The Public Company Accounting Oversight Board requires audit reports to include information about auditor tenure for fiscal years ending on or after December 15, 2017. Tanyi, Rama, and Raghunandan (2021) examine the impact of this requirement on shareholder ratification voting of auditors. Consistent with shareholders sharing the oft-expressed views of legislators and regulators that long auditor tenure may impair auditor independence and audit quality, they find that shareholder opposition to auditor ratification increases (decreases) for long-tenured (short-tenured) auditors after the tenure disclosure. Thus, the very act of public disclosure in the audit report appears to have impacted investors’ voting decisions. The results suggest that auditors should respond to the increased scrutiny of auditor tenure by proactively engaging with the audit committees of their long-tenured clients. The results are also relevant in the context of the Securities and Exchange Commission’s efforts to have such tenure-disclosure requirements in registrants’ proxy statements.
In 2017, the Public Company Accounting Oversight Board (PCAOB) issued revisions to Auditing Standard 3101, The Auditor’s Report on an Audit of Financial Statements When the Auditor Expresses an Unqualified Opinion (Public Company Accounting Oversight Board (PCAOB) 2017). Among other changes, the revised standard requires audit reports for fiscal years ending on or after December 15, 2017, to include a statement about the year in which the audit firm began serving consecutively as the firm’s auditor, located just after the auditor’s signature.1
The PCAOB justified the need for audit reports to include information about auditor tenure by asserting that the new requirement would provide useful information for shareholders when deciding to ratify the auditor selected by the audit committee (Public Company Accounting Oversight Board (PCAOB) 2016). The PCAOB relied on results from Dao, Mishra, and Raghunandan (2008) to support this action.2 After citing the Dao et al. (2008) results, the PCAOB noted that making auditor-tenure data more readily accessible by requiring their inclusion in the audit report would be useful for shareholders in their decisions, including whether to ratify the auditor selected by the audit committee.3 In support of requiring the disclosure of auditor tenure, the PCAOB stated that:
Requiring the disclosure of auditor tenure in the auditor’s report would ensure that the disclosure is in a consistent location—the auditor’s report—for all companies and would reduce search costs for investors and other financial statement users who are interested in this piece of information. (PCAOB 2016, 48–49)
The requirement to make the length of the auditor-client relationship more accessible to shareholders was not without controversy. Each of the Big 4 audit firms opposed the new requirement, largely on the grounds that including auditor tenure in the audit report would unfairly highlight this information, when the link between auditor tenure and audit quality is tenuous at best. Two of the five PCAOB board members also opposed the auditor-tenure-disclosure requirement. Then-board member Franzel (2017) was skeptical of the auditor-tenure-disclosure requirement and called for academic research on the impact and usefulness of the auditor-tenure disclosure. A recent paper by Tanyi et al. (2021) answers the call from Franzel (2017) and addresses the following question: Did the association between auditor tenure and shareholder voting on auditor ratification change after the PCAOB required the audit report to disclose auditor tenure?
II. BACKGROUND AND RESEARCH QUESTION
Legislators, regulators, and shareholder activists have expressed concerns about long-term auditor-client relationships for many years (e.g., U.S. Senate 1976, 1977; Securities and Exchange Commission (SEC) 1994, 1999). The view that long auditor tenure may impair auditor independence and audit quality forms the basis of much of the support for requiring the disclosure of auditor tenure. For example, the American Federation of Labor (AFL)-Congress of Industrial Organizations (CIO) supported the requirement and commented that their proxy voting guidelines recommend considering voting against auditor ratification if the auditor’s tenure exceeds seven years. CalSTRS and CalPERS, the large institutional investors, also supported the requirement and commented that they would consider this information when voting on auditor ratification.
