The Effects of Clawbacks on Auditors' Propensity to Propose Restatements and Risk Assessments Free

This article summarizes the previously published study titled “The Effects of Clawbacks on Auditors' Propensity to Propose Restatements and Risk Assessments” (Brink, Grenier, Pyzoha, and Reffett 2018; hereafter, “the study”). We discuss the study's motivation, the findings of three experiments, and survey and interview results. We then discuss the practical implications of the study for audit firms, client management, and stakeholders of companies that adopt a clawback.

The general intent of the Sarbanes-Oxley Act of 2002 (SOX; U.S. House of Representatives 2002) is to improve the material accuracy of public companies' financial statements by holding executives and their financial statement auditors more responsible for their actions and decisions. An important section of SOX, Section 304, contains a clawback provision, which aims to strengthen the quality of financial reporting and management's ethical behavior by requiring the Securities and Exchange Commission (SEC) to recoup any previously awarded incentive compensation based on subsequently restated financial statements due to fraud (SEC 2010). Subsequently, Section 954 of The Dodd-Frank Act of 2010 (Dodd-Frank; U.S. House of Representatives 2010) strengthened the SOX clawback provision by extending coverage to restatements due to fraud or error. Further, Section 954 included some other important changes: (1) companies must adopt and oversee their own clawback policy, (2) incentives are recoverable for up to three years, and (3) current and former executives who have control over the financial reporting process are covered by the clawback. The SEC is currently working to finalize the clawback rules from the Dodd-Frank Act.¹

Consistent with the intent of the Dodd-Frank clawback provision, most companies have been voluntarily adopting their own clawback policy while they await the SEC's final rules. Specifically, 77 percent of companies in the S&P 500 had adopted a clawback as of the end of 2015 (Equilar 2016). Prior studies find that restatements in fact have decreased for companies that have voluntarily adopted a clawback (deHaan, Hodge, and Shevlin 2013; Chen, Greene, and Owers 2015). However, Pyzoha (2015) finds an unintended consequence of adopting a clawback provision. Specifically, the author finds when a misstatement has occurred, the presence of a clawback provision increases managements' opposition to a proposed restatement in order to protect their incentive compensation, particularly when they perceive the auditor as being of lower quality. Thus, it is unclear whether the aforementioned reduction in restatements in a company imposed clawback environment is due to more faithful financial reporting or increased management opposition to restatements in an attempt to retain their compensation.

In addition to the presence of a voluntarily adopted clawback influencing management's decisions, voluntarily adopted clawbacks may also influence the auditors of the financial statements, which is the focus of the study. Specifically, after a misstatement has been identified, are auditors more, less, or equally inclined to recommend restatements when they know management will be financially impacted by the presence of a company imposed clawback provision? This effect is critical given the significant economic consequences of restatements for firm stakeholders and financial statement users. A secondary interest of the study is the examination of the effect of a voluntarily adopted clawback provision on the auditors' pre-audit assessments of the risk of material misstatement (RMM). This is important as auditors may be less vigilant in the presence of clawbacks despite Pyzoha's (2015) finding that management may not necessarily be more ethical in their reporting decisions.

The study's first hypothesis predicts: “Auditors will assess a lower risk of material misstatement in the presence versus absence of clawback provisions.” Hirst (1994) discusses how management's financial incentives to misreport impact auditors' RMM assessments, with RMM assessments increasing (decreasing) as management's incentives to misstate increase (decrease). This indicates that if auditors believe a clawback provision reduces a manager's incentive to misreport then RMM assessments from auditors likely decrease. Similarly, Cohen and Hanno (2000) and Schmidt (2014) indicate that increasing governance structures and control environments decrease RMM assessments. Therefore, if auditors believe a company imposed clawback policy is an effective governance structure or enhances the control environment, the presence of the clawback policy should decrease RMM.

