analysis may be employed to examine

Institutional background

There has been a long-standing concern over the CPA’s ability to remain independent of audit clients to whom they also provide non-audit services. The Metcalf hearings (US Senate, 1977) raised this question and the Cohen Commission (AICPA, 1978) further examined the issue. More recently, some members of The Panel on Audit Effectiveness observed that success in marketing non-audit services was a significant factor in accounting firms’ compensation system and concluded that “as long as the

Development of hypotheses

The competing positions adopted by regulators and the accounting firms in this debate suggest testable hypotheses. The regulators proposed that the potential to provide non-audit services led firms to accept risky clients (i.e., clients whose management lacks integrity). Further, they proposed that firms accepting such risky clients, who provided the opportunity for non-audit revenues, may not always have made sufficient (and potentially costly) adjustments in subsequent audit strategy.

In

Participants

The participants were 73 partners from two of the then Big 5 firms. They had a mean (σ) of 22 (6.9) years of audit experience and had been partners for a mean (σ) of 10.5 (7.1) years. All participants reported that they had prior involvement in various phases of client acceptance decisions and participated in a median of 5 (with an inter quartile range of 3–12) client acceptance decisions per year. Therefore, the participants were the appropriate pool for the task.

A coordinating partner in each

Descriptive statistics

Table 1 presents descriptive statistics on attractiveness evaluations, business risk and client screening decisions. Across subjects, the mean (σ) attractiveness rating was 6.21 (2.19). As shown in Table 1, mean attractiveness of the client is higher for the high non-audit services (μ = 7.15 vs. 5.38 for low non-audit services) and the higher management integrity cells (μ = 7.34 vs. 5.15 for lower integrity). An analysis of variance found both the non-audit services (F = 12.30, p < 0.001) and

Conclusions

There is an ongoing debate regarding the extent to which auditor should be allowed to provide non-audit services to their audit clients. While there is concern over the effect of nonaudit services on both independence in fact and appearance, the empirical evidence is equivocal. Further, there is the public policy question of whether the provision of such services might improve audit quality by allowing the auditor to gain greater insight into the client’s operations. Evidence on this debate is

Cited by (32)

• Auditor search periods as signals of engagement risk: Effects on auditor choice and audit pricing

  2017, Advances in Accounting

  How auditors manage risks associated with auditing their clients is a topic of great interest to investors, regulators and academics (Johnstone, 2000, 2001). Studies show that auditors evaluate their own firms' business risks, and their clients' business and audit risks in client acceptance decisions (Ayers & Kaplan, 1998; Cohen & Hanno, 2000; Johnstone, 2000, 2001; Asare, Cohen, & Trompeter, 2005; Schroeder & Hogan, 2013).1 Once a client is accepted, audit firms actively manage their client portfolio risk, shedding risky clients when necessary (Johnstone & Bedard, 2003).

• Section 45 of the Auditing Profession Act: Blowing the whistle for audit quality?

  2014, British Accounting Review

  They should also avoid those engagements characterised by an unacceptably high level of risk (IAASB, 2009h). In this context, ISQC 1 requires the firm to develop client acceptance and continuance procedures designed to ensure that it is capable of executing engagements to the highest standards (see also Asare et al., 2005; Cohen et al., 2002; Farag & Elias, 2011). These principles are largely consistent with King-III, which encourages the use of formal codes of conduct and effective leadership which together cultivate a culture of ethical awareness (IOD, 2009).

• Consulting revenue sharing, auditor effort and independence, and the regulation of auditor compensation

  2012, Journal of Accounting and Public Policy

  The assumption underlying the SEC’s, 2003 regulation on auditor compensation is that large non-audit services impair auditor independence. Most of the studies on non-audit services focus on the validity of this assumption (e.g., Ascioglu et al., 2005; Asare et al., 2005; Ghosh et al., 2009).6 There has been no prior study that offers a rigorous analysis of the impact of the SEC regulation on auditor behavior.

• From clear to complicated: Buying and selling accounting services post Sarbanes-Oxley

  2010, Business Horizons

  With this single provision, audit firms’ abilities to offer their portfolio of services to a client are greatly reduced; a firm can still provide consulting services to a client when they are not also the auditor, but are restricted from handling both audit and consulting work for the same client. Although it is commonly accepted that non-audit services may impair auditor independence, empirical evidence actually suggests that when auditors provide such services they better understand a client's business and are better positioned to perform the audit (Asare, Cohen, & Trompeter, 2005). Certain tax services—such as tax avoidance planning strategies—are prohibited, but most other tax services are still permitted for audit clients with pre-approval from the client's audit committee, and remain a significant source of revenue for audit firms.

Recommended articles (6)