Auditor Tenure and the Length of Earnings Misstatement

Abstract

In this paper we examine the association between audit firm tenure and the length of misstatements. Using a setting that is not subject to the endogeneity problem endemic to other studies on the topic, we find consistent evidence that longer audit firm tenure leads to a less timely discovery of misstatements and to larger magnitudes of misstated earnings. We also find evidence that the detrimental effect of long audit tenure is due to auditor complacency built up over time, but no evidence that it is due to loss of independence. Further, using the non-voluntary change of auditors following the demise of Arthur Andersen as a natural experiment, we show that misstatements by its former clients were discovered significantly faster than misstatements by other Big 4 clients that retained their auditors. These findings reinforce the benefit of a fresh look by new auditors. Our study has implications for regulators who continue to express concern regarding lengthy auditor-client firm engagement.