Household composition and infant baby food expenditure around the 2022 federal Medicaid postpartum coverage extension: A triple-difference from the Consumer Expenditure Survey

Beginning in 2022, a federal option let states extend Medicaid postpartum coverage from sixty days to a full year, lowering out-of-pocket medical exposure for new mothers. Where does the freed budget surface? Using the Consumer Expenditure Survey Diary (2013–2024) and a triple-difference comparing households with and without an infant across Section 9812 adopters and comparison states that adopt only later, I find that the infant–non-infant baby food gap widens after adoption by $3.51 in households with two or more children and $3.76 in larger households, both significant at the 5 percent level and stable across falsification, weighting, and sample-composition checks. The contribution is a household-weighted descriptive correlation and its location: the association is estimated precisely in the cells that purchase baby food most often, a pattern consistent with a budget-relaxation reading but equally consistent with the greater statistical power of those cells, and cell-level staggered estimators, once a formula-shortage base-quarter artifact is corrected, bound any state-aggregate association near zero.