Appendix E GAO Report

Appendix E: GAO Report

GAO
July 2012

United States Government Accountability Office

Report to Congressional Requesters

SUPPLEMENTAL
NUTRITION
ASSISTANCE
PROGRAM
Improved Oversight of
State Eligibility
Expansions Needed

This report was reissued on August 2, 2012 to add
an additional addressee.

GAO-12-670

July 2012

SUPPLEMENTAL NUTRITION ASSISTANCE
PROGRAM
Highlights of GAO-12-670, a report to
congressional requesters

Improved Oversight of State Eligibility Expansions
Needed

Why GAO Did This Study

What GAO Found

Over the last 10 years, participation in
the U.S. Department of Agriculture’s
(USDA) SNAP, previously known as
the Food Stamp Program, has more
than doubled, and costs have
quadrupled. Since 1999, USDA has
allowed states to expand SNAP
eligibility by adopting BBCE policies,
which make households that receive
services funded by Temporary
Assistance for Needy Families, such
as a toll-free number or brochure,
categorically eligible for SNAP. Under
BBCE policies, states are able to
increase federal SNAP limits on
household income and remove limits
on assets. Although USDA has
encouraged states to adopt BBCE to
improve SNAP access and
administration, little is known about the
effects of these policies. GAO was
asked to assess: (1) To what extent
are households that would otherwise
be ineligible for SNAP deemed eligible
for the program under BBCE? (2) What
effect has BBCE had on program
costs? (3) What are the program
integrity implications of BBCE? GAO
analyzed data from USDA, selected
states, and other national sources;
conducted site visits to 5 states; and
interviewed federal, state, and local
officials, as well as others with
knowledge of SNAP.

In fiscal year 2010, GAO estimates that 2.6 percent (473,000) of households that
received Supplemental Nutrition Assistance Program (SNAP) benefits would not
have been eligible for the program without broad-based categorical eligibility
(BBCE) because their incomes were over the federal SNAP eligibility limits. The
characteristics of these households were generally similar to other SNAP
households, although they were more likely to work or receive unemployment
benefits. BBCE removes asset limits in most states, and while reliable data on
participants’ assets are not available, other data suggest few likely had assets
over these limits. Although BBCE contributed to recent increases in SNAP
participation, other factors, notably the recent recession, had a greater effect.

What GAO Recommends
GAO recommends that USDA review
state procedures for implementing
BBCE, disseminate guidance to states
on certifying SNAP households as
eligible for school meals, and revisit its
guidance on SNAP reporting
requirements to ensure they address
all households. USDA generally
agreed with GAO’s recommendations.
View GAO-12-670. For more information,
contact Kay Brown at (202) 512-7215 or
brownke@gao.gov.

Estimated Sub-groups of SNAP Households, Fiscal Year 2010

GAO estimates that BBCE increased SNAP benefit costs, which are borne by the
federal government, by less than 1 percent in fiscal year 2010. In that year, total
SNAP benefits provided to households that, without BBCE, would not have been
eligible for the program because their incomes were over the federal SNAP
eligibility limits were an estimated $38 million monthly or about $460 million for
the year. These households received an estimated average monthly SNAP
benefit of $81 compared to $293 for other households. BBCE’s effect on SNAP
administrative costs, which are shared by the federal and state governments, is
unclear, in part because of other recent changes that affect this spending, such
as state budget and staffing reductions in the recent recession.
BBCE has potentially had a negative effect on SNAP program integrity. In recent
years, the SNAP payment error rate declined to an historic low, but evidence
suggests the decline is primarily due to changes other than BBCE. While BBCE
may improve administrative efficiency, both national data and discussions with
local staff suggest BBCE may also be associated with more errors. In addition,
BBCE has led to unintended consequences for SNAP and related programs. For
example, in implementing BBCE, some states are designating SNAP applicants
as categorically eligible without providing them with the service required to make
this determination. Further, likely because they are unaware of recent USDA
guidance, some states certify children for free school meals when their
households are determined eligible for SNAP, even though they do not receive
SNAP benefits—a result more common in states with BBCE. Finally, because of
federal guidance on BBCE, rules for reporting changes in household
circumstances now differ by household income level and may leave higherincome households without reporting requirements for several months.
United States Government Accountability Office

Contents

Letter

1
Background
While State Eligibility Changes Expanded Participation, the
Economy Likely Played a Larger Role
State Eligibility Changes Increased Benefit Costs Somewhat, but
Their Effect on Administrative Costs Is Unclear
State Eligibility Changes May Negatively Affect Program Integrity
Conclusions
Recommendations for Executive Action
Agency Comments and Our Evaluation

4
15
24
32
39
40
41

Appendix I

Objectives, Scope, and Methodology

43

Appendix II

2010 Quality Control Data Estimates and 95 Percent Confidence
Intervals

46

GAO Contacts and Staff Acknowledgments

51

Appendix III

Related GAO Products

52

Tables
Table 1: Number of States with Various Income and Asset Limits
for Programs Used to Confer BBCE, as of May 1, 2012
Table 2: Characteristics of Specified Groups of SNAP Households,
in Fiscal Year 2010
Table 3: Estimated Percent of Specified Groups of SNAP
Households Receiving Deductions and the Estimated
Average Amount of the Deduction, in Fiscal Year 2010
Table 4: SNAP Households That Would Not Have Been Eligible for
the Program without BBCE because Their Incomes Were
over Federal Limits in Fiscal Year 2010
Table 5: Characteristics of Specified Groups of SNAP Households
in Fiscal Year 2010

Page i

13
17
26
46
46

GAO-12-670 Supplemental Nutrition Assistance Program

Table 6: Benefits Provided to SNAP Households That Would Not
Have Been Eligible for the Program without BBCE
because Their Incomes were over Federal Limits in Fiscal
Year 2010
Table 7: Benefits Received by Specified Groups of SNAP
Households in Fiscal Year 2010
Table 8: Estimated Percent of Specified Groups of SNAP
Households Receiving Deductions and the Estimated
Average Amount of the Deduction in Fiscal Year 2010

48
48
49

Figures
Figure 1: Average Monthly Number of SNAP Participants per Fiscal
Year, in Millions
Figure 2: Total Federal SNAP Costs per Fiscal Year, in Billions
Figure 3: How an Applicant’s Eligibility Is Determined under
Federal SNAP Guidance
Figure 4: How an Applicant’s Eligibility Is Determined under
Traditional Cash Assistance-Related Categorical Eligibility
Figure 5: How an Applicant’s Eligibility May Be Determined under
BBCE
Figure 6: Timeline of State BBCE Policy Implementation, Fiscal
Years 2001-2012
Figure 7: SNAP Households That Would Not Have Been Eligible for
the Program without BBCE because Their Incomes Were
over Federal Limits in Fiscal Year 2010
Figure 8: Trends in SNAP Participation, Unemployment, and
Poverty, 1990-2011
Figure 9: Benefits Provided to SNAP Households That Would Not
Have Been Eligible for the Program without BBCE
because Their Incomes Were over Federal Limits, Fiscal
Year 2010
Figure 10: Selected Information on SNAP Benefits and Potentially
Related Factors, Fiscal Years 2001-2011
Figure 11: SNAP Administrative Costs, Fiscal Years 2001-2010
Figure 12: Administrative Costs per SNAP Household, Fiscal Years
2001-2010
Figure 13: Examples of How States Confer BBCE for SNAP

Page ii

5
5
8
9
11
12
16
21

25
27
30
31
35

GAO-12-670 Supplemental Nutrition Assistance Program

Abbreviations
BBCE
FNS
SNAP
TANF
USDA

broad-based categorical eligibility
Food and Nutrition Service
Supplemental Nutrition Assistance Program
Temporary Assistance for Needy Families
U.S. Department of Agriculture

This is a work of the U.S. government and is not subject to copyright protection in the
United States. The published product may be reproduced and distributed in its entirety
without further permission from GAO. However, because this work may contain
copyrighted images or other material, permission from the copyright holder may be
necessary if you wish to reproduce this material separately.

Page iii

GAO-12-670 Supplemental Nutrition Assistance Program

United States Government Accountability Office
Washington, DC 20548

July 26, 2012
The Honorable Frank D. Lucas
Chairman
The Honorable Collin C. Peterson
Ranking Member
Committee on Agriculture
House of Representatives
The Honorable Jeff Sessions
Ranking Member
Committee on the Budget
United States Senate
The Honorable Pat Roberts
Ranking Member
Committee on Agriculture, Nutrition and Forestry
United States Senate
During fiscal year 2011, the U.S. Department of Agriculture’s (USDA)
Supplemental Nutrition Assistance Program (SNAP), previously known as
the Food Stamp Program, 1 provided food and nutrition assistance to
almost 45 million individuals each month for a total of $71.8 billion in
benefits for the year. SNAP is intended to help low-income individuals
and families obtain a better diet by supplementing their incomes with
benefits to purchase food. Over the last 10 years, SNAP participation has
more than doubled and costs have quadrupled. While some of the recent
increases in SNAP participation and costs are due to the recession and
benefit increases included in the American Recovery and Reinvestment
Act of 2009 (Recovery Act), 2 state policy changes intended to improve
program access and simplify the administration of SNAP may also have
had an effect.

1

SNAP was given its present name in 2008. Food, Conservation, and Energy Act of 2008,
Pub. L. No. 110-346, § 4002(a), 122 Stat. 1651, 1853.

2

Pub. L. No. 111-5, § 101(a) and (e), 123 Stat. 115, 120.

Page 1

GAO-12-670 Supplemental Nutrition Assistance Program

Under federal law, eligibility for SNAP is based primarily on whether a
household’s income and assets fall below certain thresholds. 3 In addition,
low-income households are categorically eligible for SNAP if they receive
benefits from certain other federal programs, such as Temporary
Assistance for Needy Families (TANF). 4 Originally, categorical eligibility
was designed as a method by which households receiving cash
assistance benefits were deemed SNAP-eligible. However, since 1999,
USDA has allowed states to expand categorical eligibility to households
authorized to receive TANF-funded non-cash services by adopting broadbased categorical eligibility (BBCE) policies. 5 Under BBCE policies, states
align the SNAP household income and asset limits in federal law with the
income and asset limits for the relevant TANF-funded non-cash service.
While SNAP eligibility limits therefore differ among states that have
adopted BBCE policies, the amount of assistance eligible households
receive is determined using the same process for all SNAP applicants.
From fiscal year 2006 through fiscal year 2012, the number of states with
BBCE policies increased from 7 to 43.
Although USDA has encouraged states to adopt BBCE to improve SNAP
access and simplify administration, little is known about the impact of
BBCE on SNAP. As a result, we were asked to assess: (1) To what
extent are households that would otherwise be ineligible for SNAP
deemed eligible for the program under BBCE? (2) What effect has BBCE
had on program costs? (3) What are the program integrity implications, if
any, of BBCE?
As criteria for our review, we examined federal laws affecting SNAP, as
well as USDA regulations and guidance related to SNAP and specifically
to BBCE. To answer our research questions, we collected and analyzed
information through several methods. At the federal level, we reviewed
USDA’s information on states’ BBCE policies and interviewed department
officials. We also reviewed USDA’s data on SNAP households, program

3

7 U.S.C. § 2014(a).

4

In addition, households in which all members receive Supplemental Security Income
(SSI) or benefits from state or local general assistance programs are categorically eligible
for SNAP. 7 U.S.C. § 2014(a).

