Abstract:

In this study in Turkey in the period 2004:M01-2016:M12 according to the statistical regional units in Turkey, the presence of inflation approach between regions for 26 regions located in Level 2 was analyzed with the help of panel data analyses, with absolute beta approach, conditional beta approach, sigma approach and variation ratio approaches. The stagnation of the series; LLC, IPS and Hadri panel units were tested with root tests and the series were found to be I(1) . The presence of causality relationships between the series; Granger and Dumitrescu and Hurlin panel have been studied by causality tests. The existence of the matching relationship between the series included in the Models has been tested by the Kao panel matching test. The combination rates in the models are predicted by the PDOLS method. The results of the analysis have shown that in Turkey in the period 2004-2016 the absolute proximity hypotheses on inflation are not valid, but the conditional proximity hypotheses are valid. Analysis for sub-periods determined that the conditional proximity hypothesis was valid in the period 2004-2007 and that it was not valid in the sub-periods. According to the sigma approach and variation ratio analysis, it has also been shown that in the period 2004-2016 in Turkey there was no inflation proximity between the regions, but a low-speed inflation rate. It has also been found that the area concerned is located in the western part of the country, the average temperature, the amount of rainfall, the time of stay of the snow on the ground and the prices of oil, the one-of-the-one determiner of inflation and these variables have an impact on the inflation rate/approximation.

Keywords: