Skip to main content

House Farm Bill Would Increase Food Insecurity and Hardship

On June 21, the House passed the Agriculture Committee’s (the Committee) farm bill,[1] which includes nutrition provisions that would increase food insecurity and hardship. By adopting harmful amendments during floor debate in May, the House made the bill that ultimately passed even worse than the version the Committee originally introduced (see Appendix I). The proposed changes to SNAP would end or cut benefits for a substantial number of low-income people. The proposed changes to the Supplemental Nutrition Assistance Program (SNAP, formerly known as food stamps) would end or cut benefits for a substantial number of low-income people.

SNAP is the country’s most effective anti-hunger program, helping 1 in 8 Americans afford a basic diet, with most SNAP participants being children, seniors, or people with disabilities. Despite providing modest benefits — averaging about $1.40 per person per meal — the program combats food insecurity, alleviates poverty, and has long-term positive impacts on health as well as on children’s educational attainment. The bill would reduce SNAP’s effectiveness and put large numbers of families and individuals at increased risk of hardship.

  • The bill contains changes that would cause more than 1 million low-income households with more than 2 million people — particularly low-income working families with children — to lose their benefits altogether or have them reduced. The House would use these benefit cuts, in part, to pay for a few modest benefit enhancements. But the net effect of all these provisions on SNAP benefits would still be a significant cut overall, and a substantial number of people would lose their SNAP benefits altogether. The remaining savings from the eligibility and benefit cuts would go to expanding state and federal bureaucracies and financing various grant programs outside of SNAP, at the expense of low-income families and individuals whose basic food assistance would end or shrink.
  • In particular, the plan includes sweeping, aggressive new work requirements that would likely prove unworkable and do substantially more harm than good, fueling increases in hunger and poverty. These provisions would force states to develop large new bureaucracies, but research suggests that these requirements would do little to increase employment. This expensive and risky approach runs counter to evidence-based policy making, particularly since the results from work pilots for SNAP recipients that the 2014 farm bill established, which are well underway, aren’t yet available. Moreover, experience suggests that the bill’s proposed work requirements would leave substantial numbers of low-income people who have various barriers to employment — such as very limited skills or mental health issues like depression — with neither earnings nor food assistance.
  • The plan would also impose significant new state mandates and roll back numerous areas of state flexibility that were designed on a bipartisan basis in prior farm bills to streamline and modernize program operations and make the program easier for states to administer and for eligible households, particularly working families, to navigate. [2] The American Public Human Services Administration has raised significant concerns with multiple provisions of the bill, including those that would reinstate a benefit cliff, limit state flexibility, and introduce significant administrative burden. The organization states, “[E]stimates [that] up to two million current individual SNAP recipients, many of them children, in one million mostly working SNAP households could either lose benefits or eligibility altogether under this bill because of often unwise policy proposals [are] disturbing to us.”[3]

Many of Committee Chairman Michael Conaway’s stated themes for the nutrition package — funding job training, modernizing benefit delivery, strengthening program integrity, and supporting healthy eating — are priorities that Committee members on both sides of the aisle broadly share and could have been the basis for a meaningful bipartisan bill. In some of these areas, the bill contains modest, useful proposals. But the Chairman broke with longstanding bipartisan tradition, and he offered, and the House passed, a package from Republicans alone that’s unbalanced, untested, and likely unworkable in key areas like its sweeping work requirements — and that would put the food security and well-being of many low-income families, including working families, at risk. The Senate, by contrast, in late June passed a bipartisan farm bill with a nutrition title that strengthens SNAP and does not include harmful benefit cuts.[4]

Coming just months after a tax-cut bill that will cost $1.9 trillion over ten years (including interest costs) and lavishes tax cuts on wealthy individuals and large, profitable corporations, the House’s SNAP proposals would further widen the nation’s economic divide.

Nutrition Title of House Farm Bill Includes More Than $20 Billion in SNAP Benefit Cuts Over 10 Years
  CBO 10-Year Cost Estimate
SNAP benefit cuts   -$24.6 billion  
SNAP benefit improvements   $5.8 billion  
New administrative costs and work programs   $14.6 billion  
SNAP benefit delivery, and other program changes   -$0.7 billion  
Non-SNAP grants   $2.8 billion  
Total Nutrition Title*   -$1.9 billion  

See Table 3 for more detail. Details do not add to total because of rounding and interactions among provisions.

* Title IV total including $463 million in direct spending and $465 million in increased revenue.

Sources: Congressional Budget Office (CBO), H.R. 2, Agriculture and Nutrition Act of 2018, as ordered reported by the House Committee on Agriculture on April 18, 2018, and as passed by the House,

The House nutrition title includes 48 different provisions related to SNAP and other programs. This analysis reviews the changes in SNAP eligibility, the work requirement provision, the dual enrollment database, and the privatization provision. Appendix II includes short descriptions of the other provisions in the nutrition title.

Harsher Work Requirements Would Put Food Assistance at Risk for Millions of Working and Unemployed Households

Among its most significant provisions, the House bill would impose new, nationwide mandatory work requirements on millions of SNAP participants. It would require SNAP participants ages 18 through 59 who are not disabled or raising a child under 6 to prove — every month — that they’re working at least 20 hours a week, participating at least 20 hours a week in a work program, or a combination of the two. The typical individual who would be subject to the new requirements receives about $150 to $185 a month in SNAP benefits.

Those who can’t comply would face harsh sanctions. The first sanction for non-compliance would mean a loss of the individual’s share of the household benefit (rather than the household’s entire benefit) for 12 months; each subsequent infraction would lock individuals out of the program for 36 months. Individuals could regain their eligibility only by working at least half-time for a month or requalifying through an exemption, such as disability. The bill would provide new, though highly inadequate, funding for states to provide work and job training programs. It would save more than $14 billion over ten years from individuals losing SNAP. While the work requirements may cause a small number of individuals to leave SNAP earlier than otherwise because they have found a job, the vast bulk of the savings very likely would come from households having their benefits sanctioned — given that the House bill underfunds work programs and given the evidence from work requirements under the Temporary Assistance for Needy Families (TANF) program, as discussed below.

These sweeping, aggressive new work requirements would likely prove unworkable and do more harm than good, increasing hunger and poverty. They would force states to develop large new bureaucracies that would need to track millions of SNAP recipients, but likely would do little to boost employment, particularly given that the new funding provided in the bill for job training and work slots would amount to less than $30 per month for those recipients who need a work slot to retain SNAP benefits.[5] Moreover, experience suggests they would leave substantial numbers of low-income people who have various barriers to employment — such as very limited skills or family members with illness — with neither earnings nor food assistance.

These new requirements fly in the face of House Speaker Paul Ryan’s call for evidence-based policymaking. Given the lack of evidence about what works and what doesn’t in moving unemployed SNAP participants into jobs, policymakers on a bipartisan basis provided $200 million in the 2014 farm bill for ten major state demonstration projects to test various approaches to employment services, work programs, and work requirements for people receiving SNAP. The projects are nearing completion, and we will have results from independent evaluators in the next few years. But rather than await the results and, with them, design sound work and job training policies, the bill would mandate that all states move now to institute sweeping changes — on an unprecedented scale — that carry significant risks for poor households and would likely fail.

Background on Current Work Requirements

This approach differs from current law in several ways. For over 30 years, states have had the flexibility to set work requirements for the demographic groups that the House has identified. SNAP requires all working-age adults (with limited exceptions) to register for work and accept a job if offered. States can go further and impose very tough work requirements (up to 30 hours a week) and cut off benefits — including those for children in the household — for people who don’t comply. Importantly, states can establish criteria and provide programming for what’s appropriate in their local labor markets. The U.S. Department of Agriculture (USDA) provides $110 million a year in federal grants for employment and training programs. USDA also matches state spending for these programs above the amount of a state’s basic federal grant, as well as state expenditures for job-related costs such as participants’ child care and some transportation costs. USDA and states spent more than $700 million for SNAP employment and training programs in 2016.

Under the current structure, states can target limited resources to those who may be able to benefit most from work programs. Due to limited funding, however, some states rely on low-intensity employment services that do little to change longer-term employment and earnings. In part, the demonstration projects now underway are seeking to determine how to better use resources to secure more significant employment and earnings gains.

SNAP already has a harsh cutoff for part of this group: people aged 18 through 49 not raising minor children who are working less than 20 hours per week. SNAP is limited to just three months out of every three years for those individuals who work less than 20 hours a week (with some exemptions). States don’t have to offer people subject to this limit a work slot, and most don’t. States can waive the rule temporarily for areas of elevated unemployment[6] or can exempt a capped number of subject individuals from the requirement. Otherwise, they must impose this rule even on individuals who are actively looking for work or working less than half time. In 2016, at least a half-million unemployed individuals lost SNAP benefits due to this rule, many of them very poor individuals who receive no other form of basic assistance. (Income for those affected averages just 18 percent of the poverty line.) States, workforce training groups, and anti-hunger advocates have urged policymakers to ease this rule, not worsen it by making it even harsher or extending it to more people.

A sweeping new set of work requirements that apply to millions of recipients doesn’t reflect the fact that most working-age adult SNAP recipients are, indeed, workers, and that SNAP plays a vital role in supporting them both while they’re working and when they’re between jobs.[7]

  • SNAP is a key work support. Most working-age adults on SNAP who can work do so. Unfortunately, low-paying jobs with unreliable hours and little to no benefits are all too common. Workers in the low-wage market can’t rely on always having a steady full-time job that pays a living wage, and work requirements won’t create these jobs. Without basic benefits, a sick child, transportation snafu, or scheduling conflict can often mean a worker loses his or her job. SNAP is there to help the person who’s between jobs and searching for work. And SNAP effectively boosts wages for workers who don’t earn enough to afford a basic diet.
  • Employment is high among SNAP households that can work. SNAP has become increasingly effective at supporting work among households that can work. More than half of SNAP households with at least one working-age, non-disabled adult work while receiving SNAP. Because people often participate in SNAP when they’re between jobs, employment rates are higher over a longer timeframe: more than 80 percent of such SNAP households work in the year before or year after receiving SNAP. Work rates are still higher for families with children: more than 60 percent work while receiving SNAP, and almost 90 percent work in the prior or subsequent year.[8]

New Bureaucracy to Establish Monthly Eligibility for Millions of SNAP Participants Would Put Eligible Households at Risk

Under the House bill, individuals aged 18 through 59 who aren’t disabled and don’t have children under age 6 would have to prove on a monthly basis that they’re working 20 hours a week, participating in a qualifying job training program (or a combination of work, workfare, and job training for a total of 20 hours), or exempt from the provision. That would require states to create a massive reporting and paperwork system that would be expensive and difficult to navigate for many participants, states, and possibly those who employ SNAP participants.

Currently, most states collect detailed data on participants’ income and work status every six months or when a major change occurs that could affect the household’s overall eligibility, but small changes in a recipient’s work hours or earnings do not have to be tracked monthly. Changes to simplify reporting and tracking were made on a bipartisan basis as a part of the 2002 farm bill to reduce paperwork and office visit demands on low-wage working families and to reduce state costs as well. Those changes, along with other 2002 policies, are credited with helping to boost SNAP participation rates among low-income workers. The monthly tracking and assessment under the new work requirements would undermine the streamlined interaction between the program and participants and introduce significant opportunity for bureaucratic failure that could compromise eligible individuals’ access to benefits.

Using SNAP administrative data, we see that about 9 million individuals potentially would have been subject to work requirements in the House bill had they taken effect in 2016. (See Table 6.) And using the current Congressional Budget Office (CBO) baseline of program participation over the next ten years, which projects caseloads continuing to decline, we estimate that nearly 8 million adults in a typical month in 2021 would potentially be subject to the work requirement because they are aged 18 through 59 and not disabled or raising a pre-school-aged child. (See Table 2.)[9] Of these, nearly 2 million individuals are projected to work more than 20 hours per week and others could be exempt. We estimate each month state agencies would need to track about 7.5 million people who potentially would be required to report their work status, work program participation, or exemption qualification. [10] Low-wage workers with employment hours that vary each week would risk falling below the minimum hours requirement under the House bill and would have to be tracked. Workers with multiple employers would need to regularly submit paystubs, timesheets, or other documents, potentially from each of their employers, to ensure that they are not ruled as having fallen below the 20-hour requirement.

Under the bill language, if a worker’s hours dropped below 20 hours per week, the state SNAP agency would need to connect the participant to a work program that would provide sufficient hours for the individual to meet the requirement in that same month. However, it would be difficult, if not impossible, for participants or the state to know before a month ends whether the worker has met the mandatory minimum hours for the month. This is especially true for individuals who do shift work, as many low-wage workers do.

While challenging enough for employees, monthly reporting requirements would be even more burdensome for people who are self-employed and may lack such documents altogether, and for people who are trying to prove that they’re engaged in sufficient hours of job training, job search, or other qualifying work activities. Some individuals would qualify for different exemptions, such as a temporary inability to work due to illness or injury. Because many of those exemptions would not be permanent, they would need to be reported and verified each month.

A monthly reporting system for millions of people nationally would force states to significantly expand their administrative capacity — both their staffing and their information technology systems. They would need to change their computer systems to collect and process data on an individual’s monthly status, and they would require new staff to manage the increased workloads. If states did not build systems to handle and process this new reporting scheme properly, it could put millions of people’s benefits at risk. Properly recording and assessing monthly reports would be critical to ensuring that individuals who were eligible for exemptions, including workers, were not sanctioned incorrectly. Given how harsh the sanctions are — and how difficult it can be for those wrongly sanctioned to get their benefits restored — the stakes for such a massive system would be very high.

Furthermore, this monthly eligibility determination would create unnecessary bureaucracy that could compromise access for eligible participants. If participants did not fill out paperwork correctly or caseworkers did not process it in a timely manner, both working people and those who are supposed to be exempt from work requirements — including those who can’t work due to disability or serious illness — could lose SNAP benefits. Working people could lose coverage for any of several reasons: their work hours fluctuated from month to month and sometimes fell below required thresholds, they were between jobs but seeking work, they couldn’t navigate the new paperwork requirements and red tape, or state caseworkers made mistakes. Similarly, people with disabilities and serious illnesses could lose coverage because they didn’t meet the criteria for limited exemptions, didn’t understand that they qualify for an exemption, struggled to provide the documentation to prove they qualify — or caseworkers either didn’t process the materials on time or made mistakes in assessing recipients’ circumstances. People with serious mental illness or physical impairments could face particular difficulties meeting these requirements. Getting physician testimony, medical records, or other required documents may be hard, especially if participants don’t have health coverage while they are seeking to document that they meet the criteria for an exemption — a particular problem for many low-income people who are uninsured because they live in states that haven’t adopted the Affordable Care Act’s Medicaid expansion.

If participants improperly lost their SNAP benefits or had their household benefits reduced via a sanction, many likely would not be able to correct the reduction. Many recipients would not understand that the termination was in error or how to navigate the process to appeal a sanction. And, those who struggled to provide documentation in the first place might not be able to do so after receiving a sanction.

Proposed Sanctions Are Extreme and Not Informed by Evidence

SNAP’s current work requirement rules let states impose sanctions under a progressive and flexible structure. For example, the first time a participant fails to complete a SNAP work requirement, the state can sanction the individual for one, two, or three months. Participants who demonstrate compliance after a sanction can return to SNAP. This approach balances a punitive sanction with an incentive to encourage compliance with the work requirements. By contrast, the House bill would impose unforgiving penalties that would take food assistance away from those who don’t meet strict employment and paperwork requirements.

Under its “one strike and you’re out” provision, an individual who didn’t meet all of the new requirements after a month of receiving SNAP would be kicked off the program for an entire year. In any subsequent instance of falling below the minimum work or training hours, a participant would lose benefits for three years. (If a participant reported to her caseworker that particular barriers prevented her from meeting the minimum required work or training hours in a given month, such as lack of transportation or an illness, the state could decide to maintain her eligibility. However, the participant would need to understand the program rules and how to report such barriers.) After a sanction, individuals could only regain benefits if they found a job that consistently provided 20 hours a week or more or became exempt, for example, because of a new disability.

In multi-person households, only the benefits of the individual determined to be out of compliance with the work requirements would be cut, but any reduction to a low-income family’s food budget can increase household members’ risk of food insecurity. And creating this level of food insecurity would not improve a worker’s ability to find employment, but it could leave millions of Americans one mistake away from losing the help they need to buy food. It also ignores evidence from TANF on why participants were sanctioned. Studies of sanctions in that program showed participants lost benefits for various reasons not related to their desire to work. (See box.)

TANF Suggests Workers Lose Benefits Due to Administrative Glitches, Not Lack of Desire to Participate

In a USA Today op-ed promoting the new proposal, Chairman Conaway wrote: “Under this work proposal, only an individual who chooses not to participate in a guaranteed E&T slot will lose eligibility for SNAP.” Given the enormous administrative complexity in implementing this proposal, however, many participants would likely lose eligibility not because they chose to forgo SNAP benefits but because they couldn’t navigate the requirements to verify compliance or the reasons for an exemption.

Studies that evaluated similar work requirements in the Temporary Assistance for Needy Families (TANF) program found that in TANF, participants frequently lost benefits for various reasons not pertaining to their desire to work:

  • Some participants could not understand program rules and consequences. Notices from states frequently lacked clear guidance about policies such as exemptions from work requirements, a study from the Health and Human Services Department’s Office of Inspector General (OIG) found. This study also found that participants often didn’t understand sanctions until they were imposed on them.a
  • States did not have the capacity to ensure that participants who qualified for exemptions were not subject to sanctions. TANF clients at risk of sanctions often face barriers to work such as poor physical or mental health, low education, and a lack of transportation, and states often reported that they didn’t have the resources to identify such barriers, leaving caseworkers with no choice but to sanction clients who should have been exempted, according to another OIG report.b
  • Participants often lost benefits due to administrative glitches. Among participants in several cities who were sanctioned or had their case closed, for example, the most common reasons participants lost benefits were that they didn’t file their paperwork or they missed an appointment, according to a study that surveyed participants in three cities. Multiple studies have found that more disadvantaged participants, who may have less capacity to navigate complex administrative processes, were more likely to lose benefits.c

a June Gibbs Brown, “Temporary Assistance for Needy Families: Educating Clients About Sanctions, Office of the Inspector General, Department of Health and Human Services, October 1999,

b June Gibbs Brown, “Temporary Assistance for Needy Families: Improving the Effectiveness and Efficiency of Client Sanctions,” Office of the Inspector General, Department of Health and Human Services, July 1999,

c Andrew J. Cherlin et al., “Operating Within the Rules: Welfare Recipients’ Experiences with Sanctions and Case Closings,” Social Service Review, September 2002,

Funding for New Work Programs Would Be Highly Inadequate

States would be responsible for building new work programs on a vast scale, providing places in work programs for potentially millions of people every month. As noted above, the number of individuals states would need to track each month to determine if they’re subject to the work requirement would be larger than the group for whom the state would need to provide work or training slots. Using the number of individuals who potentially would have been subject to the requirement in 2016 — 9.4 million — and adjusting it to reflect projected caseload declines for 2021, the share working more than 20 hours a week, and available exemptions from the rule, states would need to create (under a conservative estimate) about 3.4 million slots per month in 2021.

Building and operating work programs on this scale would be unprecedented. No comparable employment and training system exists that states could leverage to generate anything close to this many work program slots for this many people. For example, the federal Workforce Innovation and Opportunity Act (WIOA) Adult Program provided job training to only 153,000 individuals nationwide in 2015.[11] In TANF in 2016, only 126,000 program recipients who satisfied their work requirement were in a work activity other than unsubsidized employment and, of those, only 36,000 were in any kind of education or training program.

Though the Chairman suggests that the millions of SNAP participants subject to these work requirements would have access to meaningful employment services to help them gain jobs, the House’s proposed funding would fall far short of what would be required to create meaningful job training for this large of a group. Programs like subsidized jobs, apprenticeships, on-the-job training, case management, and other meaningful services all would be highly unlikely to materialize, for the reasons explained below.

Many of these types of more intensive services already are options under SNAP’s existing employment and training (E&T) program, but states rarely offer them due to their costs and operational complexity. Research shows that employment and training opportunities like the ones the Chairman has cited can cost $7,500 to $14,000 per participant.[12] Even less intensive TANF work programs cost substantially more than the bill would provide. Current per-person costs for work and assessment activities for TANF program participants subject to work requirements in the median state in 2016 were nearly $5,000 per year.[13] If at least 3 million SNAP beneficiaries per month needed work slots, and costs mirrored those in TANF, it would cost more than $1.2 billion per month to provide such employment services. That translates to almost $15 billion a year or $150 billion over ten years.

By contrast, the House bill dedicates $7.3 billion over ten years in new federal grant funding (financed by benefit cuts) for the work program, or less than $30 per participant per month if 3.4 million people need work slots. The House’s approach ignores evidence that operating effective programs that promote opportunity — such as job training tied to in-demand occupations, services that help community college participants complete their degrees or certifications, and life-skills instruction that helps participants set and accomplish employment goals — costs far more than $30 per month.

Table 2 shows how CBPP derived the estimated 3.4 million slots per month that states would need to create. CBO projects that 36.6 million people will be participating in SNAP in an average month in fiscal year 2021. Assuming the share of participants who are aged 18 through 59, aren’t receiving disability benefits, and don’t have a child under 6 in the household is the same in fiscal year 2021 as it was in fiscal year 2016, almost 8 million individuals in a typical month would be subject to the requirement based on age and disability status. Of this population, approximately 6.2 million are expected to have insufficient work hours to satisfy the requirement.

We assume that 2.8 million of these individuals are likely either to live in areas with high unemployment that would be waived from the work requirements[14] or to be exempt, either under SNAP’s exemption rules (for example, because they are unable to work because of pregnancy or medically determined illness) or under a separate rule that allows states to grant individual exemptions to up to 12 percent of those subject to the work requirement. (Those exemptions would be calculated off of the estimated number of people subject to the work requirement, including those in compliance, not otherwise exempt or living in a waived area.) Given that some states do not waive high- unemployment areas or use their allotted individual exemptions under the current work and time-limit requirements, it is conservative to assume that these flexibilities will be fully utilized under the bill. Under these assumptions, approximately 3.4 million people would need a work or training slot in a typical month to comply with the requirement.

The almost certain result would be greatly under-resourced work programs that offer few, if any, decent-quality services to help people gain skills and jobs. Instead, states would spend considerable resources monitoring compliance with the requirements and creating services that largely consist of compulsory make-work slots that do little to improve employment rates. States would not be able to invest in the kinds of case management and other supports that help recipients comply with requirements; indeed, states would likely have to count on recipients not participating in such programs because states would not have enough resources to run even low-intensity services if everyone required to participate did so.

The bill’s employment and training provisions include low-cost workfare positions that provide no training or skills-building but merely allow individuals to work off the value of their benefits. Individuals participating in workfare can engage in a work-equivalent activity for a maximum number of hours that equals their benefit allotment divided by the minimum wage. Under current law, participants engaged in workfare for the maximum number of hours based on this calculation are considered to be fully complying with SNAP’s work requirement for childless adults. However, as amended, the House bill would require participants not eligible for 20 hours a week of workfare (which would be most of them, given SNAP’s relatively modest benefits) to also participate in another training activity in order to reach the 20-hour threshold. This is a breach of a decades-long federal policy permitting beneficiaries to use workfare alone to satisfy work requirements. The population that would be subject to the work requirements receives only about $150 to $185 per month in SNAP benefits, so this policy would require workfare participants to meet onerous additional requirements for a relatively modest benefit. Many likely would be unable to meet these requirements, or states would be unable to offer sufficient training hours on top of workfare participation, putting food assistance in jeopardy.

Moreover, the requirement is poorly targeted. Research finds that many of those who would be required to participate under this proposal would be employed again within a few months anyway, regardless of their participation in a work program. If, instead of squandering resources on low-intensity services and tracking for these recipients, the funds were targeted on a smaller group and used to provide real skills training, the net impact on longer-term employment and earnings could be substantially larger.

Proceeding with this massive, untested work scheme would waste billions of dollars in federal resources. The House could have provided new funding for job training to build state capacity to offer more of the kinds of opportunities that Chairman Conaway says he wants to offer participants. That would have enabled states to build programming appropriate to their local economies, their SNAP participants, and their capacity, without putting the benefits of millions of participants at risk.

Estimated SNAP Participants Potentially Subject to Work Requirement and Required Work Slots in Typical Month of 2021
  Number of SNAP Participants (in millions)
  Total SNAP Participantsa 36.6
  Estimated total adults aged 18 through 59 not receiving disability benefits, no children under 6 in householdb 7.9
Minus Those working 20+ hours per week in a given monthb 1.7


Note:  This is the number of individuals potentially subject to the work requirement

Not working 20+ hours per week in a given month   6.2

To determine how many slots are needed:

Some individuals must be exempted under employment and training exemptions (which states will implement differently). In addition, states can waive areas of high unemployment and up to 12% of those subject to the rule outside of waived areas.

