6. Summary and conclusion
We examine the earnings management behavior of firms that report internal control weaknesses (ICWs) by focusing on their use of real activities manipulation (RM). We provide evidence that firms with ICWs engage in real, beyond accrual-based, earnings manipulation. Our univariate and regression analyses indicate that firms with ICWs have lower abnormal cash flows from operations, higher abnormal production costs, and lower abnormal discretionary expenses compared to non-ICW-firms. However, the level of discretionary accruals for ICW- firms is not significantly different from non-ICW-firms. These findings provide support for the contention that the reporting of ICWs does not prevent real earnings manipulation activities from occurring. Our initial multivariate analysis demonstrates that firms reporting ICWs manage real activities and these firms do not manage discretionary accruals. These results are supportive of the position that, in the post-Sarbanes– Oxley era, firms may manage earnings with real activities manipulation because of increased scrutiny of the financial reporting practices by the external auditors. Further analysis also reveals that while ICW-firms attempt to manage real activities in order to achieve or maintain earnings targets, firm performance actually decreases in the subsequent period We apply the Heckman (1979) procedure in our study to control for self-selection bias. Our first-stage results generally support the Ge and McVay (2005) internal control weakness model that indicates that ICWs are associated with operating complexity and profitability, ICWs decrease with increasing firm size, and ICWs decrease when audited by a Big 4 audit firm. Furthermore, our second-stage results show that ICWs reported by firms are positively correlated with different measures of real activities manipulation.