Export Performance of Transition Economies

We focus on recent differences in export performance of transition countries and use recent research on trade facilitation policies and new data on export costs to examine the effect of domestic export costs and external costs to markets on the exports of transition countries. At the same time, productive capacity and domestic demand along with potential market demand are also included in our study. Our focus is on growth in exports rather than growth in GDP, but we are partially motivated by the relationship between the two. Our analysis is going to focus on specific group of transition countries: Former Soviet Union and Central and Eastern Europe countries. In our study we will try to clarify the reasons why the differences continue to exist and what are the main determinants of these variations. Along with traditional export growth determinants we will try to analyze the impact of the specific factors which are related to our sample of transitions. Specific objectives are: •To describe the differences in economic and export growths among the various transition economies in terms of natural resource endowments. •To identify the issues that cause differences in growth of exports in transition countries •To build an econometric model that estimates the impact of various trade costs on the exports of transitions •To analyze the impact of location factor, economic unions and trade agreements on the growth of exports in transition economies. Our empirical model includes a panel of the 28 transition countries for the years from 2005 through 2011. The data period goes back only to 2005 based on the availability of the Export Cost variable. We used Generalized Method of Moments (GMM) and GMM Instrumental Variables techniques for the estimation. In our estimation we use the instrument variable GMM estimation technique based on Hansen (2000, Ch. 11), Hayashi (2000, Ch. 3) and Wooldridge (2002, Chapter 8) We chose GMM based on advantages of this approach for our model. First, GMM generates efficient estimates in the presence of heteroskedasticity of unknown form. This is the case in our panel data. After applying Breusch-Pagan / Gook-Weisberg tests we observed heteroskedasticity in our data-set. Even though standard IV estimators with the robust standard errors can be consistent, they are relatively inefficient. Second, efficient GMM has an advantage of consistency in the presence of arbitrary heteroskedasticity. Both econometrically and theoretically our instrument is meeting both exogeneity and relevance conditions and was used in the previous studies (Djankov (2010) Both in the base model and as we add other determinants, based on the previous theory and theoretical relevance, estimated signs and significance levels of the important variables is expected not to change considerably. This can be considered as a type of robustness check for the empirical estimation. Expected positive and significant sign of GDP and Natural Resource variables also indicate that the transition countries with larger economies and greater endowments of natural resources tend to export more. The population variable is expected to be negative and in two out of six specifications (in preliminary results) highly significant. This is consistent with our theoretical foundation, as we were expecting negative relation between the domestic demand and the growth of exports. demand in the major exporting market stays significant as we add other control variables to our base model. Positive and significant parameter on the GDP of major export partners indicates the importance of the demand factor in the main markets. Coefficients for distance from major markets are negative, significant, and of similar magnitude across all specifications reflecting the export depressing impact of higher transportation costs. Highly significant parameter of this variable throughout the specifications indicates the robustness and importance of transportation cost for the export performance in transition countries. Lagged Foreign Direct Investment (FDI) was included in five of the specifications and the coefficients were positive, significant, and of approximately the same magnitude across all of them. Consistently with expectations, higher FDI inflows promote exports of countries through utilization of low-cost human capital and natural resources. As we add business climate and later internal export cost variables into our model all the previous determinants keep their signs and significance levels, with an exception of population variable which gets insignificant. Measure of economic competitiveness was significant and positive in the specifications. The last specifications include dummies of being a member of EU and Central European Free Trade Agreement. Being an EU Member state is estimated to have a positive and significant impact on the export growth in transition countries. On the other hand, Central European Free Trade Agreement estimated to have negative and significant impact on the exports, which is primarily because of the fact that the members of CEFTA trade organization are comparatively less developed (in terms of GDP and exports) representatives of Central and Eastern Europe. As these countries expand their economy and meet the requirements of EU membership, they become no more the members of CEFTA.