Abstract:

The purpose of this study is examine for Turkey for periods covering the years 2004-2019 according to the Classification of Statistical Territorial Units (IBBS 2) Level 2 of tax flexibility. For this purpose, Second Generation Unit Root Tests, Westternlund Cointegration and Dumitrescu Hurlin Panel Causality Tests are used. The results of the analysis show that the variables of economic growth and tax revenues have a long-term cointegration relationship. According to the panel DOLSMG test result, the elasticity coefficient for tax revenues across the panel was calculated as 0.17. According to this result for 26 regions across the country, 1% increase in economic growth creates an increase 0.17% in tax revenues in the long run. This elasticity value is of a very low quality and cannot fulfill the income generating feature of taxation. When we examine the tax flexibility according to NUTS-Level 2, among 26 regions the coefficient results of TRA2, TRC1, TRC3, TR22, TR32, TR41, TR63 and TR82 regions were over 1, positive and statistically significant. In Dumitrescu Hurlin panel causality analysis, there is a causality relationship from economic growth to tax revenues for the panel in general, and again it is seen that there is a causal relationship from tax revenues to economic growth. In other words, it can be stated that there is a bidirectional causality relationship between variables.

Keywords: