The Effects of Manager-Auditor Affiliation and PCAOB Inspection Reports on Audit Committee Members' Auditor Selection

Audit quality (AQ) is critical to capital markets, ensuring investor confidence in the financial reports of public companies for which resource allocation decisions are partly based. While there is no one universally accepted definition of audit quality (Knechel, Krishnan, Pevzner, Shefchik, and Velury 2013), most would agree that it requires auditors to comply with auditing standards, conduct the audit using a reasonable level of professional skepticism and judgment, and be objective and unbiased in their reporting. Abbott, Brown, and Higgs's (2016) paper—the basis of this summary—is based on DeAngelo's (1981) two-pronged definition of audit quality. DeAngelo (1981) defines audit quality as the joint probability that an auditor will detect (an indication of competence) and report (a reflection of independence) material misstatements.

A multitude of high-profile financial scandals in the 1990s and early 2000s generated distrust in the financial reporting process and many questioned auditors' ability to fulfill their watchdog function. Stakeholders demanded more transparency of the financial reporting process and better oversight of auditors, management, corporate boards, audit committee members (ACM), and standard setters.

The Sarbanes-Oxley Act of 2002 (SOX, U.S. House of Representatives 2002) was enacted to mitigate these concerns and to renew stakeholders' trust in financial reporting. SOX established the Public Company Accounting Oversight Board (PCAOB) to oversee public company auditors. Among other things, the PCAOB has the authority to register, inspect, and discipline auditors, as well as establish auditing standards and rules. The results of the PCAOB inspections are summarized in publicly available inspection reports. The PCAOB inspection process, including the public release of audit firms' audit engagement deficiencies (AED) and quality control defects (QCD), is a way of addressing auditor competence concerns—i.e., an auditor's ability to detect a material misstatement (Christensen, Glover, Omer, and Shelley 2016).

A major complaint pre-SOX was that auditors often appeared to be too closely aligned with management (e.g., belonging to the same private social clubs, long tenure, performing audit and non-audit services for the same client), thereby threatening their independence. SOX has several provisions to help mitigate concerns over perceived impairment of auditor independence. These include (1) audit committees (as opposed to management) are responsible for hiring, firing, compensating, and monitoring auditors; (2) auditors are required to wait 12 months (i.e., “cool off”) before accepting a key financial reporting position with their audit client; (3) audit partners are required to rotate off an engagement every five years; and (4) auditors are restricted on the types of non-audit services that they can provide for their audit clients (SOX, U.S. House of Representatives 2002).

Abbott et al. (2016) conducted a study to investigate whether ACM consider auditor competence, as measured by a PCAOB-like inspection report, and perceived auditor independence, as measured by the auditor's prior affiliation with management, when selecting the auditor, in a context in which management recommends the auditor. In that study, the authors are particularly interested in determining whether the mandatory one-year cooling-off period is sufficient to mitigate ACM independence concerns resulting from the auditor's prior affiliation with management. In this paper, I discuss the Abbott et al. (2016) hypotheses development, experimental design, primary findings, and the main implications of the results for current and prospective ACM, audit firms, company management, and regulators/policymakers.

Pursuant to SOX, the audit committee (AC) is required to assess the competence of the external auditor as part of its annual auditor selection/retention process. The PCAOB inspects public company auditors and provides an independent evaluation of the auditor's execution of generally accepted audit procedures.¹ Inspections encompass reviewing specific audit engagements and evaluating the audit work/procedures performed on them, and reviewing an audit firm's quality control system. The PCAOB summarizes and publicly publishes any engagement-specific deficiencies and quality control defects/criticisms in the PCAOB inspection report.² These detailed reports are readily accessible on the PCAOB website, making it a potential information source for audit committee members in assessing an auditor's competence in their audit selection process.

While the PCAOB encourages audit committee members to read the inspection report and discuss its findings with their auditor (PCAOB 2012a), archival research is unclear on whether ACM use the reports in the auditor selection/retention decision (Lennox and Pittman 2010; Abbott, Gunny, and Zhang 2013). The mixed results can be partially explained by auditor brand name (e.g., Big 4 auditor), costs to switch auditors, and auditors' attempts to de-emphasize inspection findings (Buslepp and Victoravich 2014; PCAOB 2012a). Additionally, some ACM may lack financial expertise and may not understand the implications of the inspection findings (Cohen, Krishnamoorthy, and Wright 2010). Also, a weak audit committee may simply defer to management's auditor recommendation. Furthermore, the PCAOB inspection is performed on a risk-based sample basis (Keyser 2015). The generalizability of any deficiencies is unclear as a result. Finally, auditors argue that the inspection findings are not representative of a deficient audit, as they are often merely documentation issues or differences of professional judgments, among a range of reasonable judgments, between the auditor and the inspection staff (PCAOB 2011a, 2011b).

