Figure 12 plots the share of each dimension in the household deprivation score for MPI-poor households across the full range of deprivation depth. This score is each household's own measure of how deprived it is overall, on the same scale used to determine whether a household is MPI-poor in the first place and differs from intensity (A) in Section 5.2, which is an average of this score across all poor households in a state or region. Although average intensity is similar across states and regions, it varies considerably among individual households, and it is this household-level variation that drives the compositional shifts examined below. Figure 12 Dimensional composition of MPI (by household deprivation score) Figure 12. Dimensional composition by household deprivation score Among MPI-poor households, education’s share rises, living standards’ share falls and health’s share remains broadly stable Education Health Living standard Share of household deprivation score Education 9% 41% Health 42% 39% Living standard 49% 20% 0.33 0.40 0.50 0.60 0.70 0.80 0.90 0.93 Household deprivation score Source: PPS 2025 (UNDP) Living standards is the dominant dimension at the lower end of the deprivation distribution, accounting for around 49 percent of the deprivation score for households with minimal poverty, and remaining substantial throughout the deprivation range. As deprivation deepens, however, living standards’ share falls steadily, reaching around 20 percent for the most deprived households. Education runs in the opposite direction: its share starts at around nine percent among less-deprived households and rises to around 41 percent the most deprived. The two dimensions cross at around a deprivation score of 0.73, at which point education’s share overtakes living standards as the largest share. Health remains comparatively stable throughout, declining only modestly from around 42 percent among less-deprived households to around 39 percent among the most deprived. This may appear to conflict with Section 5.2, where living standards' share was highest in the states and regions where survival-related deprivations dominate (Rakhine, Kayah, Chin, Kachin and Ayeyarwady), several of which also carry the highest MPI scores. The two findings operate on different axes and are not in tension. Section 5.2 reflects context: states/regions with weaker infrastructure and service environments show a high living-standards share across most of their poor households, regardless of how deep each household's own deprivation is, because baseline living-standards deficits are widespread there. Figure 12 reflects sequence: within any household, in any state or region, living standards is typically the first dimension to bind, and education becomes a larger share only as 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 39