The implication is direct: because living standards deficits are the first and most widespread to bind, service and infrastructure investment is the primary entry point for poverty reduction across all deprivation levels, addressing deficits that affect nearly all poor households. Human-capital investment becomes increasingly important for households experiencing multiple deprivations, complementing rather than replacing infrastructure priorities. 5.4 Synthesis: Similar MPI Scores, Different Poverty Dynamics Similar MPI scores do not imply similar poverty problems. Mon and Kayin record nearly identical MPI scores (0.149 and 0.153 respectively), yet their dimensional structures differ substantially: Mon's education share is 27.5 percent against Kayin's 21.8 percent, while Kayin's health share is 43.1 percent against Mon's 34.3 percent. Two states at the same poverty level face fundamentally different binding constraints. This divergence reflects the finding from Section 5.2: dimensional composition is determined by geographic context, not by poverty level alone, so two states with similar MPI scores can face structurally different poverty problems. Section 5.3 adds a second dimension to this picture. Household characteristics (employment type, income position, education of head, migration status) shift the level of poverty without shifting its dimensional structure. The implication is that compositional differences between contexts are driven primarily by geography, not by the socio-economic profile of the households within them. Area-level conditions (service environment, infrastructure and schooling access) shape what kind of poverty residents experience; household characteristics shape how deeply poor they are. Taken together, the two findings establish dimensional composition as a diagnostic that headcount alone cannot provide, and one with direct practical reading. Where living standards and health dominate, as in Rakhine, Kayah, Chin, Kachin and Ayeyarwady, service restoration and infrastructure investment carry the greatest leverage, since these are the deficits affecting the largest share of poor households regardless of how deep their poverty runs. Where education's share is comparatively higher, as in Sagaing, Mon, Mandalay and Nay Pyi Taw, human-capital investment becomes a more relevant complement, though never a substitute for infrastructure priorities given Section 5.3's finding that living standards deficits remain the most widespread starting point nationally. Mon and Kayin illustrate why this distinction matters in practice: despite near-identical MPI scores, Kayin's higher health share points toward healthcare access as a priority alongside infrastructure, while Mon's lower health and higher living-standards share point toward a different sequencing of the same two priorities. Section 6 draws out these programming implications in full. L AYE R S O F DE P R I VAT I O N : M u l t i di m en si on a l Pov er t y i n M y anmar 41