In the Abbott et al. (2016) experimental study, the authors propose ACM will use the inspection reports in evaluating the auditor's competence and recommend an audit firm with no deficiencies versus a firm with many and severe deficiencies.

ACM are also required to assess the external auditor's independence (perceived and actual) as part of its annual auditor selection/retention process. Auditor independence was a major concern in the pre-SOX era, exacerbated with reports of audit firm alumni holding key positions in some of the fallen companies (such as Enron and Global Crossing). Concern over this revolving-door phenomenon, when audit clients employ former auditors—usually managers and partners—in key financial positions, motivated regulators to enact a one-year “cooling-off” period before publicly held companies may hire their auditor's former employees or owners for key financial reporting positions (SOX Section 206, U.S. House of Representatives 2002). Academic research generally reports a negative association on the effects of the revolving-door phenomenon on financial reporting quality (Dowdell and Krishnan 2004; Menon and Williams 2004). Research also finds ACM are aware of the negative effects on auditor independence and tend to discourage the practice in an effort to enhance auditor independence in appearance (Lennox and Park 2007). Accordingly, Abbott et al. (2016) posit that ACM are more likely to select a triennial auditor with no prior affiliation with management than an auditor with a prior manager affiliation (using a 13-month cooling-off period).³

Extant archival studies in audit quality generally uses auditor brand name as a measure of audit quality, which incorporates both auditor attributes: competence and independence (Knechel et al. 2013). These studies are unable to separately test the two distinct auditor qualities. On the other hand, behavioral audit quality research generally tests either auditor competence (Brazel and Agoglia 2007) or auditor independence (Tepalagul and Lin 2015), while holding the other constant. Note, per DeAngelo's (1981) definition of AQ, the auditor must first detect the misstatement (measure of competence) before he/she can report the misstatement (measure of independence). In keeping with this definition of AQ, Abbott et al. (2016) posit ACM will be less likely to select a triennial auditor with a favorable inspection report in the presence of a manager-auditor affiliation. In other words, the auditor's independence only matters if the auditor is deemed competent.

Abbott et al. (2016) conducted a controlled experiment to test their expectations that ACM will be more likely to hire an auditor with either a favorable PCAOB-like inspection report or no prior affiliation with management, but will be less likely to hire an auditor who is affiliated with management even if the auditor's inspection results are favorable. Abbott et al. (2016) engaged 118 financially literate professionals as proxies for ACM. All participants met the SEC criteria to sit on a public board; all participants had at least one year of supervisory experience and many had board or audit committee experience.

The experiment was conducted online using Qualtrics® research software. Participants assumed the role as one of three ACM of a small hypothetical public company seeking to engage a new auditor to conduct the company's integrated audit.⁴ They were told that management had evaluated all prospective regional CPA firms that responded to the company's Request for Proposal (RFP) and recommended Firm A (a triennially inspected audit firm) to the audit committee. Participants were asked to indicate the likelihood they would select Firm A to conduct the company's upcoming integrated audit. As participants entered the experiment they were randomly assigned to one of four conditions; information provided in all conditions was the same except for the manipulated variables.⁵

Auditor competence and auditor independence were manipulated in the experiment. In the auditor competence manipulation, participants received one of two versions (favorable/unfavorable) of a PCAOB-like inspection report for Firm A, issued by a quasi-government organization similar to the PCAOB.⁶ The favorable report did not contain any deficiencies or defects (i.e., a clean report); the unfavorable report had multiple and severe audit engagement deficiencies (AED) and quality control defects (QCD).⁷

To manipulate perceived auditor independence, participants were informed that the company's new Chief Financial Officer (CFO) and Director of Internal Audit (DIA) were either (1) previously employed by CPA Firm A 13 months ago (affiliation present condition), or (2) never previously worked for Firm A (affiliation absent).⁸ The two manipulated variables yield four experimental conditions: (1) Affiliation Absent-Favorable Inspection Report (AAFV); (2) Affiliation Absent-Unfavorable Inspection Report (AAUF); (3) Affiliation Present-Favorable Inspection Report (APFV); and (4) Affiliation Present-Unfavorable Inspection Report (APUF).

