Role of energy consumption, tourism and economic growth in carbon emission: evidence from Kuwait

Figures & data

Table 1. Descriptive statistics with the correlation matrix

Dependent Variable	Mean	Median	Maximum	Minimum	Jarque-Bera	Probability	LNC	LNEN	LNG	LNR
LNC	4.228	4.34	4.642	3.51	3.0169	0.2213	1
LNEN	0.104	0.19	0.546	−0.62	2.6593	0.2646	0.9987	1
LNG	25.07	25.4	25.88	23.98	3.018	0.2211	0.9355	0.992	1
LNR	20.08	20.1	20.91	19.47	0.8988	0.638	0.7984	0.8119	0.8068	1

Table 2. Unit root test by augmented dickey fuller & Phillips Perron

		ADF		PP
		t-Statistics	Prob.	Adj. t-stat.	Prob.
LNC	I(0)	−3.01	0.051	−2.23	0.2007
	I(1)	−0.38	0.89	−3.374	0.02
	I(2)	−6.34	0.00
LNE	I(0)	−2.78	0.08	−2.51	0.13
	I(1)	0.274	0.97	−3.77	0.009
	I(2)	−7.25	0.00
LNG	I(0)	−1.53	0.5	−1.51	0.51
	I(1)	−3.48	0.02	−3.45	0.02
LNR	I(0)	−0.64	0.84	−2.05	0.26
	I(1)	−4.38	0.003	−7.49	0

Table 3. Johansen cointegration test for CARBON EMISSION as a dependent variable

S	Eigenvalue	Trace Statistics	0.05 Critical Value	Prob.*	Max-Eigen Statistics	0.05 Critical Value	Prob.*	Decision
None*	0.6027	53.951	47.856	0.012	22.152	27.584	0.213	Long Run VECM
At most 1	0.5365	31.799	29.797	0.029	18.454	21.132	0.114
At most 2	0.3021	13.344	15.495	0.103	8.631	14.265	0.318
At most 3	0.1783	4.7135	3.8415	0.029	4.713	3.8415	0.029
LNC	LNEN		LNG			LNR
1	−1.0208		−0.03796			0.133175
	0.0267		0.031			0.01988

Table 4. Granger causality pairwise tests

Observation	Null Hypothesis:	F-Statistic	Probability	Decision
25	LNEN does not Granger Cause LNC	0.63371	0.4345	EN≠C
	LNC does not Granger Cause LNEN	0.63244	0.435	C≠EN
25	LNG does not Granger Cause LNC	0.14768	0.7044	G≠C
	LNC does not Granger Cause LNG	3.68136	0.0681	C≠G
25	LNR does not Granger Cause LNC	2.0463	0.1666	R≠C
	LNC does not Granger Cause LNR	5.51336	0.0283	C→R
25	LNG does not Granger Cause LNEN	0.00019	0.989	G≠EN
	LNEN does not Granger Cause LNG	2.43806	0.1327	EN≠G
25	LNR does not Granger Cause LNEN	1.01795	0.324	R≠EN
	LNEN does not Granger Cause LNR	5.43553	0.0293	EN→R
25	LNR does not Granger Cause LNG	2.64176	0.1183	R≠G
	LNG does not Granger Cause LNR	9.94819	0.0046	G→R

Figure 1. Impulse response function.

Table 5. Variance decomposition of carbon emissions

Period	LNC	LNEN	LNG	LNR
1	100	0	0	0
2	90.1885	3.5639	4.597	1.65
3	77.957	5.04	10.315	7.287
4	67.1049	5.482	13.9117	13.502
5	62.2601	5.217	15.9627	16.56
6	60.3187	4.916	17.1767	17.588
7	59.3047	4.742	17.9103	18.04
8	58.509	4.66	18.388	18.44
9	57.8415	4.607	18.745	18.81
10	57.3127	4.558	19.037	19.092

Figure 2. Plot of CUSUM and CUSUM square test.

Table 6. Summary of diagnostics test

Variable	Breusch-Godfrey Serial Correlation LM Test			Inference
LNC	F-statistics	1.0488	0.3201	The null hypothesis of serial correlation could not be rejected.
	Obs*R-sq	1.3946	0.2376
	Breusch-Pagan-Godfrey Heteroskedasticity Test			Inference
LNC	F-statistics	0.6592	0.7187	The null hypothesis of no heteroskedasticity could not be rejected.
	Obs*R-sq	6.2427	0.6201
	Jarque-Bera			Inference
LNC	5.6228		0.06	The null hypothesis of normal distribution could not be rejected.

Data availability statement

The data will be available on the World Bank data website at https://data.worldbank.org/