CARIFORUM and UK EPA Study
Dependent Variable: LOG(AVGTRD) Method: Panel Least Squares Date: 10/17/21 Time: 09:22 Sample: 2008 2018 Periods included: 11 Cross-sections included: 14 Total panel (balanced) observations: 154
Variable
Coefficient
Std. Error
t-Statistic Prob.
LOG(EXR)
0.014091 0.055762 0.252705 0.8009 0.002263 0.000983 2.301778 0.0228 -0.015442 0.012318 -1.253640 0.2120 0.005477 0.001698 3.224896 0.0016 0.009780 0.015840 0.617465 0.5379 0.004468 0.003071 1.455078 0.1478 1.068562 0.015265 70.00072 0.0000 0.005033 0.001698 2.964539 0.0035 3.803864 0.094424 40.28475 0.0000
GDPG
INFL
INT
LOG(K)
LFG
LOG(OP)
PVI
C
R-squared
0.976093 Mean dependent var 0.974774 S.D. dependent var 0.035141 Akaike info criterion 0.179055 Schwarz criterion 301.7736 Hannan-Quinn criter. 740.0341 Durbin-Watson stat
3.822518 0.221253 -3.802255 -3.624770 -3.730161 0.389830
Adjusted R-squared S.E. of regression Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.000000
As expected, liberalization (integration), openness, liquidity, GDP growth, and productive capacity significantly affect changes in the average volume of trade. It is noteworthy that liberalization can also affect productivity through technical transfers of knowledge. In many ways, openness affects human capital, which is consistently positive in the covariance matrix for all the countries in the sample. Therefore, multicollinearity can run through human capital, openness, capital formation, and the productive capacity index. This reality accounts for the near singular matrix or the unsurprisingly high variation that has been accounted for by the changes in the average volume of trade (98%). It is equally noteworthy that the various components of the index ambiguously contribute to the overall performance of the index. Some of the components of the index show a negative relationship to the composite index; meaning that they make no meaningful contribution to productive capacity (see the covariance matrix results below).
Page 226 of 241
Made with FlippingBook Learn more on our blog