Minus those in waived areas

(Assumes all states waive all eligible areas. Likely an overestimation since some states do not and will not waive eligible areas.)

Participants in waived areas under proposed waiver rule 0.4
Equals Participants not in waived areas under the policy and still subject to the work requirements 5.8

Minus 30% for exemptions under E&T standards

(Rough assumption based on experience of states in applying exemptions under the current 3-month time limit for able-bodied adults without dependents)

Participants exempt under SNAP Employment & Training (E&T) exemptions such as medically certified as unfit for work 1.7
Equals Participants still subject to work requirements after E&T exemptions 4.1

Minus 12% individual exemptions

(Assumes all states use the full 12%. Likely an overestimation since some states do not and will not use individual exemptions.)

Participants exempt through individual exemptions 0.7
Equals Total subject to work requirements for whom the state must provide a work slot 3.4

a Total estimated average annual SNAP participants in 2021 based on CBO baseline.

b CBPP estimate based on applying the share from the 2016 SNAP Household Characteristics data to the estimated average annual SNAP participants in 2021 based on CBO baseline.

Bill Preempts Evidence-Based Efforts to Identify Interventions That Improve Employment and Earnings

The bipartisan 2014 farm bill provided $200 million to fund ten major state demonstration projects to test various approaches to SNAP employment services, work programs, and work

requirements. These projects are well underway — in red and blue states alike — with results expected in the next few years. But rather than await the results and, with them, design sound work policies, the House bill would mandate that all states move now to institute sweeping changes on an unprecedented scale without evidence to support their effectiveness. That is not evidence-based policymaking.

Moreover, the House’s proposed requirements appear to be modeled on TANF work requirements, which evidence strongly indicates have been largely ineffective at improving long-term employment outcomes. They also have left many families with neither cash assistance nor earnings, pushing their children into deep poverty.[15] A recent study of Kansas parents who left TANF due to work sanctions shows that median earnings in their first year after leaving were just $1,601, or $133 per month — less than an average family in the Midwest spends just on natural gas and electricity in a typical month. By the fourth year, this figure rose but only to $2,175 a year, still far below what a family needs to make ends meet. While work was common, it was not steady, leaving parents with no income at various points during the year.

Bill Puts Millions of Vulnerable Individuals at Risk of Losing SNAP Each Month

Chairman Conaway maintains that states would have to provide work-program slots to everyone who isn’t exempt and can’t find at least 20 hours a week of work on his or her own, and that these individuals would receive employment-related services to improve their employment outcomes. But that would be nearly impossible under the proposal. We would much more likely see the return of “sanction machines” — state work programs that are inaccessible to many SNAP participants, either by design or in practice, leaving large numbers of individuals locked out of SNAP for one to three years.

  • First, older unemployed workers are particularly vulnerable to a rigid work requirement. Restricting access to food assistance for individuals over 50 years old makes little sense. Research shows that older workers face longer bouts of unemployment after losing a job.[16] Given limited resources, older unemployed workers are also not often targeted with training opportunities, since younger workers have many more years of expected future employment.
  • Those already working but who have fluctuating and undependable schedules could lose SNAP when their hours dip below the 20-hour weekly threshold. Moreover, working recipients who don’t submit documentation — or recipients whose caseworkers err in processing their paperwork — could lose their SNAP benefits. And while a worker whose benefits are mistakenly terminated could seek redress, many likely would not understand the process to do so.
  • In addition, people with physical or mental health conditions — and those caring for family members who are elderly or have disabilities — would lose when they can’t navigate a complicated (and under-funded and under-staffed) exemption process or when a state caseworker doesn’t accurately assess their circumstances.
  • Households in crisis — such as those that are homeless or have suffered a family health emergency — also could lose benefits when they’re required to participate in work programs within a month of enrolling in SNAP and have difficulty doing so.
  • Households in which an adult may be able, with adequate supports and assistance, to benefit from job training could lose an opportunity to improve their skills when the work program doesn’t provide that support because funds are spread far too thinly to target them effectively or they are locked out of such services due to sanction.

All told, these requirements very likely would fail to adequately serve large numbers of people, many of whom would be deemed “out of compliance” and lose their basic food assistance.

From 2021 to 2028, this provision would cut SNAP benefits by more than $14 billion, the Congressional Budget Office (CBO) estimates. By 2028, in a typical month, about 1.5 million fewer individuals would participate in SNAP. Given experience with similar work requirements in TANF and the under-resourcing of employment services under the House bill, most of these people would likely be sanctioned from the program rather than leave because their earnings increased and they no longer qualified for food assistance.

CBO further estimates that the total new federal funding for the work program, given both federal funding provided by the bill and likely adjustments in state funding of employment and training programs, would be $7.3 billion over ten years, roughly a 146 percent increase above the $5 billion over ten years in current federal spending on SNAP employment and training. (Section 4015.)

Safeguards Are Insufficient to Protect the Vulnerable

The bill retains the state option to request a waiver of the work requirement in areas of high unemployment but restricts the option so that many distressed areas would not qualify. Under SNAP’s existing three-month time limit, states can request a waiver in areas with a documented lack of jobs; the standards for areas to qualify for waivers were set in the late 1990s. Under the House bill, areas with unemployment rates above 10 percent would continue to be eligible for a waiver, as would states that qualify for extended benefits or provide emergency unemployment compensation. However, the proposal would severely restrict the eligibility of areas with high unemployment relative to the national average. Under current rules, areas with an unemployment rate 20 percent above the national average over a 24-month period are eligible for a waiver, but under the bill they would no longer be eligible if their unemployment rate was below 7 percent.

The bill also would eliminate state flexibility by limiting the type of areas that can qualify for a waiver, such as excluding cities and restricting the time period that could be used to justify a waiver. The bill would apply these new waiver restrictions to the current three-month time limit in fiscal years 2019 and 2020 and to the new work requirements that would take effect in 2021.[17]

Currently, about 36 percent of the U.S. population lives in nearly 1,000 counties as well as numerous cities and reservations that are waived from the three-month time limit. Under the bill’s waiver provisions, the share of the population living in waived areas nationwide would drop to approximately 6 percent, a decline of about four-fifths. The shares of the population waived in high-unemployment states such as Alaska, New Mexico, and West Virginia would decline by nearly two-thirds. States would be much less able to respond to local labor market conditions, and many people living in areas with weak local economies would have neither sufficient work nor adequate food.

The bill would permit states to exempt a limited number of individuals from the work requirement, capped at 12 percent of those subject to the rule (including those in compliance) who are not otherwise exempt or living in a waived area. This would presumably allow a state to protect certain participants of their choice — such as certain homeless individuals or victims of domestic violence — from losing food assistance. But the number is capped. States might use some of their allowed exemptions but would likely keep some in reserve for unexpected circumstances like a hurricane, flood, or other natural disaster or a sudden closure of a major factory. SNAP’s existing three-month time limit has a similar exemption for up to 15 percent of individuals subject to the time limit, but it is drastically underutilized.

Other SNAP Benefit Cuts

Aside from the impact of the major new work requirement, the bill includes other eligibility restrictions and benefit cuts to SNAP participants that would mean $10.5 billion less in food assistance for low-income households over the next ten years. This section reviews the three major areas of benefit cuts.

#1: Reimposing Benefit Cliff and Unnecessary Bureaucracy by Restricting Categorical Eligibility Option

States currently have an option, called categorical eligibility, to raise SNAP income eligibility cutoffs and asset limits by aligning SNAP’s rules with those that states set for benefits funded through TANF. With this option, 32 states have lifted SNAP’s income limits, extending the program to more working families. Over 40 states have used the option to adopt less restrictive asset tests, i.e., the amount of financial and other assets, such as vehicles, that a household may own and remain eligible for SNAP (see Table 4).

The House bill would dramatically narrow the categorical eligibility option beginning in fiscal year 2021, setting a much more rigid standard for the types of TANF-funded services that states could use to lift SNAP income and asset (including vehicle) tests. As a result, states would largely lose this option. States have a second option under current SNAP rules that allows them to set less restrictive vehicle asset limits. The proposal would also eliminate that option. Packaged with these restrictions is a proposal to raise SNAP’s federal asset limits. Altogether, the proposal imposes stricter eligibility rules by eliminating state flexibility, with the result that eligibility would be terminated for some SNAP households, mainly working families, and states and SNAP households alike would face more paperwork and bureaucratic hassles.

Impact of Reimposing Gross Income Test

Many working families have gross income just above SNAP’s income threshold (130 percent of the federal poverty line (FPL) or $2,213 per month for a family of three in fiscal year 2018) but face significant expenses, including costly housing and child care, that can put a healthy diet out of reach. States can use the categorical eligibility option to raise their SNAP income cutoff and provide SNAP benefits to these low-income working families. Lifting the limit also helps to promote greater work effort. A household close to the income threshold that accepts a modest wage increase or more hours of work that pushes its earnings slightly above the threshold would lose all its SNAP benefits, which could actually result in some families that earn more being made worse off.[18] For example, a single mother with two children working full time at $12.50 an hour has income at 127 percent of the poverty level and receives about $152 a month from SNAP, making up about 7 percent of her total monthly income.[19] If her hourly wage rose by just 50 cents (or $87 a month), lifting her income just above 130 percent of the poverty line, the family would — without the categorical eligibility option — lose SNAP eligibility. And the loss of SNAP benefits would more than cancel out the higher earnings and make the family worse off, with its total monthly resources falling by about $65 per month.

The typical family with children that benefits from this state option receives about $100 a month in SNAP. In an average month, CBO estimates that about 400,000 low-income households receive food assistance as a result of the flexibility on the SNAP gross income limit that this option allows states. Almost all of the benefits go to working households. Many of these households would lose all their benefits under the House bill, which stands in contrast to the “pro-work” message Chairman Conaway has been using to promote the bill. (See Table 5 for a list of states that have taken this option.)

Table 5 shows the estimated number of households and individuals in each state that would lose SNAP benefits as a result of states reimposing a gross income cutoff at 130 percent of poverty. These estimates are based on 2015 and 2016 data on the characteristics of SNAP households that states and USDA collected as part of the SNAP quality control process. States may have better estimates of the effect of the proposal, particularly if they have more recent data or if they newly raised their gross income threshold and it is not fully reflected in recent years’ data.

Furthermore, children in SNAP households are directly certified to receive free school meals. Children in households that would lose SNAP eligibility due to the end of categorical eligibility would also lose access to free lunches and breakfasts at school. CBO estimates that some 265,000 children would lose free school meals.

Impact of Changes to Asset Rules

The categorical eligibility option also lets states ease the federal asset limit, thereby allowing households with assets over the federal limit to participate. Building assets helps low-income families invest in their futures and weather unforeseen financial crises. The House bill, however, would narrow states’ ability to raise asset limits for SNAP participants, effectively re-imposing SNAP’s asset test in states now taking the option. To be sure, the bill would raise the federal asset limits from $2,250 per household (or $3,500 for households with an elderly or disabled member) to $7,000 per household (or $12,000 for households with an elderly or disabled member), exclude savings accounts of up to $2,000 from the asset test, and raise the vehicle allowance, above which the market value of vehicles would count toward the asset limit.

Taken together, these asset rule changes likely would not cause many households to lose SNAP eligibility, as very few households with significant savings participate in the program. And in the ten states that have maintained the old, outdated federal asset rules, this change would represent an eligibility expansion. But reinstating asset limits in the more than 40 states that have lifted them through categorical eligibility would reintroduce significant administrative burdens. Applicant and participant households would have to submit documentation of their savings and other assets, and potentially have to document even assets that are excluded, so that caseworkers can review that information and assess households’ eligibility. States would likely need to pay for more credit checks to scrutinize and verify information that households provide. Moreover, households without savings or disposable cash would have to prove that they don’t have such resources. Consider an applicant who had previously closed his savings account because he had depleted all the funds to cover medical bills. When applying for SNAP, he might have to get a statement from the bank that the account was closed in order to prove that he no longer has it since a credit check (which conveys historical information) would likely list it as one of his assets. That change, along with several others in the bill, would undercut the substantial progress that federal and state policymakers have made in simplifying the program, particularly for working-poor households, and in reducing administrative costs.

Impact of Change to Vehicle Test

Finally, the package of asset provisions repeals another option that, in addition to categorical eligibility, enables states to adopt less restrictive rules governing the value of a vehicle that households may own and still qualify for SNAP. Federal policymakers provided this second pathway in 1999 to ensure that SNAP policy wasn’t undermining working families’ ability to own reliable transportation while participating in SNAP. The option lets states align SNAP vehicle rules with the rules they use in TANF cash or other assistance programs (such as child care). Many of these programs exclude the value of one household vehicle. States have used the option to simplify and streamline SNAP program rules and have appreciated that it enables them to convey consistent information across the two programs concerning vehicle policy.

