Gendered Determinants of Food Security Inequities Within Intersectionality Framework: Case Study From Uganda

OBJECTIVES: Food security (FS) research shows that disadvantaged groups, defined through characteristics such as gender, have less resources, and human and social capital, face more barriers to exercise their rights and are more marginalized. This study aims to determine sources of FS inequities intersecting with gender, measuring their relative significance. METHODS: We used nationally representative data from Gallup World Poll, Uganda 2019 (n = 951). A binary logit model was estimated at different equity levels before and after accounting for gender differences. Predicted probabilities of FS were used to identify differences between men and women across different determinants. We used the difference-in-difference approach to test the interaction of gender with variables exhibiting a gender difference in magnitude, direction, or significance. We tested whether a) such differences (gender gaps) are statistically significant; b) gender significantly intersects with each of those variables. We computed odds ratios that included significant intersections between gender and selected variables. RESULTS: Despite gender differences in several FS determinants, there was a significant gender gap among low-educated (p < 0.1), low-income, married, and socially supported (all p < 0.05) men and women. Further difference-in-difference analysis showed that gender significantly intersected with social support and marital status (p < 0.05). Accounting for gender variability, the final model showed that residing in the Eastern region, lacking shelter, and being a single woman decreased FS odds. More adults in the household, higher education and income, social support, and satisfaction with community infrastructures enhanced FS status. CONCLUSIONS: This is a first attempt to model and test gender differences using a difference-in-difference approach within an intersectionality framework. Results suggest that conventional FS approaches may not suffice to reduce inequities if gender is conceived as a control variable rather than a foundation to explain inequities. Gendered-centered analysis helps identify most disadvantaged groups and inform policies to target inequities. FUNDING SOURCES: Fonds de Recherche du Québec Société et Culture (FRQSC), QC Canada; McGill University.