TURKISH ECONOMY IN THE BELT ROAD INITIATIVE: A GRAVITY MODEL FOR INTERNATIONAL TRADE

Abstract

In this paper, we attempt to analyze the determinants of international trade flows between Turkey and countries in the One Belt One Road (OBOR) project and their implications for the Turkish economy through the estimation of panel data using a gravity model. In particular, we employ the Poisson Pseudo Maximum Likelihood (PPML) estimation procedure for the nine OBOR nations from the period 2000-2021. This estimator is more robust than other methods such as pooled, fixed, and random models and it is free from the problems of autocorrelation, asymptotic bias, multicollinearity, and heteroscedasticty in the residuals for estimation and inference. First, panel unit root tests show that conducted variables are stationary at the conventional level. Second, the gravity estimations revealed that remoteness, scale, land endowment variables and, dummies (Preferential Trade Agreement and Border) in the gravity model have a positive impact, while GDP Distance and Capital Endowment have an inverse effect on trade flows. Only the language dummy is statistically insignificant. For policy purposes, we concluded that it would be better to promote trade flows with the OBOR countries.