Illinois Symposium on Auditing Research October 14, 2016 A norm The Interactive Effects of Ethical Norms and Subordinate Recommendations on Accounting Decisions by Khim Kelly and Pamela Murphy Discussion by Steven Kachelmeier University of Texas at Austin
Premise: Norms shape how we (re)act Consider the norms for reviewers vs. discussants: Reviewers: Adopt a negative frame. Indicate reservations about the paper s incremental contribution. Express concerns. Discussants: Adopt a positive (or at least upbeat) frame. Say how important the topic is and how much you enjoyed reading the paper. Offer constructive feedback.
Kelly and Murphy (2016) I really enjoyed reading this paper. The research question is clearly important. I will offer comments in the spirit of constructive feedback. The subjectivity of norms Audit implications
1. Types of Norms The paper argues that norms can be descriptive or injunctive. I would argue that norms can be further classified as follows: Objective norms (reality) Perceived norms (reality through a subjective lens) Rationalized norms (egocentric reality)
Objective norms These are the actual norms that an organization cultivates. Ideally, one would measure organizational norms independently of the experiment, from the perspective of a consensus of dispassionate observers. Unfortunately, experimental participants are not dispassionate observers.
Perceived norms Two people working in the same organization can perceive different norms. Analogy: Burney L., C. Henle, and S. Widener, A Path Model Examining the Relations among Strategic Performance Measurement System Characteristics, Organizational Justice, and Extra- and In-Role Performance, AOS, April 2009. Survey participants: 242 employees from a large financial services firm. Finding: Attributes of the performance measurement system such as alignment with strategic objectives drive perceptions of organizational fairness.
Perceived norms Two people working in the same organization can perceive different norms. The key point: All 242 respondents were employed by the same company and hence observed the same performance measurement system, and yet they perceived it quite differently.
Perceived norms Two people working in the same organization can perceive different norms. What this implies: Perhaps our (re)actions are shaped not so much by norms as by our personal values that influence how we perceive norms. Why this is important: If perceived norms are partially a manifestation of different individual perspectives, change becomes much more difficult.
Perceived norms Two people working in the same organization can perceive different norms. Suggestion: The paper should acknowledge and discuss that this study elicits perceived norms, not necessarily actual norms. This is a counterargument to the limitations of manipulating assumed norms, as in the study s supplemental experiment.
Rationalized norms We feel better when we rationalize that our personal preferences are consistent with organizational norms. Related to the so-called false-consensus effect. Non-accounting example: What percentage of people prefer whole wheat to white bread? Estimate by people who prefer whole wheat: 53%. Estimate by people who prefer white: 37%. Accounting example (from Williams, J., JAR, September 2013): Financial analysts adjust insufficiently for other forecasts, as if they overestimate the extent to which their individual forecasts reflect a consensus.
Rationalized norms We feel better when we rationalize that our personal preferences are consistent with organizational norms. Related to the reverse causality threat addressed in footnote 8. Participant reasoning: I think the conservative alternative is best even though my subordinate argues for the opposite. This inconsistency makes me uncomfortable. Thus, I rationalize that the conservative approach is how we do things around here.
1. Types of Norms Takeaway: Norms are endogenous. Not only do they shape our preferences, but perceptions of norms are also shaped by our preferences. Some discussion of this point in the paper would be helpful.
2. Audit implications By far, the most provocative claim in the paper (to me) is that the most conservative accounting decisions are made in the condition where participants receive an aggressive recommendation and work in organizations with higher ethical norms (p. 3).
What I see as the primary finding: Percent choosing the conservative option 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Aggressive recommendation Conservative recommendation p = 0.21 (Fisher s Exact Test; two-tailed) Higher Norms Lower Norms
2. Audit implications Implication: The best way to maintain a norm for conservative accounting is to encounter a suggestion to be aggressive. Kelly and Murphy (2016) allude to audit implications, but this is not an auditing study per se. The subordinate and decision maker work for the same company. But what if we replace the subordinate with an audit client and the decision maker with an auditor?
Plausibility tests Conservative determination when the recommendation is aggressive? Within a company: Yes, this is plausible. Auditor-client: Yes, this is plausible. Aggressive determination when the recommendation is conservative? Within a company: Yes, this is plausible. Auditor-client: Not so sure about this one.
Could this apply to auditing? There is some evidence that auditors can be softened by a less-aggressive client stance. Example: Wolfe, C., E. Mauldin, and M. Diaz, Concede or Deny: Do Management Persuasion Tactics Affect Auditor Evaluation of Internal Control Deviations? The Accounting Review, November 2009. Findings: Auditors assess internal control deficiencies to be less severe when the client concedes there is a problem than when the client denies there is a problem. Thus, a conservative client prompts the auditor to reach a more aggressive conclusion.
This tendency is similar to what I call the remorseful child phenomenon. Dogs have this down to a science.
Could this apply to auditing? Still, I have my doubts about whether the current findings would extend to auditor-client interactions. Client manager: We ve thought about it, and we think it is best to take the more conservative alternative for recognizing revenue on this contract. Audit partner: Conservative? No way! Either you take an aggressive approach to recognizing revenue or I won t sign the audit opinion.
2. Audit implications Takeaway: It is unclear how the scenario in this study would extend to auditor-client interactions, if at all. If the authors are serious about the study s implications for auditors who attempt to curb aggressive reporting (p. 5), a deeper consideration of these points would seem appropriate. That is, if an auditor is aware of the company s preference for aggressive or conservative accounting, what exactly should the auditor do differently based on this study s findings?