Mechanisms Linking Geography Education Access to Household Welfare in Kenya: Multilevel Causal Mediation Analysis

Geography education is widely assumed to improve household welfare by equipping learners with environmental knowledge, spatial reasoning and resource-management skills, yet the mechanisms through which such access might translate into material outcomes remain theoretically underspecified and empirically untested in the Kenyan context. The proposed framework integrates a geographic information systems workflow for constructing a composite spatial index of terrain suitability, drainage density, rainfall intensity, land cover, soil permeability, elevation and distance to channels, which conditions both school placement and household livelihood strategies. Methodologically, the article specifies identification assumptions for causal mediation analysis with clustered data, including sequential ignorability, positivity and no intermediate confounding, and discusses sensitivity analysis strategies for violations of these assumptions. The contribution is a theoretically grounded, spatially explicit and causally oriented research design that clarifies the conditions under which geography education can plausibly reduce rural poverty in Kenya.