Abstract

The Supplemental Nutrition Assistance Program (SNAP) has grown substantially since the turn of the century, providing food assistance to more than 40 million individuals per month in recent years. Using data from the SNAP Quality Control dataset, I analyzed changes in SNAP households and participants from fiscal year (FY) 2000 through FY2023. The overall growth in participation has dramatically changed the types of households receiving SNAP, with a shift in participation toward households with older adults and with one person and away from households with children. These dynamics reflect not only demographic changes, such as population aging, but also differences in SNAP policies for households with elderly and disabled members compared with other households.

Introduction

The Supplemental Nutrition Assistance Program (SNAP), formerly known as the Food Stamp Program, has grown substantially since the turn of the century. Federal data show that over the past 25 years, the number of SNAP participants has more than doubled—from about 17 million to over 42 million—while inflation-adjusted total benefit costs have tripled (FNS 2026a). Even after considering population growth, SNAP participation has doubled, rising from roughly 6 percent of the population in 2000 to about 12 percent in 2025 (ERS 2025).

However, SNAP’s growth has not been uniform. Over time, demographic trends and policy choices have shifted the caseload’s composition toward older adults and single-person households and away from households with children. By fiscal year (FY) 2023 (the most recent year with available data), 36 percent of SNAP households contained an elderly person (up from 16 percent in FY2010) and only 34 percent contained children (down from 49 percent in FY2010) (Monkovic and Ward 2025, 81, table A.27). This report examines those and other changes, explains policy rules contributing to these shifts, and considers future policy reforms to slow SNAP’s growth, including ways to build on recent changes enacted through the One Big Beautiful Bill Act (OBBBA).

Trends in Overall Participation and Costs

Since the turn of the century, SNAP participation and costs have grown substantially per capita, with some variation over the business cycle (Figure 1). This trend has raised some policymakers’ concerns about the growth of federal programs and rising expenditures (Committee on Agriculture n.d.). Comparing FY2000 with FY2025, the share of the population receiving SNAP doubled from 6 percent to 12 percent, while real SNAP costs per capita increased by 179 percent—from roughly $100 per year to $279 (in 2024 dollars). Despite this growth, per capita participation and costs in FY2025 remained below their pre-pandemic peak in 2013, which coincided with the aftermath of the Great Recession and changes in eligibility and other policies stemming from the 2008 Farm Bill (Aussenberg 2008).

SNAP benefit costs generally track growth in program participation because, until recently, SNAP benefit levels adjusted only for food price inflation each year. A notable divergence happened in FY2020, when costs rose far more rapidly than participation. This was driven largely by relief efforts related to the COVID-19 pandemic and a onetime permanent increase of the maximum benefit level. SNAP benefit amounts typically phase out from a maximum level based on household size and the availability of other household income. However, in response to the COVID-19 pandemic, Congress authorized emergency allotments that provided all SNAP households with the maximum benefit regardless of income while increasing the maximum benefit temporarily through September 2021 (Aussenberg et al. 2023).

Although these measures had expired by October 2021, the US Department of Agriculture (USDA) under President Joe Biden permanently increased SNAP benefit levels by 21 percent for the average reference family through reevaluating the Thrifty Food Plan in FY2022—an increase implemented in addition to the standard annual inflation adjustment (FNS 2021). (Notably, the OBBBA returned the Thrifty Food Plan’s reevaluation to the prior precedent of cost neutrality, preventing future administrations from administratively increasing SNAP benefit levels by more than the inflation rate, but it did not reverse the FY2022 maximum benefit increase.) This onetime, permanent action explains the persistently high cost per capita after the COVID-19 pandemic.

