Application of panel quantile regression and gravity models in exploring the determinants of turkish automotive export industry

Abstract

This paper purposes to determine potential factors influencing the amount of Turkish automotive industry exports. For this purpose, the available data of 68 major trading partners of Turkey in terms of automotive industry exports were utilized for the sample period 2007-2015. Both panel quantile regression and the gravity model of trade approaches were considered to analyze the relevant data. The empirical findings of this paper revealed that the population and the distance variables were found as statistically significant for all quantiles, while the former has positive and the latter has negative signs as expected. Results also indicated that there was a statistically significant positive correlation between GDP per capita and the amount of Turkish automotive industry exports at 10 and 50% quantiles; however, it was not statistically significant at 90% quantile despite its positive sign. Among Turkey’s exporter countries, being a EU member country dummy variable was found to have a statistically significant positive impact on the amount of automotive industry exports. Real exchange rate was not found as a significant determinant of the amount of automotive industry exports. In the lights of empirical evidence obtained from this study, several recommendations were made for Turkey’s future international trade policies.