Title:
Revisiting Regional Poverty in Indonesia: Geographic Location as a Key Discriminating Factor

Authors:
Roosemarina Anggraini Rambe, Merri Anitasari, Rara Juniar Varina, Monika Anisa Putri, Indonesia

Abstract:

The purpose of this study is to analyze whether location and unemployment are factors that distinguish provinces with low poverty and provinces with high poverty in Indonesia. The data used is secondary data. The study applied the discriminant analysis method to identify the difference between groups (low poverty provinces (as group 1) versus high poverty provinces (as group 2)) with three years of observation, namely 2016, 2020, and 2024. The independent variables consist of six: location, unemployment, government spending, economic growth, population, and Gini ratio.  The results show that group 1 differs significantly from group 2. This result is supported by the classification results from 2020, which show that the accuracy of cross-validated sample classification was 60.6% in 2016, 67.6% in 2020, and 67.6% in 2024. Based on these three discriminant models, discriminant loadings are considered appropriate to measure the discriminant power of each independent variable. For the 2016 discriminant model, location (Eastern Region of Indonesia) is the most discriminating variable, whereas unemployment (with the smallest coefficient) is the least discriminating variable. The most discriminating variable after 2016 is unemployment. The least discriminating variable in 2020 and 2024 is the Location dummy. Therefore, the recommendations to reduce the poverty rate in the eastern region of Indonesia are that the government implement a poverty alleviation program and create a conducive business climate to foster business development, thereby growing the economy.

DOI: https://doi.org/10.51505/IJEBMR.2026.10407