How does quality of institutions affect the impact of trade openness on economic growth of Ghana?

Figures & data

Table 1. Unit root test results

	ADF test				P-P test
	Levels		First Difference		Levels		First difference
Variable	No Trend	Trend	No Trend	Trend	No Trend	Trend	No Trend	Trend
lnY	−3.7001***	−3.8414**	−6.9448***	−6.8257***	−3.6254**	−3.6894**	−12.7584***	−11.7062***
lnTO	−3.6019**	−2.6797	−4.9506***	−5.9471***	−3.6019**	−2.6797	−4.9206***	−5.6526***
QI	−3.4066**	−3.3024	−5.0101***	−5.3266***	−2.3479	−2.2333	−6.4976***	−7.5649***
lnFD	−2.3283	−1.8017	−6.5514***	−6.8753***	−2.3845	−1.7246	−6.6027***	−7.5788***
lnER	−3.9347***	−2.1758	−3.5409**	−4.2132**	−3.5930**	−2.1428	−3.5182**	−4.1309**
lnINF	−3.4689**	−4.3075***	−5.0388***	−4.9366***	−3.5461**	−4.1460**	−13.1187***	−15.4397***
lnK	−2.4789	−2.1356	−4.0623***	−3.5071	−3.5461**	−4.1460**	−13.1187***	−15.4397***
KPSS test
Variable	Levels					First difference
	Trend					Trend
lnY	0.1380***					0.2659
lnTO	0.2008***					0.0825***
QI	0.1442***					0.3798
lnFD	0.1912***					0.0897***
lnER	0.2034***					0.0772***
lnINF	0.0971***					0.5000
lnK	0.1537***					0.0577***

*** and ** in both ADF and P-P tests (KPSS test) denote rejection (non-rejection) of the null hypothesis at 1 and 5 percent significance level respectively.

Source: Authors’ estimation

Table 2. Cointegration test results

Model	Test statistic	Lower bound critical value	Upper bond critical value
Model 1	6.8460***	3.15	4.43
Model 2	6.2682***	2.96	4.26

***Denotes rejection of the null hypothesis of no long run relation among the sample variables at a 1 percent significance level; Model 1 is the estimation without interaction term and Model 2 is the estimation with interaction between trade openness and quality of institutions.

Source: Authors’ estimation.

Table 3. Estimated long run results

	Model 1	Model 2
Variable	Coefficient	Coefficient
lnTO	0.6017** (0.2389)	0.5627** (0.2457)
QI	0.9695* (0.5024)	1.6229* (0.8162)
lnFD	−0.4420 (0.2601)	−0.4047 (0.2674)
lnER	0.1546** (0.0695)	0.2196** (0.0953)
lnINF	−0.0977 (0.1159)	−0.0929 (0.1191)
lnK	−1.0239*** (0.2005)	−1.0682*** (0.2116)
lnTO*QI		−1.0988 (1.0318)
Constant	3.8043*** (0.8526)	3.9291*** (0.8811)

***, ** and * represent 1, 5 and 10 percent significance level respectively; standard errors are in the parentheses; Model 1 is the estimation without interaction term and Model 2 is the estimation with interaction between trade openness and quality of institutions.

Source: Authors’ estimation.

Table 4. Estimated short run results

	Model 1	Model 2
Variable	Coefficient	Coefficient
∆lnTO	0.5967** (0.2256)	0.5419** (0.2298)
∆QI	0.9615* (0.4843)	1.5629** (0.7246)
∆lnFD	−0.4384* (0.2486)	−0.3897 (0.2512)
∆lnER	−0.8866** (0.3222)	−0.8359** (0.3238)
∆lnINF	−0.0968 (0.1199)	−0.0895 (0.1195)
∆lnK	−0.0547 (0.2351)	−0.0533 (0.2340)
∆lnTO*QI		−1.0582 (0.9521)
ECM(−1)	−0.9918*** (0.1476)	−0.9630*** (0.1492)
R-square	0.7235	0.7376
Adj. R-square	0.6199	0.6236
DW-statistic	2.3991	2.3014
F-statistic	6.9793	6.4665
Prob. (F-statistic)	0.0001	0.0001

***, ** and * represent 1, 5 and 10 percent significance level respectively; Standard errors are in the parentheses; Model 1 is the estimation without interaction term whereas Model 2 is the estimation with interaction between trade openness and quality of institutions.

Source: Authors’ estimation.

Table 5. Diagnostic test results

	Model 1	Model 2
Diagnostic test	Test statistic	Test statistic
Serial correlation	2.5205 (0.1034)	1.0722 (0.3603)
Heteroskedasticity	1.0133 (0.4570)	1.3008 (0.2872)
Normality	0.4480 (0.7993)	0.2514 (0.8819)
Functional form	0.3096 (0.5833)	0.4416 (0.5132)
CUSUM	Stable	Stable
CUSUMSQ	Stable	Stable

Probability values are in the parentheses.

Source: Authors’ estimation.

Figure A1. Stability test for the model without interaction term (Model 1).

Source: Authors’ construction

Figure A2. Stability test for the model with interaction term (Model 2).

Source: Author’s construction