The study's second hypothesis predicts, subsequent to the discovery of a potential misstatement, “Auditors will be less likely to propose restatements of prior financial statements in the presence versus the absence of a clawback policy.” This prediction is based on previous auditor independence research (e.g., Hackenbrack and Nelson 1996; Kadous, Kennedy, and Peecher 2003) coupled with psychological theory on motivated reasoning (e.g., Kunda 1990). Hackenbrack and Nelson (1996) and Kadous et al. (2003), among others, find auditors exploit ambiguity in financial reporting standards to make judgments in management's favor. This motivation is a result of the desire to reach client preferred conclusions, which, in ambiguous situations, can subconsciously bias auditors' judgments. Given this tendency, the study expects, under the circumstance of an ambiguous restatement, auditors will be less likely to conclude a restatement is necessary to avoid triggering a clawback that requires client management to pay back their incentive compensation.

In addition to the two hypotheses discussed above, the study also provides evidence related to the following research question related to voluntarily adopted clawbacks: “Do the effects of clawbacks on auditors' restatement recommendations vary across higher versus lower importance clients?” The economic theory of auditor independence (DeAngelo 1981) suggests that the reluctance to restate an audit client's financial statements stemming from the presence of a clawback will be more pronounced for higher versus lower importance clients (Chung and Kallapur 2003). In several studies, client importance is perceived to decrease auditor independence (e.g., Krishnan, Sami, and Zhang 2005). On the other hand, prior archival studies have shown no association between client importance and audit quality (e.g., Chung and Kallapur 2003), which may be because engagement risk increases with client importance.

To test the study's hypotheses and research question, the authors conducted three between-participant experiments with practicing auditors to understand how the presence of a clawback provision affects auditors' judgment and decision-making. In addition, the authors conducted interviews and provide survey evidence to complement the experiments. See Figure 1 for a summary of the key results and implications for practice of the experiments, survey, and interviews.

The study's first experiment manipulates between-participants the presence of a company imposed clawback policy (non-existent, lower percentage of incentive compensation subject to clawback, higher percentage of incentive compensation subject to clawback) and client importance to the audit firm (lower, higher). Using a modified version of the experimental instrument previously developed by Pyzoha (2015), the study randomly assigns 100 auditor participants of all ranks (predominantly senior and manager), to review a fictitious client case using an online instrument.² The case incorporates a fictitious client whose CFO earned $2.5 million of compensation, $1.5 million of which was incentive based. The lower-level clawback policy mandates the CFO repay $750,000 of the incentive reward, and the higher-level policy stipulates the CFO pay back the entire $1.5 million reward in the event the previous year's financial statements required restatement. In the no clawback condition, the case clearly states that no clawback policy was in effect. The higher importance client's audit fees represent 50 percent of the local office's revenue; the lower importance client generates audit fees representing 2 percent of the local office's total revenue. After learning of the company imposed clawback policy (or lack thereof) and client importance, participants assess the risk of material misstatement on a 100-point scale. The study uses this assessment to test the first hypothesis.

The second portion of the experiment provides the participants with information regarding an aggressive accounting procedure surrounding an impairment loss that could potentially require a restatement of the client's prior year financial statements. The case informs the participants that the auditor's analysis suggests it may be prudent to restate the previous financial statements in light of a more conservative method of estimation related to the impairment. This prompts participants to answer, “How likely is it that you would recommend that [the client's] financial statements should be restated due to the impairment?” on a 100-point scale. The study uses this assessment to test the second hypothesis.

In general, the results of Experiment 1 do not support the study's predictions—the presence of a clawback feature did not affect either auditors' RMM assessments or their restatement recommendations. Specifically, inconsistent with the first hypothesis, the study finds participants' RMM assessments under the no clawback policy (M = 40.00) were not significantly different from those under the lower clawback policy (M = 41.21) nor the higher clawback policy (M = 41.39), with p > 0.10 for both comparisons. Supplemental analysis finds that while the presence of a company imposed clawback provision did not lead to lower risk assessments, auditors did perceive the financial statements to be of a higher quality in the presence of a high clawback versus no clawback.

As noted above, the study also does not find support for the second hypothesis, implying that auditors are not less likely to propose restatements due to the presence or absence of a voluntarily adopted clawback provision. The average likelihood of auditors to recommend a restatement was not significantly different under the no clawback policy (48.06 percent), versus the low clawback policy (49.09 percent), or the high clawback policy (54.72 percent), all p > 0.10. With regard to the research question, the authors find no evidence that the effects of clawbacks on RMM assessments or propensity to propose restatements depend on client importance.