5
Letter to Regional Administrators–All Regions, “FSP–Categorical Eligibility,” signed by
the Deputy Administrator of the Food Stamp Program, Food and Nutrition Service, USDA.
July 14, 1999.

Page 2

GAO-12-670 Supplemental Nutrition Assistance Program

costs, and error rates for the 50 states; Washington, D.C.; Guam; and the
Virgin Islands. We determined that these data were sufficiently reliable for
the purposes of this report. Because USDA’s data lack information on the
assets of SNAP households deemed eligible under BBCE, which is key to
understanding the effects of BBCE on participation and benefit costs, we
also reviewed additional national data sources that provide information on
household assets, such as the Survey of Consumer Finances, and
determined that these data were sufficiently reliable for the purposes of
this report. While these data sources provide estimates of the average
amount of assets owned by households of different income levels, which
can therefore be used to approximate the level of assets owned by SNAP
households, they do not directly provide estimates of SNAP households’
assets. To gather additional background information on the effects of
various factors on SNAP participation and costs, we identified and
reviewed additional studies on SNAP conducted by USDA and several
research organizations that assess programs for low-income populations.
To gather information from state and local SNAP administrators on the
effects of BBCE policies, as well as other simultaneous changes to the
program, we conducted site visits to 5 states (Arizona, Illinois, North
Carolina, South Carolina, and Wisconsin) and 18 local SNAP offices
located in both urban and rural areas in those states. We selected these
states because they varied in their BBCE adoption dates, characteristics
of their BBCE policies, and geographic locations. States selected also
had relatively large SNAP caseloads and generally high proportions of
their SNAP households deemed eligible under BBCE policies. Within
each state, we interviewed state SNAP administrators, as well as local
SNAP administrators from three to four local offices. We cannot
generalize our findings from the site visits beyond the states and localities
we visited. In addition to our site visits, we also interviewed state SNAP
administrators in Idaho and Michigan to gather information on the impacts
of the asset limits that these states added to their BBCE policies in 2011.
To gather other perspectives about the effects of BBCE on SNAP, we
also interviewed officials knowledgeable about SNAP from research and
advocacy organizations that focus on nutrition assistance policy.
We conducted this performance audit from August 2011 through July
2012 in accordance with generally accepted government auditing
standards. Those standards require that we plan and perform the audit to
obtain sufficient, appropriate evidence to provide a reasonable basis for
our findings and conclusions based on our audit objectives. We believe
that the evidence obtained provides a reasonable basis for our findings

Page 3

GAO-12-670 Supplemental Nutrition Assistance Program

and conclusions based on our audit objectives. See appendices I and II
for additional information on our objectives, scope, and methodology.

Background

SNAP is the largest of the 15 domestic food and nutrition assistance
programs overseen by USDA’s Food and Nutrition Service (FNS). FNS
jointly administers SNAP with the states. FNS pays the full cost of SNAP
benefits and pays approximately half of states’ administrative costs. 6 FNS
is also responsible for promulgating program regulations and ensuring
that state officials administer the program in compliance with program
rules. States administer the program by determining whether households
meet the program’s eligibility requirements, calculating monthly benefits
for qualified households, and issuing benefits to participants.

Participation and Costs

As shown in figures 1 and 2, SNAP participation and costs generally
increased between fiscal years 2001 and 2011, though the most
significant increases began in fiscal year 2008.

6

Generally, FNS reimburses states for 50 percent of most types of SNAP administrative
costs. However, some types of administrative costs are reimbursed at higher rates, such
as some costs related to employment and training services and computer system
development. Further, in response to SNAP participation increases during the recent
economic downturn, the American Recovery and Reinvestment Act of 2009 and the
Department of Defense Appropriations Act, 2010 made $290.5 million and $400 million,
respectively, available to states for SNAP administrative costs in fiscal years 2009-2011.
Pub. L. No. 111-5, § 101(c), 123 Stat. 115, 120 and Pub. L. No. 111-118, § 1002(a), 123
Stat. 3409, 3468-69.

Page 4

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 1: Average Monthly Number of SNAP Participants per Fiscal Year, in Millions

Figure 2: Total Federal SNAP Costs per Fiscal Year, in Billions

Page 5

GAO-12-670 Supplemental Nutrition Assistance Program

Note: “Other Costs” include the federal share of SNAP administrative costs and employment and
training programs, as well as other federal costs (e.g., benefit and retailer redemption and monitoring,
payment accuracy, electronic benefit transfer systems, program evaluation and modernization,
program access, health and nutrition pilot projects).

According to FNS, the growth in SNAP participation in recent years is
likely attributable to the economic recession, outreach efforts, and
modifications to program policy. Because households must be lowincome to receive SNAP benefits, participation and costs typically
increase during economic downturns as more people become eligible and
apply. Although the recent recession officially lasted from December 2007
through June 2009, since then, unemployment has remained above
average levels and SNAP participation has continued to grow. Further,
because federal law identifies SNAP’s main purpose as “raising levels of
nutrition among low-income households,” 7 one of the key performance
measures for the program is the rate of participation among eligible
households. As a result, for years, FNS has encouraged states to
undertake outreach efforts and adopt various modifications to program
policy to increase participation among the eligible population and increase
program efficiency. Although the participation rate varies by state, ranging
from an estimated 53 percent in California to 100 percent in Maine in
fiscal year 2009, the national rate has been about 70 percent in recent
years. 8

Determination of
Eligibility

Under federal law and regulations, eligibility for SNAP is based primarily
on a household’s income and assets. 9 A household generally includes
everyone who lives together and purchases and prepares meals
together. 10 To determine a household’s eligibility, a caseworker must first

7

7 U.S.C. § 2011.

8

To estimate the program’s participation rate, the number of SNAP participants in each
state, according to administrative data, is divided by the estimated number of eligible
persons. To estimate the number eligible, USDA applies SNAP federal eligibility rules to
household data in the U.S. Census Bureau’s Current Population Survey. This approach
does not take into account households potentially eligible for SNAP in states with BBCE
policies that modify the federal eligibility rules. If these households were included, many
states’ participation rate estimates would likely be lower.

9

7 U.S.C. § 2014.

10
7 U.S.C. § 2012(n). There are specific exceptions to the definition of a household. For
example, husbands, wives, and most children under age 22 that live together are included
in the same household, even if they purchase and prepare meals separately.

Page 6

GAO-12-670 Supplemental Nutrition Assistance Program

determine the household’s gross income, which cannot exceed 130
percent of the federal poverty guidelines, 11 and its net income, which
cannot exceed 100 percent of the guidelines (or $18,530 annually for a
family of three living in the continental United States in fiscal year 2012). 12
Net income is determined by taking into account certain exclusions and
deductions, for example, expenses for dependent care, utilities, and
housing. 13 In addition, a caseworker must determine a household’s assets
under various requirements. For example, a household’s liquid assets,
such as those in a bank account, currently cannot exceed $2,000 or, for
households with an elderly or disabled member, $3,250. 14 However,
certain assets are not counted for SNAP, such as a home, the
surrounding lot, and most retirement plans and educational savings
accounts. 15 While there are also federal SNAP provisions that limit the
value of vehicles an applicant can own and still be eligible for the
program, 16 all states have opted to modify those rules, and most exclude
the value of all household vehicles. (See figure 3 for a general depiction
of the eligibility determination process under federal SNAP rules.)

11

7 U.S.C. § 2014(c)(2). The federal poverty guidelines are updated periodically in the
Federal Register by the Secretary of Health and Human Services. 42 U.S.C. § 9902(2)
and (4).
12

7 U.S.C. § 2014(c)(1). Households that include an elderly or disabled member do not
have to meet the gross income limit but must meet the net income limit. SNAP defines
elderly members of a household as those aged 60 and older. 7 U.S.C. § 2012(j)(1).
13
7 U.S.C. § 2014(d) and (e). Exclusions include, for example, educational loans,
veteran’s educational benefits, and income earned by minor children.
14

7 U.S.C. § 2014(g)(1).

15

7 U.S.C. § 2014(g)(2)–(7).

16

7 U.S.C. § 2014(g)(2)(C).

Page 7

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 3: How an Applicant’s Eligibility Is Determined under Federal SNAP
Guidance

Federal law also allows certain households to be deemed categorically
eligible for SNAP. 17 Under statute, households receiving monthly cash
assistance from certain programs—including TANF, SSI, and state or
local general assistance programs—are categorically eligible for SNAP.
According to USDA, categorical eligibility can increase program access
and reduce administrative burden, as states assess a household’s
eligibility once for the cash assistance program rather than twice for both
the cash assistance program and SNAP. (See figure 4 for a general
depiction of the eligibility determination process under traditional cash
assistance-related categorical eligibility.)

17

7 U.S.C. § 2014(a).

Page 8

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 4: How an Applicant’s Eligibility Is Determined under Traditional Cash
Assistance-Related Categorical Eligibility

In response to welfare reforms under the Personal Responsibility and
Work Opportunity Reconciliation Act of 1996, 18 USDA advised states that
households authorized to receive non-cash services—such as case
management services, transportation subsidies, or child care subsidies—
from a program funded with TANF dollars could also be deemed
categorically eligible. 19 In order for a state to fund a non-cash service with
TANF dollars, the service generally must further one of TANF’s purposes,
which include the promotion of job preparation, work, and marriage, and
the reduction of out-of-wedlock births. 20 As set out in SNAP regulations,
households in which members are authorized to receive non-cash
services primarily funded with TANF are categorically eligible, and states
also have the option of extending categorical eligibility to households

18

Pub. L. No.104-193, 110 Stat. 2105.

19

This includes federal TANF dollars and state dollars counted as maintenance-of-effort
for TANF. 7 C.F.R. § 273.2(j)(2)(ii) (2012).
20

42 U.S.C. § 601(a) and 604.

Page 9

GAO-12-670 Supplemental Nutrition Assistance Program

receiving services that are less than 50 percent TANF-funded, with FNS
approval required in certain cases. 21 SNAP regulations also direct that the
TANF-funded non-cash services used to confer categorical eligibility be
available only to households with incomes equal to or below 200 percent
of the federal poverty guidelines.
As a result of this expansion of categorical eligibility, states have adopted
a variety of policies to deem households that receive non-cash services
from TANF-funded programs eligible for SNAP. FNS separates these
types of policies into two groups—broad-based and narrow. According to
FNS, BBCE policies make most, if not all, households that apply for
SNAP categorically eligible because they receive a TANF-funded noncash service, such as an informational brochure or toll-free number. (See
figure 5 for a general depiction of the eligibility determination process
under BBCE.) In contrast, narrow categorical eligibility policies require
households to be enrolled in certain TANF-funded programs, such as
employment assistance, or receiving child care or transportation
assistance, in order to be categorically eligible for SNAP.

21

7 C.F.R. § 273.2(j)(2)(i)-(iii) (2012).

Page 10

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 5: How an Applicant’s Eligibility May Be Determined under BBCE

Note: This figure depicts the general process for determining a household’s SNAP eligibility in a state
with a BBCE policy that removes the federal SNAP asset limit and does not include a net income
limit. Not all states have such a policy.

