Discussant Comments on 'Does the Identity of Engagement Partner Matter? An Analysis of Audit Partner Reporting Decisions'

The authors of this provocative study apply commonly used audit quality surrogate measures to a large and unique set of financial and other data on statutory audits of small private companies in Sweden. The paper has received unparalleled attention by the financial press and the PCAOB for its presumed support for regulatory intervention in standards for U.S. public company audits. In this Discussant Comment, I review the paper's content, analyze its predictive validity, and discuss its multiple implications plus, following Conference instructions, I provide constructive suggestions for improvements. Based on predictive validity analysis, I conclude that engagement partner assignment strategy is an important and acknowledged omitted variable that affects the study's internal validity via both the independent variable (partner's prior performance measure) and the dependent variable (borrower's cost of debt capital). The omission also affects construct validities and, if audit firms are applying a plausible assignment strategy, then interpretation of the study's main results would be reversed. Finally, the lack of a standards intervention noted by the authors and the extreme size and other differences between audits of Swedish private companies and U.S. public companies impair external validity and generalization to the U.S. intervention. As to improvements, I suggest that the authors (a) ask Swedish lenders to validate their presumed use of partner performance ratings in determining a borrower's interest rate, and (b) ask Swedish Big four audit firms to provide a few internal partner performance ratings for comparison with the external performance measures used in the study. This two-pronged, multi-method approach might confirm or deny critical assumptions underlying the present study and may substantively inform standards setters, evidence-based standards, and fellow researchers about the validity of commonly applied surrogates for audit quality and the study's stated conclusion