The role of spatial factors in determining unemployment in Turkey

Unemployment is a serious problem that must be prevented for an economy at regional as well as national level. The unemployment seen at different levels, as a result of affected by economic and social aspects, also varies at the level of regions and provinces. Therefore, regional clusters and decompositions are observed between provinces. Given the situation, it reveals the necessity of examining unemployment at the provincial level as important at the regional level and states the importance of determining factors effective on unemployment. It is aimed to examine the regional clustering patterns of the unemployment rate by province and province grouping and to determine the parameters that have an effect on the unemployment rate. In the study, eighty-one provinces of Turkey were analyzed within the scope of Level-3, using 2020 cross-sectional data. In this context, the spatial econometric method was used in the study. In the model applied, there are male labor force participation rate, female labor force participation rate, the share of employment in the agricultural sector, educated population ratios, vocational training, innovation, immigration and crime variables as independent variables. According to the analysis findings, the existence of a spatial relationship between eighty-one provinces was determined. In the study, the spatial distribution pattern of the explanatory variable was examined and it was observed that the unemployment rate showed dense clustering in Eastern Anatolia and Southeastern Anatolia regions of Turkey. In this context, the spatial delay model (SAR) was adopted according to the test results. As a result of the spatial regression in which the explanatory variables in the model were examined; statistically significant and negative correlations were found between the male labor force participation rate, female labor force participation rate and the share of employment in the agricultural sector in the direction of reducing unemployment. The vocational training variable was obtained as statistically significant and positive in relation. The trained population ratio has not been achieved in reducing the unemployment of innovation, migration and crime variable.