Do Remittances Buffer Household Shocks? Evidence on Food Insecurity in Kyrgyzstan and Uzbekistan

This paper asks whether the positive association between household shocks and food insecurity is weaker among remittance-receiving households in Kyrgyzstan and Uzbekistan. The analysis uses the 2019 wave of the Life in Kyrgyzstan Study and Listening to the Citizens of Uzbekistan rounds 49-82. Because the surveys differ in design, unit of observation, reference period and shock measurement, all models are country-specific and observational rather than pooled. Kazakhstan is included only as a regional benchmark context. The preferred Kyrgyzstan model estimates a negative remittance-shock interaction of -0.2140 (95% CI -0.6549 to 0.2269; p=0.3415), which is directionally consistent with a weaker shock-food-insecurity association but imprecise. The preferred Uzbekistan broad-shock model estimates a negative interaction of -0.5406 (95% CI -1.0415 to -0.0398; p=0.03437). Four-group predictions are used to interpret the interaction on the food-insecurity raw-score scale and to separate baseline remittance differences from shock-period differences. The Uzbekistan household fixed-effects estimate remains negative but is smaller and imprecise (-0.1771; 95% CI -0.5515 to 0.1973; p=0.3539). L2CU estimates are unweighted because the available popw documentation was not approved for this analysis. The findings are compatible with remittance-related moderation of household shock vulnerability, especially in Uzbekistan, but the evidence remains associational and subject to selection, timing and measurement limitations.