Do some individual auditors exhibit a systematic audit quality problem? Swedish evidence on the self-contagion effect of low-quality audits

Abstract: The purpose of this study is to test for self-contagion of low-quality audits among individual auditors in privately held companies, using both input and output proxies for audit quality. We conduct three types of empirical tests, using three types of proxies for audit quality. We perform regression analyses on audit fees and unclean audit opinions, and analyze the zero earnings threshold. Our total sample consists of Swedish data on 2,046 company observations for the period 2006-2009, of which 1,023 are observations with an individual auditor that had their license revoked during the period 2010-2012, and 1,023 are control observations. The empirical results provide both statistically and economically significant evidence of a negative relationship between individual revoked license auditors and audit fees, robust for four alternative regressions. Furthermore, our results show that revoked license auditors are associated with a higher frequency of meeting or beating the zero earnings threshold. However, no statistically significant relationship can be identified between individual revoked license auditors and the propensity to issue an unclean audit opinion. We contribute with evidence on self-contagion of low-quality audits with audit fees as a proxy for audit quality, which adds evidence on self-contagion from an input proxy perspective. Our study also provides support on self-contagion for privately held companies. Moreover, our results indicate that revoked license auditors are consistent over time in performing lower quality audits. Our study is subject to the following limitations: (i) limited generalizability to other institutional settings than Sweden, (ii) limitations in observing audit quality, and (iii) time constraints.