Although FNS issued guidance in 1999 and regulations in 2000
explaining how states could adopt BBCE policies, 22 relatively few states
implemented them early on. Between fiscal years 2001 and 2006, 7
states adopted these policies. However, when the recent economic
downturn began, and households applying for SNAP began to increase
greatly, FNS encouraged states to adopt these policies to streamline
eligibility processes and ease workload (see fig. 6).

22

65 Fed. Reg. 70,134, 70,160 and 70,198 (Nov. 21, 2000). Although promulgated in
2000, none of its provisions became effective prior to January 20, 2001.

Page 11

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 6: Timeline of State BBCE Policy Implementation, Fiscal Years 2001-2012

According to FNS, as of May 1, 2012, 43 states—including Washington,
D.C., Guam, and the Virgin Islands—had BBCE policies. 23 These policies
differ in terms of the income and asset limits used to determine eligibility,
as shown in table 1. For example, 24 states’ BBCE policies increase the
federal gross income limit for SNAP and remove the asset limit while 2
states’ BBCE policies retain the federal gross income limit and increase
the federal asset limit.

23

For this report, we reviewed information on SNAP in the 50 states; Washington, D.C.;
Guam; and the Virgin Islands. We use “states” throughout the report to refer to this group.

Page 12

GAO-12-670 Supplemental Nutrition Assistance Program

Table 1: Number of States with Various Income and Asset Limits for Programs Used
to Confer BBCE, as of May 1, 2012
Higher asset limit than
federal SNAP limits

No asset
limit

Federal SNAP gross income limit equal to
130 percent of the federal poverty guidelines

2

14a

Higher gross income limit (160-200 percent of
the federal poverty guidelines)

3

24

Source: GAO analysis of USDA data on states’ BBCE policies.
a

Nine of these states increased the federal gross income limit to 200 percent of the federal poverty
guidelines, but only for households with elderly or disabled members.

Determination of Benefits

After eligibility is established, benefits are determined based on each
household’s monthly net income, with greater benefits provided to those
with less income. SNAP expects each eligible household to spend 30
percent of its own resources on food, and therefore, each household’s
monthly SNAP benefit is determined by subtracting 30 percent of its
monthly net income from the maximum SNAP benefit for the relevant
household size. 24 All eligible one- and two-person households are
guaranteed a minimum benefit, which is $16 for households in the
continental United States in fiscal year 2012. 25 However, households with
three or more members do not receive a minimum benefit. Under federal
income eligibility limits, a household with three or more members will
typically be determined eligible for a SNAP benefit greater than $0.
However, because some states’ BBCE policies raise the SNAP income
limits, under these policies, such households are more likely to be
deemed eligible for $0 in benefits.

Certification and
Reporting Requirements

SNAP households are certified eligible for SNAP for periods ranging from
1 to 24 months, which vary based on state policy choices. Generally, the
length of the certification period depends on household circumstances,

24

7 U.S.C. § 2017(a) If a household has zero net income (that is, its deductable expenses
equal or exceed its gross income), it receives the maximum SNAP benefit based on
household size. Maximum benefits are based on USDA’s Thrifty Food Plan, which is an
estimate of how much it costs to buy food to prepare nutritious, low-cost meals for a
household. The Thrifty Food Plan estimate is changed every year to keep pace with food
prices.
25

7 U.S.C. § 2017(a). The minimum benefit for eligible one- and two-person households in
Alaska, Hawaii, Guam, and the Virgin Islands ranges from $19-$30 in fiscal year 2012.

Page 13

GAO-12-670 Supplemental Nutrition Assistance Program

but only households in which all members are elderly or disabled can be
certified for up to 24 months under federal regulations. 26 Once the
certification period ends, households reapply for benefits, at which time
eligibility and benefit levels are redetermined.
Between certification periods, households generally must report changes
in their circumstances—such as household composition, income, and
expenses—that may affect their eligibility or benefit amounts. 27 Since
early 2001, states have had the option of requiring households to report
changes only when their incomes rise above 130 percent of the federal
poverty guidelines, rather than reporting changes at regular intervals or
within 10 days of occurrence, as was required in the past. 28 According to
FNS, as of November 2010, all states except California and Wyoming use
this simplified reporting for some or all SNAP households.

Quality Control System

FNS and the states share responsibility for implementing an extensive
quality control system used to measure the accuracy of SNAP eligibility
and benefits and from which state and national error rates are
determined. 29 Under FNS’s quality control system, the states calculate
their payment errors annually by drawing a statistical sample to determine
whether participating households are eligible and received the correct
benefit amount. Because SNAP considers many factors in determining
each household’s benefit amount, any of these factors can result in a
payment error. For example, incorrect calculations of earned income or
unearned income and inaccurate accounting of the number of household
members may cause payment errors. The state’s payment error rate is
based on the sample and determined by dividing the dollars paid in error
by the total SNAP benefits issued. Once the payment error rates are
determined, FNS is required to compare each state’s performance with

26
7 C.F.R. § 273.2(f)(1) (2012). The state must have at least one contact with these
households every 12 months.
27

7 C.F.R. § 273.12(a) (2012).

28

7 C.F.R. § 273.12(a)(1)(vii) (2012) and 65 Fed. Reg. 70,134, 70,143 and 70,208 (Nov.
21, 2000) (effective Jan. 20, 2001).
29

The SNAP error rate is calculated for the entire program, as well as every state, and is a
combination of overpayments to those who are eligible for smaller benefits, overpayments
to those who are not eligible for any benefit, and underpayments to those who do not get
as much as they should.

Page 14

GAO-12-670 Supplemental Nutrition Assistance Program

the national payment error rate and impose financial penalties or provide
financial incentives according to legal specifications.

While State Eligibility
Changes Expanded
Participation, the
Economy Likely
Played a Larger Role
More Became Eligible but
Impact of Changes Likely
Limited

In fiscal year 2010, an estimated 2.6 percent 30 (approximately 473,000) of
all households receiving SNAP benefits nationwide would not have been
eligible for the program without BBCE because their incomes were
greater than the income limits defined in federal law. 31 32 However, over
half of the states did not have BBCE policies that increased SNAP
income limits above the federal limits in fiscal year 2010. In the 24 states
whose BBCE policies raised household income limits above the federal
limits in that year, around 4.8 percent of SNAP households had incomes
over those limits (see fig. 7). Those households eligible solely because of
BBCE generally had incomes that were modestly higher than the federal
limits. On average, their total monthly income was an estimated $1,965,

30

The 95 percent confidence interval for the 2.6 percent estimate is (2.4, 2.8). For the 95
percent confidence intervals for all estimates in this paragraph, see appendix II, table 4.
31

For more information about our analysis, see appendix I. As previously noted,
household eligibility is determined by local staff administering SNAP, and the accuracy of
those determinations is assessed by state and federal reviewers. We did not
independently determine households’ eligibility.
32

In April 2012, the Congressional Budget Office (CBO) released a report on SNAP that
estimates a 4.3 percent annual reduction in SNAP participants over the 2013-2022 period
if federal SNAP income and asset limits were applied to all categorically eligible
households. However, CBO’s methodology for producing its estimates differs from our
methodology in several ways. For example, although CBO indicates its estimates reflect
changes to BBCE, the Office’s estimates include both participants deemed eligible under
BBCE policies, as well as those deemed eligible under narrow non-cash categorical
eligibility policies. In addition, CBO estimates include assumptions about the share of
these households that exceed federal asset limits.

Page 15

GAO-12-670 Supplemental Nutrition Assistance Program

which is about 150 percent of the federal poverty guidelines, 33 whereas
the federal income limit for SNAP is 130 percent of the guidelines.
Figure 7: SNAP Households That Would Not Have Been Eligible for the Program
without BBCE because Their Incomes Were over Federal Limits, in Fiscal Year 2010

Note: For the 95 percent confidence intervals for these estimates, see appendix II, table 4.

Households eligible under BBCE with incomes over the federal limits had
characteristics that were generally similar to all other SNAP households;
however, they were more likely to be working or receiving unemployment
benefits (see table 2). About half of these households included a child, as
was the case for all other SNAP households, and a similar proportion of
each group of households included a single female head of household.
While a generally similar proportion of both groups of households were
elderly recipients of Social Security benefits, the average monthly amount
of Social Security benefits received by households that would have failed

33

In fiscal year 2010, 150 percent of the federal poverty guidelines equaled $1,821
monthly for a 2-person household and $2,289 monthly for a 3-person household. We
estimate that households with incomes over the federal limits in fiscal year 2010 had, on
average, 2.2 members.

Page 16

GAO-12-670 Supplemental Nutrition Assistance Program

the federal income tests was substantially higher. In two of the local
offices we visited, staff noted that BBCE may have increased the number
of elderly applicants since the policy change enabled some who were
previously ineligible because of their Social Security earnings to become
eligible for SNAP. Also, although both groups had the same proportion of
households with unearned income, a higher percentage of households
with incomes above the federal limits had members who worked or who
received Unemployment Insurance benefits. Further, the average monthly
amount they received in Unemployment Insurance was considerably
higher than that received by all other SNAP households.
Table 2: Characteristics of Specified Groups of SNAP Households, in Fiscal Year
2010
SNAP households
with incomes
over federal All other SNAP
income limits
households
Total number of SNAP households

473,381

17,895,847

Percentage of households with at least one
child

56.3

48.5

Percentage of households with a single female
as the head

30.7

25.8

Percentage of households with at least one
member receiving Social Security benefits

27.8

21.2

Average amount of Social Security benefits
received

$340

$155

Percentage of households with at least one
member with unearned income

60.0

60.3

Percentage of households with at least one
member with earned income

65.9

29.0

Percentage of households with at least one
member receiving Unemployment Insurance
benefits

18.6

6.4

Average amount of Unemployment Insurance
benefits received

$223

$55

Source: GAO analysis of USDA’s SNAP quality control data.

Note: The average amount of Social Security and Unemployment Insurance benefits received was
calculated for all households in each group, rather than only for households that received these
benefits. For the 95 percent confidence intervals for these estimates, see appendix II, table 5.

Page 17

GAO-12-670 Supplemental Nutrition Assistance Program

Available data suggest few households that qualified for SNAP under
BBCE likely had assets that would have exceeded federal asset limits. In
fiscal year 2010, 37 states had removed the federal asset limit, which was
$2,000 for most households, 34 as part of their BBCE policies. Because
asset information was therefore not collected from SNAP applicants in
these states, USDA’s data on SNAP households cannot be used to
estimate the number or share of participating households with assets over
the federal limits. However, other national data sources suggest the
number is relatively small. For example, a national survey that gathered
information on families’ assets in 2010 found that an estimated 24 percent
of families in the bottom income quintile did not have a checking, savings,
or other financial transaction account. 35 Among the estimated 76 percent
of this group that had such an account, the median balance was an
estimated $700. This survey also found that while a greater proportion of
families in the second lowest income quintile had such an account in
2010 (91 percent), their median account balance was an estimated
$1,500. For the most part, SNAP households deemed eligible under
BBCE—households with incomes under 200 percent of the federal
poverty guidelines—fall within the two lowest income quintiles. 36 A 2007
survey of families with children found that those with incomes between
100 and 199 percent of the federal poverty guidelines held median liquid
assets of around $300. 37 For those with incomes below 100 percent of the
guidelines, the median amount was estimated to be zero.
Available state-level data, as well as information shared by state and local
officials during our site visits, also suggest the value of assets held by

34

As previously noted, federal SNAP asset limits are $2,000 for most households and
$3,250 for households with an elderly or disabled member.
35

Federal Reserve Board, 2010 Survey of Consumer Finances. Transaction accounts
include checking, savings, and money market deposit accounts; money market mutual
funds, and call or cash accounts at brokerages.
36
According to U.S. Census Current Population Survey data, as of March 2010, families
with incomes of an estimated $27,379 or less were in the bottom income quintile, and
those with estimated incomes from $27,380 to $48,705 were in the second lowest income
quintile, in 2010 dollars.
37

University of Michigan, Institute for Social Research, 2007 Panel Study of Income
Dynamics. In this survey, liquid assets are defined as immediately available assets that
can be easily converted to cash. These include, for example, the value of checking and
savings accounts, as well as stocks, bonds, and cash-value of life insurance. Further, note
that families with no liquid assets are included in this analysis.

