臺大管理論叢 NTU Management Review VOL.30 NO.2

89 NTU Management Review Vol. 30 No. 2 Aug. 2020 defined as the ratio of earnings before extraordinary items deflated by prior period assets, to control for firm performance. We have no prediction pertaining to the coefficient on these two variables. 4.2.2 Factors to Discriminate between Accrual Management and Real Activities Manipulation If the accruals process leads to reversals of accruals, and if accruals adjustments reverse in the subsequent accounting period, lagged accruals will be negatively associated with current accruals (Hunt, Moyer, and Shevlin, 1996). Accordingly, we include in equation (6) lagged discretionary accruals to control for accrual reversals. We predict DA it to be negatively associated with its lagged value, DA it-1 . The primary penalty for earnings management is litigation. As mentioned earlier, auditors and regulators are less likely to scrutinize and detect real earnings management behaviors relative to accrual management. Accordingly, greater perceived litigation penalties should increase the propensity for real activities manipulation (Cohen and Zarowin, 2010). To capture potential litigation risk, we include in equation (7) LTGN it , an indicator variable equal to 1 if the firm is in a high litigation industry, and 0 otherwise. Following Barton and Simko (2002) and Cohen and Zarowin (2010), high litigation industries are SIC codes 2833-2836, 3570-3577, 3600-3674, 7370-7379, and 8731-8734, which correspond to pharmaceuticals/biotechnology, computers, and electronics industries, respectively. We expect LTGN it to be positively associated with RM it . 4.2.3 Endogeneity Considerations To test for the existence of endogeneity formally, we use the Hausman (1978) test for contemporaneous correlation between the error term and the two earnings management tools (i.e., discretionary accruals and real activities manipulation), respectively. We control for the endogeneity of management’s decision to choose particular earnings management mechanisms. 5. Sample Selection and Industry Composition We obtain the initial sample of 970 foreign firms cross-listing on the U.S. exchanges from the listings of International Registered and Reporting Companies issued by the SEC between 2007 and 2010 (SEC, 2014). We exclude firms that domicile in countries that adopt U.S. GAAP/IFRS mandatorily, because it is less likely that their filing choice is