Özet:

Export is one of the important parameters that show the economic power of a country. Each country exports according to its product range. Countries that are strong in the industrial sector export industrial products, countries that are rich in oil or underground resources export petroleum, and countries that are rich in agricultural products export agricultural products. In this study, the effects of Turkey's raw agricultural product exports and exports of products and services on the economic growth were analyzed. However, correct model selection is important in determining such relationships and making healthier predictions. For this reason, the data for Turkey between the years 1988-2018 were obtained from the World Bank. Linear and quantile regression were used as methods, and Kolmogorov-Smirnov and Shapiro-Wilk tests were used to test whether the data showed normal distribution. Using the bootstrap method, 25%, 50% and 100% bootstrap was applied to the data. In the implementations, MAD and RMSE were used to select the most suitable model among linear regression, quantile regression models. It was observed that quantile regression gave better results in the raw data and 25% bootstrap application, and linear regression in the other two applications. Thus, it is seen that OLS gives better estimates for normal distributions and quantile regression for non-normal distributions. In addition, it has been determined that the export of raw agricultural products has a significant effect on the Turkish economy. For this reason, it is of great importance not only for the sector but also for the Turkish economy to take into account the external demand as well as the domestic demand in agricultural production, and to take steps to urgently solve the existing problems by increasing efficiency.

Anahtar Kelimeler: