Unemployment is a problem that occurs in many countries and often gets special attention both from policymakers and academics. This fact is because if not addressed, it will cause socio-economic problems in the country. Therefore it is necessary to formulate the causes of unemployment by involving spatial aspects to avoid biased and inefficient estimates. This study aims to find the determinants of unemployment rates in Indonesia, including calculating the direct and indirect effect of using the spatial Durbin models (SDM) in the period 2000-2017. The results of this study indicate that the overall independent variables used significantly influence the unemployment rate in Indonesia. Besides, it turns out that the higher education variable completed by the population of a region has the most significant impact both in decreasing unemployment in a region and neighboring regions. Therefore, the policy taken should pay attention to this.
The ecological fallacy in a time series context: evidence from Spanish regional unemployment rates