SNAP is countercyclical, meaning that all else equal, the number of people receiving SNAP should rise during recessions because of increased unemployment and decline as the economy recovers. While this relationship has held to some extent since 2000, overall SNAP participation has grown faster than changes in the unemployment rate alone would predict (Rachidi and O’Rourke 2025). Over the long run, and especially since FY2020 (due to the Thrifty Food Plan’s increase in the maximum SNAP benefit), costs per capita have not returned to prerecession levels after a period of high unemployment. A range of factors explain this divergence, including rising participation rates among eligible households, especially with elderly members, and eligibility expansions (Cunnyngham 2025; Rachidi and O’Rourke 2025).

Trends in Household Composition: Households with Children, Non-Elderly Adults, and the Elderly

SNAP eligibility and benefit levels are determined at the household level, which makes assessing changes in SNAP households useful. By far the largest compositional trend over the past 25 years is a shift in participation toward households with elderly individuals and away from households with children. (SNAP defines “elderly” as age 60 or older, a convention I follow in this report.) This shift coincides with an increasing share of SNAP households comprising one person. In FY2023, the share of SNAP households containing an elderly person (36 percent) exceeded the share containing a child (34 percent) for the first time. This was a sharp departure from the early 2000s, when more than half of SNAP households contained a child and less than 20 percent included an elderly person (Figure 2).

Participant Trends: Children, Non-Elderly Adults, and the Elderly

Examining changes at the participant level, not only at the household level, is another way to assess who SNAP serves and how that population has shifted over time. In FY2000, children outnumbered non-elderly adults participating in SNAP, but by FY2023, non-elderly adults had become the more common type of recipient. Similarly, although fewer elderly adults than children and non-elderly adults participate in SNAP, this gap has narrowed—particularly between 2010 and 2023 (Figure 3). Between FY2000 and FY2023, the number of SNAP recipients increased fourfold among elderly adults while only doubling among children.

These changes affected the composition of SNAP participants. Between FY2000 and FY2023, the share of SNAP participants dropped by more than 10 percentage points (from 51 percent to 39 percent) among children while increasing by 10 percentage points (from 10 percent to 20 percent) among elderly adults (Figure 4).

General population aging contributes to these trends. However, even when considering population changes, the ratio of SNAP participants to population has steadily increased for adults age 60 and older since FY2000, while decreasing in recent years among non-elderly adults and children. (These ratios reflect the share of each population group—i.e., children, non-elderly adults, and elderly adults—receiving SNAP.) In FY2023, this ratio reached 9.7 percent among elderly adults, surpassing the ratio for adults under age 60 (9.1 percent) for the first time. The trend among children stands in contrast. After peaking in FY2013, the ratio of SNAP participants to population declined steadily among children while continuing to rise among elderly individuals (Figure 5).

Within the elderly population, the strongest relative growth has been among those age 60–69. Among adult SNAP participants, the share age 60–69 rose from about 10 percent in FY2000 to 18 percent in FY2023 (Figure 6). By FY2023, 33 percent of adult SNAP participants were age 60 or older, compared with only 21 percent in FY2000. Importantly, the shift in caseload composition by age was most dramatic after FY2010— four years after the oldest baby boomers turned 60. Although the population’s overall aging largely explains this pattern, the Great Recession’s aftermath and policy choices that treat SNAP households with elderly people differently from households with work-capable adults— the latter of which I discuss in more detail below— likely played a role.

As SNAP adults have aged, a sizable shift toward single-person SNAP households has occurred. For instance, in FY2000, 43 percent of SNAP households included only one person; by FY2023, 59 percent did. This trend toward single-person households reflects a broader US trend toward people living alone, but the change was substantially higher among SNAP households than among the whole US population (Figure 7).¹

Mirroring the shift toward older adults in the total SNAP population, the share of single-person SNAP households has shifted older over time. In FY2000, approximately 40 percent of single-person SNAP cases were over age 60, decreasing to 28 percent by FY2010. However, by 2023, the share over age 60 increased to 47 percent (Figure 8).

Why This Shift?

While SNAP remains an important resource for all low-income household types, today’s program is just as likely to serve households with an elderly person as it is to serve households with children. Moreover, the program is far more likely to serve single-person households (which composed 60 percent of all SNAP households in 2023) than to serve households with one or more children (which included 34 percent of all SNAP households in 2023). This is a sharp departure from the SNAP caseload’s composition in 2000. It contrasts especially with the caseload in 2010, when the majority of SNAP households contained at least one child and fewer than 20 percent included an elderly person (Figure 2).

This trend largely reflects an aging US population, an increase in single-person living arrangements, and declining official poverty rates among children since 2010—gains the elderly population has not similarly experienced.2 However, even when considering population changes, the ratio of SNAP participants to population for the elderly has steadily increased since FY2000—now surpassing the ratio for non-elderly adults—and narrowed the gap with children. The share of the population receiving SNAP across each age category has increased overall since 2000, but since FY2010, it has decreased by nearly 5 percentage points for children while increasing by the same amount for the elderly (Figure 5).

The growing share of SNAP households with elderly individuals also reflects evolving policy choices over the past 25 years. As the general population has aged, the implications of SNAP policies toward the elderly have extended to a larger share of adults. Currently, SNAP treats households with elderly (or disabled) individuals differently from those without in several ways. The three most consequential differences involve (1) uncapped deductions for certain expenses (such as medical care and housing) from income when determining SNAP eligibility for elderly or disabled households, which are capped or unavailable for non-elderly households; (2) longer SNAP recertification periods for elderly or disabled households, which can range from two to three years, compared with one year or less for other households; and (3) no work requirements.3 These policy differences mean households with an elderly or disabled person qualify for SNAP at higher gross income levels and can receive a higher SNAP benefit than non-elderly or nondisabled households with the same gross income, with no requirement for employment or job seeking. With rising housing (McCarty et al. 2025) and medical costs (Cox et al. 2025) in recent years, as well as changing expectations around working at older ages, these policy differences have likely become more important over time.

Table 1 summarizes the differences in how elderly and disabled households are treated compared with other households when considering SNAP eligibility. Notably, although the rules are the same for households with a disabled person as for those with an elderly person, the former have not increased as a share of all SNAP households at the same rate as the latter. The share of the US population reporting a disability has increased only slightly over the past decade, likely explaining in part this trend.4 However, as the population continues to age, it remains important to consider how program rules affect this growing share of the population differently from other households.

Five policy rules for elderly and disabled households stand out:

1. They are not subject to the gross income test. (They have only a net income eligibility limit of 100 percent of the federal poverty level.)
2. They can receive unlimited out-of-pocket medical expense deductions.
3. They have uncapped excess shelter deductions.
4. They have longer recertification periods.
5. As of July 2025, able-bodied adults without dependent children (ABAWDs) age 65 and older have no work requirement.

Essentially, elderly individuals who have high medical and shelter expenses are eligible for SNAP regardless of relatively high gross income, which can be above the gross income limit for households without an elderly person. Furthermore, once a household with an elderly individual is determined SNAP eligible, it likely does not have to recertify its eligibility for two to three years, even when its income or expenses change. SNAP rules also specify that eligibility is based on the income and expenses of only those household members who purchase and prepare meals together. This means that if an elderly individual is living with other adults but they do not claim to purchase and prepare meals together, the other adults’ income is not considered in determining the household’s SNAP eligibility.

Nearly three in four SNAP households with an elderly member receive Social Security income, and the average certification period for households with an elderly individual is just over two years (Monkovic and Ward 2025).5 When gross income is combined with allowable deductions, all these SNAP households would have net income below 100 percent of the federal poverty level, since they qualify for SNAP. Essentially, this reflects a combination of relatively modest Social Security income and high expenses—most commonly out-of-pocket medical costs and shelter expenses—leaving insufficient resources to meet basic food needs.

