STRUCTURAL EQUATION MODELING ANALYSIS ON POVERTY IN WEST KALIMANTAN WITH FINITE MIXTURE IN PARTIAL LEAST SQUARE APPROACH

Abstract

Poverty occurs when individuals or groups lack the necessary resources to fulfill their basic needs. In Indonesia, including West Kalimantan, poverty remains a significant issue influenced by various socio-economic factors. This study aims to identify valid and reliable indicators of poverty and classify regencies/cities in West Kalimantan using the 2023 data from the Central Statistics Agency of West Kalimantan and Indonesia. The analysis applies the Structural Equation Modeling approach with Finite Mixture in Partial Least Squares (FIMIX-PLS). From 19 observed indicators, only 12 were found valid and reliable based on measurement and structural model evaluation. The structural model reveals three significant relationships: the Economy significantly influences Poverty, Health influences Education, and Education influences the Economy. Based on the FIMIX-PLS results, the regencies/cities are segmented into four groups with distinct structural characteristics. Segment 1 reflects the influence of Health on Education, Segment 2 reflects the influence of Health on the Economy, Segment 3 highlights the influence of Economy on Poverty, and Segment 4 captures the influence of Education on the Economy. Detailed interpretations of each segment and their policy implications are presented in the conclusion. The results support the importance of tailored poverty alleviation strategies based on latent regional characteristics and validated model findings.