The study attributes the lack of supporting evidence for the second hypothesis to the fact that auditors feel more accountable when proposing restatements when faced with clawback policies given the nature of their consequences. An assessment of auditor accountability using an 11-point scale suggests, though statistically insignificant, that auditors perceive greater accountability in the presence of high clawback policies (M = 5.75) than in the presence of no clawback policies (M = 4.90), a result that is further investigated in Experiment 2.

The unexpected results of Experiment 1 prompted the authors to conduct additional experiments to verify the reliability of their findings. Experiment 2 features more participants per cell and exerts greater experimental control by conducting the experiment in a controlled environment rather than online. It also adds numerous process measures to assess the post hoc explanation from Experiment 1 that clawbacks promote greater auditor accountability. In addition, Experiment 2 removes the lower clawback policy variable manipulation, rendering it a 2 × 2 between-participants design. Experiment 2's participants are 98 auditors of primarily senior rank and Big 4 employment.

Similar to Experiment 1, the results of Experiment 2 do not support the first or second hypotheses. The effects examined for the research question were also not significant. Specifically, auditors do not assess lower risk of material misstatement and were not less likely to propose restatements due to the presence or absence of a company imposed clawback policy. The degree of client importance does not change auditor inclination to propose a restatement.

Using the same 11-point scale as in Experiment 1, but with higher power due to having more participants per experimental cell, to examine auditor accountability, Experiment 2's results indicate that the participants assessed auditor accountability as significantly higher in the presence of the clawback policy (M = 6.92) as opposed to the absence (M = 5.47). A mediation analysis indicates auditor accountability does explain the relationship between clawback presence and auditor likelihood to recommend a restatement, thus implying the significance of perceived accountability as a motivating factor for auditors to behave ethically and suggest restatements when necessary, despite potential management opposition.

Experiment 3 includes additional experimental design changes in an attempt to address the consistent lack of significant results of the prior two experiments. It consists of a 1 × 2 design and only manipulates the presence of a company imposed clawback between-participants. The clawback condition contains the same information as Experiment 1, but the no clawback condition is different in that participants are not provided any information on clawbacks. This modification confronts the concerns that (1) participants in the no clawback condition assumed that the company could decide to voluntarily adopt a retroactive clawback policy in the future or (2) salient non-adoption of a clawback affects auditors' judgments. Information regarding client importance is not included. Participants are 37 auditors of a variety of ranks who were recruited from three non-Big 4 firms and one of the author's contacts on a professional networking website.

In contrast with Experiments 1 and 2, Experiment 3 results do support the first hypothesis. There was significant evidence to support the claim that auditors will assess a lower RMM in the presence (M = 33.56.00) versus absence (M = 55.26) of a company imposed clawback policy (p = 0.001). A possible explanation for this difference is that clients who employ a clawback provision even when its adoption is not necessary may appear more controlled and less risky and therefore reduce the auditor's RMM assessment. The study suggests that future research be conducted regarding the effects and presence of other risk factors to further examine the discrepancies in results between the Experiments 1 and 2 versus Experiment 3. Consistent with the previous experiments, the results of Experiment 3 also do not support the second hypothesis (p = 0.273). Thus, across three experiments, the study reliably demonstrates that auditors are not less likely to propose a restatement in the presence of a clawback policy.

The study next gathered qualitative evidence to supplement the three experiments and provide insight as to whether the experiment results align with auditors' conscious opinions as to how clawback provisions impact their judgments. The qualitative evidence was captured via an expert panel discussion and interviews with auditors. The expert panel included members of the Accounting Advisory Group (AAG) of a midwestern university. The AAG panel consisted of 16 accounting professionals, 14 of whom are current or previous auditors with an average of 10.61 years of auditing experience. Panelists completed a short survey and participated in a full panel discussion facilitated by one of the study's authors.

In the survey, participants indicated to what degree they believed the presence versus absence of a company imposed clawback provision would affect an auditor's assessment of engagement risk, the control environment, RMM, substantive testing, and restatement recommendations. Responses were indicated on a scale of 0 = “Significantly decrease,” 5 = “No effect,” and 10 = “Significantly increase.” The results for each of the five questions found no significant deviation from the midpoint of no effect, revealing the lack of impact a clawback provision had over the participants' judgments. Thirteen members of the expert panel also provided written qualitative information pertaining to how they believed a clawback policy would affect an audit. In summary, the written evidence suggests company imposed clawbacks improve a firm's control environment, but the effects of this improvement on RMM would likely already be captured by auditors' risk assessments related to performance-based compensation (e.g., use of stock options).

