The Myth of Auditor Independence
Waking Up to Unconscious Bias
|Friday, March 6, 2020|
By J. Edward Ketz, PhD for CPAJournal.com
The AICPA Code of Professional Conduct requires that members in public practice be objective, free of conflicts of interest, and independent in fact and appearance (section 300.050). The SEC likewise requires independence by the external auditors who perform an audit of management’s assertions in the registrant’s financial report.
Auditor independence has been debated for decades and is in the news again. In “The Way Audits Work is About to Change,” Michael Rapoport recently wrote that regulators have found about one-third of the investigated Big Four audits to be deficient (Quartz, Jan. 13, 2020, http://bit.ly/2S0k06R). Then, in “What Goes Wrong When Accounting Firms Become Consultants,” he raised the matter of consulting, noting that in 2019 Deloitte earned 22% of its revenue from auditing but 60% from consulting (Quartz, Jan. 14, 2020, http://bit.ly/37LX55y). Rapoport asked whether this trend makes auditors more beholden to management.
Rapoport also examined activity overseas in “How the U.K. Is Trying to Prevent Accounting Scandals” (Quartz, Jan. 13, 2020, http://bit.ly/2Rxp3N0). The United Kingdom is undertaking some soul searching about its audit profession, including independence issues, after several large corporate failures and criticisms of the public accounting partnerships that audited them. These examinations have led to significant proposals by the Competition and Markets Authority in April 2019, a review of the quality and effectiveness of audits in December 2019, and revised rules by the Financial Reporting Council in December 2019 to fortify auditor independence.
The SEC, however, relaxed independence rules last June, and Chairman Jay Clayton has proposed modernizing independence rules by further loosening them (Mark Maurer, “SEC Proposes Loosening of Auditor Independence Rules,” Wall Street Journal, Dec. 30, 2019, https://on.wsj.com/3aSsy86).
Francine McKenna criticized Clayton for his remarks in her “‘Modernizing’ Auditor Independence Rules” series at her blog The Dig. In Part 1, she argued that “modernizing” auditor independence rules is just “doublespeak for capitulating to the Big 4’s dominance” (Jan. 7, 2020, http://bit.ly/2Gy4cD0). She followed this claim by investigating many examples of audit failure and relating them to violations of the independence rules. In Part 2, she related the independence issues of PricewaterhouseCoopers, which are discussed below (Jan. 8, 2020, http://bit.ly/2uEkTde). In Part 3, McKenna speculated about a potential spate of future independence violations (Jan. 9, 2020, http://bit.ly/2U4G8j6).
Independence is often referred to as the bedrock of the profession. Do these shortcomings, however, demonstrate that this presumed solidity is a myth? Independence places a presumed relationship between a business and its external auditor. Given the institutional arrangement of the audit and the conflicts of interest, one should not be surprised by this inference.
Recent Examples of Independence Violations
There are many examples of auditors who were not independent. One example is PricewaterhouseCoopers’s noncompliance with this principle. In an administrative release dated September 23, 2019, the SEC documented 19 engagements for 15 SEC registrants in which the firm violated its auditor independence rules (http://bit.ly/37vqcKt). These activities involved designing and implementing multiple software projects and an internal audit co-sourcing engagement. The SEC also maintained that PricewaterhouseCoopers employees wrongly characterized these consulting assignments as audit work. This marks another instance of misconduct in consulting work, despite the constant reassurance that consulting has no influence on auditor independence.
Another example was in 2016, when the SEC brought charges against Gregory Bednar and Ernst & Young because of independence issues (http://bit.ly/2tZ8dxl). Bednar had too close of a relationship with the CFO of an unidentified firm; they and their families attended professional football and hockey games together, played golf together, and vacationed together. Ernst & Young, through Bednar, incurred at least $109,000 in entertainment costs on this CFO. The SEC felt the relationship so close that it revoked Ernst & Young’s independence for the 2012, 2013, and 2014 audits. The SEC cited the firm as well because it had evidence of this lavish spending by Bednar and did not act on it.