In their Appendix A, Tanyi et al. (2021) review prior studies examining auditor tenure. From this review, they conclude that most academic research does not indicate that longer auditor tenure is associated with lower audit quality. Consistent with this assessment, in commenting against the proposal, Deloitte & Touche LLP reviewed the academic literature and stated:
If audit quality problems do occur, they are less likely to occur later in the auditor’s term; the literature, however, does not support a demonstrable link between auditor tenure and audit quality or independence. (Deloitte & Touche 2016, 3)
PricewaterhouseCoopers LLP commented as follows:
Including auditor tenure in the audit report would create the false impression such a relationship exists and would give undue prominence to this information. (PricewaterhouseCoopers (PwC) 2016, 6)
Tanyi et al. (2021) provide the example of Chart Industries Inc. (GTLS) as anecdotal evidence of shareholders reacting to the disclosure of a long auditor tenure and of an audit committee responding to the results of the auditor-ratification vote. In the case of GTLS, 0.59 percent of shareholders voted against auditor ratification in the year prior to the auditor-tenure-disclosure requirement. After the tenure-disclosure requirement, GTLS’s audit report disclosed that the auditor had served since 1991; the votes against auditor ratification increased to 2.05 percent. Consistent with the evidence in Barua, Raghunandan, and Rama (2017) that even small increases in votes against auditor ratification can result in dismissal of the auditor, GTLS’s audit committee changed the auditor the next year.
Tanyi et al. (2021) provide two competing theories to explain a change in shareholders’ ratification voting in response to the disclosure of auditor tenure.4 Their first theory proposes that, if shareholder voting is more consistent with the oft-expressed views of legislators and regulators that long auditor tenure may impair auditor independence and audit quality, then they expect to find an increase in the association between shareholder dissatisfaction and auditor tenure after the mandatory disclosure of auditor tenure. On the other hand, their second theory puts forward that, if shareholder voting is more consistent with the academic literature showing that audit quality is typically lower in an auditor’s initial years, then they expect to find a decrease in the association between shareholder dissatisfaction and auditor tenure after the mandatory disclosure of auditor tenure. Tanyi et al. (2021) examine which of these two theories is supported by the data.
III. RESEARCH METHOD AND RESULTS
Tanyi et al. (2021) use data from 2,718 Securities and Exchange Commission (SEC) registrants that had shareholder voting for the same auditor in the fiscal years immediately before and after the PCAOB’s auditor-tenure-disclosure requirement was effective. Since the PCAOB’s requirement was effective for fiscal years ending on or after December 15, 2017, the first year with the auditor-tenure disclosure consists of fiscal-year ends from December 15, 2017, through December 14, 2018. The auditor-tenure disclosure in the audit report would be available to shareholders when the financial statements are filed, typically 30–90 days after fiscal-year end. Shareholders would have this information when later voting to ratify the auditor selected to audit the financial statements for the next fiscal year.5
In their first analysis, Tanyi et al. (2021) estimate a regression model to examine the effect of disclosing the auditor’s tenure on shareholders’ auditor-ratification voting, specifically the effect on shareholder dissatisfaction. Following prior research, they use two measures of shareholder dissatisfaction. The first measure includes shareholder votes against ratification and those who abstained from voting compared with the total votes for, against, and abstaining from auditor ratification.6 The second measure only includes votes against ratification compared with the total votes for and against auditor ratification.
