Skip to main content

House Farm Bill’s SNAP Changes Are a Bad Deal for States and Low-Income Households

The House Agriculture Committee farm bill would impose new mandates on states, limit state flexibility, and make the Supplemental Nutrition Assistance Program (SNAP) more complicated to administer.  These changes would increase state costs and significantly undermine almost two decades of progress in simplifying, streamlining, and modernizing SNAP (which until the 2008 farm bill was known as food stamps).  The bill would take away SNAP benefits from recipients across the country who need them while failing to achieve the proponents’ stated goals of improving employment outcomes.The bill would take away SNAP benefits from recipients across the country who need them while failing to achieve the proponents’ stated goals of improving employment outcomes.  The House Agriculture Committee bill would:

  • Force states to collect monthly information from up to 7 million SNAP recipients about their hours worked, their hours of participation in work programs, and the reasons they may not be working or participating in a work program.  The proposed expansion of SNAP’s already strict work requirements would impose substantial administrative costs on states and reverse 15 years of efforts by federal policymakers and states to make access to the program easier for working families that are juggling work and family obligations.  In recent years more than 70 percent of working families eligible for SNAP have participated — in 2002 just 43 percent did.  Reversing course would make SNAP less effective as a work support.
  • Require states to run much larger employment and training (E&T) programs without the flexibility — or funding — to decide whom to target for services and which types of job training would be most effective.  As a result, states would be directed to waste enormous resources on ineffective, poorly targeted employment services rather than being allowed to invest resources in effective job training efforts.  Many states would not be able to continue the innovative SNAP E&T approaches they have undertaken in recent years.
  • Take away options states have used to reduce paperwork and make it easier for people who need food assistance to get SNAP.  States have used existing options to limit the documentation households need to submit to prove that they don’t have much money in the bank, their car isn’t worth too much, and they have utility bills they have already shown to another agency.  And they have used options to make sure that working families with high expenses like child care don’t lose SNAP due to a small increase in earnings. The bill takes away these options, making the program more paperwork-heavy, more expensive for states to operate, and harder for families to access.
  • Mandate that states impose child support requirements that most states have chosen not to impose based on evidence that they would be costly to implement, would not result in significant increases in child support collections, and would risk cutting SNAP to families that need it.  States currently have the option of taking away SNAP from parents who don’t participate in the child support program. Only six states have taken the option, but this bill would mandate it.  Utah studied the option and concluded that it would do more harm than good.

Most state SNAP officials — in states led by Democrats and Republicans alike — share these concerns about the bill.  The American Public Human Services Administration, which represents the heads of state human services agencies, raised concerns about many of these issues in a preliminary response to the bill (see Appendix).

House Bill Would Undermine Decades of Progress in Simplifying SNAP and Increasing State Flexibility

One of SNAP’s longstanding strengths has been its state-federal partnership.  The federal government largely sets eligibility and benefit rules and funds all of SNAP’s food assistance benefits.  States administer SNAP eligibility determinations, issue SNAP benefits, operate the program’s  Quality Control (QC), Employment and Training (E&T), and Nutrition Education components, and are reimbursed for about half of their administrative costs.

For more than 15 years ― under Republican and Democratic administrations ― Congress, the U.S. Department of Agriculture (USDA), and states have engaged in efforts to streamline and modernize the program in ways that have given states more options, maintained a commitment to program integrity, and made it easier for households that qualify to access SNAP.  These efforts have resulted in higher participation among eligible households and strong payment accuracy.  

The efforts to streamline the program to make it easier for states to administer and households to access came after participation in the program declined precipitously following the enactment of the 1996 welfare law.  That law included significant SNAP cuts and a new three-month time limit on participation for certain unemployed childless adults, but the decline in participation was far larger than anticipated. Policymakers of both parties became concerned that households that were eligible for SNAP and faced difficulty affording food were falling off the program or failing to apply — including many families with children that left cash assistance and were working, but remained poor.  One reason these families were losing SNAP was because of program rules that made it hard for families with earnings to participate, such as requirements that participants had to report to the state even very small changes in monthly earnings, which are frequent in low-wage jobs.

Aggravating these problems, some states instituted administrative practices in those years that had the unintended effect of making it even harder for many working-poor parents to participate, largely by requiring them to take too much time off from work for repeated visits to SNAP offices at frequent intervals, such as every three months, to reapply for benefits. 

Falling participation among households that struggled to make ends meet and were eligible for SNAP prompted many analysts and state policy officials from across the political spectrum to call for reforms that would improve low-income working families’ access to SNAP.  They called for changes that would reduce paperwork requirements on families and strike a more appropriate balance between ensuring that SNAP benefits were calculated correctly and recognizing that households’ circumstances change frequently and capturing small changes to avoid an error in the benefit calculation can do more harm than good. 

Both the Clinton and the George W. Bush administrations acted to address these issues.  In addition, Congress enacted significant, although relatively modest, changes in 2002 and 2008 to lessen barriers to SNAP participation by, among other things, reducing low-wage working households’ number of trips to the SNAP office and reporting and documentation of modest changes to their circumstances.  States also made changes to their own procedures — including online applications, greater use of telephone call centers in lieu of in-person office visits, and procedures that made it easier to document basic information — to help make the program easier for households to navigate. 

These efforts paid off.  The share of eligible individuals in low-income working families that receive SNAP rose from 43 percent in 2002 to more than 70 percent in recent years.  These gains were retained during and after the recent deep recession.  (See Figure 1.)  Table 1 in the Appendix shows that SNAP participation rates have increased in every state since 2002, especially for households with earnings.

The House Agriculture Committee bill would put this progress at risk by imposing new mandates and costs, increasing paperwork requirements, and curtailing state flexibility.

Bill Would Impose Costly New State Mandates

The bill would impose costly new state mandates in areas where states currently have programmatic flexibility.[1]  First, it would require states to impose work requirements on a broad set of SNAP recipients where they previously had flexibility to target employment services and requirements and would require them to expand employment and training programs on a scale that is both unprecedented and unrealistic.  The bill also would require states to impose new requirements on parents to participate in the state-run child support program, even when states determine that the requirement would not be cost-effective and could harm children. Currently states have policy options in this area and just six states have opted to implement this kind of mandate. 

As a result of these new state mandates, the bill would increase state administrative costs and would reduce food assistance to many households, including households with children, despite research that shows that SNAP is linked to improved long-term health and education outcomes for children.[2]

Work Requirement Mandate

The bill includes new sweeping, prescriptive nationwide work requirements on certain adults who receive SNAP that would force states to develop large new bureaucracies to track millions of SNAP recipients on a monthly basis and would limit the types of employment and training services states can offer.  The changes run counter to the research and state experience, which suggest that these types of requirements do little to increase employment. And, the vast system required to track millions of SNAP recipients’ employment, work program participation, and exemptions each month would once again make the program harder for states to implement and families to access.

It is important to note that most working-age adult SNAP recipients are, indeed workers, and SNAP plays a vital role in supporting them both while they’re working and when they’re between jobs.[3] 

SNAP currently has several interrelated provisions related to work and training that provide states a degree of flexibility. 

  • First, for over 30 years, states have had flexibility to offer a broad range of employment and training services to SNAP participants who might benefit from such services and who opt to participate.  They can tailor the services to match their local labor markets and target populations with the types of services that they deem most appropriate. 
  • Second, states may also elect to impose very tough work requirements (up to 30 hours a week) on working-age adults (age 18 through 59, with limited exceptions), may require such individuals to participate in an employment and training program, and may sanction benefits for those who do not comply. 
  • Finally, states must limit SNAP to just three months out of every three years for individuals who are aged 18 through 49 who are not raising minor children and who are working less than 20 hours per week (with limited exceptions).  States do not have to offer people subject to the time limit a work slot that would qualify them to continue participating in SNAP, and most don’t.[4]  As a result, in most states people subject to the time limit are cut off SNAP after three months.  This is the most prescriptive of the requirements; states must impose the time limit unless the area in which the individual lives has high unemployment and the state chooses to seek a waiver from the time limit for the area.

SNAP provides $110 million a year in federal grants for employment and training programs and matches additional state spending for employment and training programs and for job-related costs such as child care and transportation.  USDA and states spent more than $700 million on SNAP employment and training programs in 2016. 

