ERS’s Updated SNAP Policy Database Facilitates Research on Program Participation and Outcomes

USDA’s Supplemental Nutrition Assistance Program (SNAP) is the Nation’s largest food assistance program. It provides benefits, redeemable for groceries at authorized stores, to more than 20 million low-income households. For much of the program’s history, the administration of SNAP was largely uniform across States. However, legislative and regulatory changes since the early 2000s, including the 2002 Farm Bill, increased States’ administrative discretion over some program components, such as asset limits, application processes, and reporting requirements.

The SNAP Policy Database offers a central source for information on a broad range of policy options adopted by States as a result of the increased administrative discretion. The database can be used to document changes in SNAP implementation and facilitate research on the program. USDA’s Economic Research Service (ERS) updated the database in February 2024, and it now provides State-level monthly information from January 1996 to December 2020.

SNAP Is a Legal Entitlement Program With Many Features Established by Federal Regulations

SNAP is administered jointly by Federal and State agencies. To receive benefits, households must qualify based on their income, expenses, and assets. Federal eligibility criteria include gross income at or below 130 percent of the Federal poverty level, net income (gross income minus certain deductions) at or below 100 percent of the Federal poverty level, and assets, such as bank account balances, at or below $2,750 (as of fiscal year 2024). Households with senior (age 60 years or older) or disabled members are not subject to a gross income test and have a higher asset limit under Federal eligibility criteria. Households in which everyone receives cash assistance from certain other safety net programs, such as Supplemental Security Income (SSI), are deemed “categorically eligible” and considered to meet the gross income and asset tests. Households with lower net income receive higher benefits up to a maximum based on their household size.

When a household applies for SNAP, a State agency determines its eligibility and sets the length of its certification period. Although households are required to report some income changes within their certification period, they do not have to go through the process of recertification—which typically involves an application, an interview, verification, and eligibility determination—until the end of their certification period.

Unlike most other programs serving low-income households, SNAP is not restricted based on family structure, age, or disability status, so benefits reach a broad range of economically disadvantaged households. There are, however, some eligibility restrictions, such as those for legal noncitizens, as well as additional work-related requirements for working-age adults.

While Federal guidelines establish basic program requirements, States can alter aspects of program administration within certain bounds. For example, States can relax income and asset tests through a policy option called broad-based categorical eligibility (BBCE). States that adopt BBCE can extend categorical eligibility for SNAP to households who qualify for noncash benefits financed out of the Temporary Assistance for Needy Families program or closely related funds. BBCE can be used to raise the gross income test up to 200 percent of the Federal poverty level, eliminate the net income test for eligibility (net income is still used to calculate benefits), exclude vehicles from countable resources, and/or relax or eliminate the asset test. BBCE policies can be used to expand eligibility beyond the Federal guidelines, but they cannot be used to limit eligibility.

Updated SNAP Policy Database Provides More Detail Into How State Broad-Based Categorical Eligibility Implementation Varied From 2005 to 2020

States began adopting broad-based categorical eligibility in 2000. By September 2005, 11 States were using it. The pace of adoption picked up after 2006, and 41 States were using BBCE by September 2020. Most States used BBCE to eliminate or loosen asset limits, and many also increased the gross income limits applied to households under their policy. In some States, households with senior or disabled members were subjected to a gross income limit to qualify for SNAP under BBCE, although if these households were eligible under the Federal criteria (which did not subject them to a gross income limit), they retained eligibility regardless of the State BBCE policy. From 2005 to 2020, the number of States with a gross income limit of 185 percent of the Federal poverty level or higher increased from 9 States to 25 States. Some States also applied BBCE differently to households with different characteristics. For example, as of December 2020, 12 States applied higher gross income limits to households with senior or disabled members under their BBCE policy than they applied to the general population.

Prior editions of the SNAP Policy Database included information on whether States had adopted BBCE and the asset and income limits that applied to the majority of SNAP households in those States. In some States, the income or asset limits under BBCE have differed for subgroups of households, such as those with children or those with disabled or senior members. The most recent edition of the database introduces new variables that provide greater detail on how the income and asset limits under BBCE vary over time for these subgroups.

The SNAP Policy Database Documents How States Lengthened Recertification Periods, Decreasing Administrative Burden for Participants

In addition to having discretion over some aspects of SNAP eligibility criteria, States have been allowed since the early 2000s to simplify the process of applying for and remaining enrolled in SNAP. For example, States have lengthened certification periods or the number of months that could elapse before a SNAP household had to recertify eligibility. Doing this reduces the administrative burden involved with program participation, particularly for working members of households who may need to take time off from work to complete the recertification process.

