To estimate the effects of the social inclusion programme PROSPERA on food insecurity (FI) in Mexican households during 2012 and 2016.

Quasi-experimental study using cross-sectional data from 2012 to 2016 National Household Income and Expenditure Survey – Socioeconomic Conditions Module (in Spanish, ENIGH-MCS).

Data were used from a 2012 sample of 56 888 Mexican households (representative of 31 206 819 households) and a 2016 sample of 70 263 Mexican households (representative of 33 445 353 households). Severity of FI was estimated with the Mexican Food Security Scale (in Spanish, EMSA). The statistical analysis estimated a differences in differences (DD) model weighted by propensity score to compare program beneficiary and non-beneficiary households in 2012 than in 2016. We estimated the effect on households with and without children (< 18 years of age). We also compared this model to a DD model without propensity score weighting.

Mexican households.

FI among all beneficiary households decreased 8·0pp as compared to non-beneficiary households over the study period. In beneficiary households with children, this decrease was 6·0pp and for beneficiary households without children, this decrease was 12·9pp (for all, P-value < 0·001).

The PROSPERA program had a positive effect on FI reduction at the household level through increasing food access, which usually improves nutritional outcomes in vulnerable Mexican populations.