The Agriculture Committee bill would require individuals aged 18 through 59 who are not disabled and do not have children under 6 to prove every month that they’re working 20 hours or more a week, participating in a qualifying job training program for 20 hours or more a week (or a combination of the two that totals at least 20 hours a week), or that they should be exempted from the provision.  An individual would be subject to a 12-month sanction after just one month of not meeting this requirement.  The second month of non-compliance would trigger a three-year sanction.  The only way an individual could end a sanction would be to work 20 hours a week or become exempt.[5] 

The bill would require states to offer all individuals who would be subject to the work mandate employment and training services that meet the 20-hour-a-week requirement as well as case management services (including comprehensive intake assessments, individualized service plans, progress monitoring, and coordination with service providers).  While the bill would provide $1 billion a year in new federal funding and retain the current matching program, those resources would be far less than what’s needed to implement services the bill envisions. 

  • States would need to track work status, participation in work programs, and exemptions for up to 7 million individuals.  For all individuals aged 18 through 59 who do not receive disability benefits and have no children under age 6 in their household, states would need to determine every month whether the individual worked 20 hours a week, participated in a qualifying job training program, or should have been exempted from the work requirement, for example, because of a temporary disability.[6]  Currently, under a state option called “simplified reporting” that USDA and Congress made available in the early 2000s, most states collect detailed data on participants’ income and circumstances every six months or when a major change occurs that could affect the household’s eligibility, but do not track small changes in work hours or earnings — a key change that made the program more accessible for working families. 

    Nationally in 2021, states would need to track about 7 million people in this group monthly.  About 30 to 40 percent of SNAP households would include at least one member subject to the requirements.  The mandate would be expensive. In addition to tracking work and participation hours and exemption status, states would need to hire staff to process the information, respond to participant questions, and make decisions about exemptions and sanctions.  The net result would be a more expensive program that was less accessible to households that need food assistance.  (See Table 3 for estimates of the number of people who could have been subject to the requirements in each state based on data for 2016, the most recent year for which data are available.)

  • States would be required to build new work programs on a vast — and untested — scale.  States would be required to offer a slot in a work program to every SNAP participant who is subject to the work requirement and not working 20 hours or more a week (and not exempted or living in a waived area).  A conservative estimate is that this would amount to approximately 3 million slots nationally in a typical month of 2021, the first year the rule would go into effect.[7]  Building and operating work programs on this scale would be an unprecedented undertaking for states. 

    Currently the SNAP E&T program serves about 700,000 individuals at some point over the course of a year, and about 200,000 individuals in a typical month, according to states’ data reports.[8]  So, the bill would require a more than ten-fold increase in the number of work and training slots.  It would be very difficult for states to piece together useful work opportunities for individuals who need to participate in an activity for multiple months. 

  • States wouldn’t get the money they need for the work slots required by the bill.  The committee’s bill dedicates about $30 per participant per month based on the $1 billion in federal funding the bill adds, if 3 million people need work slots.[9]  If states operate programs that cost more they can be reimbursed at a 50 percent match, but they would need to pay half of the additional cost.

    Chairman Conaway has suggested that the millions of SNAP participants subject to the work requirements would have access to meaningful employment services to help them gain jobs, but the programs envisioned, like subsidized jobs, apprenticeships, on-the-job training, case management, and other services, are more expensive than the current E&T services most states offer.  These types of more intensive services are already options under SNAP E&T, but states rarely offer them due to their cost and operational complexity and because many people who apply for SNAP are likely to be employed again within a few months anyway, so states view targeted E&T spending to be more efficient. 

