CARIFORUM and UK EPA Study

VAR Granger Causality/Block Exogeneity Wald Tests Date: 09/23/21 Time: 13:09 Sample: 1 182 Included observations: 180

Dependent variable: LOG(AVGTRD)

Excluded

Chi-sq

df

Prob.

LOG(EXR)

0.297837 6.551721 1.151780 3.052384 1.648305 35.60186 3.085179 1.913199

2 2 2 2 2 2 2 2

0.8616 0.0378 0.5622 0.2174 0.4386 0.0000 0.2138 0.3842

GDPG

INFL

INT

LOG(K)

LFG

LOG(OP)

LOG(EXP01)

All

62.97457

16

0.0000

The income growth -export specification shows a rather weak correlation; partly because aggregate spending on foreign goods or imports are ignored. Net exports or the volume of trade serve as better alternative measurements of trade. The table below shows a weak correlation between the variables and the growth rate of GDP. Of course, the growth rate can be a biased option (without generality) because of its weak performance in the region (see Figure 2). We fail to reject the null that the variables do not Granger-Cause GDP growth.

Dependent variable: GDPG

Excluded

Chi-sq

df

Prob.

LOG(AVGTRD)

2.417598 2.607051 3.569110 1.332679 0.350021 0.504338 2.486892 2.004385

2 2 2 2 2 2 2 2

0.2986 0.2716 0.1679 0.5136 0.8394 0.7771 0.2884 0.3671

LOG(EXR)

INFL

INT

LOG(K)

LFG

LOG(OP)

LOG(EXP01)

All

14.58538

16

0.5552

The variables show a very strong interaction with the general price level or GDP deflator. We fail to reject the null that they do not collectively correlate with inflation. Exports, openness, the growth

Page 215 of 241