Also in 2016, the SEC instituted proceedings against Robert Brehl, Pamela Hartford, Michael Kamienski, and Ernst & Young for lack of independence (http://bit.ly/38SyGM9). Specifically, Brehl was Chief Accounting Officer for an unspecified company at the same time as he had a romantic relationship with Hartford, who served as the engagement partner on the audit. Kamienski was the coordinating partner on the engagement. When he suspected the romantic relationship, Kamienski did nothing; however, Brehl’s supervisors learned about the affair and informed Ernst & Young. Because of this entanglement, the SEC said that Ernst & Young was not independent during 2012 and 2013.
In 2019, the SEC cited RSM US LLP for violating independence rules when it provided nonaudit services to various firms; in addition, a partner had a prohibited employment relationship with a client (http://bit.ly/2GxlsrW). The PCAOB censured Marcum LLP for promoting some of its audit clients to potential investors (http://bit.ly/37LbUWp), and it sanctioned KPMG Bermuda for re-executing independence confirmations, backdating them, and not telling PCAOB inspectors that they were not the originals (http://bit.ly/2GwhsrO).
There have been many commissions that have examined independence and worked to strengthen it, including the Kirk Advisory Panel on Auditor Independence and the O’Malley Panel on Audit Effectiveness.
More violations could be named, but this sample should be sufficient to declare that the profession has a problem. Either practitioners have biases that they do not recognize, which requires institutional changes; or they are unaware of the independence rules, which raises the issue of professional competence; or they are breaking the independence rules, which damages the perception of the profession’s integrity. While competence and integrity are at the heart of some violations, most problems arise because of biases.
Expressed Concerns about Independence
Concerns about independence are not new. Harvey Hendriksen wrote that when he was at the SEC, he noticed that auditors called meetings with the chief accountant when they were unable or unwilling to exercise professional judgment (“Relevant Financial Reporting Questions Not Asked by the Accounting Profession,” Critical Perspectives on Accounting, October 1998, http://bit.ly/36BSi5k). At a minimum, the accountants lacked the power or the will to confront management. Hendriksen cited observations in 1936 by then–SEC Chairman James M. Landis that auditors displayed their loyalties to management rather than to investors. In 1994, Chief Accountant Walter Schuetze criticized auditors’ cheerleading for their clients, inferring a lack of independence; in 1997, Chief Accountant Michael Sutton echoed these thoughts as he asked the profession’s leaders to address independence issues. In addition, there have been many commissions that have examined independence and worked to strengthen it, including the Kirk Advisory Panel on Auditor Independence and the O’Malley Panel on Audit Effectiveness. None has succeeded.
Steve Zeff has noted the historical changes in auditing. “By 1980,” he concluded, “a deterioration in professional values appears to have set in” (“How the U.S. Accounting Profession Got Where It Is Today,” Accounting Horizons, December 2003, http://bit.ly/2O2YHjG). With the growth of consulting came a concomitant decrease in the space separating auditor from client, as auditors felt pressure from their clients to advance their causes. Art Wyatt, a former FASB member, noted in a 2003 speech to the American Accounting Association that auditors no longer presented their independent assessments to the board but argued for accounting treatments desired by their clients (http://bit.ly/3aJ3TD0). Furthermore, the firms engaged in various consulting activities that endangered their independence.
Accountants who perform external audits typically counter these criticisms by stating that they comply with the requirements of the SEC and the profession, meticulously avoid loans with and stock ownership in client firms, and avoid personal relationships that are too close. Furthermore, they argue that consulting has no influence on their audit work, and any transgressions are sporadic and generally inconsequential. One wonders how many violations must occur before some practitioners admit that they are regular and consequential.
How can one reconcile the examples of independence shortfalls and the criticisms by Landis, Schuetze, Sutton, Zeff, and Wyatt with the rebuttals put forth by external auditors? The answer lies in the distinction between conscious and unconscious bias. While some CPAs deliberately decide to trample the rules, as did Friehling & Horowitz, the auditor for Madoff Investment Securities, most auditors seem careful to avoid conscious corruption or even its appearance. The profession remains quite susceptible, however, to unconscious bias.