  • States would lose flexibility to decide which types of employment services they can provide.  The bill narrows the types of E&T services that states can offer SNAP recipients.  For example, job search, which now is the most common use of SNAP E&T funding, would have to be supervised and in a state-approved location.  Workfare would no longer be allowed.  Combined, these two E&T services now make up about 60 percent of all SNAP E&T services offered.  States providing these services would have to develop different services and find new providers.  It is not clear if current providers would be able or willing to change focus, as many rely on funding streams other than SNAP E&T. 
  • The mandate would likely force states to waste money on ineffective programs that would not be targeted on those who can most benefit.  Because the bill requires states to provide a sufficient number of 20-hour-a-week work programs for every individual subject to the requirement who is not working, states would likely need to create low-cost placements rather than the targeted, skills-based, more intensive training that states are increasingly turning to.  States would face difficult decisions on whether or how to continue to support intensive and comprehensive training programs like those in the Food and Nutrition Service’s (FNS) current “SNAP-to-Skills” initiative that many believe hold promise for improving long-term employment outcomes.  
  • There is no evidence that the bill’s approach would improve employment or earnings outcomes.  The 2014 farm bill included funding for ten state pilots to test different employment and training approaches and evaluate their effects.  Yet the House Agriculture Committee proposes to institute new work requirements on a vast scale before these evaluations are completed.  Moreover, studies that evaluated similar work requirements in the Temporary Assistance for Needy Families (TANF) program found that participants frequently lost benefits for various reasons not pertaining to their desire to work, such as not understanding program rules and consequences, not being granted exemptions for which they qualified, or administrative glitches.[10] 

Child Support Enforcement Mandate

The bill would require low-income parents or guardians who do not live with the child’s other parent to cooperate with child support enforcement (CSE) in order to receive SNAP benefits.  Parents whom states judge as not cooperating would be sanctioned and their share of the household’s food benefit would be cut.  The mandate would apply both to custodial parents (who have physical or legal custody) and to non-custodial parents.  Since 1996 states have had the option to mandate child support cooperation for both groups, but only six states have adopted the option for custodial parents and only one state for non-custodial parents.[11]

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.

  • Mandating cooperation with CSE would be costly for states and the federal government.  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.[12]  Similarly, the Congressional Budget Office’s cost estimate of the bill indicates that the new administrative costs associated with this proposal would amount to almost $11 billion over ten years, of which the federal government would cover only about $7 billion; in order to draw down these federal funds, states would need to pay the additional $4 billion.[13]
  • A CSE mandate wouldn’t be likely to generate significant increases in child support payments, but would increase children’s risk of food insecurity.  The potential negative impact on children has appropriately given states pause.  Some 72 percent of custodial families with incomes below 200 percent of the federal poverty level already access CSE services, according to Census estimates.[14] 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 and some additional families had other formal and informal support arrangements.[15]  Imposing a child support cooperation requirement in SNAP would take a significant number of adults off food assistance and would result in few receiving additional funds from child support payments, a dynamic that would leave children more vulnerable.

Bill Would Increase Paperwork Beyond Work Requirement Burdens

As mentioned, the House SNAP changes undermine more than 15 years of progress that states together with USDA and Congress have made on simplifying SNAP’s administration and modernizing the program.  In addition to the mandates on states to track SNAP work status and exemption status on a monthly basis under the work requirement as discussed above, the bill also would:

  • Reinstate federal asset tests, which would add substantial paperwork burdens on states and households.  The bill would eliminate a state option (known as broad-based categorical eligibility) that more than 40 states have used to lift the SNAP asset test.[16]  The bill substantially raises the amount of assets people can own and still qualify for SNAP, so few households would be likely to lose SNAP as a result of the change, but reinstating asset limits would reintroduce significant administrative burdens.  Even though very few households that apply for SNAP have substantial assets, with an asset test states need to ask about and verify asset information for virtually every applicant SNAP household.  Households would need to submit documentation of their (almost always very small) savings and other assets for review so that state workers could assess households’ eligibility.  It can be especially hard for households without savings or other disposable assets to prove that they don’t have such resources.
  • Eliminate the state option on the value of vehicles SNAP households may own.  Federal SNAP rules require states to count a car’s fair market value toward the SNAP resource test to the extent that the value exceeds $4,650.  The House bill raises the amount substantially, but it eliminates the options states have had since 2000 to set their own vehicle asset rules by aligning with a rule used in a state TANF program or through broad-based categorical eligibility.  Every state has used this flexibility on the vehicle asset rules to let low-income households, especially working families, own a reliable means of transportation and still access food assistance.  Like with other assets, documenting and verifying car ownership and the value of vehicles reintroduce administrative complexity and burden.
  • Impose new verification requirements on households that receive Low Income Home Energy Assistance Program (LIHEAP) assistance.  Currently households that receive energy assistance from LIHEAP can automatically qualify for a SNAP income deduction that is tied to a household’s expense for utilities like heating and electricity.  The House bill eliminates this simplification, in effect requiring all such households to provide documentation of utility costs, increasing administrative burdens on states and households alike.
  • Require small errors to count toward SNAP’s Quality Control (QC) error rate.  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.  The Committee’s proposal would eliminate the error threshold altogether and set the amount at zero.  This would increase pressure on states to focus on small errors that do not represent a serious threat to program integrity.  Many states would likely respond by requiring more paperwork from households.

