Measuring Food Security with U.S. Federal Data

By Kari Williams & Isabel Pastoor

The U.S. Department of Agriculture (USDA) defines a household as being food secure when all household members at all times have access to “enough good for an active, healthy life;” it sets a minimum threshold for food security of “ready availability of nutritionally adequate and safe foods” and the “assured ability to acquire acceptable foods in socially acceptable ways” (USDA Economic Research Service, 2025). The USDA provides survey modules for assessing food security in the U.S. (see Table 1), which are used in a number of federal surveys.

Following the recent announcement by the USDA that they plan to cease data collection for the Food Security supplement fielded as part of the December Current Population Survey, we are highlighting data sources for studying food security in the U.S. Table 2 provides an overview of a number of federal data sources that can be used to study aspects of food security in the U.S. This list of data sources is not exhaustive; we have prioritized data available through IPUMS and other long-running and large-scale population surveys. Additional sources covering shorter time periods or more specific focal populations can be found from the USDA’s Food Security in the United States Documentation page and the Food Access Research Atlas.

Table 1. Characteristics of USDA Food Security Modules (adapted from the USDA Food Security Survey Tools webpage)

ModuleTotal ItemsTypical Number of Items (with screener)Reference PeriodExplicit Coverage of Children's Food SecurityComparison to Household Module
Household183Last 12 months*Yesn/a
Adult103Last 12 months*NoNo specific information on food security of children
Short-form66Last 12 months*
NoDoes not measure most severe levels of food insecurity†; less precise; omits conditions of household children
Self-administered Youth 99Last monthYesAdapted from the household module for self-administration by children ages 12 and older
*USDA provides guidance on adapting the module to use a reference period of the “last 30 days.”
† Children’s food intake is likely to be reduced at the most severe levels of food insecurity.

Table 2. Summary of Food Security Coverage in Select U.S. Federal Data Sources

Data SourceFirst Year of Food Security CoverageSample Size
(households)
Geographic DetailUSDA Food Security ModulesSupplemental Coverage Related to Food Security
Current Population Survey (CPS)*199550,000-
60,000
State, county, metro area, cityHouseholdSNAP, WIC, school meal subsidy, food expenditures, minimum necessary food spending, community assistance (e.g., food pantry), places to buy food, food that meets dietary needs/food preferences
National Health Interview Survey (NHIS)*201130,000-
60,000
Census regionAdult (30 day reference period)
SNAP, WIC, school meal subsidy
Medical Expenditure Panel Survey (MEPS)* †201610,000-
14,000
Census regionAdult (30 day reference period)
SNAP, grocery stores, hard to pay for basics like food
American Community Survey (ACS)*§
20001,300,000-
1,500,000
State, metro area, county, PUMA, cityNoneSNAP, WIC‡, lunch subsidy‡
American Time Use Survey (ATUS)*20148,000-
12,000
State, county, metro areaNoneSNAP, WIC, places to buy food, preferred foods, meal preparation
National Survey of Children’s Health (NCHS)201620,000-
55,000
State, CBSANoneSNAP, WIC, expenditures, community assistance (e.g., food pantry)
Panel Study of Income Dynamics (PSID)199720,000-
26,000
StateHouseholdSNAP, WIC, expenditures, community assistance (e.g., food pantry)
Survey of Income and Program Participation (SIPP)199814,000-
52,000
StateShort-formSNAP, WIC, school meal subsidy
* Harmonized version of data available via IPUMS.
† While IPUMS MEPS includes measures related to SNAP included in the MEPS-HC files, it does not yet include the separately released food security data files available directly from AHRQ.
§ IPUMS disseminates ACS data from the public use microdata sample (PUMS) via IPUMS USA and the summary files via IPUMS NHGIS.
‡ These measures are imputed and not directly reported by ACS respondents.

The CPS Food Security supplement is uniquely robust for measuring food security in the U.S. context. Beyond using the full household module that dedicates eight questions to children’s food insecurity, the CPS provides three decades of data, large sample sizes, state identifiers, and important variables that further describe and contextualize the food security module data (e.g., food expenditures, minimum necessary food spending, receipt of public benefits, and information on accessing food).

While the NHIS has similar sample sizes to CPS, the adult module used in the NHIS does not explicitly cover children. The public use NHIS data do not include state-level identifiers and do not include a full household roster for 2019-forward, which severely limits detail about others in the household. The NHIS also offers fewer years of food security data; food security coverage begins in 2011. Future collection of the module is unclear. The public documentation from January 2025 of the sample adult questionnaire, the section of the NHIS where food security measures are collected, categorizes USDA-sponsored food security questions as annually sponsored content. However, a July 2025 version of the documentation indicates that food security questions are only collected for selected years (2019-2024).

The CPS also has advantages for measuring food security as compared to other data sources available through IPUMS. The MEPS food security files use the adult module (like NHIS), but MEPS sample sizes are smaller than NHIS and food security module coverage is limited to a handful of years (2016-2017; 2020-2022). The ACS and ATUS do not include full USDA modules on food security.

Among other long-running, large-scale population data sources that cover food security, only the PSID includes the full 18-question household food security module. The PSID collects these data frequently, but covers fewer years than the CPS and has a considerably smaller sample size – this makes it much harder to use the PSID to examine narrow geographic areas or population subgroups. The SIPP only includes the short-form module of the food security questions, which does not capture conditions of children in the household or the most severe levels of food insecurity. The sample size of the SIPP is also much smaller than the CPS. The NSCH has a large sample-size to ensure representativeness for children at the state level, but it does not include a full food security module and does not include households without children.

If you are interested in learning more about the termination of data collection for the CPS Food Security supplement, you may be interested in this webinar hosted by dataindex.us. They are also collecting comments through October 21 on how people use the CPS Food Security supplement.