Simulation of Budget Allocation for Stunting Reduction Programs in West Aceh Regency using a Linear Regression Model

Abstract

Stunting remains a national priority issue in Indonesia, and West Aceh Regency is one of the regions facing this challenge. The success of stunting reduction programs heavily depends on effective and well-targeted budget allocation. This study aims to develop a budget allocation simulation model for stunting reduction programs in West Aceh Regency using Multiple Linear Regression. Historical program and budget data from the template-data-kabupaten-aceh-barat-stunting.csv dataset were used to train the model. Features such as year, program category (Infrastructure, Empowerment, Health, Other), program name, and a priority score were analyzed to predict the budget amount. The model fitting results show very high performance, with an Adjusted R-squared of 0.989 and an F-statistic of 298.6 (p < 0.001), indicating the model significantly explains the variability in budget allocation. The prioritas (priority) variable was found to be the most statistically significant predictor (p = 0.000). This model was then used to simulate an optimal budget allocation for 2025, with a total recommended budget of IDR 36.50 Billion. The main recommendation focuses on the "Program for Fulfilling Health Efforts" with a simulated allocation of IDR 33.75 Billion. This research demonstrates the potential of predictive modeling as a data-driven decision-making tool for local governments in planning budgets for stunting interventions. Furthermore, this study underscores the critical importance of shifting from historical-based budgeting to evidence-based allocation methods to accelerate national stunting reduction goals effectively.