Page 18

GAO-12-670 Supplemental Nutrition Assistance Program

SNAP households is low. For example, according to state officials in
Idaho and Michigan, both states initially removed the federal asset limits
as part of their BBCE policies but reinstated an asset limit of $5,000
during 2011. Officials indicated that the new limits had a very small
impact on overall caseloads. For example, during the 9 months following
Idaho’s reinstatement, approximately 850 new applicants and existing
recipients seeking recertification were denied benefits because their
assets exceeded the asset limit. This represented less than a 1 percent
reduction in the total number of SNAP households in that state during that
period. 38 Similarly, during the month following Michigan’s reinstatement of
the asset limit, about 1 percent of the state’s existing SNAP cases were
closed due to assets. 39 Further, during our site visits, caseworkers in all of
the offices we visited said they believe the value of assets held by SNAP
households is usually very low or $0. Several caseworkers said that while
they may have served SNAP applicants that held assets greater than the
federal limits, they believe such instances are rare. Many caseworkers
noted it is common to hear from applicants that they have exhausted a
significant portion of their available assets before applying for SNAP.

Economic Downturn a
Major Cause of Recent
Trends

While implementation of BBCE by many states has enabled more
households to receive SNAP, the nation’s recent economic downturn has
likely played a larger role in the increases in participation during the past
decade. As shown in figure 8, increases and decreases in SNAP
participation often coincide with similar changes in unemployment and
poverty. A 2002 USDA study found that during past economic recessions,
a 1 percentage-point increase in the national unemployment rate has
been associated with an increase in the number of SNAP participants of 1

38

For the 9 full months after reinstatement, the average monthly SNAP caseload in Idaho
was approximately 101,000 households.
39

However, according to Michigan officials, this figure includes both cases closed because
they had assets over the $5,000 limit, as well as those closed because they had vehicles
worth more than the $15,000 limit that the state simultaneously established. Also, since
both Idaho and Michigan’s new $5,000 asset limits are higher than the federal SNAP
asset limits, it is likely that these estimates are lower than they would be if the states had
reinstated the federal asset limits.

Page 19

GAO-12-670 Supplemental Nutrition Assistance Program

to 3 million. 40 This relationship also existed during the most recent
economic recession of 2007-2009, 41 which was marked by a steep rise in
the nation’s unemployment rate and an increase in the proportion of
families living in poverty. Between fiscal years 2007 and 2010, the
number of SNAP participants rose by around 14 million (or approximately
54 percent), while the unemployment rate increased by 5 percentage
points. This relationship was also noted by staff administering SNAP in all
18 local offices we visited who cited the economic downturn, and related
unemployment, as the primary cause of the increases in SNAP
participation in their localities.

40

Kenneth Hanson and Craig Gunderson, How Unemployment Affects the Food Stamp
Program, USDA Economic Research Service, FANRR-26-7 (Washington, D.C.:
September 2002). GAO and others have also noted the relationship between
unemployment and SNAP. See GAO, Domestic Food Assistance: Complex System
Benefits Millions, but Additional Efforts Could Address Potential Inefficiency and Overlap
among Smaller Programs, GAO-10-346 (Washington, D.C.: April 15, 2010); and James
Mabli and Carolina Ferrerosa, Supplemental Nutrition Assistance Program Caseload
Trends and Changes in Measures of Unemployment, Labor Underutilization, and Program
Policy from 2000 to 2008 (Cambridge, MA: Mathematica Policy Research, Inc., October
18, 2010).
41

According to the National Bureau of Economic Research, the U.S. economy was in
recession from December 2007 to June 2009.

Page 20

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 8: Trends in SNAP Participation, Unemployment, and Poverty, 1990-2011

Note: Numbers of people in poverty are based on data for March of the following year (except the
2010 figure is based on June 2010 data). Numbers of SNAP participants are based on the average
monthly value for the fiscal year. Unemployed people are based on data as of June of that year.

Federal changes to SNAP, as well as those initiated by individual states,
and a shift in public perception of the program, have also likely
contributed to increases in participation during the past decade. For
example, the Recovery Act implemented a 13.6 percent increase in
maximum monthly SNAP benefits, 42 which likely made participation in the
program more attractive to eligible households. In addition, the simplified
reporting option, which most states have implemented since it became
available in 2001, has been linked to increased participation, likely
because it reduces the administrative burden for SNAP households and
lengthens certification periods. Further, USDA expenditures targeted to

42

§ 101(a), 123 Stat. 120.

Page 21

GAO-12-670 Supplemental Nutrition Assistance Program

state and community outreach efforts, as well as relaxed limits on vehicle
ownership, have been linked to increased SNAP participation. During our
site visits, officials noted that the Recovery Act’s suspension of the 3month time limit for able-bodied adults without dependents also caused a
noticeable increase in SNAP participants. 43 In addition, individual states
have implemented program changes that may have increased
participation, such as taking steps to make it easier to apply for SNAP.
For example, staff in most of the states we visited cited implementation of
online applications and phone interviews, instead of in-person interviews,
as improving access to SNAP and shifting the public’s perception of the
program. Some local caseworkers noted that being able to apply without
going to a public assistance office lowers the stigma associated with
receipt of government assistance. These changes may also be
encouraging participation among specific age groups, as local
caseworkers across several states we visited described an increasing
trend of single people aged 22 applying as their own SNAP households. 44
Several studies have examined the impact of various changes on SNAP
participation, though it is difficult to measure the precise impact of any
single change. 45
Other studies, our own analysis of USDA data, and information we
obtained during our site visits indicate the impact of BBCE on SNAP
participation is likely small, and the extent to which the policy directly
encouraged eligible households to participate is uncertain. While several
studies have concluded that BBCE policies have contributed to increases

43

§ 101(e), 123 Stat. 121. The Recovery Act eliminated the SNAP time limit for these
adults without dependents during the period April 1, 2009, through September 30, 2010,
but most states have since received waivers to continue the time limit suspension through
fiscal year 2013.

44
At age 22, a SNAP applicant can be declared a separate household, even if living with
friends or family. 7 C.F.R. § 273.1(b)(ii). Caseworkers noted that they have seen some
nd
people apply for SNAP on their 22 birthdays, and they believe that online information
about program rules, as well as online applications, may be affecting this trend.
45
Caroline Ratcliffe, Signe-Mary McKernan, and Kenneth Finegold, The Effect of State
Food Stamp and TANF Policies on Food Stamp Program Participation (Washington, D.C.:
The Urban Institute, March 2007); James Mabli, Emily Sama Martin, and Laura Castner,
Effects of Economic Conditions and Program Policy on State Food Stamp Program
Caseloads, 2000 to 2006 (Washington, D.C.: Mathematica Policy Research, Inc., August
2009); and Jacob Alex Klerman and Caroline Danielson, Determinants of the Food Stamp
Program Caseload (Washington, D.C.: RAND, January 2009).

Page 22

GAO-12-670 Supplemental Nutrition Assistance Program

in SNAP participation, 46 they yield inconclusive results concerning the
magnitude of the impact. Our own analysis of changes in SNAP
participation in the 17 states that adopted BBCE policies during fiscal
year 2009 revealed those states had a slightly larger increase in
participation between fiscal years 2008 and 2010 than states without
BBCE. However, because of the many other factors influencing SNAP
participation during these years, we cannot attribute this increase entirely
to BBCE. State officials and local caseworkers in the five states we visited
were also uncertain of BBCE’s effect on SNAP participation. According to
those we spoke to, BBCE had a noticeable, but relatively small, effect on
SNAP participation, and local staff in all 18 of the offices we visited said
BBCE’s impact on SNAP participation was considerably less than that
caused by the economic downturn. Those in one office said they had
been alerted by their state office to prepare for a significant spike in
applications once the BBCE policy went into effect, but the subsequent
increase was considerably less than expected. Others noted that the
participation increases they noticed after BBCE implementation may also
have been due to some other simultaneous cause, such as seasonal
increases.

46

Jacob Alex Klerman and Caroline Danielson, “The Transformation of the Supplemental
Nutrition Assistance Program,” Journal of Policy Analysis and Management, vol. 30, no. 4,
(2011); Janna Johnson, The Dynamics of SNAP Participation and the Increase in SNAP
Caseloads during the Recovery of 2003-2007 (Madison, WI: Institute for Research on
Poverty, November 30, 2011); Mabli, Martin, and Castner, Effects of Economic Conditions
and Program Policy on State Food Stamp Caseloads, 2000 to 2006; and Ratcliffe,
McKernan, and Finegold, The Effect of State Food Stamp and TANF Policies on Food
Stamp Program Participation.

Page 23

GAO-12-670 Supplemental Nutrition Assistance Program

State Eligibility
Changes Increased
Benefit Costs
Somewhat, but Their
Effect on
Administrative Costs
Is Unclear
Benefit Costs Increased
Less Than 1 Percent

Although SNAP households that had incomes over the federal limits
made up an estimated 2.6 percent of the SNAP caseload in fiscal year
2010, this group received an estimated 0.7 percent 47 of all SNAP benefits.
These benefits totaled an estimated $38.3 million a month, or
approximately $460 million annually. 48 In the group of states that
increased the federal SNAP gross income limit with their BBCE policies,
benefits provided to households that had incomes over the federal limits
were an estimated 1.5 percent of all SNAP benefits (see fig. 9). Due to
data limitations, these estimates represent minimums, as they do not
include benefits provided to SNAP households deemed eligible under
BBCE with assets over the federal SNAP asset limits. 49

47

The 95 percent confidence interval for the 0.7 percent estimate is (0.6, 0.8). For the 95
percent confidence intervals for all estimates in this paragraph, see appendix II, tables 4
and 6. For more information about our analysis, see appendix I.

48

In April 2012, CBO released a report on SNAP that estimates $1.2 billion as the average
annual savings in spending over the 2013-2022 period if federal SNAP income and asset
limits were applied to all categorically eligible households. CBO considers this estimate to
be equivalent to a 1.6 percent annual savings for SNAP. However, CBO’s methodology for
producing its estimates differs from our methodology in several ways. For example,
although CBO indicates its estimates reflect changes to BBCE, the Office’s estimates
include both benefits provided to households deemed eligible under BBCE policies, as
well as benefits provided to households deemed eligible under narrow non-cash
categorical eligibility policies. In addition, CBO estimates include assumptions about the
share of these households that exceed federal asset limits and the benefits they receive.
49

The national SNAP household data that we used for our analysis does not include
reliable information on assets owned by households deemed eligible under BBCE. As a
result, our estimates represent minimums that do not include benefits provided to
households who had assets over the federal SNAP asset limits.