Abbott et al. (2016) captured participants' responses for the primary dependent variable—the likelihood they would recommend Firm A to conduct the company's audit—on a seven-point scale with endpoints labeled 1 = extremely unlikely to 7 = extremely likely. Participants were also asked to assess their level of confidence in CPA Firm A. As predicted, ACM consider auditor competence and independence in their auditor selection decisions. As illustrated in Figure 1—duplicated here for convenience from Abbott et al. (2016, 10, Figure 2)—Firm A is most likely to be selected if they have a favorable inspection report and no prior affiliation with management (mean = 5.67) and least likely to be selected if they have an unfavorable inspection report and are perceived to be affiliated with management (mean = 2.48). In general, the main effects show auditors with clean inspection reports (management-auditor affiliation absent) are favored over auditors with unfavorable reports (management-auditor affiliation present). The interaction results show the affiliation effect holds only when the auditor is deemed competent—that is, when the inspection report is favorable. Interestingly, the results suggest the mandatory 12-month cooling-off period appears to be insufficient to mitigate perceived auditor independence concerns.

The results of Abbott et al. (2016) provide insight into some of the factors ACM consider in their auditor selection decision—in a scenario where management recommends the auditor. SOX increased audit committees' responsibilities and tasked them with hiring, firing, compensating, and monitoring auditors of public companies. Yet, stakeholders are still skeptical as to whether management is still “pulling the strings” and ACM are merely acting as figureheads (Cohen et al. 2010). The results of the study imply ACM use the PCAOB inspection report to measure auditors' competence and consider the auditors' affiliation with management as a measure of impaired independence. Given the efforts of the PCAOB in educating ACM on its auditor oversight role, and the reputation and litigation incentives imposed by SOX, it is not surprising to find that the PCAOB inspection report is considered in the audit procurement process even when management recommends an auditor.

The bigger contribution of the Abbott et al. (2016) study is the demonstration that the mandatory cooling-off period does not appear to be sufficient to create an appearance of independence. One of the most interesting aspects from the study is the variation of participants' perceptions on the length of time clients should wait before hiring their former auditors for key financial positions. Eleven percent of participants thought no cooling-off period was necessary, while others (19 percent) thought clients should never hire their former auditors for these positions. Twenty-two percent of participants thought the one-year cooling-off period was sufficient, but the majority (79 or 67 percent) felt it was not enough. As the pool of participants were selected from the financially literate general public, participants' perceptions are most likely reflective of the perceptions of other stakeholders.

Therefore, Abbott et al.'s (2016) study has several implications for practitioners.⁹ First, it may benefit audit committee members who might need to more fully understand the auditor hiring process to be more effective at their job. The ACMs' express oversight responsibility on auditors imposed by SOX is considered by Chairman Doty to be SOX's second most significant reform, next to the establishment of the PCAOB (PCAOB 2015). The Blue Ribbon Committee (BRC 1999, 37) states, “the proper functioning of an audit committee relies … specifically on the audit committee members' attitude toward their own role. If an audit committee is determined to be diligent in its oversight role, a sure sense of appropriate action will follow.” The study may also remind ACM that they cannot blindly accept management's recommendation and must consider how that recommendation would be perceived by financial statement users and other stakeholders. Despite the increased focus on corporate governance and audit committee effectiveness pursuant to SOX, academic research suggests that managers, not audit committees, may have a primary role in appointing auditors, suggesting a possible lack of diligence on the part of directors (Cohen, Krishnamoorthy, and Wright 2008; Cohen et al. 2010; Beasley, Carcello, Hermanson, and Neal 2009).

Second, the evidence provided by Abbott et al. (2016) may help auditors who are bidding on jobs better understand the importance of the inspection process and cooling-off period for hiring. Some auditors devalue the PCAOB inspection findings as trivial, claiming they are primarily documentation issues, mere differences of professional judgment between auditors and the inspection staff (PCAOB 2011b), and not representative of the average audit engagement, as the inspection is risk based. As such, the PCAOB is concerned that auditors are dismissing the results of the inspection reports when selling their services to clients, and urges audit committee members to become more acquainted with the PCAOB inspection process and discuss the report findings with their auditor (Buslepp and Victoravich 2014; PCAOB 2012a). Auditors should be aware that financially literate ACM might react negatively to these kinds of denigrating statements. Additionally, effective ACM will also look for perceived impairment of audit independence in assessing audit quality.

