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

Do Government Initiatives and Reasons for Auditor-Client Realignment Influence Audit Fees and Audit Quality? Evidence from China 234 For the purpose of our examination, we conduct a difference-in-differences analysis. As shown in Table 3, all observations are divided into the pre-2007 and post-2007 time periods. We also group observations according to state ownership as SOEs and NSOEs. As illustrated in the table, the difference in SOEs disclosing their realignment reasons before and after 2007 is significant (difference = 18.9%, p -value < 0.001). Moreover, the difference in NSOEs disclosing their realignment reasons before and after 2007 is significant as well (difference = 43.5%, p -value < 0.001). Comparing this difference between SOEs and NSOEs during the pre-2007 time period, we find that the between- group difference in disclosing realignment reasons is significant (difference = 15.15%, p -value = 0.001). This difference is also significant during the post-2007 time period (difference = -9.45%, p -value = 0.004). Taking these results together, it is evident that companies included in our analyses tend to support the government initiative to expand and strengthen accounting firm practices, regardless of state ownership. Moreover, NSOEs appear to be more aggressive regarding supporting this initiative, which is consistent with the statistics in Panel D of Table 1. Table 3 Di ff erence-In-Di ff erences Analysis on Reasons for Disclosing Auditor- Client Realignment POST_2007 Pre-2007 Post-2007 Difference p -value STATE SOEs 32.69% 51.59% 18.90% <0.001 NSOEs 17.54% 61.04% 43.50% <0.001 Difference 15.15% -9.45% -24.60% <0.001 p -value 0.001 0.004 With these findings in mind, we then conduct a univariate analysis, whereby we treat the managerial decision to disclose the reasons for realignment as the dependent variable in order to further analyze the joint effect of state ownership ( STATE ) and of the initiative ( POST_2007 ) on the managerial decision to disclose the realignment reasons. Referring to Table 4, the preliminary evidence obtained from these analyses shows that the decision to disclose the reason for auditor-client realignment may be driven by state ownership ( p -value = 0.088). The result is consistent with the findings reported from the difference- in-differences analysis shown in Table 3. However, the decision to make the disclosure of the realignment reasons could be influenced by the government initiative ( p -value < 0.001). Similar results presented in Table 3 are in line with the findings demonstrated in Table 4. In addition, we investigate the relationship between MAO and REASON_ DISCLOSURE . The results reveal that firms receiving MAO are less likely to disclose their reasons for realigning with auditors ( p -value = 0.003).