The proposal to repeal the vehicle asset test option comes with a provision to lift the federal vehicle asset limits. Current federal policy requires states to count a car’s fair market value (FMV) towards the SNAP resource test to the extent that the FMV exceeds $4,650. The proposal would lift that level to $12,000. The proposed level is less, however, in inflation-adjusted terms than the level the program set when this limit was first established in 1977. That limit of $4,500, when indexed for inflation, would be more than $18,500 in 2018.

A household with a modest car of slightly higher market value than the proposed $12,000 limit would lose SNAP benefits entirely. Cars such as a 2015 Toyota Corolla, 2012 Ram Truck, or a 2011 Ford Explorer all exceed this limit. And participants may well have outstanding loans on their cars so that their equity value is much less than the car’s market value. Taken together, the vehicle policy proposals would undermine the goal of letting low-income households, especially working families, own a reliable means of transportation and would reintroduce an eligibility restriction and complexity into the program that most states essentially eliminated years ago.

On balance, CBO estimates, the proposal to roll back broad-based categorical eligibility, repeal the state option to align vehicle policy with TANF assistance, and raise the federal asset and vehicle limits would decrease federal spending on food assistance by $3.8 billion over ten years. (Sections 4006, 4012, 4013, and 4014.)

#2: Eliminating Connection Between SNAP and LIHEAP, Causing Some SNAP Households With Out-of-Pocket Utility Expenses to Lose Benefits

The House bill proposes to cut SNAP benefits to certain SNAP households with out-of-pocket utility expenses. Except for households with an elderly or disabled member, it would eliminate an administrative simplification through which states use households’ receipt of benefits under the Low Income Home Energy Assistance Program (LIHEAP) to qualify them for a SNAP income deduction that is tied to a household’s expenses for utilities like heating and electricity.

When calculating a household’s available income for food, SNAP provides deductions for certain necessities. One of the most important, the shelter deduction, is available to households whose housing and utility costs consume more than half of their income. In administering the shelter deduction, states can establish and use Standard Utility Allowances (SUA) that reflect typical utility bills for households incurring utility costs in that state or local area, rather than require each household to provide — and caseworkers to review — the household’s monthly utility bills when the household applies for benefits.

Because LIHEAP goes to low-income households who can’t afford to pay all their energy bills and covers only part of a household’s utility expenses, states have long considered a household’s LIHEAP receipt as evidence that the household incurs utility expenses. For decades, federal law has made households that receive LIHEAP eligible for the SUA. This connection between these two programs reduces unnecessary paperwork for states and poor households alike. But this provision of the bill would end this simplification and require these households to provide documentation of utility costs to continue receiving the SUA. Some households would struggle to produce verification of their utility bills and, as a result, would receive a smaller SNAP benefit.

The Chairman’s materials indicate that this change is meant to address the concern that some states provide a small $21 LIHEAP payment to households that may not actually incur heating or cooling costs and therefore should not qualify for an SUA. Only 14 states have LIHEAP programs that offer $21 payments that can produce this outcome for some SNAP households. The proposal would mean benefit cuts and require more documentation in every state and, though it wouldn’t affect households with elderly or disabled members, many of the households that would lose benefits have members with disabilities or are working-poor households.

CBO estimates that this change would cut SNAP assistance by $2.9 billion over ten years. (Section 4010.)

#3: Putting Children’s Food Security at Risk With Unproven Child Support Cooperation Mandate and Limiting State Flexibility

The House bill would take two unpopular state options related to child support and mandate them as federal policy. First, it would require that low-income parents or guardians who do not live with the child’s other parent cooperate with child support enforcement (CSE) in order to receive SNAP benefits. That would apply both to custodial parents (who have physical or legal custody) and to noncustodial parents. In a welcome move, the bill also would eliminate a state option to sanction noncustodial parents who are in arrears on child support payments — an option that hurts poor noncustodial parents who can’t meet their child support obligations. Second, the proposal makes some technical changes to how child support payments by a noncustodial parent are treated when determining the noncustodial parent’s SNAP benefits, requiring states to exclude these payments from the payor’s gross income. That change would have little impact on SNAP eligibility but would reduce state flexibility and require some states to make costly systems changes.

CBO estimates that these changes would together increase federal spending by $3.5 billion over ten years. (Section 4011.) This includes the estimated costs of mandating cooperation with child support enforcement and of mandating changes to how child support payments are treated.

  • CBO estimates a $3.4 billion increase in federal costs from mandating that custodial and noncustodial parents applying for or participating in SNAP cooperate with child support enforcement. This consists of:
    • A $7.2 billion increase in federal administrative expenses over ten years, and
    • A $3.8 billion reduction in SNAP benefits over ten years due to sanctions for noncompliance and lower SNAP benefits for households receiving increased child support collections. This also includes increased federal SNAP costs from increased benefits to households that pay monthly child support and deduct those payments from their income, which qualifies them for larger benefits.
  • CBO estimates a $124 million increase in federal costs from excluding noncustodial parents’ child support payments from gross income.

Mandated Child Support Cooperation

The 1996 welfare law gave states the option to condition SNAP eligibility on whether a custodial or noncustodial parent is cooperating with CSE. Funded and overseen at the federal level by the Department of Health and Human Services, CSE is a partnership among federal, state, tribal, and local governments to assist families in establishing formal child support orders that require regular, ongoing payments by a parent for the financial benefit of his or her child. CSE helps locate the absent parent, establish parentage, put a support order in place, and enforce an order’s execution.

In states that take the mandatory cooperation option in SNAP, parents who are judged as not cooperating are sanctioned and their share of the household’s food benefit is cut. Few states have adopted the option.[20] The goal of the policy is to punish parents who do not cooperate in hopes that the sanction will change their behavior. States have had serious concerns about the high costs associated with implementing the option, the limited evidence of its impact on child support collections, and the risks to children.

In 2014, Utah commissioned a study to examine the value and potential impact of mandatory child support cooperation in SNAP. It found that implementation would be expensive, requiring $3.2 million to $3.6 million for systems changes and more staff, including a substantial increase in child support personnel to work newly opened cases.[21] Similarly, CBO’s preliminary cost estimate of the bill indicates that the new administrative costs associated with this proposal would amount to $10.9 billion over ten years, of which the federal government would cover only about $7.2 billion and another $3.7 billion would need to be paid for by states.[22] Administering a mandatory cooperation requirement would entail expensive investments to ensure automated systems are seamlessly coordinated across public assistance programs, as well as additional CSE staff to work newly opened cases.

For example, if a custodial parent is obliged to open a new case under the requirement and isn’t in contact with the noncustodial parent, CSE may have to spend considerable time and resources tracking down the noncustodial parent and establishing parentage, before even beginning the process to put a child support order in place. And, custodial parents often don’t pursue child support when they know the noncustodial parent can’t pay much or anything — so if CSE tracks down the noncustodial parent, establishes parentage, and puts an order in place, the pay-off may be very low. In addition, the mandate would require enhancements in information sharing between state child support and SNAP eligibility workers and processes for notifying SNAP applicants and participants.

CBO estimates that, on net, the provision would reduce SNAP benefits by $3.8 billion over ten years through a combination of lower SNAP benefit levels due to increased income from child support collections and cuts due to sanctions for noncompliance. (This figure also reflects increases in SNAP benefits to households that would newly pay and deduct monthly child support as a result of this mandate.) CBO did not provide details on how much more child support income it expects or how many families would be sanctioned.

Along with the proposal’s high costs, there is no evidence that it would generate more child support payments to custodial households without significantly increasing children’s risk of food insecurity. The potential negative impact on children has appropriately given states pause. When parents lose food assistance for non-cooperation, children stand to get hurt. Sanctioning parents’ SNAP benefits reduces a family’s overall food budget and puts children at increased risk of food insecurity and inadequate nutrition. Research indicates that additional food assistance to a household improves not only children’s food security but also the nutritional value of their diets.[23] Consistent access to adequate nutrition in childhood is associated with important longer-term outcomes, including greater educational achievement and better health.[24] The Utah study projected that imposing a child support cooperation requirement in that state’s SNAP program would take a significant number of adults off food assistance but few of them would end up receiving additional funds from child support payments, a dynamic that would leave children more vulnerable.[25]

These very high administrative costs and the risks to food assistance for families with children raise serious questions about this proposal. Though it’s important to address, the gap in eligible, low-income families unserved by child support is relatively modest. Most low-income custodial families already receive child support services. According to Census estimates, 72 percent of custodial families with incomes below 200 percent of the federal poverty level access CSE services.[26] When Utah assessed the costs and benefits of implementing mandatory cooperation, the state similarly found that nearly 70 percent of custodial parents receiving SNAP already had an open child support case.[27] Further, some custodial parents without formal child support cases receive support from noncustodial parents through informal arrangements.

Even if the proposal increased the number of open child support cases for those not currently participating, it likely would not substantially increase the resources available to the child, since low-income noncustodial parents often face significant financial hardship and can make only limited (if any) payments. Lack of regular, reliable income or accrued debt can make it hard for many poor noncustodial parents to pay child support regularly. Mandatory cooperation does nothing to promote their economic stability. Evidence suggests that about 70 percent of child support debt is held by those with incomes of $10,000 a year or less.[28] Many custodial parents who don’t seek formal child support orders opt out because they believe the noncustodial parent could not afford to pay or already provides what he or she can.[29] In these cases, mandatory cooperation would make neither the custodial nor the noncustodial parent better able to support their child.

Furthermore, SNAP already includes an important incentive to encourage low-income noncustodial parents to establish a child support order and make child support payments, which is another potential reason why very few states have adopted a punitive mandate. States can deduct child support payments from the payor’s gross income when calculating his or her household’s SNAP benefits or can exclude the payments from the household’s gross income altogether. (As discussed below, the House bill would mandate that all states exclude child support payments from the payor household.) The deduction or exclusion of child support payments increases the SNAP benefits that the payor household can receive, recognizing that resources paid to support a child living in another household aren’t available to buy food for the noncustodial parent’s household. These policies also acknowledge and reward the responsible behavior of noncustodial parents who can make their payments by ensuring that SNAP provides more resources to them than to those who fail to meet their obligations. These policies also create an incentive for low-income noncustodial parents who already provide informal financial support to establish a child support order, so their payments can be appropriately counted to enhance their food assistance.

By contrast, making cooperation with CSE a condition of SNAP eligibility could put poor noncustodial parents’ food at risk if they were unable to make payments and were deemed refusing to cooperate. Taking away SNAP benefits from a parent struggling to meet his basic needs would likely make him even less capable of providing financial support to his child, no matter how much he may want to meet his obligations. Given the lack of evidence about the policy’s costs and effectiveness in states that have implemented it, the National Child Support Enforcement Association, which represents state child support agencies, opposes a national mandate.[30]

Policymakers should consider the reasons why states have not adopted this mandate: the high administrative costs required to implement it relative to the modest possible benefits to SNAP households, and the risks to children that SNAP serves. These realities argue against imposing this mandate on states that have already chosen not to use this approach, based on their assessment of the costs and benefits.

Removal of Penalty on Noncustodial Parents in Arrears

Under current law, states also can make poor noncustodial parents with child support arrears ineligible for SNAP, which is an option that only Mississippi currently implements. This option punishes those who have been unable, rather than unwilling, to pay their child support obligations. The bill would wisely eliminate this harsh option, which does not increase the noncustodial parent’s ability to pay child support.

Mandate That States Adopt Child Support Exclusion Over Deduction

Finally, the proposal would eliminate another state option related to program simplification. It’s one of numerous proposals to restrict state flexibility and impose new mandates on states.

In 2002, the farm bill gave states the option to exclude child support payments from the payor household’s gross income rather than factor it in as a deduction from gross income that’s applied after the household has met or failed the gross income test. Since it was created, only 14 states have adopted this option.[31] The remainder deduct child support payments made to non-household members from the payor’s gross income when calculating SNAP benefits. The change makes no difference in the amount that an eligible household receives in monthly benefits. The gross income exclusion approach does expand eligibility, albeit very modestly, because some noncustodial parents with gross income above SNAP’s federal income cutoff have monthly gross income under the eligibility cutoff if child support is excluded from income before the gross income test is applied. As a result, the exclusion is slightly preferable to the deduction. But few households fall into this category, and most states have chosen to stick with the deduction option.