Patronage and Conflicts of Interest
The sociology of professions undertakes a critical examination of professions like accounting to analyze their historical development and their place in modern societies. Terence Johnson, for example, is a sociologist who has theorized various professional arrangements and their power relationships. He calls one of these arrays “the patronage system,” in which a corporation recruits professionals to perform various services; accountancy is identified as the classic example. The patron chooses the professional, and the patron possesses most of the power, as it controls far greater resources and has the power to hire and to fire. The preferred auditor must not only have the desired technical competence, but also, and perhaps more importantly, values that are socially acceptable to the patron. This social acceptability is important if the professional desires to be hired and to sustain the relationship.
This uneven power relationship and the necessity for the professional to be acceptable to the patron lead to conflicts of interest. The Oxford English Dictionary defines a conflict of interest as an “incompatibility between the concerns or aims of different parties.” The incompatibility is evident inasmuch as the external auditor serves the public interest, especially investors and creditors, while simultaneously being bound to corporate management. Calling the corporation the client when in fact shareholders, creditors, and the investing public are the true clients exemplifies the inherent conflict of interest.
Consider the following example: a faculty member at Penn State teaches advanced accounting and has a reputation for giving low grades. A wealthy student in the class asks him for tutoring, offering $200 per hour. The faculty member, who has impeccable ethical standards, accepts the offer and tutors the student 50 hours during the semester, at the end of which the student receives an A. When asked about the arrangement, the professor displays the student’s exams to administrators, who observe high marks on each exam. Is there a conflict of interest? Of course there is! The professor has a responsibility to the school, the profession, and the other students to grade all exams equally, while at the same time he has the aim of tutoring well the student he coaches, especially if he would like similar engagements in the future. At a minimum, the situation is rife with unconscious bias.
There is an inherent conflict of interest in the auditor-client relationship because the auditor naturally moves toward agreement with the client’s views in an effort to strengthen and maintain the relationship.
In their article “Why Good Accountants Do Bad Audits,” Max H. Bazerman, George Lowenstein, and Don A. Moore point out that the real trouble is the profession’s “vulnerability to unconscious bias” (Harvard Business Review, November 2002, http://bit.ly/311XDSi). External auditors seem unaware of the underlying nexus between client and auditor and do not address these challenges to independence. The engagement fee may be characterized as a periodic cash flow in an annuity stream; the value to the audit firm is the present value of this perpetuity. Partners are evaluated positively on the business they keep and the business they bring in, but evaluated negatively on lost clients. In such a setting, there is pressure—even if only subconsciously—for the partners to accept a little misbehavior here or there to keep or to win the business.
In another article, “The Impossibility of Auditor Independence,” Bazerman, Loewenstein, and Kimberly P. Morgan review the psychology literature, which shows that bias enters the judgment process unconsciously and unknowingly (MIT Sloan Management Review, January 1997, http://bit.ly/2U5PZ8m). Human subjects seem to choose the outcome they prefer and backfill the thought process to justify the conclusion. They allow preferences to drive decisions, a process of which they are not consciously aware. As the article says, “Despite the auditors’ best efforts to place the external users’ interests above the client’s and to maintain objectivity, they may be unable to overcome cognitive or psychological biases that make them arrive at marginal decisions in the client’s favor.” Thus, there is an inherent conflict of interest in the auditor-client relationship; this is true because the auditor naturally, if subconsciously, moves toward agreement with the client’s views in an effort to strengthen and maintain the relationship.
Proposals to Deal with the Independence Problem
What might be done to change incentives or institutions to reduce independence violations? There are at least six possible recommendations. Details about them may be found in Bazerman et al. (2002), as well as Eleanor Bloxham’s “Should Companies Eliminate Audits?” (Fortune, June 18, 2015, http://bit.ly/2O4F0bc), this author’s Hidden Financial Risk: Understanding Off–Balance Sheet Auditing (Wiley, 2003), and Joshua Ronen’s “Corporate Audits and How to Fix Them,” (Journal of Economic Perspectives, Spring 2010, http://bit.ly/3aSzguQ).
