Modeling Illiteracy Rate in Indonesia with Spatial Regression

Abstract

The illiteracy rate (ILR) serves as an indication of educational attainment, representing the percentage of individuals aged 15 and older who lack reading and writing skills. Despite a reduction in Indonesia's ILR to 3.33% in 2024, the objective of eliminating it entirely remains a priority to fulfill the fourth aim of The Sustainable Development Goals (SDGs) by 2030. This research seeks to estimate Indonesia's rate of illiteracy by examining relevant elements, including the number of individuals living in poverty, mean years of schooling, and gross enrollment ratio. The data is obtained from the BPS-Statistics Indonesia. This study employs spatial regression, utilizing an area-based methodology to capture spatial impacts among regions and analyze them with R software. The analysis results indicate that a) the chosen weighted matrix is k-Nearest Neighbor, b) the selected spatial model for the illiteracy rate in Indonesia is the Spatial Durbin Model (SDM), and c) mean years of schooling and gross enrollment ratio within a province significantly affect the illiteracy rate in that province, which may indirectly elevate the illiteracy rate in neighboring regions.