Page 55 - Azerbaijan State University of Economics
P. 55
Javid Seyfullali: Tax Revenue and Economic Growth in Resource-Rich Country:
Empirical Evidence from Azerbaijan
Table 4 presents the computed F-statistics from F-bounds tests. Instead of critical
values compiled by Pesaran (2001), Narayan (2005) critical values are used as
Narayan (2004) argues that Pesaran (2001) critical values are not suitable for small
sample sizes. Rejecting the null hypothesis means that there is a cointegration
relationship between independent variables.
Table 4: F-bounds test results (Null hypothesis: No level relationship)
Critical value Test statistic Significance I(0) I(1)
source value
Model 1
Pesaran (2001) F-statistic = 1% 3.29 4.37
Narayan (2005) 27.41 1% 3.37 4.72
Model 2
Pesaran (2001) F-statistic = 1% 3.29 4.37
Narayan (2004) 14.22 1% 3.37 4.72
Source: Compiled by author based on Eviews calculations
As you can see from the table above, all series are cointegrated, thus the long-run
coefficients can be estimated. The estimated long-run coefficients are presented in
Table 5:
Table 5: Long-run estimate results and robustness check results
Panel A: Long-run estimate results
Variables Model 1 Model 2
Dependent variable Real GDP growth Real non-oil GDP growth
-0.67** -0.58*
Constant
(-2.45) (-1.69)
-0.10*** -0.16***
LRGOVREV
(-2.97) (-4.31)
0.09*** 0.13***
LRCAPINV
(3.97) (2.65)
0.10*** 0.11**
LRTOTTRADE
(2.72) (2.65)
0.32*** 0.33***
OILPRCG
(3.91) (3.02)
Panel B: Robustness check results
Model 1 Model 2
R-squared 0.71 0.72
Adjusted R-squared 0.66 0.68
Observations 69 69
2
2
Serial correlation (1 lag) X SC = 0.42 (0.52) X SC = 1.95 (0.16)
2
2
Serial correlation (4 lags) X SC = 1.12 (0.89) X SC = 3.83 (0.43)
Normality JBN = 71.15 (0.00) JBN = 0.07 (0.97)
*** **
*
Notes: , and denote statistical significance at 1%, 5% and 10% level, respectively. Numbers
in brackets in Panel A represent t-Statistics. Numbers in brackets in Panel B represent P-values.
Source: Compiled by author based on Eviews calculations
54