Tanyi et al. (2021) include the following control variables that could potentially influence shareholders’ votes to ratify the auditor: client size, nonaudit fee ratio, auditor type (Big 4 and industry specialist), internal control weakness disclosure, financial restatement during the preceding year, and whether the client received a modified going-concern opinion from the auditor. Based on prior research, shareholder dissatisfaction is expected to increase with (1) higher nonaudit fees (measured as a proportion of total fees paid to the auditor), (2) disclosure of material weakness in internal control, (3) a financial restatement, or (4) a modified going-concern opinion from the auditor. On the other hand, shareholder dissatisfaction is expected to be lower for Big 4 auditors and industry specialists.7
The results of the first regression analysis show that, for a given tenure, a higher percentage of shareholders voted not to ratify the auditor after public disclosure of auditor tenure than before the public disclosure. These findings suggest that the public disclosure of auditor tenure influenced shareholders in their auditor-ratification voting. In terms of effect size, for the median level of auditor tenure (12 years), the tenure disclosure appears to increase the shareholder votes not to ratify the auditor by 10 percent of the median. As expected, they find that shareholder dissatisfaction increases with higher nonaudit fees, suggesting that, as nonaudit fees increase, shareholders are more likely not to ratify the auditor. They also find that shareholder dissatisfaction increases with an adverse internal-control report, a going-concern modified audit opinion, and client size, but decreases with the use of a Big 4 auditor.8
Next, Tanyi et al. (2021) estimate a similar regression model using the same control variables, but partition the auditors’ tenure into three subgroups: short, medium, and long. Tanyi et al. (2021) define short tenures as auditor tenures less than or equal to five years and long tenures as auditor tenures of 15 or more years. The medium tenure (between six and 14 years) is used as the benchmark for comparing the other two groups. The results show that the percentage of shareholders voting not to ratify a short-tenured auditor is lower after tenure information is disclosed. Conversely, the percentage of shareholders voting not to ratify a long-tenured auditor is higher after tenure information is disclosed. Regarding effect size, the percentage of shareholders voting not to ratify the auditor decreases by 15 percent for short-tenured auditors after the tenure disclosure. For long-tenured auditors, the percentage of shareholders voting not to ratify the auditor increases by 37 percent after the tenure disclosure. These results demonstrate that the disclosure of auditor tenure impacts short-tenured and long-tenured auditors contrarily, such that shareholder opposition to auditor ratification decreases (increases) for short-tenured (long-tenured) auditors after the tenure disclosure.
Tanyi et al. (2021) split their sample between Big 4 auditors and non-Big 4 auditors to separately analyze the effect of auditor-tenure disclosure on shareholder ratification voting for the two types of audit firms. They find that the positive association between auditor tenure and shareholder dissatisfaction with the auditor becomes stronger after the public disclosure of auditor tenure for both types of auditors.
As part of their additional analyses, Tanyi et al. (2021) also split their sample based on client size and based on institutional ownership.9 Their findings suggest that smaller clients and clients with lower institutional ownership primarily drive the strengthening of the positive association between auditor tenure and shareholder dissatisfaction after the public-disclosure requirement. Tanyi et al. (2021) posit that auditor-tenure disclosure might not impact large institutional investors as much as smaller retail investors because the former already had access to databases that contained auditor-tenure information. In contrast, smaller retail investors were likely to have greater difficulty obtaining information on auditor tenure before its disclosure was required.
Tanyi et al. (2021) perform three sensitivity analyses. First, they perform separate analyses for the three types of filers (i.e., nonaccelerated, accelerated, and large accelerated) and control for audit-report lag. Second, they consider that shareholder dissatisfaction with the auditor may be associated with shareholder dissatisfaction with the directors. Thus, they include the average proportion of shareholder votes against the election of directors as a control variable in their regression model. Third, they consider that the company’s stock-price performance could influence shareholder voting, so they include controls for the market performance of the company’s shares in their regression model. For all three sensitivity analyses, the overall inferences remain unchanged: public disclosure of auditor tenure influenced shareholders in their auditor-ratification voting.
IV. IMPLICATIONS FOR PRACTICE
The evidence in Tanyi et al. (2021) suggests that disclosure of auditor tenure in the audit report resulted in greater shareholder dissatisfaction with long-tenured auditors. Given the evidence from prior academic studies showing that audit quality is, in fact, lower in the initial years of an audit engagement, the evidence in Tanyi et al. (2021) provides empirical support for the concerns expressed by the Big 4 audit firms and others that highlighting auditor-tenure disclosure may lead to unwarranted focus by shareholders on auditor tenure. Thus, when regulators require prominent disclosure of a data item that was already available (albeit with some effort), the very act of such disclosure by itself could impact investors’ decisions.