To illustrate this point: SNAP program data show that the average monthly out-of-pocket medical expense deduction for households with an elderly individual was $25 in FY2023, compared with only $2 for households with children. (Medical expense deductions are not allowed for non-elderly households, so this reflects the average among households with both elderly individuals and children.) The average excess shelter deduction was $449 for households with an elderly individual and $346 for households with children (Monkovic and Ward 2025, 72, table A.18). While the topic is beyond the scope of this report, further analysis is needed to better understand the relative contributions of changes in Social Security income, medical expenses, and shelter costs to the rising share of elderly households eligible for and receiving SNAP.

What is clear is that households with an elderly individual qualify for SNAP at higher gross income levels and receive larger income deductions than do households without an elderly or disabled member. This reflects a policy choice recognizing that the elderly may face higher expenses than the non-elderly. However, the availability of large income deductions for households with an elderly member means that, when SNAP benefits are added to net income, elderly households participating in SNAP likely have more total resources available to them than do households with non-elderly adults or children.

Figure 9 confirms this pattern. Net income represents the resources available to a household after considering allowable expenses. When average net income is combined with average SNAP benefits per person (in 2023 dollars), households with an elderly individual have more total resources than other SNAP households have. This pattern of more total resources in elderly SNAP households than in other SNAP households has remained consistent over time: Average net income (adjusted for inflation) plus SNAP benefits per person has been relatively stable across household types since FY2000.

Why Might This Contribute to a Shift Toward Elderly Adults’ Participation?

Alongside a generally aging population, these policies likely moved SNAP toward serving a larger share of elderly households in several ways. One involves outreach and awareness. Throughout the 2000s and 2010s, the USDA’s Food and Nutrition Service—the federal agency that oversees SNAP—allowed waivers and demonstration projects aimed at increasing elderly adults’ access to SNAP. A review in 2020 found that a sizable share of states had implemented a targeted intervention to increase the elderly population’s access to SNAP (Levin et al. 2020, 13, exhibit I-3). A 2020 study found that of five selected interventions, those to increase awareness and extend recertification periods for the elderly ultimately increased SNAP elderly caseloads (Levin et al. 2020).

As awareness increased, elderly households likely understood they were eligible for SNAP at higher gross incomes than other households. Similarly, the elderly likely became more aware of the availability of income deductions, including that high out-of-pocket medical expenses and excess shelter costs could make them SNAP eligible even when their gross income was above the statutory limit.

Another factor is likely the value of the benefit itself. The average SNAP benefit per person in elderly households has increased by 160 percent over time, even after adjusting for inflation (FNS n.d.; author’s calculations).6 Importantly, because certain expense deductions are uncapped for elderly and disabled households, these households receive a SNAP benefit increase as their expenses increase, unlike non-elderly households, which face capped expense deductions. Research suggests that increasing the benefit’s real value leads to higher participation (Li and Çakır 2024).

Additionally, because the certification period averages more than two years for households with an elderly person (compared with 10 months for households with a child) (Monkovic and Ward 2025, 68, table A.15), elderly households remain on SNAP longer than do other households even when their income rises, boosting their relative share of the SNAP caseload.

Policy Considerations

SNAP is the largest it has ever been in terms of absolute participants and expenditures per capita. In FY2025, annual federal SNAP expenditures topped $279 per US resident (Figure 1) and $2,260 per participant (adjusted to 2024 dollars) as total SNAP expenditures remained near $100 billion per year. More than 42 million individuals received SNAP in the average month in FY2025, and, when considering population growth, participation was near a record high, with 12 percent of the population receiving SNAP that year. The total share of the US population receiving SNAP has doubled since FY2000 (from 6 percent to 12 percent), and real benefit costs have tripled (from $28 billion to $95 billion in 2024 dollars). These trends raise questions about SNAP’s long-term trajectory and size, including whether this growth is sustainable given the United States’ current fiscal situation.

In absolute numbers, SNAP growth since the turn of the century has occurred across various household types, but the strongest relative growth has been among households with elderly adults and with one person, while households with children have declined as a share of total SNAP households over time. Today, SNAP households with an elderly member outnumber those with children. Policies aimed at properly targeting future SNAP growth must therefore confront the reality of population aging and past policy choices that have allowed elderly households to qualify for SNAP at higher gross income levels, resulting in elderly households having more resources when SNAP benefits are counted than do non-elderly households, including those with children.

