Abstract
The objective of this study is to re-analyze the relationship between natural gas consumption (NGC) and economic growth (GR) for Turkey in a multivariate framework by including capital and labor as additional variables because several papers suggest that a bivariate model can suffer from omitted-variables bias. As compared to the findings of Isik (2010), who previously investigated the short- and long-run relationships between GR and NGC using a bivariate model, we find that the magnitude of the coefficient estimate of NGC become substantially smaller in the long-run and the sign of short-run estimate of NGC shift to negative after accounting for capital and labor as well. In addition to that covered by Isik (2010), we investigate the direction of causality between GR and NGC using the vector error correction model Granger causality approach, and reveal the evidence of feedback hypothesis for Turkey.
Notes
1 It is commonly accepted that countries wish to have higher rates of economic growth.
2 The data are available at http://data.worldbank.org.
3 The Microfit 5.0 strtistical package is used to obtaion the estimated coefficients in ther ARDL approach. For a detailed explanation about normalization, please see the MIFT5.0 manual prepared by pesaranand pesaran (Citation2009).
4 Note that the null hypothesis of including a unit root can be rejected when the absolute value of the ADF-test statistic is greater than the absolute value of the critical point.
5 The straight lines stand for the critical bounds at the 5% significance level.