The House bill would eliminate the decades-old child support payor deduction and instead require the 39 states not currently using the exclusion approach to shift to treating child support payments as an exclusion, making the current state option mandatory. States would have to reprogram their computers and change their forms to do so, with likely little impact on household eligibility.

#4: Denying Food Assistance to Individuals
with Certain Criminal Convictions and Their Families

Current law denies SNAP eligibility to individuals who were found guilty of a violent crime if, after release from prison, they violate their parole or the terms of their release. The House bill would go further, terminating food assistance for all individuals with convictions for certain violent crimes even if they are working hard to avoid recidivism and become productive members of their communities. It would take away SNAP from people who committed a single crime years or even decades ago and have since completed their sentence and complied with all terms of their release. And, given incarceration patterns in the United States, the provision likely would have a sharp racial bias, disproportionately compromising adequate nutrition for low-income African Americans, including poor elderly African Americans convicted of a single crime decades ago by segregated southern juries.

Taking food away from those with criminal records flies in the face of bipartisan efforts to support individuals reentering society after release from incarceration. Access to reentry services can help individuals while they work to attain self-sufficiency and avoid rearrest and reincarceration. SNAP is a critical part of a reentry infrastructure, providing basic food assistance and supplementing inadequate income. Some evidence suggests that it reduces recidivism.[32] The formerly incarcerated already face significant barriers to rehabilitating and finding adequate employment; this proposal would create more barriers. Without the necessary assistance to get back on their feet, some persons leaving prison might be more likely to engage in illicit activity to obtain food and meet their basic needs.

Furthermore, the provision would hurt families of barred individuals, including children, who would see their benefits cut as well. Although a person with a criminal record would be prohibited from participating in SNAP, his or her income would still be counted towards the household’s income, which could significantly reduce that household’s benefits. Low-income families should not be punished for having a formerly incarcerated individual in their household or for helping their family member succeed with their rehabilitation by providing them shelter.

CBO estimates that this change would cut SNAP benefits by $23 million over ten years. (Section 4039.)

Modest Benefit Enhancements Are Much Smaller Than Benefit Cuts

The House proposal would use $5.8 billion of the $24.6 billion in savings over ten years from cutting benefits to pay for a few modest benefit improvements. The net effect would be a significant reduction in food assistance to low-income households. The benefit enhancements include:

  • Capping exclusion for basic allowance for housing. Currently, payments to military families to cover their housing costs when living off-base, called the basic allowance for housing (BAH), are counted as income in determining eligibility for SNAP and a household’s benefit level. The proposal would exclude up to $500 of the BAH from the household’s monthly income. It also would allow the household to claim excess shelter costs toward SNAP’s shelter deduction only to the extent that they exceed the family’s allowance.

    CBO estimates that this provision would increase SNAP benefits by $116 million over ten years. (Section 4007.)

  • Mandating homeless housing allowance. This provision would require all states to include a simplified deduction for homeless households with shelter costs and to index the amount of this deduction to inflation beginning in 2019. The 2002 farm bill gave states the option of providing a flat deduction of $143 for households without permanent housing who are not receiving free shelter throughout the month and who incur shelter expenses. Households experiencing homelessness often have expenses such as paying for a hotel or motel for some nights, paying relatives with whom they are staying, telephone charges, or charges for a vehicle if they’re living in a car. This deduction reduces the burden of producing documentation for, and verifying, expenses, which is especially significant for homeless people with transient living arrangements who may not receive or be able to consistently produce receipts or other forms of documentation for their shelter costs.[33] In 2016, over half of states took this option.[34] By deducting these costs, these households can qualify for higher SNAP benefits, which more accurately reflect the income they have available to buy food. The deduction also simplifies the process of verifying these costs for states. Extending the policy to all states would be a positive policy improvement.

    CBO estimates this provision would increase SNAP benefits by $76 million over ten years. (Section 4009.)

  • Mandating transitional benefits. The 2002 farm bill created an option to let states provide up to five months of transitional SNAP benefits to families that leave TANF cash assistance without requiring the family to reapply or submit additional paperwork or other information. That helps to ensure that families leaving a state’s cash assistance program don’t suddenly also lose access to SNAP. In 2016, 22 states had transitional benefits of differing periods of up to five months.[35] Under the House proposal, all states would be required to provide transitional benefits for these households for five months. In 2016, about 5 percent of all SNAP households and 11 percent of SNAP households with children received TANF cash assistance.

    CBO estimates that this provision would increase SNAP benefits by $895 million over ten years. (Section 4024.)

  • Enhancing earned income deduction. SNAP’s rules give preferential treatment to earned income over unearned income (like Social Security or cash assistance). The rules disregard 20 percent of earned income when calculating benefits, which is meant to reflect income spent on work-related expenses like transportation and clothing that is, thus, not available to buy food. Due to the earnings deduction, a household with earnings will receive a larger SNAP benefit than a household of the same size and gross income in which income comes from unearned sources. So, an unemployed worker who can replace his or her income from unemployment insurance or cash assistance with a comparable level of earnings will take home more total income from earnings and SNAP than from unearned income and SNAP.

    The proposal would increase the earned income deduction from 20 percent to 22 percent of earnings. About three-quarters of households with earnings would receive a modest benefit increase of $10 per month on average.[36] The benefit increase would be tied to household earnings — households with higher earnings would get a slightly larger increase, and vice versa.

    CBO estimates that this provision would increase SNAP benefits by about $500 million a year, or about $4.6 billion over ten years. (Section 4008.)

New Program Integrity Provision Unnecessarily Puts Personal Information for Tens of Millions of Americans at Risk

The House bill would establish a duplicative enrollment database by expanding the National Accuracy Clearinghouse (NAC) to a nationwide, mandatory effort. The NAC, now implemented under a state-initiated contract between Mississippi and LexisNexis, lets five participating southern states (Alabama, Florida, Georgia, Louisiana, and Mississippi) share data on their SNAP participants to identify and prevent SNAP participation in more than one state. An evaluation of the NAC found that less than 0.2 percent of SNAP participants were dual participants (i.e., more than 99.8 percent were not).[37] CBO estimates that the NAC will save $588 million over ten years. (Section 4001.)

Expanding the NAC could strengthen SNAP’s program integrity and is a sound concept, though the House bill is flawed. The most serious flaw relates to the privacy of current and past SNAP participants. The provision requires that USDA build an enormous database of all SNAP participants that would include income and other personal information (including Social Security numbers and income and asset information) that’s well beyond what’s needed to limit duplicate participation. The data, which states would share with USDA each month, would be stored in the database for many years, if not indefinitely. When all states participate, the NAC would include personal information on every SNAP recipient ― some 40 million people at any point in time and potentially tens of millions more over time as others participate in SNAP and then leave the program. This raises concerns about data security, privacy, and other potential uses of the information. Given that the current pilot database is operated by a private vendor, LexisNexis, it’s unclear whether the House envisions giving a commercial entity access to government information. The proposal does not specify USDA’s costs to implement security and privacy protections that match those of the Social Security Administration and Internal Revenue Service for the similar information they store about millions of Americans. Nor does it appear to provide any dedicated new funding for such measures, or to authorize USDA to slow or pause implementation under the aggressive time frames required, if the Agriculture Secretary believed that USDA could not protect the personal data of Americans.

This approach does not reflect current best practices for large data matches. To avoid becoming a target for hackers and to protect private information, data security experts recommend that a match like the one required to prevent dual participation under the NAC “ping” (that is, send an individual query for whether an individual is already in enrolled in SNAP) to the various individual state datasets, rather than have all the data aggregated in one enormous database. The Federal Data Services Hub that the Department of Health and Human Services oversees under the Affordable Care Act facilitates matches between states and federal agencies and private vendors rather than gathering all the data together. This provision is another example of how the bill includes sweeping, untested proposals that would create risks for low-income Americans if the government does not execute them well.

Committee materials indicate that the language in the House bill intends for USDA to create a longitudinal database on all SNAP participants over time to monitor their length of SNAP participation and work histories. The Committee materials suggest that no such data set exists. Actually, there is a longitudinal survey of income and program participation that includes much of the information that the House seeks. USDA uses data from it to publish its report on the details of SNAP participation.[38] In addition, some states maintain longitudinal data on program participants (without individual client identifiers) for research purposes. There is no reason to put the personal information of millions of people at risk for this research purpose, particularly when it’s not clear that the bill’s drafters thoroughly explored the alternatives.

Provision Would Allow States to Outsource SNAP Operations, Risking Access to Benefits

Currently, SNAP requires that civil service employees perform core SNAP eligibility functions. This decades-old requirement ensures workers’ aim is to effectively implement program rules unhindered by profit motives. It may be appropriate to give some SNAP operational functions to private contractors (such as computer systems, building maintenances, custodial services, or debit card issuance) in order to leverage their competitive advantage in these areas. But the government must retain others, like determining eligibility. Some SNAP clients, including many elderly persons, have complex cases that require trained, professional civil service workers to dedicate significant time to appropriately screen and verify their information and ensure they receive the correct benefit levels. When these functions are turned over to for-profit companies, they may focus on the bottom line rather than providing comprehensive support to the needy.

Furthermore, privatization could compromise the security of participants’ data. SNAP collects detailed information about applicants and participants, including Social Security numbers, household composition, and income and employment information. Handing private data of millions of individuals over to private companies raises serious concerns about their ability to keep it secure and not use it for other purposes.

When states have experimented with privatization in the past, the results have been disastrous. During the early 2000s, Texas and Indiana contracted for-profit companies to perform key parts of the eligibility process, including accepting applications, advising clients on program requirements and eligibility, and verifying eligibility. Serious problems resulted in both states. In Texas,[39] for example, thousands were unable to apply or were given misinformation; many received incorrect benefit allotments or were wrongly denied benefits. Long backlogs materialized. Individuals’ private information was released, compromising their security. And taxpayer dollars were wasted, as none of the promised performance improvements or cost savings (which were slated to come from closing eligibility offices) were realized.

CBO estimates this provision would have a negligible effect on federal spending. (Section 4043.)

Other Proposals

The House bill includes numerous other provisions related to SNAP operations and issuance as well as funding for grants, pilot projects, and programs outside of SNAP. In total, they would cost $2.8 billion over ten years. See Appendix II for summaries of each of these provisions. As noted, the bill would pay for them by reducing benefits for a substantial number of low-income households, resulting in a net cut to the nutrition title of almost $2 billion over ten years.


SNAP is a highly effective program targeted to households that need its help to meet basic food needs. With an average benefit of just $1.40 per person per meal, it lifts millions out of poverty, and it has demonstrated long-term benefits for children who participate, including better health and higher educational achievement. While its enrollment and spending are falling as the economy improves,[40] it provides vital assistance to over 40 million low-income Americans.

The House nutrition title includes benefit cuts and harsh new work requirement that, among other problems, would hurt many poor working families, children, and individuals who struggle to find stable employment. Coming just a few months after the tax-cut law, which will mainly benefit the wealthy and profitable corporations, this proposal represents the wrong vision for the country, one that would increase hardship and further widen inequality.

CBO Estimate of Nutrition Provisions (Title IV) of H.R. 2
  CBO 10-Year Cost Estimate
SNAP benefit cuts   -$24.6 billion  
Benefit cut from expanded work requirements   -$14.1 billion  
Rollback of expanded categorical eligibilitya   -$3.8 billion  
Repeal of Standard Utility Allowance (SUA) simplification   -$2.9 billion  
SNAP benefit cuts from mandating child support cooperationb   -$3.8 billion  
SNAP benefit improvements   $5.8 billion  
Mandate of income exclusion of child support paidb   $0.1 billion  
Exclusion of military basic housing allowances   $0.1 billion  
Increase in earned income deduction from 20 to 22 percent   $4.6 billion  
Mandate of simplified homeless housing deduction   $0.1 billion  
Mandate of transitional benefits for households leaving TANF   $0.9 billion  
New administrative costs and work programs   $14.6 billion  
Grants to states for work programs   $7.3 billion  
Administrative costs (Department of Health and Human Services) associated with mandating child support cooperation   $7.2 billion  
Funds for USDA implementation   $0.2 billion  
SNAP benefit delivery and other program changes   -$0.7 billion  
Duplicative enrollment database (National Accuracy Clearinghouse)   -$0.6 billion  
Mobile technologies   Less than $0.1 billion  
National electronic benefit transfer (EBT) Gatewayb   $0.1 billion  
SNAP benefit transfer transaction data report   Less than $0.1 billion  
Adjustment to percentage of recovered funds retained by states   $0.1 billion  
Repeal of SNAP high-performance bonuses   -$0.4 billion  
Grants for Non-SNAP Programs and Pilots   $2.8 billion  
Retailer-funded incentives pilots   $1.2 billion  
Gus Schumacher Food Insecurity Nutrition Incentive Program   $0.5 billion  
Nutrition education   $0.6 billion  
Emergency food assistance   $0.5 billion  
Total nutrition title   -$1.9 billion  

Details do not add to total because of rounding and interactions among provisions.

a Net effect of categorical eligibility cut (including child nutrition effects) and asset improvements.

b Includes prohibition on EBT processing fees (section 4018) and direct spending and revenues from the National Gateway (section 4022).