    The 2002 farm bill reformed SNAP’s QC system to strike a better balance between precision of benefit amounts and improving program access, which in turn was a major contributing factor to rising SNAP participation rates among eligible households, especially working families.  The combination of mandates that increase paperwork and the elimination of the QC threshold would thus pose a serious risk for access to SNAP.  Working households would be more likely than others to lose out on SNAP benefits for which they qualify.  Many households that receive SNAP to supplement their earnings work in low-wage jobs with variable hours, or are self-employed.  If states are required to focus more on smaller errors they are likely to increase paperwork disproportionately for workers.

Other Provisions That Would Limit State Flexibility

Several other provisions of the House bill would eliminate flexibility that states have used to improve the program for working families and seniors and to introduce innovative approaches to reaching eligible low-income households.  The bill would:

  • End an option states have used to eliminate benefit “cliffs” on working families that get a small increase in their earnings. More than 30 states have also used the option known as broad-based categorical eligibility to prevent working families whose overall income rises just above the eligibility cutoff, but who have significant expenses such as child care that make it hard for them to afford food, from abruptly losing SNAP benefits.  This proposal would take away $100 in SNAP benefits per month from the typical family affected.  For a parent earning $13 per hour, that’s a significant hit to her budget.
  • End funding that states can use to improve program access.  The bill would eliminate all state performance bonuses, including bonuses states can receive for serving a high proportion of eligible individuals and others for providing SNAP benefits within federal time standards to a high share of applicant households.  States must reinvest the bonus money in activities that improve SNAP operations, a requirement that Congress added in the 2014 farm bill.  The bill would also prohibit states from using certain grants for projects that simplify SNAP application and eligibility systems and/or improve access to benefits to fund projects aimed at improving program access.  These changes could hinder innovative uses of technology that address barriers to participation or improve state operations.  It is common for other states to pick up on innovative approaches after another state has tried it and shared lessons learned.


SNAP is the country’s most effective anti-hunger program, helping 1 in 8 Americans afford a basic diet. The House Agriculture Committee broke with longstanding bipartisan tradition by passing a solely Republican farm bill that’s unbalanced, untested, and likely unworkable.  The bill turns its back on SNAP’s state-federal partnership by imposing numerous new mandates on states and undermining years of progress in simplifying and streamlining SNAP that has improved access to SNAP for millions, including working families.


APHSA Opposes or Voices Concerns About Numerous SNAP Provisions of the House Farm Bill

The American Public Human Services Association (APHSA) is the bipartisan, nonprofit membership organization that represents state and local health and human services agencies.  Despite the fact that it represents diverse states with different approaches, its letter of May 2, 2018 to the Agriculture Committee chairman and ranking member includes concerns about many of the House’s SNAP provisions.[17]  Here we excerpt from that document on the provisions we highlight in this paper.  APHSA supports numerous provisions of the bill and opposes others; even for the below items we have cut material due to space constraints.

Workforce Solutions 

[W]ithout further debate and amendment, we cannot offer support at this time for a number of the provisions in the work solutions section of the bill as currently outlined.  The Workforce Solutions section of the bill contains by far the most complex, contentious, and speculative elements of this proposed legislation. Member states will have differences of opinion on the mandatory requirements for SNAP E&T based on the fact that under current state options, they run both mandatory and voluntary programs and sometimes a combination of both.  They also have the latitude to enforce compliance through sanctions and some do so currently. States also want the flexibility to design programs and interventions according to their labor markets, available employment opportunities, and what interventions in skill development will be most suitable for employers and potential workers. . . .  What states do not want or need to be effective are highly prescriptive instructions and rigid, administratively cumbersome federal reporting requirements that often measure the wrong performance indicators and divert staff time from focusing on getting people employed. . . .

