Economic Resilience of Indonesian Farmers: Statistical and Machine Learning Approach for Sustainable Agribusiness Strategy

Digital Agribusiness, Faculty of Food Security, Universitas Negeri Surabaya, Indonesia

^* Corresponding author: achmadfitro@unesa.ac.id

Abstract

Farmers' economic resilience plays a critical role in sustaining agricultural systems and national food security. This study analyzes Indonesia's Farmer Terms of Trade (NTP) from 2022-2025 using a hybrid methodology that integrates descriptive-inferential statistics, cluster analysis, and machine learning. Monthly data for NTP, the price index received (IT), and the price index paid (IB) were examined alongside key input-cost components: fertilizer, transportation energy, and agricultural wages. Statistical tests confirm a significant post-2022 recovery, with NTP rising by 4.76% in 2023 and 6.36% in 2024. K-Means clustering identifies three resilience regimes: a cost-shock phase in 2022, a stabilization phase in 2023, and a commodity-driven boom in 2024. A Random Forest model (R² = 0.887) highlights transportation costs as the strongest determinant of NTP fluctuations (39.1% importance). The findings show that cost-price dynamics, particularly fuel-related expenses, dominate short-term resilience, while commodity-price cycles shape medium-term adaptive capacity. This study provides an evidence-based framework for strengthening farmer resilience through targeted input-price stabilization, adaptive agribusiness strategies, and long-term structural transformation.