Source: Congressional Budget Office (CBO), H.R. 2, Agriculture and Nutrition Act of 2018, as introduced April 12, 2018, and as passed by the House,

State-by-State Impact of Proposed SNAP Cuts in H.R. 2: Eliminating Expanded Categorical Eligibility and Time-Limit Waivers
  Has the State Adopted the Expanded Categorical Eligibility Option?a Current or Past Waiver from the Three-Month Time Limit?
State/Territory Asset Test Income Test  
Alabama Yes   Past
Alaska     Current
Arizona Yes Yes Current
Arkansas     Past
California Yes Yes Current
Colorado Yes Yes Current
Connecticut Yes Yes Current
Delaware Yes Yes N/A
District of Columbia Yes Yes Current
Florida Yes Yes Past
Georgia Yes   Current
Hawaii Yes Yes Current
Idaho Yes   Current
Illinois Yes Yes Current
Indiana Yes   Past
Iowa Yes Yes Past
Kansas     Past
Kentucky Yes   Current
Louisiana     Current
Maine Yes Yes Past
Maryland Yes Yes Current
Massachusetts Yes Yes Current
Michigan Yes Yes Current
Minnesota Yes Yes Current
Mississippi     Past
Missouri     Past
Montana Yes Yes Current
Nebraska Yes   Past
Nevada Yes Yes Current
New Hampshire Yes Yes Current
New Jersey Yes Yes Current
New Mexico Yes Yes Current
New York Yes Yes Current
North Carolina Yes Yes Past
North Dakota Yes Yes Current
Ohio Yes   Current
Oklahoma Yes   Past
Oregon Yes Yes Current
Pennsylvania Yes Yes Current
Rhode Island Yes Yes Current
South Carolina Yes   Past
South Dakota     Current
Tennessee     Current
Texas Yes Yes Past
Utah     Current
Vermont Yes Yes Current
Virginia     Current
Washington Yes Yes Current
West Virginia Yes   Current
Wisconsin Yes Yes Past
Wyoming     Past
Guam Yes Yes Current
Virgin Islands Yes Yes Current
United States 43 have adopted 32 have adopted 36 have in 2018

a These states have adopted broad-based categorical eligibility. Additional states have narrow categorical eligibility (beyond cash assistance, but not affecting large numbers of households) and may also have some households that would be cut off from SNAP.

Sources: USDA, Food and Nutrition Service, Broad-Based Categorical Eligibility Chart, and additional CBPP research; see and dependents-abawds

Table 5
Estimates of Households and Individuals Losing SNAP Due to Rolling Back State Option for a Higher Gross Income Limit Under Expanded Categorical Eligibility
(For states that have adopted the option, based on number participating in 2015 and 2016a)
State/Territory Gross Income Limit Households Individuals
Arizona 185% FPL 12,000 30,000
California 200% 37,000 85,000
Colorado 200% n/a n/a
Connecticut 185% 11,000 24,000
Delaware 200% 3,000 6,000
District of Columbiab 200% Less than 1,000 1,000
Florida 200% 60,000 130,000
Hawaii 200% 3,000 8,000
Illinoisb 165% 3,000 7,000
Iowa 160% 6,000 16,000
Maine 185% 3,000 8,000
Maryland 200% 15,000 37,000
Massachusetts 200% 17,000 45,000
Michigan 200% 20,000 45,000
Minnesota 165% 6,000 20,000
Montana 200% 2,000 4,000
Nevada 200% 6,000 13,000
New Hampshire 185% 2,000 7,000
New Jersey 185% 12,000 35,000
New Mexico 165% 4,000 10,000
New Yorkb 150%/200%c 15,000 40,000
North Carolina 200% 20,000 50,000
North Dakota 200% 1,000 3,000
Oregon 185% 15,000 37,000
Pennsylvania 160% 30,000 65,000
Rhode Island 185% 4,000 9,000
Texas 165% 45,000 125,000
Vermont 185% 3,000 7,000
Washington 200% 25,000 60,000
Wisconsin 200% 13,000 30,000
Guam 165% 1,000 2,000
Virgin Islandsb 175% Less than 1,000 1,000
United States 32 states have adopted About 400,000 About 960,000

a These states have adopted broad-based categorical eligibility. Additional states have narrow categorical eligibility (beyond cash assistance, but not affecting large numbers of households) and may also have some households that would be cut from SNAP.

b Numbers should be viewed with caution because of small sample sizes.

c In New York, households with dependent care expenses are eligible up to 200 percent FPL and households with earnings are eligible up to 150 percent FPL.

Sources: USDA, Food and Nutrition Service, Broad-Based Categorical Eligibility Chart and CBPP analysis of FY 2015 and 2016 SNAP Household Characteristics data. See

Illustrative Example of People Potentially Subject to Work Requirement and Employment and Training Grant Had Work Requirement Applied in 2016
State/Territory Number of non-disabled adults without children under 6 in a typical month of FY2016   Number of non-disabled adults without children under 6 not working 20 hours per week in a typical month of FY2016 Estimated SNAP Employment and Training Granta
Alabama  186,000  147,000 $19,960,000  
Alaska  22,000  19,000 $2,524,000  
Arizona  218,000  181,000 $24,561,000  
Arkansas  88,000  71,000 $9,591,000  
California  1,193,000  1,015,000 $137,656,000  
Colorado  90,000  66,000 $8,959,000  
Connecticut  98,000  77,000 $10,510,000  
Delaware  30,000  22,000 $3,038,000  
District of Columbia  38,000  34,000 $4,583,000  
Florida  769,000  595,000 $80,695,000  
Georgia  363,000  299,000 $40,596,000  
Hawaii  32,000  25,000 $3,389,000  
Idaho  26,000  18,000 $2,379,000  
Illinois  501,000  397,000 $53,896,000  
Indiana  143,000  105,000 $14,272,000  
Iowa  74,000  56,000 $7,634,000  
Kansas  44,000  30,000 $4,123,000  
Kentucky  157,000  135,000 $18,308,000  
Louisiana  192,000  155,000 $21,087,000  
Maine  29,000  20,000 $2,744,000  
Maryland  174,000  143,000 $19,381,000  
Massachusetts  127,000  98,000 $13,310,000  
Michigan  352,000  269,000 $36,466,000  
Minnesota  80,000  62,000 $8,352,000  
Mississippi  132,000  101,000 $13,737,000  
Missouri  165,000  131,000 $17,797,000  
Montana  25,000  18,000 $2,447,000  
Nebraska  32,000  24,000 $3,213,000  
Nevada  107,000  90,000 $12,195,000  
New Hampshire  12,000  9,000 $1,173,000  
New Jersey  146,000  106,000 $14,330,000  
New Mexico  104,000  76,000 $10,258,000  
New York  544,000  395,000 $53,570,000  
North Carolina  351,000  268,000 $36,286,000  
North Dakota  9,000  6,000 $869,000  
Ohio  344,000  263,000 $35,664,000  
Oklahoma  125,000  97,000 $13,213,000  
Oregon  200,000  161,000 $21,806,000  
Pennsylvania  353,000  265,000 $35,957,000  
Rhode Island  35,000  27,000 $3,637,000  
South Carolina  154,000  126,000 $17,060,000  
South Dakota  17,000  13,000 $1,700,000  
Tennessee  281,000  230,000 $31,191,000  
Texas  571,000  402,000 $54,452,000  
Utah  31,000  23,000 $3,118,000  
Vermont  13,000  9,000 $1,266,000  
Virginia  150,000  112,000 $15,158,000  
Washington  233,000  188,000 $25,537,000  
West Virginia  87,000  73,000 $9,861,000  
Wisconsin  148,000  105,000 $14,208,000  
Wyoming  6,000  4,000 $598,000  
Guam  9,000  7,000 $916,000  
Virgin Islands  7,000  6,000 $770,000  
United States  9,415,000  7,374,000 $1,000,000,000  

a We allocated the $1 billion funding of employment and training programs for fiscal year 2021 to each state based on its share of non-disabled adults without children under 6 who are not working 20 hours per week in a typical month of FY2016.  These estimates exclude households with gross income greater than the federal limit via the categorical eligibility option.

Note: Individual state totals may not add up to the U.S. total due to rounding.

Source: CBPP analysis of 2016 SNAP Household Characteristics data.

Appendix I: Amendments Approved to Committee Bill on the House Floor

On May 18, the House adopted several amendments to the original, Agriculture Committee-approved bill before voting it down. On June 21, the House reconsidered and passed the bill as amended. While a few changes made improvements to protect vulnerable households, the amendments as a whole made the bill even more harmful for low-income households, including by increasing benefit cuts from about $17 billion to nearly $19 billion.[41] The amendments would:

  • Delay the restriction of categorical eligibility. The bill provision that severely narrows state options under categorical eligibility would take effect in fiscal year 2021 instead of immediately upon enactment. CBO estimates this change would increase federal costs by $1.07 billion over ten years relative to the original bill.
  • Retain an administrative simplification for households with a disabled member that have out-of-pocket utility expenses. For households that include a person with disabilities, this would allow states to continue using the receipt of benefits under LIHEAP to qualify them for a SNAP income deduction tied their expenses for utilities like heating and electricity. (The bill eliminates this option for all households without a disabled or elderly member.) As a result, fewer individuals who were unable to demonstrate their utility costs would lose benefits. CBO estimates this change would increase federal costs by $2.32 billion over ten years relative to the original bill.
  • Make it even harder for states to protect vulnerable individuals from harsh work provisions. First, the bill would make it even harder than the original bill for states to get federal waivers for high-unemployment areas. The original bill made areas with unemployment 20 percent higher than the national average ineligible for waivers if their unemployment was below 6 percent. An adopted amendment raised this floor to 7 percent. And while the original bill only applied the bill’s more restrictive waiver provisions to new work requirements (which would take effect in 2021), the amended bill would also apply them to the current three-month time limit on childless adults in 2019 and 2020. As a result, 600,000 more people would lose SNAP in 2021 than under the original bill.

Second, the same amendment cuts, from 15 to 12 percent, the share of adults subject to the work requirements that states could exempt. It also changes the underlying pool of participants used to calculate the maximum percentage that states can exempt, which can be interpreted as expanding the pool. The net effect of these two changes on the number of exemptions available to states is unclear.

Third, the amendment adds workfare as an activity that counts towards the bill’s minimum work or training hours requirement.

CBO estimates that these and other changes to the work provisions would decrease federal spending on SNAP benefits by $4.9 billion over ten years relative to the original bill.

  • Cut the bill’s new funding for job training, which already was grossly inadequate. The House adopted an amendment that would require states to return unspent E&T funds to the federal government. Under current law, unspent funds can be redistributed between states to help those with higher capacity to absorb these funds and expand their programs. CBO estimates this will decrease federal spending by $350 million over ten years relative to the original bill.
  • Open the door for states to give key SNAP functions to for-profit companies. The provision to permit states to contract with private companies for eligibility and enrollment services was not in the original bill but was adopted on the House floor. CBO estimates this provision would have a negligible effect on federal spending.
  • Deny food assistance to persons with certain criminal convictions who were released from prison and are working to reintegrate into society. The provision to restrict eligibility for individuals with certain criminal convictions was not in the original bill but was adopted on the House floor. CBO estimates this change would cut SNAP benefits by $23 million over ten years.
  • Make multivitamin-mineral and other dietary supplements eligible for purchase with SNAP benefits. The provision would let participants use their SNAP allotment to purchase dietary supplements. CBO estimates this provision would have a negligible effect on federal spending in fiscal year 2019.
  • Add provisions to review benefit use in group living facilities; adjust Commodity Supplemental Food Program certification period; review certain child nutrition program regulations; and study food costs and food assistance in Puerto Rico.

