Spatial Regression Approach in Determining Factors Affecting Percentage of Human Trafficking Victims inWest Java.

Abstract

The spatial dimension plays a key role in many social phenomenons such as poverty, pollution, disease, crime and others. Social phenomenon in certain areas is influenced by the social phenomenon in other locations. One of the interesting social phenomenons that will be studied in this research is human trafficking in West Java. The aims of this study are to identify the spatial effects on percentage of human trafficking victims, and to determine the factors which affected percentage of human trafficking victims in West Java in 2010. The existence of spatial relationship information between district/municipality led to need to accommodate the spatial variability in the model, so that the model used were spatial regression models. Based on the analysis, spatial effect influential significantly the percentage of human trafficking victims. The hotspot for human trafficking victims were Cimahi Municipality, Bandung Municipality and Bandung District. The best model to accommodate spatial effect in this study was General Spatial Model (GSM) with AIC-value=-361.46 and R2a=71.34%. Based on GSM obtained the indepen-dent variables or factors which significantly affected percentage of human trafficking victims in West Java in 2010 for α=0.05 as follows: the average of population with divorce status, the average of population with local migrant status, and the average of population with maximum education in Junior High School.