The Role of Audit Committee Interlocks in the Dissemination and Contagion of Accrual-Based and Real Earnings Management

Abstract

We examine the effect of audit committee interlocks on the dissemination and contagion of accrual-based earnings management (AEM) and real earnings management (REM) for the pre- and post-SOX periods. We use “dissemination” to denote information association and “contagion” to designate causal information transmission. We first document that interlocked firms exhibit AEM dissemination (but not REM dissemination) for the pre-SOX period and REM dissemination (but not AEM dissemination) for the post-SOX period. To alleviate the concern that our results might be driven purely by the enactment of SOX, we perform a placebo-controlled falsification test, which shows that neither AEM nor REM dissemination exists in the absence of audit committee interlocks. To address the endogeneity concern that two similar firms with parallel accounting practices appoint the same director, we use an interlock initiation analysis and establish an AEM contagion effect for the pre-SOX period but no REM contagion effect for the post-SOX period. We also construct an AEM-REM substitution index. When information-sending firms substitute REM for AEM, information-receiving firms similarly switch. Collectively, while ease of detection and an increased level of scrutiny are listed as reasons for the shift from AEM to REM after the enactment of SOX, we contribute to the literature by showing that audit committee interlocks could be a viable mechanism in converting these motivations into actions and are a plausible factor in facilitating this transition from AEM to REM.