Appendix II: Other Provisions in the House Nutrition Title

The House bill includes numerous other provisions related to SNAP operations and issuance as well as funding for grants, pilot projects, and programs outside of SNAP. In total, the other SNAP provisions would reduce SNAP spending by about $600 million over ten years, and the grants and pilots outside of SNAP would cost $2.8 billion over ten years, according to CBO. The bill would pay for the additional amount needed to fund the grants and pilots by cutting benefits to low-income households, thus reducing their resources to buy food. While several of these proposals offer useful policy changes to important operational issues, others raise concerns or are not sufficiently explained in the Chairman’s or Committee’s materials.

SNAP Operations and Benefit Issuance

Altering benefit accrual and expungement procedures. SNAP benefits that go unspent need to be returned to the federal government, and existing law has a sound policy to deal with this. Currently, states may make unused benefits inaccessible by moving them offline after six months (and restore them at the beneficiary’s request) and may cancel (or expunge) unused benefits after 12 months. Even though only a very small share of total SNAP benefits sit unused — 97 percent are redeemed within a month of issuance[42]— the House bill would abbreviate these timeframes, allowing states to make unused benefits inaccessible after three months and cancel unused benefits after six months.

Very little is known about why some benefits go unused and what the appropriate policy change should be. Unused benefits may reflect lack of access issues rather than lack of need. Some households, particularly seniors and those with intellectual disabilities, may face barriers to redeeming their assistance. They may need help understanding how to use their benefits and when they are replenished, or they may not have regular access to transportation to get to a store. Alternatively, accruals may reflect that a household is economizing a modest benefit. Recipients of small SNAP monthly benefits may allow their benefits to accrue over several months so that they can cover household food costs entirely for one shopping trip or one month.

This proposal also would likely increase costs associated with electronic benefit transfer (EBT) contracts between states and EBT vendors. These new requirements could necessitate a change in the terms of service and may result in an increased charge by the companies that support EBT.

CBO estimates this provision would have a negligible effect on federal spending. (Section 4020.)

Eliminating the error tolerance threshold. When households apply for SNAP, states conduct a complex process to determine their eligibility based on income, expenses, and other household characteristics. Households must report their income and other relevant information, and states verify the accuracy of this information using data matches or paper documentation. In addition, SNAP has a quality control (QC) process to ensure the accuracy of household eligibility and benefit amounts. States must sample a representative number of cases each month and state QC staff thoroughly review the accuracy of the original eligibility determination and benefit level. USDA further reviews a subset of these reviews and, based on these reviews, annually reports state overpayment and underpayment error rates; states can be assessed significant penalties for above-average error rates.

To encourage states to focus their payment integrity efforts on the costliest types of errors, SNAP’s QC rules have long had a “tolerance” level below which error amounts do not count toward the state’s error rate. The current level, set by the 2014 farm bill, is $37, with annual adjustments for inflation. From 2000 to 2009, the QC tolerance level was set at $25 (with no adjustments for inflation). In 2009, Congress temporarily raised the tolerance level to $50 as a part of the Recovery Act to acknowledge states’ concern that the Recovery Act’s temporary increase in household benefits would raise states’ error rates without indicating fault in program administration. The House bill would eliminate the error threshold altogether and set the amount at zero.

By definition, this would increase SNAP’s payment error rate by including more small errors in the calculation. [43] It would also impede states from focusing on larger errors by including any small error that results from a minor mistake. Some states may respond by requiring more paperwork and imposing other rules that aim to reduce errors but are more likely to create administrative burden and deter access to the program or reduce benefits.

CBO estimates this provision would have a negligible effect on federal spending. (Section 4028.)

Increasing state retention of recovered overpayments. When states discover that they’ve made an overpayment to a SNAP household, they must (with limited exceptions) file a claim against the household and attempt to recover the overpaid funds. Under current law, states can retain 35 percent of the funds they recover from overpayments that they determine were not due to agency error. This provides a modest incentive to states to pursue the claims, as SNAP benefits are 100 percent federally funded and would otherwise be fully returned to the Treasury. The House bill would raise the state share to 50 percent and require states to use these funds for SNAP administration and operation.

It is unclear whether this proposed change would increase states’ efforts to pursue outstanding claims or whether states would just keep a greater share of the money they already collect.

CBO estimates this provision would increase federal spending by $102 million over ten years. (Section 4027.)

Eliminating state performance bonuses. The House bill would eliminate $48 million a year in SNAP performance bonuses beginning in 2018, which CBO estimates would cut SNAP spending by $432 million over ten years. These modest, targeted payments reward states for high or improving performance related to program error rates and payment accuracy, application processing timeliness, and overall program access.

A 2015 report from USDA’s Office of Inspector General (OIG) suggested that some states, motivated in part by the bonus incentives, were introducing bias into their QC measures regarding program error and payment accuracy. USDA reviewed all states, executed corrective action plan agreements, and revised QC policy guidance and practices and procedures. Food and Nutrition Service (FNS) Administrator Brandon Lipps testified before the Senate Agriculture Committee that he believed states and FNS oversight of the QC system were on track to correct for the measurement error.[44] It is premature to eliminate these bonuses before giving the agency and states time to address the concerns raised by the OIG report.

Eliminating all performance bonuses, as opposed to only those for exemplary or improved payment accuracy and error rates, goes beyond any concerns raised by the OIG. This proposal would remove important incentives for states to strengthen program administration, including by achieving high participation rates and providing benefits within federal timeliness standards. (Of the $48 million that states received in performance bonuses in 2015, $18 million was for high or improved performance in timeliness or access.) Further, states use these bonus payments to re-invest in SNAP, for example through improvements in technology, administration, and program integrity.

CBO estimates this provision would cut federal SNAP spending by $432 million over ten years. (Section 4029.)

Imposing a new state requirement for EBT replacement cards. The 2014 farm bill set national policy regarding replacement EBT cards. Federal rules now require a household requesting more than four replacement cards in one year to explain why they have requested so many replacements; the SNAP agency must also review program rights and responsibilities with the household. The House bill would require the state to review program rights and responsibilities after two card replacement requests, but would not change the rule requiring an explanation from the household only after the fourth replacement request.

CBO estimates this provision would not change federal spending. (Section 4019.)

Expanding USDA access to state systems. Currently, state SNAP agencies are responsible for maintaining records about the certification of applicant and participating households and about EBT issuance, and must make those records available for federal audit. In response to issues raised by FNS and USDA’s Office of Inspector General about not having full and timely access to SNAP records, the House bill would amend this provision to also make state information systems (i.e., computer systems) that house SNAP records open for federal audit.

While USDA needs to be able to access such records, this provision may provoke some concerns from state agencies. They may be concerned about whether the requirement would contradict state privacy laws and policies, and states with systems that are integrated with other public benefits programs may be concerned about protecting non-SNAP client information.

CBO estimates this provision would not change federal spending. (Section 4023.)

Prohibiting grant funds from being used to improve access to benefits. Currently, Congress funds USDA to provide grants to states (and other entities) for projects that simplify SNAP application and eligibility systems and/or improve access to benefits. The House bill would eliminate USDA’s authority to fund projects aimed at improving program access.

This could hinder USDA’s ability to support innovative uses of technology that address barriers to participation. USDA has used this authority to support innovative projects to create mobile apps that make applying for SNAP easier but that also provide clients with real-time information about their SNAP benefits, making it easier for them to ensure they are complying with program requirements. USDA has also provided grants to improve communication between SNAP agencies and clients by utilizing text messaging and automated voice messaging.

CBO estimates this provision would not change federal spending. (Section 4025.)

Requiring all EBT transactions to be routed through a national gateway and eliminating EBT processing fees. The House bill would establish a national mechanism (or “gateway”), which USDA would oversee, through which all SNAP EBT transactions would be routed to benefit issuers for validation and settlement. Currently, they are routed through third-party commercial processors. Routing EBT transactions through a single national gateway would give the federal government more oversight into transactions and could lower costs. The provision directs the Secretary to conduct a feasibility study in advance of national implementation.

The proposal also would expand and clarify that EBT processors (the companies involved in various aspects of EBT transactions) may not charge certain “interchange” fees or other fees for “switching” or “routing” EBT transactions.

The proposal would require the Secretary to establish fees that benefit issuers and third-party processors would pay to the gateway that would offset some of the gateway’s federal costs.

CBO estimates these provisions, on net, and after accounting for interactions, would increase federal spending by about $150 million over ten years. (Sections 4018 and 4022.)

Expanding the state option for EBT cardholder identification. Current law allows states to require photographs on EBT cards, which has proven expensive and ineffective. The House bill would allow the use of other technologies to confirm the identity of the cardholder, like biometrics, which likely would create barriers to access and raise potential privacy concerns.

CBO estimates this provision would not change federal spending. (Section 4016.)

Authorizing the use of mobile technologies in stores to redeem SNAP benefits. Currently, retail food stores can redeem SNAP benefits through mobile technologies (i.e., smartphone apps in lieu of cards), provided they establish privacy protections for recipients and pay implementation costs. The House bill would shift responsibility for privacy protections and implementation costs to state agencies and requires up to five state pilot demonstration projects to test the technology before the Secretary authorizes implementation.

CBO estimates this provision would increase federal spending by $12 million over ten years. (Section 4017.)

Enabling SNAP benefits to be redeemed online. Currently, online merchants are not included in the definition of retail food stores. The House bill would update this definition to enable online merchants to accept SNAP benefits.

CBO estimates this provision would not change federal spending. (Section 4021.)

Establishing an annual transaction data report. The House bill would allow the Secretary to collect store transaction data that would include the cost and description of items purchased with SNAP benefits; USDA would use these data to compile a summary report on SNAP benefit use that is available to the public. While the report would not identify individual retail food stores or household members, the proposal does not include safeguards to protect the data or prohibit their use for other purposes. Collecting and reporting the data could be a significant burden for retailers, particularly small enterprises such as individually owned stores or farmers’ markets, which may not have or maintain individual customer purchase data.

CBO estimates this provision would increase federal spending by $30 million over ten years. (Section 4026.)

Providing funding for USDA implementation. The House bill would provide $150 million to USDA to carry out the SNAP changes in the bill. (Section 4036.)

Reviewing SNAP Operations in group living facilities. The bill requires the Secretary to conduct a review of the use of SNAP benefits in certain group living arrangements, including senior centers, and report results to Congress.

CBO estimates this provision would not change federal spending. (Section 4038.)

Extending authorization of appropriations to prevent retailer and recipient trafficking. The bill extends through 2023 the authorization for appropriations of up to $5 million per year to support USDA’s efforts to prevent trafficking of SNAP benefits.

CBO estimates this provision would not change federal spending, (Section 4034.)

Making multivitamin-mineral dietary supplements eligible for purchase with SNAP benefits. The bill would allow SNAP participants to use their benefit allotment to purchase dietary supplements, despite a scientific consensus that nutritional needs be met primarily through food consumption.[45] Vitamin and mineral supplements, as their name suggests, are meant to complement a food diet. However, for SNAP recipients who have limited budgets, expenditures on dietary supplements would likely substitute for food. A growing body of research suggests that the modest SNAP benefit — which averages just $1.40 per person per meal — already may be inadequate to fully meet the needs of eligible households.[46] Therefore, diverting food spending to supplements, which can be costly and are not regulated in a way that guarantees efficacy, could put participants’ food security at risk without improving nutrition.

CBO estimates this provision would have a negligible effect on federal spending in fiscal year 2019. (Section 4037.)

Other Food Assistance Programs

Increasing funding for the Emergency Food Assistance Program (TEFAP). TEFAP supplements the diets of low-income Americans, including seniors, by providing emergency food through local agencies, primarily food banks. The House bill would increase funding for this critical program. Currently, TEFAP is expected to receive about $290 million in entitlement funding under SNAP in fiscal year 2019; the proposal would increase this amount by $45 million, with inflation adjustments in later years.

CBO estimates this provision would increase federal spending by $499 million over ten years. (Section 4032.)

Making modest changes to the Food Distribution Program on Indian Reservations (FDPIR). FDPIR provides food packages to low-income households on reservations and to Native American families living in designated areas near reservations and in Oklahoma. Currently, some program funds are dedicated to the purchase of traditional and locally grown foods. The House bill would also permit these funds to be used for the purchase of regionally grown foods. It would also extend the availability of FDPIR funds from one fiscal year to two.

CBO estimates this provision would not change federal spending. (Section 4005.)

Reauthorizing the Commodity Distribution Program. The House bill would extend through 2023 the Secretary’s authorities to purchase and distribute agricultural commodities, to implement the Commodity Supplemental Food Program, and to continue to distribute surplus commodities to the agency’s nutrition programs.

CBO estimates these provisions would not change federal spending. (Sections 4101, 4102, and 4104.)

