The Impact of Anecdotal Data in Regulatory Audit Firm Inspection Reports

Abstract:     
A critical and pervasive component of firm-specific audit firm inspection reports is the inclusion of vivid and detailed descriptions of the deficiencies uncovered by inspection teams. Prior research in psychology indicates that people are likely to focus on such anecdotal information without adequately considering the statistical context provided (e.g., the number of audits that the regulator inspected to find those deficiencies), thereby leading to misperceptions regarding audit firm quality. In this study, managers and other business professionals assumed the role of an audit committee member tasked with evaluating two audit firms. Participants were provided with firm-specific inspection reports where we manipulated both the number of deficiencies reported as well as the number of audits that were inspected at each firm. Our results indicate that participants made decisions consistent with having ignored, or underweighted, the implications of the statistical data provided. That is, participants exhibited an anecdotal bias by focusing on the number of deficiencies reported without appropriately considering the statistical context. This finding is important as it indicates that the common practice of including lists of deficiencies in firm-specific statutory inspection reports can lead to misperceptions of audit firm quality. In addition, we test and provide evidence that two easily implemented decision aids can help to mitigate this problem. Our findings should be of particular interest to audit regulators that currently include, or are considering including, lists of deficiencies in firm-specific reporting. Our study should also be of interest to investors, audit firms, audit committees, managers, researchers, and other stakeholders interested in auditor oversight, auditor reputation and measures of auditor quality.

Number of Pages in PDF File: 43

Keywords: PCAOB, Auditor Quality, Anecdotal Bias, Judgment Orientation

JEL Classification: M41, M49

working papers series