Do Shareholder Ratification Votes Impact Subsequent Audit Outcomes? Free

The Securities and Exchange Commission (SEC) and the Public Company Accounting Oversight Board (PCAOB) have justified many disclosure-related rules and regulations using the argument that the data provided in such disclosures would be useful for shareholders in their decisions to ratify the auditor selected by the audit committee. For example, in Financial Reporting Release (FRR) No. 56, the Securities and Exchange Commission (SEC) (2000) justified rules about the disclosure of audit and non audit fees as follows:

Later, the SEC (2003) revised the rules related to such disclosures and noted:

In 2017, the Public Company Accounting Oversight Board (PCAOB) (2017) issued a new rule that changed the format of the audit report. This rule required disclosure of critical audit matters (CAMs) and disclosure about the length of the auditor–client relationship. The PCAOB (2017) noted as follows:

Thus, a common theme running through many SEC and PCAOB actions related to auditor-related disclosures is that such data will be useful to investors in their auditor ratification voting decisions. The assumption in such SEC and PCAOB statements is that such voting can, in turn, influence auditor actions and lead to enhanced audit quality.

Activists, regulators, and others have asserted this assumption between shareholder voting on auditor ratification and audit quality more explicitly elsewhere. For example, the U.S. Department of the Treasury (DoT) formed an Advisory Committee on the Auditing Profession to examine “the condition and future of the auditing profession, with emphasis on the sustainability of a strong and vibrant profession” (U.S. Department of the Treasury (DoT) 2008, II: 1). One of the recommendations of the Advisory Committee on the Auditing Profession (ACAP) was that the SEC should take action to require all public companies to have shareholder ratification of the auditor (DoT 2008). Specifically, Recommendation VIII-5 of the ACAP was adopt annual shareholder ratification of public company auditors by all public companies. The ACAP’s justification was that “ratification allows shareholders to voice a view on the audit committee’s work” and that such voting “can enhance the audit committee’s oversight” (DoT 2008, VIII: 20).

Although the SEC and the PCAOB have acted upon many ACAP recommendations, the SEC has not acted on the shareholder voting recommendation of the ACAP. To the contrary, SEC staff have often issued “no action letters” to companies wanting to exclude activists’ proposals related to greater say for shareholders in auditor selection.¹ Such actions by SEC staff led to public dissent by SEC Commissioner Aguilar (2012, 3), who noted:

Yet there is little empirical evidence about the association between auditor ratification voting and subsequent audit outcomes. Tanyi et al. (2020) shed light on this issue by addressing the following question: are levels of shareholder dissatisfaction associated with subsequent audit outcomes, including audit quality or audit fees? Their study provides empirical evidence about a question that has hitherto been unexamined yet asserted as fact by the SEC and PCAOB to justify disclosure rules.

Corporate governance theory suggests that shareholders elect a board of directors who then monitor management and hire an independent auditor on behalf of shareholders. Section 301 of Sarbanes-Oxley Act of 2002 (U.S. House of Representatives 2002) vested the authority to hire the auditor with the audit committee. However, managers retain considerable influence over auditor hiring (Mayhew and Pike 2004; Dao, Raghunandan, and Rama 2012). Hence, some regulators and governance activists have highlighted the need for shareholders to have a say in auditor hiring (e.g., CalSTRS 2008; Aguilar 2012).

U.S. Federal or state securities laws do not require shareholder ratification of the auditor. Hence, companies voluntarily providing shareholders the opportunity to vote on auditor ratification signals a commitment to enhanced corporate governance. This changed in 2009, when the SEC granted permission for the New York Stock Exchange (NYSE) to change listing Rule 452. The SEC rule change increased the incentive for NYSE-listed companies to include shareholder voting just for the purposes of achieving a quorum.² One argument is that, as shareholder voting becomes more widespread, shareholders, auditors, and audit committees would become more sensitive to such voting. Conversely, if all companies are including such voting, then the signaling element of such a vote is lower. Nevertheless, it is noteworthy that, even after 2009, the SEC and PCAOB have continued to use shareholder voting on auditor ratification as a justification for requiring audit-related disclosures.

An earlier study by Dao et al. (2012) tested the ACAP’s assertion that audit quality will be higher at companies that provide shareholders the opportunity to vote on auditor ratification. This study examined a basic question: are there differences in audit quality at companies that provide shareholders the opportunity to vote on auditor ratification compared with companies that do not have such voting? Dao et al. (2012) use data from 1,382 firms for the year ending December 31, 2006 and find that audit fees are higher in firms with shareholder voting on auditor ratification. In addition, after controlling for other factors, firms that started having a shareholder vote pay higher fees than firms that stopped having a shareholder vote. More interestingly, Dao et al. (2012) find that, in firms with shareholder voting on auditor selection, (1) subsequent restatements are less likely and (2) abnormal accruals are lower.³

The results from Dao et al. (2012) do not address a more important question that arises based on the assertions of the SEC, PCAOB, and others. Can the results from shareholder votes on auditor ratification influence auditors’ subsequent actions?

