Timothy Loyo2026-09-062026-09-062026-06-11https://hdl.handle.net/20.500.12311/3645UndergraduateFood insecurity remains a persistent development challenge in Uganda, affecting approximately four in ten households. While previous studies have identified income, education, household size, and asset ownership as important determinants, they have largely relied on caloric intake or hunger-experience indicators (e.g., HFIAS) that do not capture dietary diversity and micronutrient adequacy. Furthermore, most existing studies report average effects without testing for heterogeneity across regions or between urban and rural areas, and none have directly compared the explanatory power of Sen’s Entitlement Approach (income and assets) versus Becker’s Household Production Theory (education) within a single model. This study examines the socio-economic determinants of household food security in Uganda using the Food Consumption Score (FCS), a multidimensional measure of dietary diversity, and the latest Uganda National Household Survey (UNHS) 2023/24 sub-sample of 1,527 households (weighted population 926,574). A generalised ordered logit (partial proportional odds) model was estimated with cluster-robust standard errors and sampling weights to account for the complex survey design. Average marginal effects (AMEs) were computed for the probability of achieving acceptable food security.The results show that 73.6% of households have acceptable food security, 21.4% borderline, and 5.0% poor. Income is the only main independent variable statistically significant at the 5% level: a 1% increase in income is associated with a 0.000128 (0.0128 percentage point) increase in the probability of acceptable food security (p < 0.001), supporting Sen’s Entitlement Approach. Secondary education is marginally significant at the 10% level (p = 0.080), and its association appears indirect via income. Adult-equivalent household size shows a positive borderline association (p = 0.052), contrary to the resource dilution hypothesis. Asset ownership is not significant (p = 0.293). Married heads, female heads (after income control), and urban residence show positive supplementary associations. Regional disparities are enormous: the predicted probability of acceptable food security ranges from 89% in Ankole to 43% in Karamoja.The study concludes that income is the most robust correlate of dietary diversity, while education and assets operate indirectly. Policies should prioritise income-raising interventions combined with complementary programmes and geographically targeted strategies, especially for Karamoja. The Parish Development Model should consider regionally differentiated allocations and support for vulnerable unmarried heads and female-headed households. Keywords: Food security, Food Consumption Score (FCS), socio-economic determinants, household income, Uganda, UNHS 2023/24, ordered logit, marginal effects .enDissertation