Page 24

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 9: Benefits Provided to SNAP Households That Would Not Have Been
Eligible for the Program without BBCE because Their Incomes Were over Federal
Limits, per Month in Fiscal Year 2010

Note: For the 95 percent confidence intervals for these estimates, see appendix II, table 6.

Because SNAP benefits are calculated based on income and expenses,
and provide greater benefits to those with fewer means, those with
incomes over the federal limits tend to be eligible for fewer benefits. On
average, these households received an estimated $81 average monthly
SNAP benefit in fiscal year 2010 compared to an estimated $293 average
monthly benefit received by all other SNAP households in that year. 50
These households also disproportionately received the minimum benefit
of $16. An estimated 44 percent of these households received the
minimum benefit compared to 3 percent of all other households.
Households eligible solely because of BBCE had higher average
deductions in certain categories—including dependent care and child
support expenses—than other households in fiscal year 2010 (see table
3), and deductions increase monthly SNAP benefits. However, in general,

50

For the 95 percent confidence intervals for all estimates in this paragraph, see appendix
II, tables 7 and 8.

Page 25

GAO-12-670 Supplemental Nutrition Assistance Program

the higher incomes of households eligible solely because of BBCE seem
to have had a greater impact on their SNAP benefits than their
deductions, given the relatively low average benefits they received.
Table 3: Estimated Percent of Specified Groups of SNAP Households Receiving
Deductions and the Estimated Average Amount of the Deduction, in Fiscal Year
2010

Deduction

Percentage of SNAP
households with incomes over
the federal limits receiving
deduction (average amount)

Percentage of all other
SNAP households
receiving deduction
(average amount)

6.7 ($20)

1.9 ($4)

Child support expenses
Dependent care

14.3 ($50)

3.5 ($8)

Earned income

65.9 ($243)

28.9 ($57)

13.0 ($19)

3.5 ($5)

72.2 ($220)

70.5 ($265)

Medical
Excess shelter
Source: GAO analysis of USDA’s SNAP quality control data.

Note: The average amount of the deduction was calculated for all households in each group, rather
than only for households that received the deduction. All percentage and numerical estimates in this
table are significantly different at the 95 percent confidence level except for the estimated percentage
of each group receiving the excess shelter deduction. See appendix II, table 8 for the 95 percent
confidence intervals for these estimates.

Both the cost of total SNAP benefits and the average benefit per
household increased over the last decade while many states were
implementing BBCE; however, other factors likely had a greater effect on
benefit costs (see fig. 10). The annual adjustment made to the Thrifty
Food Plan—which is the basis for the maximum SNAP benefit amounts,
as well as changes in the economy, demographics, and policies affecting
deductions, outreach, and eligibility can all affect total spending on SNAP
benefits. In recent years, the recession drove increased benefit costs,
both by changing household circumstances and by increasing the benefit
cost per household. For example, because household benefits are
primarily determined based on each household’s monthly income,
increases in the poverty and unemployment rates likely correlate with
increases in the average benefit provided to households. In addition, as
previously noted, the Recovery Act implemented a 13.6 percent increase
in the maximum monthly SNAP benefit per household. During our site
visits, some officials cited these changes as key factors that impacted
household benefits in recent years. Officials we spoke to also noted that
the slow economic recovery has led to SNAP households remaining on
the program for longer time periods than before the recession, which can
lead to increases in total benefit costs.

Page 26

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 10: Selected Information on SNAP Benefits and Potentially Related Factors, 2001-2011

Because many factors impact SNAP benefit costs, the full extent of
BBCE’s impact is unclear, though evidence suggests other factors played
a more important role in recent years. Although BBCE may impact SNAP
benefit costs because the policy both expands who is eligible for the
program and streamlines the process for receiving benefits, state and
local officials we met with consistently indicated that they did not think

Page 27

GAO-12-670 Supplemental Nutrition Assistance Program

BBCE had a significant impact on benefits. Further, our analysis of SNAP
household data suggests factors beyond BBCE typically have a greater
impact on benefits. For example, in our review of SNAP benefits in the
group of 17 states that implemented BBCE during fiscal year 2009, we
found that the average monthly benefit per household significantly
increased in all of these states between fiscal years 2008 and 2010.
However, for most of these states, the increases were likely primarily
related to the increase in maximum benefits implemented under the
Recovery Act, as we found no significant differences in the two factors
used to determine benefit amounts—net income and household size—for
those years. 51

Administrative Costs
Affected by Many Factors

Many factors affect SNAP administrative costs, and state BBCE policies
are one factor that may help reduce such costs. Studies have shown that
factors ranging from a state’s economy and demographic characteristics
to its SNAP policies, administrative processes, staff salaries, and the use
of technology all impact state administrative spending to varying degrees.
As we previously reported, because categorical eligibility policies simplify
the eligibility determination process by creating consistency in income
and resource limits across programs, these policies can save resources,
improve productivity, and help staff focus more time on performing
essential program activities. 52 During our site visits, staff in many of the
local offices we visited stated that, before BBCE was implemented,
verifying assets often took a considerable amount of time, and state
officials added that it could be costly, as banks sometimes charge SNAP
offices a fee to provide account documentation. As a result, staff in almost
all of the local offices we visited said BBCE’s removal of the SNAP asset
limit helped streamline case processing, and some noted that
streamlining occurred both because SNAP households did not have to

51
We found no differences in these two factors in 13 states. However, when comparing
SNAP household characteristics for fiscal years 2008 and 2010, we found significant
increases in SNAP households’ average net income in 3 states that implemented BBCE in
fiscal year 2009. In these states, average net income increases appear to be related to
higher net incomes in households deemed eligible under BBCE. One state also had a
significant decrease in household size.
52
See GAO, Human Service Programs: Demonstration Projects Could Identify Ways to
Simplify Policies and Facilitate Technology Enhancements to Reduce Administrative
Costs, GAO-06-942 (Washington, D.C.: Sept. 19, 2006).

Page 28

GAO-12-670 Supplemental Nutrition Assistance Program

provide documentation of assets and caseworkers did not need to verify
asset information.
Consistent with annual increases in SNAP participation and benefit costs
between fiscal years 2001 and 2010, SNAP administrative costs generally
increased annually during this period, though at a lower rate. Certification
costs—a sub-set of SNAP administrative costs that include the cost of
staff determination of household eligibility for benefits—also generally
increased over this period (see fig. 11). Cost increases in recent years
are likely directly related to the $690.5 million in extra federal funding for
SNAP administrative costs provided to states through the Recovery Act
and the Department of Defense Appropriations Act, 2010, in response to
the national economic recession. 53 However, despite this additional
federal funding, because administrative costs increased at a lower rate
than SNAP participation, administrative costs per SNAP household
declined during this period (see fig. 12).

53
Although the federal government generally funds half of states’ administrative costs for
SNAP, this additional funding was provided to the states without a requirement that they
match it with their own funds.

Page 29

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 11: SNAP Administrative Costs, Fiscal Years 2001-2010

Note: These data include administrative costs paid by both federal and state governments.

Page 30

GAO-12-670 Supplemental Nutrition Assistance Program

Figure 12: Administrative Costs per SNAP Household, Fiscal Years 2001-2010

Note: These data include administrative costs paid by both federal and state governments.

While many states implemented BBCE during this period, the largest
decreases in these costs occurred in recent years when the economic
recession was also a factor. Specifically, during the recent recession,
states faced budgetary constraints to funding SNAP administrative
expenditures. Because states pay for approximately half of these
expenditures, when state tax revenues decrease during recessions, state
balanced budget requirements and other constraints affect state and local
governments’ ability to provide services at the same time that demand for
services increases. 54 In our site visits to five states, officials frequently
54
Although the most recent national recession affected states to different degrees, in part
depending on their tax structures and economic characteristics, state and local
governments generally experienced more severe and long-lasting declines in revenue
during this recession when compared to past recessions. For more information on the
effect of recessions on state and local governments, see GAO, State and Local
Governments: Knowledge of Past Recessions Can Inform Future Federal Fiscal
Assistance, GAO-11-401 (Washington, D.C.: Mar. 31, 2011).

Page 31

GAO-12-670 Supplemental Nutrition Assistance Program

noted how overwhelmed local SNAP caseworkers have been with the
increased workload during the recent recession. They noted that
workload increases have been driven by increases in SNAP participation
and the amount of time households remain on the program, as well as
budget constraints that hinder their offices’ ability to hire additional staff.
Across the seven local offices we visited in states that adopted BBCE
during the recent recession, staff noted that while BBCE helped
streamline the processing of individual cases, these improvements were
offset by the increased workload. However, some staff indicated they
believe reinstating the federal SNAP asset test that was removed under
BBCE would make their workload unmanageable.
In addition to the recession, other changes that states have made to
simplify and ease program administration during the last decade make it
difficult to determine BBCE’s full impact on administrative costs. For
example, state and local officials frequently cited the implementation of
reduced reporting requirements under the simplified reporting option, the
conversion of case files from paper to electronic formats, the
implementation of online SNAP applications, and increased use of phone
interviews as changes that also helped to ease staff workloads. Officials
in one state we visited noted that while these changes may have helped
to reduce administrative expenditures over time, some, like BBCE, may
have resulted in increased spending in the short-term due to the need for
training and modifications to computer systems. Further, while most state
SNAP officials we met with during our site visits felt that BBCE likely
decreased administrative expenditures to some extent, they did not know
the policy’s actual impact because of other changes.

State Eligibility
Changes May
Negatively Affect
Program Integrity
Payment Errors May Be
Negatively Affected

In recent years, the SNAP payment error rate declined to an historic low
while multiple program changes occurred, including BBCE, but evidence
suggests that factors other than BBCE may have played a larger role in
the decline. Between fiscal years 2000 and 2010, USDA reported that the
national payment error rate—the percentage of SNAP benefits paid in
error, including underpayments and overpayments—fell from 8.91 percent

Page 32

GAO-12-670 Supplemental Nutrition Assistance Program

to 3.81 percent as the number of states with BBCE policies increased
from 0 to 39. 55 Because most states’ BBCE policies eliminated the need
to confirm that SNAP household assets fall below certain limits, BBCE
effectively removed the potential for asset-related errors in these states.
However, USDA data indicate that most errors have been caused by
factors other than assets in recent years. In fact, fewer than 4 percent of
all error cases nationally have been caused by asset errors since 2000. 56
Therefore, it is likely that other factors had a greater impact on error rates
during this time. For example, the number of states adopting the
simplified reporting option for at least some SNAP households increased
during this period. Because this option eliminates substantial paperwork
requirements for participants and states, and reduces the number of
times income is verified, states experience fewer related errors. In
addition, states we visited reported that they had also made other
changes during this period to help lower their error rates, such as
incorporating the use of technology with new case management models
or digital files.
Further, both our analysis of USDA data and our discussions with SNAP
staff suggest that BBCE may, in fact, contribute to more payment errors.
Although BBCE has been promoted by USDA as a possible means to
reduce errors, we found that a greater percentage of SNAP households
eligible under BBCE that had incomes over the federal limits had payment
errors than other households (17.2 percent compared to 6.7 percent) in
fiscal year 2010. 57 This may be related to the fact that these households
were significantly more likely to have earned income, and income is a
frequent cause of SNAP payment errors. In addition, while most states’
BBCE policies removed a potential source of error by eliminating asset
limits, SNAP caseworkers we spoke with told us that a reduction in the
level of verification they perform may actually increase the potential for
errors as well as fraud. For example, staff in two states reported that

55

The 95 percent confidence interval for the 8.91 percent estimate is (8.56, 9.26) and the
95 percent confidence interval for the 3.81percent estimate is (3.61, 4.01).
56

Income-related errors have consistently been the largest contributor to payment errors,
representing 49 to 55 percent of all errors since 2000.
57

The 95 percent confidence interval for the 17.2 percent estimate is (14.3, 20.2), and the
95 percent confidence interval for the 6.7 percent estimate is (6.4, 7.0). These estimates
differ from the SNAP payment error rate, as the error rate measures benefit dollars
provided in error rather than the percentage of SNAP households with errors.