Additionally, the panel stated auditors would uphold their ethics and be unlikely to avoid proposing a necessary restatement simply because a company adopted a clawback provision. Lastly, the study's co-authors interviewed three audit partners about their perceptions of the effects of clawback policies on auditor risk assessment and propensity toward restatement. The findings of these interviews are consistent with those of the experiments and qualitative evidence.

The results of the study's three quantitative experiments and qualitative evidence uniformly indicate, inconsistent with expectations, that there is not a negative relationship between the presence of a voluntarily adopted clawback policy and auditor propensity to propose a restatement. Though some evidence suggests a clawback feature may decrease auditors' assessments of control risk, these effects are likely already captured by auditor's risk assessments related to performance-based compensation. The lack of support for the original hypotheses appear related to the fact that auditors recognize even greater accountability for proposing necessary restatements when such a proposal may result in managerial forfeiture of incentive-based compensation due to a clawback. This decision is a critical choice in an audit engagement, and consequently, likely is viewed by auditors as too important to be influenced by the presence or absence of a clawback provision.

The study's findings are important for the auditing profession, regulators, and stakeholders of companies that voluntarily adopt a clawback provision or for companies that will be statutorily required to adopt one once the Dodd-Frank Act's clawback rules are finalized by the SEC. The auditing profession can have comfort that a company's clawback policy will not impair auditor's ethical judgments and their likelihood to recommend a restatement once a potential misstatement is identified. This suggests auditors may not need to alter their audit approach due to the presence of clawbacks. Based on the study's findings, it does not appear auditors change their perceptions of reporting quality in response to a clawback provision. In fact, the study shows that clawback provisions strengthen auditor accountability for higher quality financial reporting, which mitigates any negative effects that the study expected to find related to auditors' ethical judgments. This should be of note for regulators and policy makers due to the increased prevalence of voluntarily adopted clawback policies in practice. Specifically, although SOX, Section 304 applies to all public companies, recent data shows that three in four S&P 500 firms have also adopted company-specific clawback policies in response to the pending SEC clawback rules based on Dodd-Frank (Equilar 2016). In sum, results of the study should give practitioners a degree of comfort that their quality control structures around auditor independence and engagement performance review are sufficiently rigorous to prevent auditors from either consciously or unconsciously comprising their ethics when client management would monetarily suffer from a restatement, even for highly important clients.

For regulators, the purpose of imposing clawback provisions is clear in that the SEC believes that reporting quality will increase if the incentives for misreporting decrease. The increase in reporting quality should result in a decrease of future restatements. However, it is important for policy makers and regulators to consider that there may be multiple reasons why clawbacks could decrease restatements, some of which are not indicative of higher quality financial reporting such as: (1) greater management opposition to restatements (Pyzoha 2015); (2) reduced substantive audit testing because auditors perceive a decreased risk of material misstatement; and (3) auditors being less likely to pursue a restatement once a potential misstatement is discovered to avoid client conflict. The study focuses on the second and third possible scenarios and finds limited evidence that auditors will decrease their assessment of the risk of material misstatement and no evidence that auditors are less likely to propose a restatement to avoid triggering a company imposed clawback. These are important findings in light of Pyzoha (2015) finding that management may be less willing to restate the financial statements in a clawback environment. Taken together, evidence from the study should provide comfort to regulators and the stakeholders of companies that employ a clawback provision that audit professionals will maintain their ethics when considering whether to propose a restatement. This is especially important due to the significant economic consequences that come along with restatements as well as the SEC's work in finalizing the clawback rules based on the Dodd-Frank Act's clawback provision.

The study also prompts future researchers to consider examining the possible effects of other critical factors as they relate to clawback provisions and auditors' judgments and decision-making. For example, future research may consider the length of the audit firm's tenure, the riskiness of the client, or whether the client voluntarily adopted a new clawback policy during the current year audit engagement. Although the study addresses some critical impacts of the application of clawbacks and auditor behavior, many questions remain unanswered about the effects of clawback provisions for future studies to build upon.