Nadia Savitri1, Maria B. Theresia2, Rahma Fitriani2

Abstract: Unemployment is a condition in which a person is in the labor force but does not work and is still looking for a job. Unemployment data in East Java can be accessed as a data panel. This data is the result of a combination of cross-section data and time series data. Will the data be observed at intervals once, over time will have correlated or interdependent data. Data from year to year, then between data also correlate each other. If the response variable states many unemployed in the city or district in East Java in 2006 until 2015 the spreads that belong to the family are exponential and contain autocorrelation. Verbekke and Molenberghs (2005) for data models containing autocorrelation and response variables did not spread normally. This study aims to analyze the data panels in East Java with the Generalized Integrated Mix Model, forecasting the number of unemployed and identification of the predictor variables on the number of unemployed.

In the Generalized Linear Merger Model. Estimation of model parameters using Maximum Likelihood (ML) method for estimation of fixed effect and. For estimation of random effects. In GLMM modeling, many labor force, large MSEs, economic growth, GRDP, inflation and region are. Various labor force, MSEs, economic growth, GDP and inflation over many unemployed in 38 districts and cities in East Java. Based on research data, it can be concluded that GLMM can be used as an approach to model unemployment panel data in East Java.

Keywords: panel data, GLMM, unemployment in East Java

Full Text: PDF