Page 33

GAO-12-670 Supplemental Nutrition Assistance Program

removing asset verification under BBCE has reduced their ability to
investigate other applicant information for possible inconsistencies.
Specifically, while asset verification often took considerable time to
perform, they noted that previously reviewing bank accounts gave them
the ability to identify regular deposits that may be income to ensure those
were reported by the applicant. Beyond changes due to BBCE,
caseworkers in several states we visited suggested there has been a
cultural shift towards an overall reduction in the level of verification and
investigation they perform, in part because of the increased participation
and workload related to the recent recession. They expressed concern
about maintaining a balance between providing assistance to those who
need it and ensuring program integrity, noting they worry about losing
access to information to help ensure integrity.

Limited Oversight May
Contribute to Unintended
Consequences

While federal rules provide states with considerable flexibility in designing
their BBCE policies, gaps in federal oversight may contribute to some
unintended consequences for SNAP and related programs’ integrity. We
found unintended consequences relating to three key areas:
•
•
•

TANF-Funded Service Not
Consistently Provided

provision of a TANF-funded service, 58
direct certification for free school meals, and
requirements for categorically eligible households to report changes
(in household circumstances).

Our visits to states suggest that SNAP applicants are not consistently
receiving the TANF-funded information required to confer categorical
eligibility and that the extent to which this information is TANF-funded is
unclear. According to USDA, BBCE policies make most households that
apply for SNAP categorically eligible because they receive a TANFfunded service, such as an informational brochure or toll-free number, as
long as the household’s income is within the state’s specified income limit
(see fig. 13). 59 However, in one state we visited, some local SNAP
caseworkers told us they did not consistently provide the guide to

58

For this report, we use “TANF-funded service” to refer to a TANF-funded non-cash
benefit or service that can be used by states to confer BBCE, per USDA guidance. Per
federal regulations, states may grant categorical eligibility to any household authorized to
receive a TANF-funded benefit or service. 7 C.F.R. § 273.2(j)(2)(ii) and (iii) (2012).
59

Applicants may be granted categorical eligibility solely through receipt of such a
document, regardless of whether or not they access or receive additional services.

Page 34

GAO-12-670 Supplemental Nutrition Assistance Program

services brochure to all applicants. In another, staff said that at the
applicant’s request, and/or if caseworkers think there is a need, they will
provide referrals to services. In a third state we visited, applicants were
directed on the SNAP application to call a toll-free number to receive an
informational brochure on services; however, we were unsuccessful in
obtaining this brochure after repeated (5) attempts to call the number
listed. Further, according to USDA, states must use TANF funds to pay
for either the document households receive or the services mentioned in
the document. If states use TANF funds to cover at least 50 percent of
the cost, they do not need to obtain USDA approval of their BBCE
policies. While SNAP officials in three of the states we visited confirmed
that the documents used to confer categorical eligibility are partially
TANF-funded, they did not know the exact percentage of TANF dollars
used to fund them.
Figure 13: Examples of How States Confer BBCE for SNAP
Massachusetts
A Non-Public Assistance food stamp household which consists of adults between the
ages of 19 and 59 is subject to the food stamp gross and net income tests and is no
longer subject to an asset test. Receipt of the Help for Those in Need: A Resources
Brochure confers a TANF Program service.
Nevada
All SNAP applicant households receiving the revised “This is Your Copy” page from the
back of the application are categorically eligible for benefits. There will be no resource
or gross/net income tests applied to these households. This page was revised to
contain the following information about services available from the Nevada Public
Health Foundation (NPHF):
Utilizing TANF funds, DWSS through the Nevada Public Health Foundation (NPHF),
has developed a class to target pregnant and parenting teens…
Washington
Basic Food Assistance Units with countable income up to 200% of the federal poverty
guidelines are eligible to use the department’s Online CSO website. This website
provides information about our programs as well as referrals to resources in the
community. This web-based information and referral service is partly funded with TANF
and TANF Maintenance Of Effort funds. Because of this funding, we use this service to
make Assistance Units categorically eligible for Basic Food if they have countable
income at or under 200%. Clients are notified and authorized to receive this service
through a text block on their approval letter for Basic Food.
Source: Commonwealth of Massachusetts, Executive Office of Health and Human Services, Department of Transitional Assistance,
Field Operations Memo 2008-27, “Maximized Categorical Eligibility for NPA Food Stamp Households,” (May 30, 2008); State of
Nevada, Department of Human Resources, Welfare Division, Policy Transmittal, “SNAP Categorical Eligibility – Receipt of TANF
Benefit” (March 9, 2009); and Washington State Department of Social and Health Services, Eligibility A-Z Manual, “Categorical
Eligibility for Basic Food,” (revised February 28, 2012).

Note: We did not conduct site visits to these 3 states, and we did not conduct an independent legal
analysis to verify this information.

Page 35

GAO-12-670 Supplemental Nutrition Assistance Program

Gaps in USDA’s oversight of states’ procedures for implementing BBCE
may contribute to the inconsistencies we found in providing qualified
applicants with the TANF-funded information or service that confers
BBCE. While USDA has issued guidance over the past 3 years in
response to various state questions about BBCE, the agency’s
documentation requirements for states that adopt it are limited. According
to agency guidance, while states must document that a household was
determined categorically eligible, USDA does not require states to
document that the TANF-funded service was received by applicants. As a
result, in a state where a document, such as a brochure, is used to confer
eligibility, the state does not have to verify that it has provided the
document to applicants as part of the eligibility determination process. In
addition, headquarters and regional USDA officials told us the agency
does not request documentation from states on the extent of TANF
funding used, even though that information is necessary to determine
whether a BBCE policy would require agency approval. Agency officials
added that the burden is on the states to let USDA know if approval is
needed. Agency officials also told us that while they provide technical
assistance to states, as needed, on the development of their BBCE
policies and collect summary information on states’ BBCE provisions,
they do not approve state BBCE policies.

Inappropriate Certification for
Free School Meals

Because states have flexibility to decide how to treat SNAP households
deemed eligible for $0 in benefits—an outcome more likely under
BBCE—some children have been inappropriately certified for free school
meals, including in two states we visited. Under SNAP, states have been
allowed to decide whether to deny eligibility to households who qualify for
$0 in benefits or whether to certify these households SNAP-eligible
without benefits. For school meals programs, statute indicates states
must certify children in households that receive SNAP benefits eligible for
free school meals—a process called direct certification that is designed to
ease administrative burden when certifying children for multiple
assistance programs with similar eligibility criteria. 60 Many states rely on
data matches between their SNAP program and district-level school data

60

42 U.S.C. § 1758(b)(4). School districts have been required to directly certify SNAP
recipients since school year 2008-2009. According to USDA, once a child is directly
certified for free school meals, eligibility lasts the entire school year regardless of a
change in family circumstances. Households may voluntarily report a change in family
circumstances; however, because of the year-long duration of eligibility, households are
not required to report changes in their categorical eligibility status.

Page 36

GAO-12-670 Supplemental Nutrition Assistance Program

to identify children eligible for direct certification, and beginning in school
year 2012-2013, all states are expected to do so. However, because a
state can certify families receiving $0 in SNAP benefits as eligible in its
SNAP data system, it can directly certify children in such families for free
school meals, even though they do not receive SNAP benefits. 61 This
practice occurred in two states we visited. SNAP officials in one state told
us the state adopted BBCE, in part, to potentially enable more children to
become eligible for free school meals. Local caseworkers in that state
similarly said that they believe parents apply for SNAP specifically
because they know their child(ren) are eligible for free school lunch even
if they are deemed eligible for $0 in SNAP benefits.
In recognition of this practice, in 2011, USDA issued guidance for states
through its regional offices reiterating that children in households
receiving $0 in benefits are not categorically eligible for free school meals
and therefore should not be directly certified; however, officials in the
states we visited were unaware of this guidance. In its October 2011
memorandum, USDA further suggested that state SNAP agencies work
with their school meal agency counterparts to ensure that children from
$0 benefit SNAP households are excluded from direct certification as
soon as possible. According to USDA, school meal agencies were to be
in compliance with this guidance by July 1, 2012. 62 Agency officials in
USDA’s regional offices representing the states we visited told us they
routinely transmit guidance and policy changes to states from the national
office. This guidance was also made available on USDA’s Web site.
However, in June 2012, we followed up with the two states we visited that
had been directly certifying children from $0 benefit SNAP households,
and state officials indicated the practice was still occurring, as they were
not aware of this guidance from USDA.

61

This may also impact other programs. For example, some school districts that receive
federal funds under Title I, Part A of the Elementary and Secondary Education Act of 1965
(20 U.S.C. § 6301-6339) use the number of children eligible for free and reduced-price
meals in each of their schools as a proxy measure of school poverty in order to
appropriately distribute those funds to their schools.
62
Letter to Program Directors-All Regions, “National School Lunch Program and Direct
Certification with SNAP,” signed by the Director of the Program Development Division,
FNS, USDA. October 25, 2011. The practical result is that direct certification of students
from families eligible for SNAP but entitled to $0 benefit should not continue in the 201213 school year.

Page 37

GAO-12-670 Supplemental Nutrition Assistance Program

Direct certification of children in categorically eligible SNAP households
creates another unintended consequence—one of effectively increasing
the income eligibility limit for free school meals for some children. While
the federal gross income-eligibility limit for SNAP aligns with that of the
school meals programs—providing free meal benefits to children in
households at or below 130 percent of the federal poverty guidelines 63—
the programs no longer align in states with BBCE policies that have
raised the SNAP gross income limit. In the 27 states with BBCE gross
income limits between 160 and 200 percent of the federal poverty
guidelines, children in categorically eligible households may receive free
school meals when, under traditional federal rules, they would not qualify
for free meal assistance. 64 In short, through their BBCE policies, some
states have effectively increased the income eligibility limits for two key
federal nutrition assistance programs.

