Modeling Poverty Rates in Indonesia Using Spline Nonparametric Regression

Abstract

Poverty is a complex and multidimensional issue that remains a major focus in Indonesia as a developing country. Data show that the poverty rate in March 2024 reached 9.03%, exceeding the government’s target of 6.5–7.5%. This study aims to analyze the factors that influence poverty levels in Indonesia by using secondary data obtained from the Central Statistics Agency (BPS) in 2023. The analytical method employed is Spline Nonparametric Regression, which is considered appropriate for processing social and economic data that tend to be sparse and non-stationary, such as Indonesia’s poverty data. The results of this study indicate that among the four variables analyzed, only the Mean Years of Schooling has a significant effect on the percentage of the poor population in Indonesia. Other variables, namely GDP growth at constant prices, the open unemployment rate, and life expectancy, were not proven to have a significant effect on provincial poverty levels. The best model obtained uses Spline Nonparametric Regression with two knot points, as it has the smallest GCV value compared to the one-knot and three-knot models. Therefore, the findings of this study are expected to provide input for the government in formulating more targeted poverty alleviation policies, particularly through improvements in education.