Factors Influencing open unemployment rates: a spatial regression analysis

Abstract

The present study employed spatial regression analysis as a methodological approach to get insights into the unemployment rates across Indonesian provinces in the year 2016. The official website of the Bureau of Labor Statistics (BPS) offers secondary data pertaining to several socio-economic indicators, including the Total Open Unemployment Rate, Economic Growth Rate, Human Development Index, Severity of Poverty Index, and School Participation Rates. The investigation employed the Geoda software package and encompassed Ordinary Least Squares (OLS) regression, Dependency/Correlation investigation, and Spatial Autoregressive Model. The data presented in the study revealed the existence of three distinct provincial groupings characterized by varying levels of unemployment rates. In the context of unemployment variance, the traditional regression model accounted for 30 percent of the observed variation. However, the spatial regression model used spatial dependencies to enhance accuracy in capturing the phenomenon. The aforementioned findings have the potential to assist policymakers in formulating strategies to address unemployment in regions characterized by distinct spatial attributes, hence offering a potential blueprint for other nations.