Inequities in Reporting
Requirements

In states that have adopted BBCE, requirements for reporting changes in
household income or household size can vary, resulting in unequal
treatment of households. Under simplified reporting rules adopted by
nearly all states, 65 households are required to report changes in income
between scheduled reporting periods only if income exceeds the federal
SNAP gross income limit—130 percent of the federal poverty guidelines.
Because of BBCE, however, 27 states have, in effect, changed their
SNAP gross income limit to levels greater than 130 percent. USDA issued
SNAP guidance on change reporting requirements clarifying that, in
states with simplified reporting, categorically eligible households with
gross incomes over 130 percent of the federal poverty guidelines at the
time of certification have no federal SNAP reporting requirements until
they recertify or file a periodic report. 66 While guidance further indicates
that states may choose to require these households to report when their
gross income exceeds the income limit of the TANF program that confers

63

42 U.S.C. § 1758(b)(1)(A).

64

Under school meals rules, children in families with gross household incomes between
130 and 185 percent of the federal poverty guidelines, while not eligible for free meals, are
eligible to receive reduced price meals. 42 U.S.C. § 1758(b)(1)(A).
65

As of November 2010, all states, except California and Wyoming, used simplified
reporting for at least some SNAP households.
66
One of the recent documents that provides this guidance is the Memo to Regional
Directors, “Categorical Eligibility Questions and Answers,” signed by the Director of the
Program Development Division, FNS, USDA. January 26, 2010.

Page 38

GAO-12-670 Supplemental Nutrition Assistance Program

categorical eligibility, they are not required to do so. Two states we visited
do not require households with incomes above 130 percent of the poverty
guidelines to report changes in income between reporting periods. This
results in lower-income SNAP households having a greater reporting
burden than higher income SNAP households in order to retain their
benefits.
While USDA has issued guidance to states in this area, its guidance
relies on TANF reporting requirements that do not exist. USDA officials
told us that TANF rules require categorically eligible SNAP households to
report to TANF when their incomes exceed the income limit of the TANF
service used to confer BBCE. However, because BBCE households are
often authorized to receive a TANF-funded service through a brochure or
toll-free telephone number given to them by a SNAP office, they may not
be aware of any related TANF reporting requirements. Further, as we
have previously reported, state TANF agencies are not required to collect
data on many recipients of TANF-funded services, which include BBCE
households. 67 Accordingly, a state TANF agency would not seek
information on these households’ income changes in order to share that
information with the SNAP agency.

Conclusions

In response to the recent economic downturn and prolonged recovery,
the Supplemental Nutrition Assistance Program has grown to provide
unprecedented numbers of low-income households with benefits for food
assistance. While the substantial increases in SNAP participation led to
concerns that the large number of states adopting BBCE policies in
recent years may have been a driver of those increases, these policies
have had only a modest impact on program participation. Further, SNAP
generally continues to serve households with the same types of
characteristics it always has, and is intended to.
As federal and state governments face mounting fiscal pressures and
confront limited resources, ensuring the integrity of SNAP and other
programs spending public dollars is critical. While USDA touted BBCE as
a way to improve program integrity and administrative efficiency, state
adoption of BBCE has created unintended consequences that may

67
GAO, Temporary Assistance for Needy Families: Implications of Caseload and Program
Changes for Families and Program Monitoring, GAO-10-815T (Washington, D.C.: Sept.
21, 2010).

Page 39

GAO-12-670 Supplemental Nutrition Assistance Program

weaken both SNAP and related programs’ integrity and introduce
inequities. First, because gaps exist in USDA’s review of states’
procedures for implementing BBCE, some states are deeming
households eligible under BBCE without following the required steps to
do so. In addition, it is not known whether states are following the funding
requirements associated with these policies. Second, because USDA’s
guidance clarifying children’s eligibility for free school meals when their
families receive $0 in SNAP benefits—an outcome likely more common
because of BBCE—has not reached all states, school meal programs are
vulnerable to overpayments and abuse. Finally, USDA’s guidance on
SNAP reporting requirements has resulted in lower-income households
eligible under federal SNAP rules having to do more to retain their
benefits than higher-income SNAP households eligible solely because of
states’ BBCE policies.
While these unintended consequences of BBCE on SNAP program
integrity are potentially significant, they may also be easily addressed by
those overseeing and administering the program. At a time when the
economy has left more in need of assistance, SNAP continues to help
low-income households obtain adequate nutrition. As a result, any
changes to BBCE should carefully weigh the potential benefits and costs,
which at this time include the increased burden on state and local staff
who are already stretched thin as a result of decreased budgets and staff
resources.

Recommendations for
Executive Action

To improve SNAP program integrity and oversight, we are recommending
that the Secretary of Agriculture require FNS to take several actions:
•

•

•

Review state procedures for implementing BBCE, specifically those in
place for providing the relevant TANF-funded service to all SNAP
applicants deemed eligible under BBCE, as well as ensuring the
relevant service is funded with TANF dollars.
Disseminate the agency’s October 2011 guidance clarifying that
children in households certified as eligible for $0 in SNAP benefits
should not be directly certified to receive free school meals directly to
state agencies administering SNAP.
Revisit agency guidance on change reporting requirements to ensure
that all households, including those deemed eligible under BBCE with
incomes above the federal gross income limit, are treated equitably.

Page 40

GAO-12-670 Supplemental Nutrition Assistance Program

Agency Comments
and Our Evaluation

We provided a draft of this report to USDA for review and comment. On
July 16, 2012, the Associate Administrator for SNAP and other FNS
officials provided us with the agency’s oral comments. Officials stated that
they were in general agreement with the findings and recommendations
presented in the report and offered technical comments that we have
incorporated as appropriate. Officials also discussed the positive impacts
BBCE has had on SNAP, including state administrative relief and cost
savings, and emphasized our finding that BBCE policies have generally
not changed the characteristics of SNAP households. As a result, the
program continues to serve those it is intended to. Officials also noted
their agreement with our conclusion that BBCE’s benefits should be
considered when assessing changes to these policies. Concerning our
finding on the percentage of SNAP households with incomes over the
federal limits that had payment errors, officials noted that these
households may be more likely to have benefit errors than other SNAP
households because they have greater earned income and deductions—
factors that have been found to increase the likelihood of errors. We
agree, and our findings on the characteristics of this sub-group of
households support that conclusion. Further, officials suggested that the
total amount of benefit dollars provided in error to this sub-group of
households is likely relatively small because the average monthly benefit
provided to these households is much smaller than the average benefit
provided to all other SNAP households. Because of this, officials believe
that errors in these households impact the overall SNAP payment error
rate to a small extent, which is supported by the fact that the program’s
error rate has been relatively constant in recent years while the number of
states with BBCE has increased. While we agree that it is likely that the
total amount of benefit dollars provided in error to this sub-group of
households is relatively small, we did not develop such an estimate
during our analysis of the SNAP quality control data.
As agreed with your offices, unless you publicly announce the contents of
this report earlier, we plan no further distribution of it until 30 days from
the report date. At that time, we will send copies of this report to the
appropriate congressional committees, the Secretary of Agriculture, and
other interested parties. In addition, this report will be available at no
charge on GAO’s website at http://www.gao.gov.
If you or your staff have any questions concerning this report, please
contact me at (202) 512-7215 or brownke@gao.gov. Contact points for

Page 41

GAO-12-670 Supplemental Nutrition Assistance Program

our Offices of Congressional Relations and Public Affairs may be found
on the last page of this report. GAO staff who made key contributions to
this report are listed in appendix III.

Kay E. Brown
Director, Education, Workforce, and Income Security Issues

Page 42

GAO-12-670 Supplemental Nutrition Assistance Program

Appendix I: Objectives, Scope, and
Methodology
Appendix I: Objectives, Scope, and
Methodology

U.S. Department of
Agriculture (USDA)
Data Analysis
Quality Control Data

To determine the prevalence and characteristics of households deemed
eligible under states’ broad-based categorical eligibility (BBCE) policies
that had incomes over the federal Supplemental Nutrition Assistance
Program (SNAP) eligibility limits in fiscal years 2008 and 2010, we
analyzed the Food and Nutrition Service’s (FNS) quality control (QC)
system data of active SNAP cases. 1 Per federal SNAP requirements,
state officials draw monthly random samples of SNAP cases and review
them to determine the extent to which households received benefits to
which they were entitled. FNS officials in its regional offices and
headquarters perform a secondary review of a sub-set of each state’s
sample of cases. The weighted analyses of the QC data produce
nationally representative results.
To identify which households were deemed eligible under BBCE, and the
sub-set of BBCE households that had incomes over the federal SNAP
eligibility limits, we took several steps. First, we identified which states
had BBCE policies in place in fiscal years 2008 and 2010 using an FNS
compilation of BBCE policy implementation dates. Based on our
discussions with FNS officials, Mathematica Policy Research, Inc. staff, 2
and state and local staff we spoke to during our five site visits, we
assumed that once BBCE was enacted by a state, it was used as the
default SNAP eligibility policy. Therefore, in states with BBCE policies in
the fiscal year analyzed, we considered BBCE households to be those
denoted in the QC data as categorically eligible in which all members did
not receive cash assistance from another means-tested program. From
this group, we determined the sub-set of BBCE households that had
incomes over the federal SNAP eligibility limits.
We obtained the QC data directly from the QC database, which is made
available to the public via Mathematica Policy Research, Inc.’s Web site.

1
We reviewed information on SNAP in the 50 states; Washington, D.C.; Guam; and the
Virgin Islands. We use “states” throughout the report to refer to this group.
2

FNS contracts with Mathematica Policy Research, Inc. to maintain the SNAP QC data.

Page 43

GAO-12-670 Supplemental Nutrition Assistance Program

Appendix I: Objectives, Scope, and
Methodology

To analyze the data, we reviewed the technical user’s manual for both the
2008 and 2010 QC public release data sets and evaluated the sampling
methodology used to produce the data. We also reviewed the
documentation for the internal review and coding process that FNS
follows to prepare the QC data. Further, we checked the variables used in
our analysis for out-of-range values or outliers. To produce weighted
frequencies, weighted percents, and weighted dollar estimates for QC
variables at the state and national level, we used the household weight
variable provided in the public release QC data set. Because the records
in the SNAP QC data are from a random sample, data analysis results
are weighted estimates for a population of eligible households and thus
are subject to sampling errors associated with samples of this size and
type. The QC sample is only one of a large number of samples that states
might have drawn. As each sample could have provided different
estimates, we expressed our confidence in the precision of our QC data
estimates as a 95 percent confidence interval (e.g., plus or minus 10
percentage points). To produce 95 percent confidence intervals around
our weighted estimates, we used a statistical software package and an
appropriate variance estimation method suitable for the sample design of
the QC data. (Appendix II provides the estimates and 95 percent
confidence intervals for the data we present in the body of this report.)
Through our analysis of the QC data, a review of the technical
documentation, and interviews with FNS officials and Mathematica
statisticians, we determined that the QC public release data were
sufficiently reliable for the purposes of our audit.
In addition to conducting our own analysis of the QC data, we reviewed
national-level data on SNAP payment error rates—the percentage of
SNAP benefits paid in error—available from FNS for fiscal years 2000 to
2010. We also reviewed the primary sources of payment errors from
2000-2010 to help identify the extent to which payment errors were
attributable to assets or another source.

Other SNAP Data

In addition to the QC data, we reviewed other data on SNAP participation
and costs from USDA. Specifically, we analyzed data on average monthly
SNAP participation in recent years obtained from USDA reports. In
addition, we obtained data on total benefit costs and the average monthly
SNAP benefits per household from USDA’s Web site and the annual
SNAP State Activity Reports for fiscal years 2001-2011, as well as data
on the proportion of households receiving the maximum SNAP benefit
from the annual Characteristics of SNAP Households reports for fiscal

Page 44

GAO-12-670 Supplemental Nutrition Assistance Program

Appendix I: Objectives, Scope, and
Methodology

years 2001-2010. To assess administrative costs, we obtained data on
federal and state outlays and obligations for fiscal years 2001-2010 from
USDA’s National Data Bank. These data are annually reported by states
on the Standard Form 269 in specific cost categories designated by
USDA. In addition, we obtained data on state expenditures of the federal
Recovery Act funds provided for state administrative expenses in fiscal
years 2009 and 2010, as well as the related funds provided through the
Department of Defense Appropriations Act, 2010.