These trends raise difficult questions about future SNAP reforms. Congress recently enacted changes through OBBBA to partly address concerns about SNAP’s growth and long-term trajectory.7 The Congressional Budget Office estimated that OBBBA would reduce SNAP spending by approximately $18.7 billion per year on average over the next decade and reduce participation by about 2.4 million people in the average month (roughly 5 percent), primarily due to changes in work requirements.8 Notably, only one provision in OBBBA directly affected what the program considers elderly individuals: the expansion of the work requirement to adults age 60–64.

If Congress seeks to slow SNAP’s growth further, it may need to reconsider how the program defines and treats elderly households, not only in the context of work requirements. For example, Congress should consider increasing the age at which the SNAP program classifies individuals as elderly for eligibility purposes, similarly to how OBBBA raised the age threshold for work requirements to 64. Alternatively, Congress could consider matching the definition of “elderly” for SNAP-eligibility purposes to how the Social Security administration defines the “full retirement age,” which is 67 for those born after 1960.

Policymakers might also consider limits on certain deductions that significantly expand eligibility and benefit levels for elderly and disabled households, such as out-of-pocket medical expenses and excess shelter deductions. Congress should consider reducing the certification periods for households with elderly individuals to 12–18 months, especially when medical and shelter expenses are not fixed, and change the “purchase and prepare” rule to assume all individuals in a household are part of a SNAP case unless they can document otherwise.

SNAP remains an important resource for many low-income households. However, efforts to control its long-term growth must confront the realities of population aging and the past policy choices that today shape how elderly households increasingly qualify for and receive benefits.

About the Author

Angela Rachidi is a senior fellow and Rowe Scholar in opportunity and mobility studies at the American Enterprise Institute, where she studies poverty and the effects of federal safety-net programs on low-income individuals and families in America.

References

Aussenberg, Randy Alison. 2008. Domestic Food Assistance and the 2008 Farm Bill. Congressional Research Service. August 22. https://www.everycrsreport.com/files/20080822_RL33829_8fdf54239fe319fd934046c51eefe20c13a1146c.pdf.

Aussenberg, Randy Alison. 2025. Supplemental Nutrition Assistance Program (SNAP) and Related Nutrition Programs in P.L. 119-21: An Overview. Congressional Research Service. August 15. https://www.congress.gov/crs-product/R48552.

Aussenberg, Randy Alison, Kara Clifford Billings, and Gene Falk. 2023. USDA Nutrition Assistance Programs: Response to the COVID-19 Pandemic. Congressional Research Service. January 24. https://www.congress.gov/crs-product/R46681.

Committee on Agriculture. n.d. Restoring Accountability & Prioritizing Work. US House of Representatives. https://agriculture.house.gov/uploadedfiles/reconcilation_overview_one_pager.pdf.

Cox, Cynthia, Jared Ortaliza, Emma Wager, and Krutika Amin. 2025. “Health Care Costs and Affordability.” In Health Policy 101, ed. Drew Altman. KFF.

Cunnyngham, Karen. 2025. Reaching Those in Need: Estimates of State Supplemental Nutrition Assistance Program Participation Rates in 2022. US Department of Agriculture and Mathematica. February. https://www.fns.usda.gov/research/snap/state-participation-rates/2022.

DHDD (Division of Human Development and Disability). n.d. Disability and Health Data System. Disability Prevalence by Age, United States. Centers for Disease Control and Prevention, National Center on Birth Defects and Developmental Disabilities. https://dhds.cdc.gov/LP?CategoryId=DISEST&IndicatorId=STATTYPE&ShowFootnotes=true&View=Table&yearId=YR1&stratCatId1=CAT1&stratId1=BO1&responseId=Q6DIS1&dataValueTypeId=AGEADJPREV&MapClassifierId=quantile&MapClassifierCount=5.