Prior studies also suggest that audit committees react to even relatively low levels of shareholder dissatisfaction with the auditor when such dissatisfaction is higher than expected (Barua et al. 2017; Tanyi, Rama, Raghunandan, and Martin 2020). Thus, although the evidence in Tanyi et al. (2021) suggests that the mandated public disclosure of auditor tenure made shareholders more sensitive to long auditor tenures, an implication of this evidence taken together with the evidence in Barua et al. (2017) and Tanyi et al. (2020) is that auditors should also become more sensitive to long audit tenures.
Although auditors should make efforts to increase audit quality for all clients, the increased scrutiny of auditor tenure suggests that auditors may have more to do in that regard for clients where their tenure is longest. For example, auditors could increase internal inspections of the audits of their longest-tenured clients. Through their internal inspections, auditors could ensure they have an effective system of quality control that addresses the threats to independence and audit quality that arise due to longer auditor tenure. Some suggested procedures might include implementing a policy on the periodic rotation of key engagement-team members and the engagement-quality-control reviewer. Improving audit quality at long-tenured clients is not only important for individual audit firms, but also for the auditing profession as a whole. If scrutiny of auditor tenure were to increase further due to, for example, negative reports by regulators or a major accounting scandal at a client with a long auditor tenure, this would likely result in renewed calls for mandatory audit-firm rotation like there were prior to the enactment of the Sarbanes-Oxley Act.10 In other words, a further increase in the scrutiny of long auditor tenures could result in long auditor tenures becoming a thing of the past.
The results in Tanyi et al. (2021) are also relevant because the SEC is considering adding auditor-tenure disclosure to proxy statements (Securities and Exchange Commission (SEC) 2015). Adding auditor tenure to the proxy statement could affect shareholder ratification even more than disclosing it in the audit report because it is the proxy statement that notifies shareholders about the vote on the auditor; further, the proxy-statement filing occurs closer to the date of the vote. Drawing even more attention to auditor tenure in the proxy statements could result in even greater shareholder dissatisfaction with long-tenured auditors.
Example of Audit Report before Auditor-Tenure-Disclosure Requirement
Chart Industries, Inc. Report of Independent Registered Public Accounting Firm
Form 10-K filed February 23, 2017 for the fiscal year ended December 31, 2016
Example of Audit Report after Auditor-Tenure-Disclosure Requirement
Chart Industries, Inc.
Report of Independent Registered Public Accounting Firm
Form 10-K filed February 22, 2018 for the fiscal year ended December 31, 2017
Auditor tenure disclosure highlighted for emphasis.
See Appendix A for examples of audit reports before and after the revised standard became effective. The revised standard also requires the audit report to communicate critical audit matters.
Prior to the new requirement, some companies voluntarily disclosed their auditor’s tenure, but the location of this disclosure varied.
Tanyi et al. (2021) also recognize that it is possible that the mandatory disclosure of auditor tenure has no influence on shareholder ratification voting.
See Tanyi et al. (2021, Appendix B) for a timeline that provides a more detailed illustration and description of the various time periods relevant to the study.
Tanyi et al. (2021) state that the first measure is more appropriate for analyses of shareholder voting on auditor ratification per the detailed arguments provided by Barua et al. (2017) based on Delaware corporate law. Nonetheless, Tanyi et al. (2021) provide results using both measures. Inferences are substantively similar under both measures; therefore, we focus on the first measure when discussing the results.
Tanyi et al. (2021) do not have a prediction for client size, but based on prior research, they include it as a control variable in their model.
Tanyi et al. (2021) did not find significant results on the effect of financial restatements in the preceding year or the use of an industry specialist on shareholder dissatisfaction.
Tanyi et al. (2021) partition large and small clients based on the median total assets of observations in their sample, and they partition high and low institutional ownership based on the median institutional ownership of observations in their sample.
The study of mandatory audit-firm rotation by the Government Accountability Office (GAO) (2003) concluded that mandatory audit-firm rotation was inefficient.