Dynamic inter-relationships among tourism, economic growth and energy consumption in India

Abstract

This study attempts to explore the dynamic causal and inter-relationships among tourism, economic growth and energy consumption in India. This study covers the annual data from 1971 to 2012. This study applies the cointegration and generalised variance decomposition methods to verify the relationship. The bounds testing approach to cointegration and the Gregory–Hansen test for cointegration with structural break consistently reveal that energy consumption, tourism and economic growth in India are cointegrated. We find that tourism and economic growth strongly affects energy consumption in the long-run. Additionally, we also find that tourism and economic growth in India are inter-related, but the causal effect of tourism on economic growth is stronger than the other way around in both the short- and long-run. Therefore, this study concludes that the tourism-led growth hypothesis is valid but the energy-led growth hypothesis is invalid in India. With such findings, we can confirm that tourism is an important catalyst of growth to the Indian economy. Therefore, policy-makers should promote and expand tourism industry in order to sustain the process of economic growth and development in India.

Keywords:

• Cointegration
• economic growth
• energy consumption
• tourism
• variance decomposition

Acknowledgement

The authors would like to thank the three anonymous reviewers for their constructive comments and suggestions to improve the earlier draft of this research. The usual disclaimer applies.

Notes

1. Apart from the issue of insignificance of tourism variable, they also found that any model that include tourism variable tends not to pass the diagnostic tests, particularly the stability test.

2. One may doubt about the use of level VAR to assess the Granger-causal chain among the variables, especially when the variables are cointegrated. Engle and Granger (1987) narrated that the VAR system at level can be used if the variables are cointegrated because the long-run constraints will be satisfied asymptotically. Fanchon and Wendel (1992) added that forecasting model with VAR in level yield the best forecast compared to Bayesian VAR and vector error-correction model (VECM). As the generalised VD method is to assess the dynamic inter-relationships among energy consumption, tourism and economic growth but not to obtain the parameters, the use of VAR in level remain appropriate and should not be worried, especially when the variables are cointegrated as the case of the present study. Based upon our reading, we notice that many of the earlier studies also used VAR in level model for cointegrated variables (e.g., Ibrahim, 2005; Ramaswamy & Slok, 1998).