The mandate that states report monthly is unmanageable and unacceptable.  For instance, how would states manage exemptions and other matters if they have to be reviewed monthly?  The answer is they will have to spend additional administrative funds of their own beyond what will be made available in the bill.

APHSA thinks the rigidity of this seemingly one size fits all structure does not leave states the latitude to implement diverse approaches that respect their individual labor markets and caseload composition.

We appreciate the recognition that in order to achieve any real success in moving individuals into jobs, there must be a significant increase in resources.  But depending on the size of the work eligible pool, we are skeptical that even the 100 percent increased SNAP E&T funding to $1 billion by year three will be sufficient to provide slots to all work eligible SNAP recipients as the bill requires.

Restricting Categorical Eligibility

Removing the BBCE option is troubling, as it will reinstate a new benefit cliff and limit state flexibility.  It also will create . . . new QC issues in case review, particularly around asset verification, that we outline in detail below in our response to Section 4012.  Given that the majority of APHSA member states have availed themselves of the option, we do not support this provision….

Mandatory Child Support Cooperation

[W]e have serious reservations about the time it would take to implement this provision, the significant systems cost it would require, the caseload burdens it would place on an already overstretched IV-D system, whether or not it would actually increase child support payments, and the fact that it almost certainly will result in otherwise eligible needy families losing SNAP benefits and jeopardizing their food security. . . .

States already have the option to decide on child support cooperation rules in SNAP and penalize households for clear non-compliance and some states already take advantage of this option. We see no valid reason to mandate cooperation nationally — instead it should remain a state option.

Adjustment to Asset Limitations

APHSA member states appreciate these long overdue increases in asset limits in principle for both non-elderly and elderly households to avoid losing SNAP benefits as a result of modest savings, as they have not been updated in years. However, our concern is with verifying asset levels as is required under this section, as it poses a significant administrative burden on state staff and QC systems.   Contacting banks is usually the only viable way to do so — other technology that could assist is very expensive.  Banks are often entirely non-responsive to such verification requests or take a lengthy time to respond that goes beyond the required time that an application or recertification must be completed. When they do cooperate, banks often charge states fees for such asset verifications...

Updated Vehicle Allowance 

While APHSA appreciates the intent of establishing a national vehicle value ceiling adjusted annually for licensed drivers, it remains burdensome for states and localities to check blue book values.  It would be far simpler and more reasonable to state that a vehicle to be used by a licensed driver to seek work or for those in SNAP already working be exempt — TANF allows this already and Medicaid imposes no such asset test. This ongoing disconnect between programs is unnecessary.

Availability of Standard Utility Allowances Based on Receipt of Energy Assistance

[T]his provision goes further than just eliminating the ability of states to make nominal LIHEAP payments to households to generate access to the SNAP SUA.  Instead it completely delinks the receipt of LIHEAP by non-elderly households to the SNAP SUA and requires actual receipts from all non-elderly households in order to receive the SNAP SUA.  This link to LIHEAP, unlike the nominal payment, is not a work around but a longstanding method for simplified determination of eligibility for the SUA. . . .

This provision would end the simplification and require these households to provide documentation of utility costs to continue receiving the SUA, which again is a significant administrative burden.  This provision is an unnecessary overreach and should be removed from the bill.