Auditor ratification votes routinely elicit about 98 percent support. So why should small changes in such votes matter? Quoting senior audit partners and an ACAP member, Dao et al. (2012,174) note that “everyone expects the auditor to receive 98 or 99 percent approval from the shareholders, so even if you get 90 or 95 percent approval, there are bound to be questions from the audit committee.” This suggests that auditor ratification votes can influence auditor behavior, even if only a small proportion of votes are against (or abstain from) auditor ratification. Many prior studies show that small changes in shareholder votes can elicit significant changes in corporate governance.⁴

Tanyi et al. (2020) use data related to 10,326 auditor-ratification votes spanning the years 2010–2015. Using a company with December 31 fiscal year for example, if shareholder voting occurred on May 15, 2013, they examine the audit quality for the year ended December 31, 2013.

In their primary analysis, they focus on unexpected levels of shareholder votes not voting to ratify the auditor (i.e., against or abstaining from auditor ratification).⁵Tanyi et al. (2020) first estimate a model that predicts shareholder dissatisfaction given the nature of the relationship between the auditor and the client.⁶ Based on prior research, Tanyi et al. (2020) use the following variables to predict shareholder dissatisfaction in a regression model: non audit fee ratio, auditor tenure, SOX 404 internal control weakness disclosure, financial restatement during the preceding year, client size, and auditor type (Big 4 or not). Shareholder dissatisfaction is expected to increase with (1) higher non audit fees (measured as a proportion of audit fees), (2) higher auditor tenure, (3) disclosure of material weaknesses in internal control, or (4) a financial restatement.⁷ Conversely, shareholder dissatisfaction is expected to be lower for Big 4 auditors.⁸

Tanyi et al. (2020) obtain the “unexpected” level of shareholder dissatisfaction as follows. First, for each observation, they obtain the “expected” or predicted level of shareholder dissatisfaction based on the results of the regression model that is estimated within each industry and year. By using the coefficients from the regression model together with the actually observed values of the explanatory variables, one can obtain the “predicted” value of each observation. In the second step, Tanyi et al. (2020) calculate the unexpected level of shareholder dissatisfaction. This is obtained by subtracting from the actual value (i.e., the actual vote results) for a given observation the expected value for that observation, calculated as above. As part of additional analysis, they also examine changes in shareholder dissatisfaction in comparison to the prior year; such analysis assumes that the best predictor of shareholder dissatisfaction this year is the voting from the prior year.

Next, Tanyi et al. (2020) use regression analysis to examine subsequent restatements and include the unexpected shareholder dissatisfaction measure as the explanatory variable of interest. In such analyses, Tanyi et al. (2020) use many other control variables based on prior research. The results show that higher-than-expected shareholder dissatisfaction is associated with lower likelihood of restatement of the following period’s financial statements. The marginal effects of a one standard deviation increase in unexpected shareholder dissatisfaction is about a 7 percent reduction in the likelihood of a future restatement. As part of additional analysis, Tanyi et al. (2020) include a second measure of financial reporting quality: abnormal accruals. The results show that higher than expected shareholder dissatisfaction is associated with lower abnormal accruals in the next period. The marginal effect of a one standard deviation increase in unexpected shareholder dissatisfaction is a reduction of 4.8 percent in abnormal accruals.

Tanyi et al. (2020) also examine audit fees, since there are multiple reasons to expect that unexpected shareholder dissatisfaction can be associated with audit fees. First, audit committees would respond to higher than expected shareholder dissatisfaction by demanding a higher-quality audit. Second, unexpectedly high shareholder dissatisfaction can also influence the auditor to restructure the audit process and exercise more effort during the audit engagement. Together, these factors suggest that audit fees will be relatively higher after unexpectedly high shareholder dissatisfaction.

In such analyses, Tanyi et al. (2020) use ordinary least squares regression and control for many factors shown to be determinants of audit fees. The results show that higher levels of shareholder dissatisfaction are associated with higher audit fees in the subsequent period. In terms of effect size, a one standard deviation increase in unexpected shareholder dissatisfaction is associated with a 3.5 percent increase in audit fees. That is, if we compare two clients A and B that are otherwise identical, and if client A (client B) has shareholder dissatisfaction that is one-half standard deviation more (less) than the sample average, then the audit fees for client A will be 3.5 percent more than the audit fees for client A.

Since audit fee increases can be due to greater effort and/or higher price per unit of effort, Tanyi et al. (2020) also examine audit report lags (measured as the number of days between the fiscal year-end and the audit report date). Prior studies typically use audit report lags as a surrogate for the year-end audit effort.