ERS (Economic Research Service). 2025. “Supplemental Nutrition Assistance Program (SNAP)—Key Statistics and Research.” US Department of Agriculture. July 24. https://www.ers.usda.gov/topics/food-nutrition-assistance/supplemental-nutrition-assistance-program-snap/key-statistics-and-research.

FNS (Food and Nutrition Service). 2021. Thrifty Food Plan, 2021. US Department of Agriculture. August. https://www.fns.usda.gov/cnpp/thrifty-food-plan-2021.

FNS (Food and Nutrition Service). 2025a. “Broad-Based Categorical Eligibility (BBCE).” US Department of Agriculture. December 29. https://www.fns.usda.gov/snap/broad-based-categorical-eligibility.

FNS (Food and Nutrition Service). 2025b. “SNAP Special Rules for the Elderly or Disabled.” US Department of Agriculture. September 30. https://www.fns.usda.gov/snap/eligibility/elderly-disabled-special-rules.

FNS (Food and Nutrition Service). 2026a. SNAP Data Tables, National Level Annual Summary, Participation and Costs, 1965–2025. US Department of Agriculture. March 24. https://www.fns.usda.gov/pd/supplemental-nutrition-assistance-program-snap.

FNS (Food and Nutrition Service). 2026b. “Supplemental Nutrition Assistance Program Participation and Costs.” US Department of Agriculture. March 13. https://fns-prod.azureedge.us/sites/default/files/resource-files/snap-annualsummary-3.pdf.

FNS (Food and Nutrition Service). n.d. Supplemental Nutrition Assistance Program Quality Control Data, Fiscal Years 2000–23. US Department of Agriculture and Mathematica. https://snapqcdata.net/.

Levin, Madeleine, Marian Negoita, Annelies Goger, et al. 2020. Evaluation of Alternatives to Improve Elderly Access to SNAP. Social Policy Research Associates and Mathematica. May. https://spra.com/wp-content/uploads/2022/06/Elder-SNAP-Full-Report.pdf.

Li, Qingxiao, and Metin Çakır. 2024. “Estimating SNAP Purchasing Power and Its Effect on Participation.” American Journal of Agricultural Economics 106(2): 779–804. https://doi.org/10.1111/ajae.12399.

McCarty, Maggie, Mary Daniels, and Mark P. Keightley. 2025. Housing Cost Burdens in 2023: In Brief. Congressional Research Service. March 11. https://www.congress.gov/crs-product/R48450.

Monkovic, Mia, and Ben Ward. 2025. Characteristics of Supplemental Nutrition Assistance Program Households: Fiscal Year 2023. US Department of Agriculture, Food and Nutrition Service, Evidence, Analysis, and Regulatory Affairs Office. https://www.fns.usda.gov/research/snap/characteristics-fy23.

NBER (National Bureau of Economic Research). n.d. Business Cycle Dating, Unemployment Rate and Recessions Since 1948. https://www.nber.org/research/business-cycle-dating.

Rachidi, Angela, and Thomas O’Rourke. 2025. How Large Would SNAP Be? Simulating the Size of SNAP Based on Changes to the Unemployment Rate. American Enterprise Institute. July 22. https://www.aei.org/research-products/report/how-large-would-snap-be-simulating-the-size-of-snap-based-on-changes-to-the-unemployment-rate/.

Shrider, Emily A. 2024. Poverty in the United States: 2023. US Census Bureau. September 10. https://www.census.gov/library/publications/2024/demo/p60-283.html.

US Census Bureau. n.d.-a. “Figure HH-4: The Rise of Living Alone.” https://www.census.gov/content/dam/Census/library/visualizations/time-series/demo/families-and-households/hh-4.pdf.

US Census Bureau. n.d.-b. Population and Housing Unit Estimates, 2000–10 National Intercensal Totals, 2010–20 National Intercensal Totals, and 2020–25 National Population Totals. https://www.census.gov/programs-surveys/popest.html.