SNAP Participation Rates Increased in Every State from 2002 to 2015
USDA estimates of the share of all eligible individuals that received SNAP, by state, 2002 and 2015
State 2002 2015 Change
Alabama 55% 85% +30%
Alaska 59% 87% +28%
Arizona 58% 70% +12%
Arkansas 59% 72% +13%
California 47% 70% +23%
Colorado 46% 76% +30%
Connecticut 56% 94% +38%
Delaware 49% 100% +51%
District of Columbia 63% 98% +35%
Florida 48% 92% +44%
Georgia 59% 86% +27%
Hawaii 77% 84% +7%
Idaho 49% 80% +31%
Illinois 59% 100% +41%
Indiana 66% 83% +17%
Iowa 54% 92% +38%
Kansas 51% 71% +20%
Kentucky 64% 77% +13%
Louisiana 66% 80% +14%
Maine 66% 90% +24%
Maryland 48% 92% +44%
Massachusetts 38% 82% +44%
Michigan 60% 100% +40%
Minnesota 58% 83% +25%
Mississippi 57% 83% +26%
Missouri 70% 89% +19%
Montana 48% 78% +30%
Nebraska 56% 76% +20%
Nevada 41% 81% +40%
New Hampshire 48% 75% +27%
New Jersey 45% 74% +29%
New Mexico 53% 100% +47%
New York 51% 87% +36%
North Carolina 47% 83% +36%
North Dakota 48% 62% +14%
Ohio 57% 87% +30%
Oklahoma 60% 78% +18%
Oregon 76% 100% +24%
Pennsylvania 53% 90% +37%
Rhode Island 52% 99% +47%
South Carolina 57% 82% +25%
South Dakota 50% 90% +40%
Tennessee 70% 95% +25%
Texas 47% 70% +23%
Utah 47% 69% +22%
Vermont 59% 99% +40%
Virginia 53% 75% +22%
Washington 56% 100% +44%
West Virginia 70% 85% +15%
Wisconsin 49% 96% +47%
Wyoming 43% 59% +16%
United States 54% 83% +29%

Note: The earliest state-level SNAP participation rate estimates are for 1994 and the most recent are for 2015.  The participation rates for the two years are not directly comparable because of changes in USDA’s methodology over the years, but these differences do not substantially affect the 2002 to 2015 increase.

Sources: USDA, “Trends in USDA Supplemental Nutrition Assistance Program Participation Rates: Fiscal Year 2010 to 2015,” June 2017; “Reaching Those in Need: Estimates of State Supplemental Nutrition Assistance Program Participation Rates,” January 2018, and earlier reports in the series.

SNAP Participation Rates Among Working Households Increased in Every State from 2002 to 2015
USDA estimates of the share of all eligible individuals in working households that received SNAP, by state, 2002 and 2015
State 2002 2015 Change
Alabama 50% 74% +24%
Alaska 54% 70% +16%
Arizona 48% 62% +14%
Arkansas 53% 64% +11%
California 33% 57% +24%
Colorado 37% 63% +26%
Connecticut 41% 69% +28%
Delaware 42% 85% +43%
District of Columbia 37% 62% +25%
Florida 41% 77% +36%
Georgia 48% 74% +26%
Hawaii 62% 74% +12%
Idaho 42% 80% +38%
Illinois 51% 82% +31%
Indiana 60% 74% +14%
Iowa 44% 83% +39%
Kansas 44% 64% +20%
Kentucky 59% 70% +11%
Louisiana 69% 72% +3%
Maine 59% 79% +20%
Maryland 38% 74% +36%
Massachusetts 23% 62% +39%
Michigan 62% 85% +23%
Minnesota 39% 72% +33%
Mississippi 51% 74% +23%
Missouri 64% 73% +9%
Montana 45% 73% +28%
Nebraska 43% 71% +28%
Nevada 24% 77% +53%
New Hampshire 39% 65% +26%
New Jersey 27% 65% +38%
New Mexico 48% 97% +49%
New York 41% 79% +38%
North Carolina 40% 74% +34%
North Dakota 49% 57% +8%
Ohio 50% 77% +27%
Oklahoma 58% 64% +6%
Oregon 77% 93% +16%
Pennsylvania 51% 78% +27%
Rhode Island 37% 83% +46%
South Carolina 55% 75% +20%
South Dakota 47% 85% +38%
Tennessee 65% 79% +14%
Texas 38% 67% +29%
Utah 36% 63% +27%
Vermont 51% 85% +34%
Virginia 44% 67% +23%
Washington 39% 82% +43%
West Virginia 76% 83% +7%
Wisconsin 46% 90% +44%
Wyoming 40% 55% +15%
United States 43% 72% +29%

Note: The earliest state-level SNAP participation rate estimates for individuals in working households are for 2002 and the most recent are for 2015.  The participation rates for the two years are not directly comparable because of changes in USDA’s methodology over the years, but these differences do not substantially affect the 2002 to 2015 increase.  For these estimates the numerator is all participating individuals in SNAP households with earnings while receiving SNAP, and the denominator is all SNAP-eligible individuals in households with earnings.  Working households that are temporarily unemployed are not counted.

