Modeling resource conflict and inequality in fragile systems: A spatial agent-based framework with a Uganda case study.

Abstract

This study develops a spatial agent-based framework for examining how unequal resource access, environmental stress, institutional support, movement, and adaptive cooperation jointly shape conflict pressure and resource inequality. A Uganda case study initializes spatial heterogeneity from WorldPop 2020 population data and CHIRPS v2.0 October-November 2025 rainfall. The primary experiment uses an 80×80 grid, 250 agents, 120 time steps, and 20 baseline seeds. Parameter dependence is evaluated using 128 Latin-hypercube designs with five stochastic replicates per design, time-aware cluster-robust regressions, partial rank correlations, dense one-at-a-time sweeps, structural ablations, and grid/agent-count robustness checks. In the baseline ensemble, the Gini coefficient rises from 0.192 to 0.868 while mean cooperation falls from 0.547 to 0.118; conflict pressure stabilizes near 0.207 after an early transient. Institutional support increases cooperation and mean resources and reduces inequality over the tested range, while also altering the interaction-level conflict index, showing that lower modeled conflict is not necessarily a better system state. Removing the resource-disparity term reduces but does not eliminate conflict, and suppressing behavioral adaptation raises final conflict by about 40% relative to the baseline. The framework is therefore best interpreted as a transparent mechanism-exploration and stress-testing tool rather than an event-forecasting model.