The SNAP Policy Database includes a set of variables documenting the mean length, median length, and distribution of eligibility recertification periods (the share of households with periods less than 3 months, 4–6 months, 7–12 months, or 13 or more months) for the following three subsets of households:

• those with earned income (such as from a wage or salary),
• those with no earned income and at least one senior member, and
• those with no earned income and no senior members.

With the update, the database extends the recertification period variables through September 2019, the latest data available at the time of the database release.

The prevalence of short recertification periods (1–3 months) increased from the mid-1990s to the early 2000s, particularly for SNAP households with earnings. After the mid-2000s, the use of short recertification periods declined and were applied to less than 3 percent of SNAP households. The majority of households with earners had to recertify every 7–12 months by 2019, representing a shift since the mid-2000s when the majority had to recertify every 4–6 months. Among SNAP households with no earnings, there has been an increase in the share of those who must recertify every 13 months or more, particularly among those with a senior member.

The SNAP Policy Database Highlights Trends in Similar Policy Adoption Across States

In addition to broad-based categorical eligibility and recertification periods, the SNAP Policy Database provides information on other State policy options such as the adoption of electronic benefit transfer (EBT) cards, program outreach spending, Combined Application Projects, and the exclusion of vehicles from asset limit tests. Previous ERS research created indices aggregating several of these policies related to eligibility, administrative burden or the transaction costs of participation, stigma, and outreach from 1996 to 2014. This research finds that over time, States tended to adopt more policies that made participation easier or less costly in terms of stigma or administrative burden. The recent update of the database indicates this trend continued through 2020 among a wide variety of State policy options.

In the late 1990s, States began issuing benefits via electronic benefit transfer cards in place of traditional paper vouchers. This transition was meant to reduce stigma for participants but also to improve security of benefits and prevent fraud. EBT adoption meant households could receive and spend benefits through an electronic benefit transfer card, much like a debit card, as opposed to physical “food stamps.” By 2004, every State had implemented EBT.

Similarly, different States increased the adoption of policies aimed at relaxing restrictions on eligibility and increasing participation in the 2000s. The adoption of the policy of excluding one or more vehicles from the asset test grew beginning in 2000 and now is used in nearly every State. BBCE adoption grew similarly, with more growth occurring in the late 2000s. Combined Application Projects, which offer a streamlined SNAP application process for Supplemental Security Income recipients, grew more slowly. Combined Application Project adoption plateaued in 2010, and, as of fiscal year 2020, only one-third of States had adopted the program. The State yearly average outreach spending, which is used to ensure vulnerable populations are aware of the availability, eligibility requirements, application procedures, and benefits of SNAP, also has increased over time.

The SNAP Policy Database Enables Research on SNAP

The variation in SNAP policies adopted by States and the timing of their adoption provides researchers with the opportunity to estimate their effects on household participation in the program. For example, ERS and academic researchers have found that policy options related to eligibility increase SNAP participation, as do policies aimed at reducing administrative burden. However, potential stigma-inducing policies decrease SNAP participation. These effects, however, differ by household characteristics, such as between senior and nonsenior households.

The data from the SNAP Policy Database can be used in future research to estimate the expected effects of proposed changes to SNAP that may arise from the Farm Bill and other legislative changes to the program. The additional BBCE variables included in the latest update to the database allow researchers to better identify and capture treatment differences in subgroups of participants, such as households with seniors, children, or disabled members.

The database also provides researchers with an opportunity to use the “natural experiment” created by State variation in policy to compare program outcomes, such as household food security, between households in States that adopted a policy relative to those in States that did not adopt the policy. Models may be defined to identify relationships between differences in policy and differences in outcomes across two otherwise similar populations. These natural experiments improve the potential for researchers to pinpoint whether program participation leads to improvements in measures of well-being, such as household food security.

The SNAP Policy Database is part of a suite of ERS data products and web resources that provide information related to SNAP. For example, the SNAP Distribution Schedule Database provides information on the monthly distribution schedule for SNAP benefits in each State and Washington, DC, from 1998 to 2018. The SNAP topic page and the Food and Nutrition Assistance Landscape report series provide an overview of trends in program participation and spending as well as information on significant changes in food assistance programs. Other relevant data, such as the Food Environment Atlas and National Household Food Acquisition and Purchase Survey (FoodAPS), include information relevant to SNAP and its participants.

Alone or in conjunction with these data and reports, the policy variables in the SNAP Policy Database can facilitate research on factors that influence SNAP participation and on SNAP's effects on programmatically important outcomes, such as food insecurity, health, and dietary intake. The database provides opportunities for quasi-experimental research methods. More information on the database is available on the ERS website, and more details on the variables covered here, other variables, and their construction are available for download on the ERS website.