Surprising Effects of Mandatory Auditor Rotation on Audit Quality
Kendall O. Bowlin, University of Mississippi; Jessen L. Hobson, University of Illinois at Urbana-Champaign; and M. David Piercey University of Massachusetts – Amherst | December 15, 2015 | 4
As new auditor rotation mandates are debated and adopted or rejected worldwide, a new research study takes a different approach to assessing the effects of these mandates on audit quality.
So far, the debate over mandatory auditor rotation has been framed as two competing arguments. On one hand, proponents of mandatory rotation are concerned about the risks that long-term auditor-client relationships pose to the auditor’s mindset. By limiting that relationship, rotation will ostensibly improve audit quality by helping ensure that auditors remain professionally skeptical and do not become overly trusting of their clients’ assertions. On the other hand, opponents of mandatory rotation argue that the knowledge acquired by an auditor about the specifics of a company will be lost with each rotation, ultimately harming audit quality.
A new study, “The Effects of Auditor Rotation, Professional Skepticism, and Interactions with Managers," looks more closely at the presumed benefits of mandatory rotation and concludes that those benefits sometimes do not exist. In fact, in some cases, mandatory rotation may actually worsen the very issue that it is presumed to improve. Specifically, the basic assumption has been that auditors become more trusting of clients over time. Therefore, it is further assumed that they would be less trusting of their clients if they have to rotate.
This study calls these assumptions into question, demonstrating that they are only half true. On one hand, if auditors are focused on assessing the likelihood that their clients’ claims are true, then the study’s findings suggest that auditors in long-term relationships become too trusting of their clients’ claims, and rotation helps reduce this tendency. On the other hand, if auditors are focused on assessing the risk that their clients’ claims are false (e.g., fraud risk assessments), auditors benefit from being in a long-term relationship with their clients, and rotating auditors are the ones who become too trusting, to the detriment of audit quality.
Why does this occur? In a long-term relationship, auditors gain confidence in assessing the truth or falsity of their clients’ claims. So when auditors are assessing the likelihood that their clients’ claims are true, they are able find plenty of psychological support for that possibility. On the other hand, when they are assessing the risk that those claims are false, they are also able to find more psychological support for that possibility. Rotation undermines this confidence. Rotating auditors are at a disadvantage in finding psychological support for either the possible truth or falsity of their clients’ claims. As a result, this study suggests that the basic assumption underlying this debate—that rotation would improve auditor efforts to be skeptical—is not valid. Skepticism appears to benefit from a long-term auditor-client relationship.
This is a fundamental psychological effect that occurs among people in general. In order to demonstrate its findings, the study used a carefully controlled laboratory experiment. Participants in the experiment interacted with one another in anonymous pairs over a computer network. In each pair, one participant took the role of an auditor, and the other took the role of the manager. They had to make choices that were analogous to choices both clients and managers regularly make. That is, managers had to choose aggressive or conservative financial reporting, and could make non-binding claims to the auditors. Auditors could decide whether to believe those claims, and whether to engage in higher-cost or lower-cost auditing of that particular assertion. Although these roles and choices were described using generic terms to prevent participants guessing the study’s purpose, the experiment captured the key dynamic of the choices auditors and managers face. In the interests of audit efficiency, auditors’ best choice is lower-cost auditing as long as the manager is reporting conservatively. Otherwise, lower-cost auditing is the worst choice for auditors. For managers, reporting aggressively is the best choice as long as the auditor has selected lower-cost audits. But aggressive reporting is the worst choice if auditors engage in higher-cost audits.
Why use this experimental approach to study auditors? Nearly all of the prior research on the effects of rotation uses archival databases of historical data. As regulators have pointed out, these studies therefore suffer from an inherent limitation. Specifically, they have primarily examined whether long-term auditor-client relationships are correlated with better or worse earnings quality. For example, if long-term relationships result in better earnings quality, the assumption is that the knowledge that auditors gain over time must trump the threats that the relationship poses to their skepticism. If true, rotation would undermine audit quality, which would appear to grow over time. The problem with these correlational studies, though, is that this correlation could occur for completely different reasons. For example, auditors tend to drop their more aggressive clients over time (i.e., client renewal decisions), and more aggressive financial reporters tend to go through different auditors (i.e., “opinion shopping”). Both of these effects would drive a positive correlation between longer auditor-client relationships and earnings quality, but this says nothing about whether audit quality itself improves over time.
By using this experimental approach, the study controls for these alternative explanations in ways that correlational data cannot. For example, in this experiment, aggressive financial reporters were randomly assigned to rotation or no-rotation conditions, and so simply being an aggressive financial reporter could not artificially drive the relationship between rotation and audit quality. In addition, by using an experiment, the study can see important “internal” measures of audit quality and earnings quality that archival databases are blind to. For example, the study can see exactly how often managers lie, how often auditors believed them, and how often they selected lower-cost audits while managers were reporting aggressively.
The article concludes “[a]n implication of our study is that focusing auditors on a skeptical assessment frame rather than mandating auditor rotation may be a less costly way to reduce low-effort audits and aggressive reporting.” The study appears in the July issue of The Accounting Review, published by the American Accounting Association.