Third, the study could be useful to CFOs who might serve on the selection committee, as it demonstrates how financially literate people view PCAOB inspection reports and the mandated cooling-off period. If effective ACM view the inspection report as a useful source in assessing auditors' competence and are concerned over stakeholders' perception of auditors' independence, they could view management negatively if management attempts to persuade the ACM to hire a “friendly” auditor or an auditor that is not competent. Regulators and standard setters have enhanced auditing standards (e.g., Auditing Standard No. 16) to improve communication between auditors and the audit committee, effectively reducing the influence of management on auditor judgments (PCAOB 2012b, 2012c).

Fourth, regulators/policymakers who want to assess the value of the inspection process will also find the study useful. In the PCAOB's 2014 annual report, Chairman Doty stated, “I believe our work demonstrates that the PCAOB continues to be the essential oversight body that Congress envisioned” (PCAOB 2015, 3). The PCAOB has over 800 employees covering over 1,300 domestic registered firms in the U.S. and almost 1,000 accounting firms in 90 jurisdictions outside the U.S. The inspection program accounts for almost half (48.5 percent or $125.1 million) of the PCAOB's 2016 total budgeted outlay ($257.7 million) (PCAOB 2015, 2016). Consequently, validating the usefulness of inspection reports is important in justifying the regulatory cost of one of the PCAOB's primary functions. This study demonstrates the PCAOB's efforts are working to enlighten the public's awareness of the PCAOB inspection report. It also provides evidence on the effectiveness of the mandatory cooling-off period.

Fifth, policymakers who may wish to rethink the cooling-off period or require additional disclosures about former auditor relationships will also benefit from this study. While the one-year cooling-off period rule does not differentiate among an auditor's rank, some academic research finds the time between leaving the audit firm and being employed by the former client, as well as the rank of the auditor (e.g., partner or staff), influences users' perception of auditor independence (Dowdell and Krishnan 2004; Menon and Williams 2004). Other studies find the hiring of ex-employees of a company's external auditors does not impair auditor independence, in appearance (Geiger, North, and O'Connell 2005; Daugherty and Dickens 2010). Policymakers must find the optimum balance between too short and too long of a cooling-off period. If the period is too long, the accounting profession may lose its attractiveness as the auditors' relevant client knowledge fades. Further, firms may not be able to readily hire qualified industry specialists if they are prohibited from being hired indefinitely—reducing the pool of qualified applicants and eliminating a viable source of income (a second career) for audit partners with younger retirement (e.g., Big 4 firms have mandatory retirement for partners at age 60 or below).

Post-SOX research indicates that not all of the reforms have worked as hoped. For example, Daugherty, Dickins, Hatfield, and Higgs (2012) show that mandatory partner rotation, while improving independence in appearance, may do so at the cost of auditor competence. Additionally, research on whether non-audit prohibitions are working document mixed results (Schneider, Church, and Ely 2006). More recently, Church and Zhang (2011) report users' perception of auditors' provision of non-audit services differs across decision contexts and is interpreted opportunistically. Abbott et al. (2016) contribute to the post-SOX literature stream by providing insight on two other SOX reforms. The authors report (1) the inspection program (and thus the PCAOB) is valued and appears to be an effective source for informing audit committees about audit quality, and (2) the results indicate that the one-year cooling-off period may not be sufficient to alleviate stakeholder concerns about independence in appearance.

While the implications of Abbott et al.'s (2016) study are insightful, the study is subject to limitations. First, as an experimental study, the question of external validity exists. However, behavioral research provides a greater advantage over archival research in that it can control for factors (such as the interactions between auditors, management, and audit committee members) that are not easily accomplished with archival-designed studies. Second, while all participants met the basic requirements of an ACM, only 33 percent (48 percent) reported they had audit committee (board) experience. Thus, the results may not be generalizable. Additional analyses revealed no significant differences across groups in the study. Third, the focus of the study is on triennially inspected auditors, limiting the generalizability of the results to larger audit firms and clients. Fourth, the study sets good (favorable) and bad (unfavorable) inspection reports at extremes, making the differences obvious to participants. Finally, as the study used a hypothetical regulator instead of the PCAOB brand name, it is possible that participants did not view the inspection reports as equivalent to those of “actual” PCAOB inspection reports.