*Appendix A.2. Estimated Coefficients*

**Table A2.** Estimated coefficients by FMOLS and OLS.


Note: \*, \*\* and \*\*\* for statistically significant at the 0.10, 0.05 and 0.01 levels, respectively. Source: FMOLS and OLS estimation results for data panel in Eview.

Table A2 above shows that adjusted R squared for all estimators is quite high, except Pooled OLS. Significance in F-statistics of Pooled OLS, FEM and FEM are all at 0.01 level. The Redundant Fixed Effects test in the FEM estimate gives significance in the Cross-section F statistic at the 0.01 level, so it allows a strong rejection the null hypothesis that the effects are redundant and shows that the cross-section fixed effects are statistically significant. This means that FEM is more reliable than Pooled OLS estimation. The Chi-square statistical significance of Cross-section random in Hausman Test does not reach 0.05 level, and the Null hypothesis cannot be rejected, so the REM model is selected. These results show that both the FMOLS and REM estimators are reliable.
