Did the 2005 Deferred Prosecution Agreement Adversely Impact KPMG's Audit Practice?

Abstract

SUMMARY We examine the consequences of misconduct in a Big 4 firm's nonaudit practice for its audit practice. Specifically, we examine whether KPMG's audit practice suffered a loss of audit fees and clients and/or a decline in factual audit quality following the 2005 deferred prosecution agreement (DPA) with the Department of Justice for marketing questionable tax shelters. We find little evidence that the DPA adversely impacted KPMG's audit practice by way of either audit fees or the likelihood of client gains/losses, suggesting little or no harm to KPMG's audit reputation. We also find that the DPA had no effect on the firm's factual audit quality, even for those audit clients that dropped KPMG as their tax service provider. Collectively, our findings suggest that there was no spillover effect from the DPA to KPMG's audit practice. Data Availability: All data are publicly available.

Do Client Bankruptcies Preceded by Clean Audit Opinions Damage Auditor Reputation?*

“…There is the potential that poor research design and/or sample limitations may cause a failure to reject null hypotheses although associations truly exist. Following previous archival studies (Baugh et al ; Blaylock ; Craswell et al ), I examine variation in the estimated coefficients, and conclude that the lack of statistical significance in the auditor dismissal and audit fees analyses is more likely due to a lack of practical significance in an underlying association than a lack of statistical power. However, this remains a limitation of the study.…”

“…The former is consistent with no underlying association, while the latter suggests a lack of statistical power. To evaluate which situation holds for these analyses, I follow previous research, and calculate the hypothetical estimated coefficient that would have achieved statistical significance given the standard error of the reported estimated coefficient (Baugh et al ; Blaylock ; Craswell et al ).…”

“…However, this remains a limitation of the study. Previous studies that document non–statistically significant results acknowledge similar limitations that design and/or data limitations may cause them to incorrectly fail to reject the null hypothesis (Baugh et al ; Blaylock ; Boone et al ; Callaghan et al ; Chung and Kallapur ; DeFond et al ; Knechel and Sharma ; Koh et al ; Lennox , ; Reynolds et al ; Ruddock et al ).…”

Sins of the Father: The Effect of a Parent Firm's Accounting Misconduct on Current and Former Subsidiaries.

“…This setting allows us to examine a) the effect of a parent firm management's reputation on subsidiaries and b) the extent of information transfer between a parent and subsidiaries. To our knowledge, although auditor reputation has been studied with mixed results (e.g., Nelson, Price, and Rountree 2008;Weber, Willenborg, and Zhang 2008;Baugh, Boone, Khurana, and Raman 2019;Donelson, Ege, and Leiby 2019), management reputation has received little attention due to the limitation that managers typically only oversee one publicly traded firm at a time. 1 While information transfer has been studied extensively (e.g., Foster 1981;Baginski 1987;Gleason, Jenkins, and Johnson 2008, among many others), these studies often use broad groupings to examine information transfer.…”

“…Several other studies similarly find mixed evidence of reputation effects for auditors (e.g.,Baugh et al 2019;Donelson et al 2019). For the 12/31/2000 year end, KMP was audited by PwC, Houston, Texas; NBP was audited by Andersen, Omaha, Nebraska; EOTT and EOG were both audited by Andersen, Houston, Texas.…”