Table 5 Amplification effect of each shock type on predicted MPI poverty Shock type Prevalence (% households exposed) Predicted MPI poverty if not exposed Predicted MPI poverty if exposed Amplification effect (pp) 95% CI Aggregate contribution to national MPI (pp) Conflict 10% 30.6% 44.8% 14.2 9.9 – 18.5 1.4 Economic 57% 25.6% 36.7% 11.1 9.5 – 12.7 6.3 Natural hazard 20% 31.3% 35.0% 3.8 1.0 – 6.5 0.8 Source: PPS 2025 (UNDP) Conflict exposure carries the largest amplification effect at 14.2 percentage points, raising predicted MPI headcount from around 31 percent for unexposed households to 45 percent for those exposed, comparable to the 43.5 percent observed among households headed by someone with no formal education. Economic exposure produces an amplification effect of 11.1 percentage points, somewhat smaller per household but still substantial. Natural hazard shocks add 3.8 percentage points, a statistically significant but more modest effect, partly because much of the vulnerability associated with environmental hazards is already absorbed into the structural baseline through agricultural employment, low income and rural location. However, the shocks that affect the most households are not those that damage most severely per exposed household. Economic shocks are the most pervasive, affecting 57 percent of households nationally; conflict shocks are the most concentrated, reaching 10 percent; and natural hazard shocks fall between the two at 20 percent. This inverse relationship between scale and severity points toward different programming responses. Economic shocks, operating through market and employment channels, warrant broad-based interventions because they reach the largest share of households. Conflict shocks warrant the most intensive and targeted support because they cause the deepest poverty among those affected, acting simultaneously across livelihoods, mobility and service access. Natural hazard shocks are concentrated among agriculture-dependent households and operate through production and asset-level channels, namely crop loss, physical damage and livelihood disruption, that overlap substantially with those of economic shocks. Although the survey instrument distinguishes the two, for farming households a drought or flood that destroys a harvest is simultaneously an income loss, an asset depletion and a food price pressure. For both shock types, addressing the underlying risk, not only responding once it materializes, requires investment beyond the household level: climate-resilient infrastructure, market stabilization and employment protection operate at community, market and policy level and are addressed further under Section 6. Cumulative shock exposure erodes coping capacity The amplification effects in Table 5 capture the impact of each shock type individually. But households rarely experience shocks in isolation: economic, natural hazard and conflict pressures often co-occur and compound over time. Figure 6 shows how predicted MPI poverty rises with total shock count, holding structural characteristics constant.14 14 26 The cumulative trajectory is estimated from a Firth-penalized logistic regression fitted on the full sample, using a count of all shocks reported by the household (natural hazard, economic and conflict combined) as the independent variable. The model includes structural controls, namely household size, sex of head, education of head, employment of head, per capita income quintile, migrant status, together with state/ region and urban/rural indicators. Predicted probabilities are averaged across the full sample at observed values of the controls for each shock count value from zero to ten. LAYERS OF DEP RIVATION : Mu ltid im e n s io n a l Pov er t y i n M y a n m a r