BETA REGRESSION MODELING ON POVERTY DATA IN INDONESIA 2019 - 2022

Muhammad Arib Alwansyah (Department of Mathematics Education, Faculty of Mathematics and Natural Sciences, Universitas Negeri Jakarta, Indonesia)
Sigit Nugroho (Department of Mathematics, Faculty of Mathematics and Natural Sciences, Universitas Bengkulu, Indonesia)
Ramya Rachmawati (Department of Mathematics, Faculty of Mathematics and Natural Sciences, Universitas Bengkulu, Indonesia)

BAREKENG: Jurnal Ilmu Matematika dan Terapan
Vol 20 No 3 (2026): BAREKENG: Journal of Mathematics and Its Application

The Central Statistics Agency (BPS) reported that the percentage of poor people in Indonesia increased from 2019 to 2021, reaching 10.14 percent. This condition highlights the need for an analytical approach capable of accurately modeling percentage data that are naturally bounded between 0 and 1. This study introduces a new approach by applying the Beta regression model to analyze the factors influencing poverty levels across Indonesian provinces. The novelty of this research lies in the application of the Beta regression model to panel data on poverty, which remains rarely explored in empirical studies on Indonesia’s socio-economic indicators. The model was chosen because it provides more efficient and unbiased parameter estimates than the ordinary least squares (OLS) method, especially when the dependent variable exhibits asymmetry and heteroskedasticity. Parameter estimation was conducted using the Maximum Likelihood Estimation (MLE) method with the Newton–Raphson iterative algorithm to ensure convergence and estimation efficiency. The data used in this study are provincial-level poverty data sourced from official publications by the BPS. The analysis results indicate that the model meets the model suitability criteria for 2019 and 2020. Factors that significantly influenced the percentage of poor people in both years included the percentage of the population with health insurance and the literacy rate. Meanwhile, in 2021 and 2022, factors that significantly influenced the percentage of the poor population included the average years of schooling, the percentage of the population with health insurance, and the literacy rate. This study contributes to the field of applied statistics by demonstrating that the Beta regression model offers a novel and robust alternative for analyzing bounded and asymmetric socio-economic data. Furthermore, it provides new empirical insights into the statistical modeling of poverty in Indonesia, offering a methodological advancement over traditional regression approaches.