Sources: USDA, “Trends in USDA Supplemental Nutrition Assistance Program Participation Rates: Fiscal Year 2010 to 2015,” June 2017; “Reaching Those in Need: Estimates of State Supplemental Nutrition Assistance Program Participation Rates,” January 2018, and earlier reports in the series.

Illustrative Example of People Potentially Subject to Work Requirement and Employment and Training Grant Had Work Requirement Applied in 2016a
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 Annual SNAP Employment and Training Grantb
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 The figures in this table are for 2016, the most recent year for which data are available.  SNAP caseloads have declined since 2016 and under CBO’s projections are expected to continue to fall.  As a result, the numbers of individuals who would be subject to the work requirement after it went into effect (in 2021 and later years) would be somewhat lower under CBO’s assumptions.

b 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.


End Notes

[1] The Congressional Budget Office (CBO) finds that the SNAP provisions of the farm bill would impose unfunded mandates on state governments under the definition of mandates Congress uses under the Unfunded Mandates Reform Act (UMRA).  The unfunded state mandates would come because, CBO concludes, “the bill’s requirements would increase the workload of state agencies in areas where they have limited flexibility to amend their responsibilities and offset additional costs.” CBO points to six provisions that contribute to unfunded mandates under the UMRA definition.  See

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

[3] For resources on SNAP and work, see

[4] States can waive the rule temporarily for areas of elevated unemployment and may exempt a limited number of individuals who would otherwise be subject to the time limit, though many states do not use these exemptions.

[5] In other words, during a period of a sanction, complying by participating in a job training program 20 hours a week would not requalify an individual for SNAP.

[6] Certain SNAP recipients would be exempt from the provisions, including those medically certified as mentally or physically unfit for employment and certain college students.  In addition, the state would be allowed to exempt up to 15 percent of the individuals otherwise subject to the requirement and certain (limited) areas with high unemployment could be waived.  States likely would not need to track individuals who live in waived areas as they would typically be waived for the full fiscal year, but all of those working and many of those exempted would need to be tracked each month.

[7] For a discussion of how CBPP derived the 3 million slots figure, see Ed Bolen et al., “House Agriculture Committee’s Farm Bill Would Increase Food Insecurity and Hardship,” Center on Budget and Policy Priorities, updated May 10, 2018,  The House Agriculture Committee reportedly projects a similar number of slots will be needed — up to 2.5 million.  See

[8] See Statement of Kathryn Larin, Director Education, Workforce, and Income Security Issues, U.S. Government Accountability Office, “Supplemental Nutrition Assistance Program: Observations on Employment and Training Programs and Efforts to Address Program Integrity Issues,” testimony before the Subcommittees on Healthcare, Benefits and Administrative Rules and Intergovernmental Affairs, Committee on Oversight and Government Reform, House of Representatives, May 9, 2018, p. 9,

[9] CBO estimates a $7.65 billion net increase in federal administrative costs over ten years from the work requirements, which includes the new federal grant funding for employment and training, an offsetting reduction in federal matching funds for E&T services in the near term, and the federal share of additional general state administrative costs for tracking compliance with work requirements and exemptions.

[10] See Bolen et al., box on p. 8.

[11] The six states that have adopted the option for custodial parents are Florida, Idaho, Kansas, Michigan, Mississippi, and South Dakota.  Mississippi is the only state that has adopted the option for non-custodial parents.

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

[13] Child support enforcement administrative expenses are matched at 66 percent by the federal government. The CBO estimated detailed cost breakdown for the child support cooperation provision appeared in supporting documents.

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

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

[16] Current SNAP federal asset rules require households without an elderly or disabled member to have countable assets of $2,250 or less, and households with an elderly or disabled member to have countable assets of $3,500 or less.  Generally, countable assets include those that could be available to the household to purchase food, such as amounts in bank accounts.  Items that are not accessible, such as the household’s home, personal property, and retirement savings, do not count.  Most automobiles do not currently count.

[17] “Preliminary Response of the American Public Human Services Association: Response to the Sections in the Title IV Nutrition Portion of the Proposed House Farm Bill Reauthorization,” May 2, 2018,