Page 58 - Azerbaijan State University of Economics

Seda Ozekicioglu, Yilmaz Bayar: Tax revenues, corruption and governance in OECD
countries: a panel regression analysis

4.2. Results of Panel Unit Root Test

The integration levels of the variables were examined with CIPS unit root test of Pesaran (2007)
and the test results were displayed in Table 4. The results revealed that all the variables were not
stationary at their levels, but became stationary after first-differencing.

Table 4: Results of Panel CIPS Test

Test TAXREV COC GE RQ ROL
CIPS 5,902* 7,851* 7,426* 5,084* 6,077*
* it is significant at 5% significance level

4.3. Model Selection

Chow (1960) and Breusch-Pagan (BP) (1980) tests were employed to determine which
estimation method we use in the panel regression analysis Chow test is used for the
determination of common significance of country and time specific effects among the panel data.
The null hypothesis shows that pooled OLS is effective, while alternative hypothesis shows that
fixed effects model (FEM) is effective under the Chow test. BP test is used to determine whether
we use pooled OLS or random effects model (REM) and the null hypothesis shows that pooled
OLS is effective, while alternative hypothesis shows that REM is effective under BP test. Both
Chow and BP tests were implemented and the results were presented in Table 5. Chow test
dictated us to use FEM model, while BP test dictated us to use REM. So Hausman test will be
used to select between two models in the next stage.
Table 5: Results of the Estimation Method Test of Panel Regression

Test p value Decision

Chow (F) test 0,015 Accept
-
0
BP c test 0,002 Accept
-

Finally, Hausman test is used to make a selection between FEM and REM. The null hypothesis
asserts that REM is efficient, while the alternative hypothesis shows that FEM is effective. We
applied Hausman test and the results were presented in Table 6. So the null hypothesis was
accepted and therefore we decided to use REM model considering the results of Hausman test.

58