Tanyi et al. (2020) first examine all observations in a single regression. However, since accelerated and non accelerated filers have different filing deadlines, they then examine the two groups of observations in separate regressions. In all of the models, the results show that audit-reporting lags are higher after higher than expected shareholder dissatisfaction, suggesting that auditors react to higher-than-expected levels of shareholder dissatisfaction by increasing the quality and/or quantity of audit effort. However, although this association is statistically significant, the economic or practical significance is relatively low; the marginal effects of a one-standard-deviation increase in shareholder dissatisfaction is associated with only a 1.4 percent increase in audit report lag. That is, if we compare two clients A and B that are otherwise identical, and if client A (client B) has shareholder dissatisfaction that is one-half standard deviation more (less) than the sample average, then the audit report lag for client A will be 1.4 percent more than the audit fees for client B. Thus, if company A has an audit report lag of 70 days, company B will have an audit report lag of 71 days.

Audit committees may be less inclined to purchase non audit services from the incumbent auditor, given the long-standing controversy surrounding auditor independence and non audit services, after unexpectedly high shareholder dissatisfaction. Coupled with the increase in audit fees, the above argument also suggests that the non audit fee ratio, which itself has often been the focus of regulators and legislators (SEC 2000, 2003; U.S. House of Representatives 2002), would be lower in the period following unexpectedly high shareholder dissatisfaction. As part of additional analyses, Tanyi et al. (2020) find that unexpectedly high shareholder disapproval of the auditor is associated with significantly lower non audit fees and non audit fee ratio.

Instead of relying on unexpected values of shareholder dissatisfaction, Tanyi et al. (2020) also use a more straightforward analysis. The question examined in such analysis is the following: are changes in the levels of shareholder dissatisfaction associated with changes in audit quality, audit fees, and reporting lag (when each of these metrics is compared with the values for the prior year)? The results show that increases in shareholder dissatisfaction are associated with improved audit quality, higher audit fees, and higher audit reporting lag in the subsequent period.

As part of additional analysis, Tanyi et al. (2020) restrict the analysis to two types of severe financial misstatements: (1) misstatements with a negative effect on net income and (2) misstatements that (a) involve an accounting fraud, (b) are associated with an SEC investigation, or (c) are associated with a class action lawsuit. Similarly, for the accruals-based tests, Tanyi et al. (2020) split the sample into two groups: observations with positive accruals and observations with negative accruals. Overall, the inferences are unchanged in such additional analyses.

Tanyi et al. (2020) perform a number of sensitivity tests, including (1) using only votes against auditor ratification (i.e., discarding abstain votes), (2) an alternative measure of the explanatory variable that represents the change in shareholder dissatisfaction from the prior year, and (3) financial and stock price performance of the client, as additional variables to predict the expected level of shareholder dissatisfaction. The inferences remain unchanged in such tests: higher than expected shareholder dissatisfaction is significantly associated with higher audit quality, higher audit fees, and greater audit report lag.

The key takeaway for auditors is that audit firms appear to react to shareholder votes on auditor ratification even when the absolute levels of shareholder dissatisfaction are very low. This is important when viewed together with evidence from another study that audit committees are more likely to dismiss the incumbent auditor when there is higher than expected levels of shareholder votes not to ratify the auditor (Barua et al. 2017). Thus, auditors should pay close attention to shareholder voting on auditor ratification and proactively discuss with audit committees ways to strengthen audit quality when there is higher than expected levels of votes not to ratify the auditor. When faced with higher levels of shareholder dissatisfaction, auditors should ask the audit committee to give another chance to address any issues related to shareholder perceptions.

The results in Tanyi et al. (2020) also are relevant for regulators, such as the SEC and PCAOB, who have used shareholder voting on auditor ratification as a justification for many audit-related disclosure rules. The findings in Tanyi et al. (2020) support such assertions by regulators. Hence, the auditing profession can expect the SEC and PCAOB to use the results from Tanyi et al. (2020) to justify other disclosure-related proposals and rules in the future. For example, the PCAOB relied on the results of Dao et al. (2008) to justify the new rule related to auditor tenure disclosure in audit reports.

Shareholder voting on auditor ratification continues to be voluntary in the U.S. However, it is mandatory in some other countries. For example, in the United Kingdom and other commonwealth countries, even if the audit committee selects the auditor, the law requires that shareholders must vote on that decision and ratify the auditor. Further, NYSE’s rule change in 2009 about voting for shares held in “street name” and quorum-related rules made it expedient for NYSE-listed firms to offer this option to shareholders. However, many public companies listed in other U.S. exchanges (such as the National Association of Securities Dealers Automated Quotations (NASDAQ)) do not provide shareholders the opportunity to vote on auditor ratification. As pressure from activists grows to mandate such shareholder participation, it may be prudent for auditors to suggest to the audit committees of their clients the benefits of providing such opportunity for shareholders.