Examining disparities in food accessibility among households in Columbus, Ohio: an agent-based model

The objectives of this study were to evaluate the effect of complex interactions among household and environmental-level factors on household-level food availability via a simulation model, the Food Accessibility Agent-based Model in Central Columbus, Ohio (FAAMC) and to test impacts of novel interventions for reducing disparities in food availability. FAAMC simulates food shopping patterns of households based on the actual location of homes and food stores, transportation network, household income, vehicle ownership, and distance to food stores. Policy interventions, which were evaluated as single or combined interventions, included: (1) reducing preference for convenience stores/partial markets; (2) increasing food availability in stores; and (3) increasing household income through a guaranteed basic income supplement program. The FAAMC estimated that mean food availability for food insecure households is 23% (95% Confidence Interval (CI): 22-24%) lower than for food secure households. Increasing household income among the poorest households may lead to a 14% (95% CI: 13-18%) increase in monthly food availability for food insecure households. Implementing multiple interventions would lead to a 41% (95% CI: 40-43%) increase in monthly food availability among food insecure households. This study exemplifies how a systems science approach may serve as an effective and efficient tool for evaluating What if? scenarios for improving household-level food security.