Keep Factually independent
Whether you agree or disagree with our analysis, these conversations matter for democracy. We don't take money from political groups - even a $5 donation helps us keep it that way.
Fact check: What percentage of SNAP recipients were children in 2020 versus 2023?
Executive Summary
The clearest, directly reported figure in the provided material is that children comprised about 39% of SNAP participants in fiscal year 2023, a statistic repeated across multiple 2025 summaries [1] [2] [3]. By contrast, there is no directly stated percentage in the provided 2020 materials that says “X% of SNAP recipients were children”; instead the 2020 documents report the share of households that included children (about 65%), which is a different metric and cannot be equated to the share of individual participants who were children [4] [5] [6]. The net: you can state 2023’s child share (~39%) with confidence from the supplied analyses, but you cannot reliably state a like-for-like 2020 percentage of child recipients from these sources because they report household composition rather than individual-age distribution [4] [1].
1. Why 2023’s figure is the clearest headline — a direct participant-age breakdown
The material labeled as fiscal year 2023 explicitly reports the age distribution of SNAP participants and states that children accounted for about 39 percent of all participants, a figure presented consistently across the 2025 write-ups [1] [2] [3]. This is an individual-level statistic: it counts SNAP recipients and classifies them by age, then reports the share that are children. Because the 39% number is framed as a distribution of participants rather than households, it is suitable for direct comparison with other participant-level distributions when those other years provide comparable individual-age data. The 2023 figure therefore stands as the best available datapoint from the supplied analyses for stating “what percentage of SNAP recipients were children” in that year [1].
2. Why 2020’s documents don’t answer the same question — household vs. individual measures
The 2020 materials in the packet do not present an individual-age distribution analogous to the 2023 documents; instead they emphasize that 65% of SNAP participants lived in households with children or that over 62% of participants were in families with children in later summaries [4] [6] [5]. Those statements describe household composition — the presence of a child in the household — not the share of program participants who themselves are children. A single household with children may include adults and children receiving benefits, so household-level proportions can substantially overstate or understate the share of recipients who are children when compared with participant-level percentages [4] [6].
3. The crucial methodological distinction that changes interpretation
The apparent discrepancy between the 65% household-with-children figure for 2020 and the 39% child-participant figure for 2023 stems from different units of analysis: households versus individuals. Household measures count whether any child is present; participant measures count whether each SNAP beneficiary is a child. Those are not interchangeable metrics; converting one into the other requires additional data on household size, age composition within households, and how benefits are distributed among household members. The supplied sources do not provide that crosswalk, so any direct numeric comparison across 2020 and 2023 without adjusting for these differences would be misleading [4] [1].
4. What can be defensibly reported given the supplied analyses
Based solely on the supplied materials, the defensible public statement is that in fiscal year 2023, about 39% of SNAP participants were children, but the supplied 2020 reports do not provide an equivalent participant-age percentage and instead report that a large majority of SNAP households included children (about 65%). A careful communicator should therefore avoid asserting a percent-for-percent change between 2020 and 2023 because the underlying metrics differ; instead the accurate comparison is that household presence of children was high in 2020 while participant-level child share is documented at ~39% in 2023 [4] [6] [1].
5. Next steps and how to get a like‑for‑like comparison
To produce a rigorous year‑over‑year comparison you need a source that reports participant-level age distributions for both 2020 and 2023 or household‑level breakdowns converted to participant shares using age-by-household-size microdata. The supplied packet suggests where the gap lies — the FY2020 household characteristics report versus the FY2023 participant distribution — but does not contain the necessary cross-tabulations to reconcile them. For a definitive change estimate, request participant-age tables for FY2020 from the same dataset that produced the FY2023 distribution or ask for microdata that permits constructing identical measures across years [4] [1].