The first proposal is to mandate auditor rotation, for example every four years. The point is to reduce the coziness of the relationship, thereby reducing the power of management. Audit rotations do, however, have the drawback of a installing a learning curve upon each rotation.
A second suggestion is to prohibit all consulting assignments, transforming CPA firms into accounting-only firms. Here too, the impetus is to weaken the social bonds between auditor and client and transfer some of the power to auditors so they can oppose managers when necessary. Prohibition of consulting reduces a large number of the sources of unconscious bias so it should be relatively effective. Recommendation 3 in the 2019 Competition and Markets Authority (CMA) market study report recommends “an operational split between the audit and non-audit practices of the Big Four,” which accomplishes the same end (http://bit.ly/2tXylc4). Furthermore, the CMA suggests that profit sharing between the audit business and the nonaudit business should be eliminated.
A third possibility is to increase auditors’ liability when there is an audit failure, perhaps by repealing the Private Securities Litigation Reform Act of 1995 or by invoking a doctrine of strict liability to auditors. The rationale is to increase the losses when an audit failure occurs. This proposal indirectly addresses the independence issue by adding disincentives for bad behavior; however, it might not change the business model very much. Instead, bankruptcy might become a more likely scenario in case of severe audit problems. A variation of this proposal is to criminalize egregious violations of independence.
Recently, a fourth option has surfaced—to use technology such as blockchain or AI to let machines determine the audit opinion. Such a belief shows incredible faith that technology can embrace every aspect of what audit partners do. The aggregation of audit evidence into an opinion about the financial report, including all necessary disclosures, is far too complex for today’s technology. Worse, it only seems possible in a relatively static world. As new issues, new regulations, new financial instruments, and new theories of finance continually arise, blockchain and AI will be challenged to keep up.
When the securities acts were initially proposed, some argued for government auditors, and a few have wondered whether that is viable today. This fifth proposal is problematic, as it raises its own questions of auditor independence in fact and appearance. The audit process could be come subject to political pressure as well as widespread accusations of political bias, whether real or imagined. A variation of this proposal is for some government agency, such as the SEC, to choose auditors, but this would be similarly problematic.
The last proposal to consider is Joshua Ronen’s conception of insurance for financial accounting distortions (“Post-Enron Reform: Financial Statement Insurance, and GAAP Revisited,” Stanford Journal of Law, Business, and Finance, 2002-2003, http://bit.ly/36FN60E; Ronen and Julius Cherny, “A Prognosis for Restructuring the Market for Audit Services,” The CPA Journal, May 2003, http://bit.ly/2u0gQI1). Ronen and Cherny’s system would work like this: Companies would be required to buy policies that would insure against financial statement misrepresentations. A company would contact a financial statement insurance carrier. It would negotiate the coverage the carrier would be willing to provide, the premium, and the cost and scope of the audit. The insurance carrier would send an agent, most likely the same auditor it will employ, to evaluate its insurance risk. Once a deal was struck, the carrier would have its agent conduct the audit. If the opinion were clean, then coverage would commence.
From Myth to Reality
Much criticism has been leveled against the audit profession over the last several decades; as demonstrated above, it has intensified in recent years. This crescendo of grievances is not surprising, given the increase in accounting scandals, the sizes of those scandals, and the transformation of the audit profession into a consulting profession.
This is not just an optics problem. The patronage problem, usually with other labels, has been criticized. Some professionals have argued that everything was fine as long as firms and partners were independent in fact, but the patronage system induces cognitive biases, and these cognitive biases contraindicate independence in fact.
The merits of the six proposals above depend upon one’s view of the problem. Financial statement insurance would combat unconscious bias due to the patronage system because it changes the power structure sufficiently to solve the problems. Partial solutions include auditor rotation and a consulting ban. Increased civil penalties or the introduction of criminal penalties may be needed to combat incompetence or integrity issues.
This author views the problem of independence violations mostly as one of unconscious biases arising from a patronage system that rewards cozy relationships between auditors and clients and discourages strict adherence to GAAP. This patronage system needs to go away—the sooner, the better.