Defining certification period for Commodity Supplemental Food Program (CSFP) participants. The bill mandates a minimum certification period for seniors participating in CSFP of not less than one year and allows the Secretary to approve state requests for longer certification periods.

CBO estimates this provision would not change federal spending. (Section 4103.)

Conducting studies on Puerto Rico’s Nutrition Assistance Program (NAP). Unlike most of the United States that participates in the SNAP entitlement, Puerto Rico receives a block grant to operate NAP, its nutrition assistance program for low-income households. Because of its capped funding structure, eligibility and benefit levels for NAP are well below those for SNAP. The bill would require USDA to conduct two new studies about the cost of a healthy diet in the Commonwealth and the feasibility of implementing SNAP in Puerto Rico.

CBO estimates these provisions would not change federal spending. (Sections 4040 and 4042.)

Reviewing child nutrition regulations. The bill requires the Secretary to review and reissue regulations on the meal pattern standards for the national school breakfast and lunch programs and standards for all other foods sold in schools. The provision specifies that the new standards must be based on research on school-age children and must not increase the costs of the programs beyond the amount reimbursed to states by the federal government. The current meal pattern standards for these programs are based on the most recent scientific evidence. Imposing other contingencies, like cost-neutrality, would undermine the evidence-based nature of these standards and potentially compromise the nutritional value of meals served to children during the school day.

CBO estimates this provision would not change federal spending. (Section 4205.)

Making changes to the services of traditional foods in public facilities. The bill makes minor changes to USDA and the Food and Drug Administration’s ability to accept donated traditional foods, including by explicitly allowing such foods to be served in federally funded child nutrition programs and senior meal programs. (Section 4041.)

Programs and Initiatives to Support Healthy Eating

Changing SNAP’s nutrition education program. The SNAP Nutrition Education and Obesity Prevention Grant Program (SNAP-Ed) is an evidence-based program that provides grants to help SNAP households eat well and live healthier lives on a limited budget. The House bill would have the grant program run through the National Institute of Food and Agriculture, in consultation with FNS, and would make the grants available to land-grant institutions. The proposal increases the grants to states for nutrition education by $57 million in 2019, adjusted in later years for inflation.

CBO estimates this provision would increase federal spending by $632 million over ten years. (Section 4033.)

Establishing a retailer-funded incentive pilot. The House bill directs USDA to establish a pilot program through which food retailers can provide financial incentives for SNAP participants’ purchase of fruits, vegetables, and fluid milk. USDA would reimburse retailers 25 percent of the cost of bonuses provided to SNAP customers up to $120 million per year.

This provision raises questions about whether funds would be used to create new retail incentives or rather to buy out retailers for a share of existing store coupon programs. Furthermore, this approach is less efficient than providing food assistance directly to households to purchase healthy foods.

CBO estimates this provision would increase federal spending by $1.2 billion over ten years. (Section 4002.)

Reauthorizing the Food Insecurity Nutrition Incentive Program (FINI). The House bill would reauthorize FINI grants, which provide incentives to improve nutrition and test the effectiveness of those incentives. It would limit incentives to financial incentives; establish a training, evaluation, and information center for FINI grantees; and update program priorities. Funding would be provided for ten years, beginning at $45 million for 2019 and rising by $5 million a year to $65 million for 2023 and later years.

CBO estimates this provision would increase federal spending by $472 million over ten years. (Section 4003.)

Creating new pilots on public-private partnerships. The House bill would establish up to ten pilot projects to test the effectiveness of public-private partnerships in addressing food insecurity and poverty.

The bill would authorize the appropriation of $5 million to conduct these pilots. (Section 4030.)

Reevaluating the Thrifty Food Plan. USDA designs four food plans meant to reflect an adequate diet at different cost levels. The lowest cost of these is the Thrifty Food Plan (TFP). The House bill would require USDA to update the TFP every five years, beginning in 2020, based on “current food prices, food composition data, and consumption patterns.”

CBO estimates this provision would not change federal spending. (Section 4004.)

Reauthorizing miscellaneous nutrition programs. The House bill includes a subtitle of miscellaneous provisions that include language to (1) extend the Fresh Fruit and Vegetable Program for schools and service institutions; (2) extend the Senior Farmer’s Market Nutrition Program; (3) extend the authorization of appropriation for the Healthy Food Financing Initiative; and (4) amend the Fruit and Vegetable School Lunch Program to allow not only fresh fruits and vegetables to be provided, but also those that are canned, dried, frozen, or pureed.

CBO estimates these provisions would not change federal spending. (Sections 4201, 4202, 4203, and 4204.)


End Notes

[1] H.R. 2: the Agriculture and Nutrition Act of 2018,

[2] Dottie Rosenbaum, “House Farm Bill’s SNAP Changes Are a Bad Deal for States and Low-Income Households,” May 15, 2018, Center on Budget and Policy Priorities, May 15, 2018,

[3]American Public Human Services Association, Letter to House Committee on Agriculture re: H.R. 2, May 2, 2018,

[4] “Senate’s Bipartisan Farm Bill Strengthens SNAP and Avoids Harming SNAP Households, Center on Budget and Policy Priorities, updated July 30, 2018,

[5]This estimate is calculated by taking $1 billion a year in employment and training funding and dividing it by our estimate of the number of work slots that the bill would require states to create. States could augment the available funding by contributing their own funds and drawing down a federal match. See Table 2 for more details on this estimate.

[6] Ed Bolen and Stacy Dean, “Waivers Add Key State Flexibility to SNAP’s Three-Month Time Limit,” Updated February 6, 2018, Center on Budget and Policy Priorities,

[7] For resources on SNAP and work, see

[8] Brynne Keith-Jennings and Raheem Chaudhry, “Most Working-Age SNAP Participants Work, But Often in Unstable Jobs,” Center on Budget and Policy Priorities, March 15, 2018,

[9] To derive this estimate, we assumed that the share of participants subject to the work requirement relative to the total SNAP caseload will remain constant between 2016 and 2021.

[10] States would not need to track all of the nearly 8 million participants potentially subject to work requirements because some would live in areas waived for the full fiscal year. For exemptions based on individual circumstances, states may need to collect information to recertify an exemption each month, or USDA could determine that states have the authority to provide exemptions for longer periods ― for example, a health care provider may document that an illness will prevent a participant from working for several months.

[11] This does not include all individuals who received WIOA-funded case management services and other individualized or basic services, as distinguished from job training services.

[12] LaDonna Pavetti, “Opportunity-Boosting Job Preparedness Takes Significant Investment, Evidence Shows,” Center on Budget and Policy Priorities, April 12, 2018,

[13] Elizabeth Wolkomir and Stacy Dean, “Will Chairman Conaway’s SNAP Proposal Match His Rhetoric?” Center on Budget and Policy Priorities, April 12, 2018,

[14] The bill would allow states to waive areas with elevated unemployment from the work requirement. We estimate that these proposed waivers would protect about half as many people as the current geographic waivers from SNAP’s three-month time limit for able-bodied adults without dependents. For more information see:

[15] LaDonna Pavetti, “Work Requirements Don’t Cut Poverty, Evidence Shows,” Center on Budget and Policy Priorities, June 7, 2016,

[16] Connie Wanberg et al., “Age and Reemployment Success After Job Loss: An Integrative Model and Meta-Analysis,” Psychological Bulletin 2016, Vol. 142, No. 4, 400-426.

[17] As written, the provision seems to remove states’ authority to request geographic waivers from the current three-month time limit in 2019 and 2020 altogether. However, as interpreted and scored by CBO, the provision seems intended to impose the revised waiver parameters on the time limit in fiscal years 2019 and 2020 before the new work requirements take effect.

[18] Elizabeth Wolkomir and Lexin Cai, “The Supplemental Nutrition Assistance Program Includes Earnings Incentives,” Center on Budget and Policy Priorities, updated March 6, 2018,

[19]Under the categorical eligibility option, states may not raise the gross income limit above 200 percent of the poverty line, though most households that are helped have gross income just above 130 percent of poverty. The calculation cited here assumes that the family receives only earned income, claims the $160 SNAP standard deduction and 20 percent earned income deduction, and has $1,188 in monthly shelter costs (the median value for working households of three with children that have incomes at or above 125 percent of poverty based on the fiscal year 2016 SNAP Household Characteristics data, inflated to fiscal year 2018 dollars).

[20] Six states — Florida, Idaho, Kansas, Michigan, Mississippi, and South Dakota — use this option.

[21] Rodney W. Hopkins, “Food Stamp Child Support Cooperation Study,” Social Research Institute, University of Utah, August 29, 2014.

[22] Child Support Enforcement administrative expenses are matched at 66 percent by the federal government.

[23] Collins et al., “Summer Electronic Benefits Transfer for Children (SEBTC) Demonstration: 2012 – Final Report,” U.S. Department of Agriculture, August 2013,

[24] Steven Carlson et al., “SNAP Works for America’s Children,” Center on Budget and Policy Priorities, September 29, 2016,

[25] Rodney W. Hopkins, “Food Stamp Child Support Cooperation Study,” Social Research Institute, University of Utah, August 29, 2014.

[26] Kye Lippold and Elaine Sorensen, “Characteristics of Families Served by the Child Support (IV-D) Program: 2010 Census Survey Results,” The Urban Institute, July 2013,

[27] Social Research Institute, “Food Stamp Child Support Cooperation Study,” University of Utah, August 29, 2014,

[28] Elaine Sorensen et al., “Assessing Child Support Arrears in Nine Large States and the Nation,” The Urban Institute, July 11, 2007,

[29]Timothy Grall, “Custodial Mother and Fathers and Their Child Support: 2013 – Current Population Reports,” U.S. Census Bureau, January 2016,

[30] National Child Support Enforcement Association, Letter to House Committee on Agriculture re: H.R. 2, May 14, 2018,

[31] The states that use the child support income exclusion in lieu of the deduction are: California, Colorado, Delaware, Illinois, Iowa, Louisiana, Maine, Michigan, New Jersey, New York, Rhode Island, South Dakota, Vermont, and Washington.

[32] Cody Tuttle, “Snapping Back: Food Stamp Bans and Criminal Recidivism,” March 29, 2018; Crystal S. Yang, “Does Public Assistance Reduce Recidivism?” American Economic Review, May 2017,

[33] Ty Jones, “SNAP’S Homeless Shelter Deduction Can Provide Much-Needed Help for Homeless Individuals and Families,” Center on Budget and Policy Priorities, December 2, 2011,

[34] United States Department of Agriculture, “State Options Report: Thirteenth Edition,” August 15, 2017,

[35] Ibid.

[36] CBPP analysis of 2016 SNAP Household Characteristics data. About a quarter of households with earnings would not see a benefit increase because they receive either the program’s maximum benefit or minimum benefit.

[37] See National Accuracy Clearinghouse (NAC) Evaluation, Final Report, PCG, October 2015, p. 9,

[38] U.S. Department of Agriculture, “Dynamics of Supplemental Nutrition Assistance Program Participation from 2008 to 2012,”

[39] Center for Public Policy Priorities, “Updating and Outsourcing Enrollment in Public Benefits: The Texas Experience,” November 2006,; Center for Public Policy Priorities, “News Release: New Report on Texas’ Troubled Outsourcing Experiment Tells Cautionary Tale for Sister States,” November 13, 2006,

[40] Dottie Rosenbaum, “SNAP Caseloads and Spending Declines Track CBO Projections,” Center on Budget and Policy Priorities, May 22, 2017,

[41] On net, the amendments would result in even deeper cuts to SNAP benefits and affect more households and individuals. However, given some amendment provisions that marginally enhance benefits for some participants relative to the original bill, the overall impact would still be to cut or reduce benefits for more than 2 million individuals in over 1 million households.

[42] Laura Castner and Juliette Henke, “Benefit Redemption Patterns in the Supplemental Nutrition Assistance Program,” Food and Nutrition Service, Office of Research and Analysis, February 2011,

[43] Government Accountability Office, “Supplemental Nutrition Assistance Program: Policy Changes and Calculation Methods Likely Affect Improper Payment Rates, and USDA is Taking Steps to Help Address Recipient Fraud,” July 6, 2016, GAO-16-708T,

[44] Statement of Brandon Lipps, Acting Deputy Under Secretary, Food, Nutrition and Consumer Services Administrator, Food and Nutrition Service, United States Department of Agriculture, Before the Senate Committee on Agriculture, September 14, 2017,

[45] U.S. Departments of Agriculture and Health and Human Services, The 2015-2020 Dietary Guidelines for Americans,

[46] Brynne Keith-Jennings, “SNAP Benefits Do Much But Are Often Too Small,” Center on Budget and Policy Priorities, December 9, 2015,