4.2.2. Results Hypothesis 2—Employee Representatives

The regression results indicate that employee representation has a direct negative effect on earnings management, as expected in the hypothesis. The finding may be due to several causes. In line with Fauver and Fuerst (2006) analysis on German companies, the result could imply that employee representation provides a credible channel for information to the board of directors. Supported by the findings of Chen et al. (2015), this could improve the quality of managerial monitoring and board decision-making since employee representation provides a richer information environment. Moreover, one could argue that the operational information provided by the employee representatives helps to decrease the control issue of information asymmetry. In line with the findings of Gul et al. (2002), Peasnell et al. (2005) and Beatty and Harris (1999), the assumed increased monitoring quality and decreased information asymmetry brought to the board by employee representation is seemingly effective in mitigating agency costs and earnings management.


**Table 4.** Regression results of model (3).

**Notes:** The equation used to test the hypotheses': absDAit = β<sup>0</sup> + β1(BISEit) + β2(DERit) + β3(SODit) + β4(MJSit) + β5(BAit) + β6(ACit) + β7(FSit) + β8(ROAit) + β9(ROEit) + εit(3). \*\*\*, \*\* and \* indicate the significance level at 1%, 5% and 10%, respectively (two-tailed). All numbers reported in NOK million. Robust standard errors in parentheses, \*\*\* *p* < 0.01, \*\* *p* < 0.05, \* *p* < 0.1.
