Dynamic relationship between international tourism, economic growth and environmental pollution in the OECD countries: evidence from panel VAR model

Figures & data

Figure 1. Contribution of tourism to service exports, selected OECD countries, 2018.

Table 1. Variables definition.

Variable	Definition	Sources
GDP	GDP per capita (constant 2010 US$)	World Development Indicators
TOUR	International inbound tourists	World Development Indicators
CO2	Carbon dioxide emission (per capita, million tons)	BP Statistical Review
EC	Primary energy consumption (per capita, Exajoules)	BP Statistical Review
GFC	Gross Fixed Capital Formation (% of GDP)	World Development Indicators
TRD	Sum of exports and imports (% of GDP)	World Development Indicators

Source: Authors.

Table 2. Descriptive statistics.

Variables	Obs.	Mean	Std dev.	Min	Max
GDP	700	35,366.43	21,595.65	3,911.57	105,454.7
TOUR	700	25,900,000	35,300,000	211,000	183,000,000
CO2	700	9.136	4.818	1.209	27.489
EC	700	192.104	114.428	26.587	683.428
GFC	700	22.518	3.614	13.247	37.489
TRD	700	83.771	55.645	16.390	408.362

Source: Authors.

Table 3. Correlations and CD test results.

Correlations	LGDP	LTOUR	LCO2	LEC	LGFC	LTRD	CD test
LGDP	1.000						97.203 [0.000]
LTOUR	−0.163***	1.000					97.180 [0.000]
LCO2	0.715***	−0.009	1.000				96.596 [0.000]
LEC	0.845***	−0.204***	0.830***	1.000			97.173 [0.000]
LGFC	−0.038	−0.083	0.039	0.044	1.000		97.099 [0.000]
LTRD	0.229***	−0.271***	0.155***	0.208***	−0.111***	1.000	97.153 [0.000]

Note: ^*** and ^** indicate statistically significant correlation at 1% and 5% level. [.] is p-value.

Source: Authors.

Table 4. Panel unit root test results.

Level	Constant	Constant and trend
LGDP	−1.925	−2.373
LTOUR	−1.844	−2.089
LCO2	−1.813	−2.407
LEC	−1.221	−2.684
LGFC	−1.875	−2.098
LTRD	−1.871	−2.174
First differences	Constant	Constant and trend
LGDP	−3.315***	−3.482***
LTOUR	−4.035***	−4.118***
LCO2	−4.645***	−4.710***
LEC	−4.891***	−4.929***
LGFC	−3.909***	−4.103***
LTRD	−4.047***	−4.249***

Note: ^*** indicates stationarity at the 1% significance level.

Source: Authors.

Table 5. Lag length selection.

Lag	CD	J test	MBIC	MAIC	MHQIC
1	0.230	53.881 [0.028]	−170.131	−18.118	−77.747
2	0.293	29.759 [0.325]	−138.250	−24.240	−68.962
3	0.297	16.982 [0.524]	−95.023	−19.017	−48.831
4	0.759	13.689 [0.133]	−42.314	−4.310	−19.218

Note: [.] is p-value.

Source: Authors.

Table 6. Granger causality test results.

Variables	DLGDP	DLTOUR	DLCO2
DLGDP does not cause	–	0.088 [0.957]	3.026 [0.220]
DLTOUR does not cause	2.579 [0.275]	–	4.263 [0.119]
DLCO2 does not cause	0.402 [0.818]	2.965 [0.227]	–

Note: The figures in the square bracket show the probability to reject the null hypothesis. D indicates the first differences of the variables.

Source: Authors.

Figure 2. Impulse-responses analysis results. Note: The results show cumulative responses. Shaded areas are 95% confidence interval. Monte Carlo simulation with 1000 draws is used to obtain a confidence interval.

Source: Authors.

Figure 3. The responses of carbon emissions and GDP to a tourism shock.

Source: Authors.