Site Visits

To better understand the effects of state BBCE policies on SNAP, as well
as other factors impacting SNAP, we conducted site visits to 5 states and
18 local offices responsible for administering SNAP in those states,
during January and February 2012. The states and localities we visited
were Arizona—Maricopa, Pima, and Pinal counties; Illinois—Cook and
Lake counties; North Carolina—Cabarrus, Gaston, Lincoln, and
Mecklenburg counties; South Carolina—Greenville, Laurens, and Pickens
counties; and Wisconsin—Kenosha, Milwaukee, and Racine counties. We
selected these states because they varied in their BBCE adoption dates,
in the characteristics of their BBCE policies, and in their geographic
locations. States selected also had relatively large SNAP caseloads and
generally high proportions of their SNAP households deemed eligible
under BBCE policies.
In each state, we interviewed state officials responsible for administering
SNAP, as well as local SNAP administrators and caseworkers at three or
four local offices. The local offices we visited ranged from urban to rural
areas. During the interviews we collected information about the state’s
BBCE policy and its application. In addition, we collected information
about recent trends in SNAP participation, benefit amounts,
administrative workload, and program errors, as well as BBCE’s impact
on each. We also collected information on other economic and noneconomic factors that have impacted SNAP. Also, at each local office we
observed the office’s general process for serving SNAP applicants,
including the forms, documents, and technological systems used, and we
gathered information on how BBCE was applied during the process.
Lastly, we conducted interviews with federal officials at the USDA
regional office associated with each state in order to discuss their role in
the oversight of SNAP. We cannot generalize our findings beyond the
states and localities we visited.

Page 45

GAO-12-670 Supplemental Nutrition Assistance Program

Appendix II: Quality Control Data Estimates
and 95 Percent Confidence Intervals
Appendix II: Quality Control Data Estimates
and 95 Percent Confidence Intervals

Table 4: SNAP Households That Would Not Have Been Eligible for the Program
without BBCE because Their Incomes Were over Federal Limits, in Fiscal Year 2010

Estimate

95 percent
confidence
interval

2.6%

2.4% - 2.8%

473,381

438,088 508,675

18,369,228

18,332,348 18,406,108

SNAP households deemed eligible under BBCE with
incomes over federal limits as a percentage of SNAP
households in states with BBCE policies that increased
the gross income limit

4.8%

4.4% - 5.2%

SNAP households deemed eligible under BBCE with
incomes over federal limits in states with BBCE policies
that increased the gross income limit

405,927

374,464 437,390

8,423,615

8,398,837 8,448,393

$1,965

$1,919 $2,010

151%

149% - 153%

SNAP households deemed eligible under BBCE with
incomes over federal limits as a percentage of all SNAP
households
SNAP households deemed eligible under BBCE with
incomes over federal limits
All SNAP households

All SNAP households in states with BBCE policies that
increased the gross income limit
Average total income of SNAP households deemed
eligible under BBCE with incomes over federal limits
Average total income of SNAP households deemed
eligible under BBCE with incomes over federal limits as a
percentage of the federal poverty guidelines
Source: GAO analysis of USDA’s SNAP quality control data.

Table 5: Characteristics of Specified Groups of SNAP Households, in Fiscal Year
2010

Estimate

95 percent
confidence
interval

SNAP households with at least one child as a percentage
of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

56.3%
48.5%

52.5% - 60.1%
47.9% - 49.1%

SNAP households with a single female as the head as a
percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

30.7%
25.8%

27.2% - 34.2%
25.3% - 26.3%

Page 46

GAO-12-670 Supplemental Nutrition Assistance Program

Appendix II: Quality Control Data Estimates
and 95 Percent Confidence Intervals

Estimate

95 percent
confidence
interval

SNAP households with at least one member receiving
Social Security benefits as a percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

27.8%
21.2%

24.3% - 31.4%
20.8% - 21.7%

Average amount of Social Security benefits received by:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

$340
$155

$293 - $388
$151 - $159

SNAP households with at least one member with unearned
income as a percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

60.0%
60.3%

56.3% - 63.7%
59.7% - 60.9%

SNAP households with at least one member with earned
income as a percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

65.9%
29.0%

62.2% - 69.5%
28.4% - 29.5%

SNAP households with at least one member receiving
Unemployment Insurance benefits as a percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

18.6%
6.4%

15.6% - 21.6%
6.1% - 6.7%

Average amount of Unemployment Insurance benefits
received by:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

$223
$55

$185 - $262
$52 - $58

Source: GAO analysis of USDA’s SNAP quality control data.

Page 47

GAO-12-670 Supplemental Nutrition Assistance Program

Appendix II: Quality Control Data Estimates
and 95 Percent Confidence Intervals

Table 6: Benefits Provided to SNAP Households That Would Not Have Been Eligible
for the Program without BBCE because Their Incomes were over Federal Limits, per
Month in Fiscal Year 2010

Estimate

95 percent
confidence
interval

SNAP benefits provided to households
deemed eligible under BBCE with incomes
over federal limits as a percentage of all
SNAP benefits

0.7%

0.6% - 0.8%

SNAP benefits provided to households
deemed eligible under BBCE with incomes
over federal limits

$38,333,145

33,972,894 –
42,693,395

$5,273,937,192

5,227,633,694 –
5,320,240,691

1.5%

1.3% - 1.6%

$33,987,767

$30,127,143 $37,848,392

$2,308,594,258

$2,277,254,048 $2,339,934,469

Total amount of benefits provided to all SNAP
households
SNAP benefits provided to households
deemed eligible under BBCE with incomes
over federal limits as a percentage of all
SNAP benefits in states with BBCE policies
that increased the gross income limit
SNAP benefits provided to households
deemed eligible under BBCE with incomes
over federal limits in states with BBCE
policies that increased the gross income limit
Total amount of benefits provided to all SNAP
households in states with BBCE policies that
increased the gross income limit
Source: GAO analysis of USDA’s SNAP quality control data.

Table 7: Benefits Received by Specified Groups of SNAP Households, in Fiscal
Year 2010

Estimate

95 Percent
Confidence
Interval

Average monthly SNAP benefit received by:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

$81
$293

$74 - $88
$290 - 295

SNAP households receiving the minimum benefit as a
percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

44.3%
2.7%

40.3% - 48.1%
2.5% - 2.9%

Source: GAO analysis of USDA’s SNAP quality control data.

Page 48

GAO-12-670 Supplemental Nutrition Assistance Program

Appendix II: Quality Control Data Estimates
and 95 Percent Confidence Intervals

Table 8: Estimated Percent of Specified Groups of SNAP Households Receiving
Deductions and the Estimated Average Amount of the Deduction in Fiscal Year
2010
95 Percent
Confidence
Estimate Interval
SNAP households receiving child support deduction as a
percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

6.7%
1.9%

4.6% - 8.9%
1.7% - 2.0%

Average amount of child support deduction received by:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

$20
$4

$13 - $27
$4 - $5

14.3%
3.5%

11.6% - 17.0%
3.3% - 3.7%

Average amount of dependent care deduction received by:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

$50
$8

$37 - $62
$7 - $8

SNAP households receiving earned income deduction as a
percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

65.9%
28.9%

62.2% - 69.5%
28.4% - 29.5%

$243
$57

$227 - $260
$56 - $59

SNAP households receiving medical deduction as a
percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

13.0%
3.5%

10.3% - 15.8%
3.3% - 3.8%

Average amount of medical deduction received by:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

$19
$5

$14 - $25
$5 - $6

SNAP households receiving dependent care deduction as a
percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

Average amount of earned income deduction received by:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

Page 49

GAO-12-670 Supplemental Nutrition Assistance Program

Appendix II: Quality Control Data Estimates
and 95 Percent Confidence Intervals

95 Percent
Confidence
Estimate Interval
SNAP households receiving excess shelter deduction as a
percentage of:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households
Average amount of excess shelter deduction received by:
•
SNAP households deemed eligible under BBCE with
incomes over federal limits
•
all other SNAP households

72.2%
70.5%

68.7% - 75.6%
70.0% - 71.1%

$220
$265

$205 - $235
$263 - $268

Source: GAO analysis of USDA’s SNAP quality control data.

Page 50

GAO-12-670 Supplemental Nutrition Assistance Program

Appendix III: GAO Contacts and Staff
Acknowledgments
Appendix III: GAO Contacts and Staff
Acknowledgments

GAO Contact

Kay E. Brown, (202) 512-7215 or brownke@gao.gov

Staff
Acknowledgments

Kathy Larin (Assistant Director), Rachel Frisk (Analyst-in-Charge), Avani
Locke, and David Perkins made significant contributions to all aspects of
this report. Also contributing to this report were Carl Barden, Marquita
Campbell, Susannah Compton, Heather Dunahoo, Greg Kutz, Jean
McSween, Mimi Nguyen, Susan Offutt, Rhiannon Patterson, Kathy
Peyman, Almeta Spencer, Craig Winslow, and Jill Yost.

Page 51

GAO-12-670 Supplemental Nutrition Assistance Program

Related GAO Products
Related GAO Products

State and Local Governments: Knowledge of Past Recessions Can
Inform Future Federal Fiscal Assistance. GAO-11-401. Washington, D.C.:
March 31, 2011.
Temporary Assistance for Needy Families: Implications of Caseload and
Program Changes for Families and Program Monitoring. GAO-10-815T.
Washington, D.C.: September 21, 2010.
Supplemental Nutrition Assistance Program: Payment Errors and
Trafficking Have Declined, but Challenges Remain. GAO-10-956T.
Washington, D.C.: July 28, 2010.
Domestic Food Assistance: Complex System Benefits Millions, but
Additional Efforts Could Address Potential Inefficiency and Overlap
among Smaller Programs. GAO-10-346. Washington, D.C.: April 15,
2010.
Food Stamp Program: FNS Could Improve Guidance and Monitoring to
Help Ensure Appropriate Use of Noncash Categorical Eligibility.
GAO-07-465. Washington, D.C.: March 28, 2007.
Human Service Programs: Demonstration Projects Could Identify Ways to
Simplify Policies and Facilitate Technology Enhancements to Reduce
Administrative Costs. GAO-06-942. Washington, D.C.: September 19,
2006.
Food Stamp Program: States Have Made Progress Reducing Payment
Errors, and Further Challenges Remain. GAO-05-245. Washington, D.C.:
May 5, 2005.
Food Stamp Program: Farm Bill Options Ease Administrative Burden, but
Opportunities Exist to Streamline Participant Reporting Rules among
Programs. GAO-04-916. Washington, D.C.: September 16, 2004.
Food Stamp Program: Steps Have Been Taken to Increase Participation
of Working Families, but Better Tracking of Efforts Is Needed.
GAO-04-346. Washington, D.C.: March 5, 2004.
Food Stamp Program: States’ Use of Options and Waivers to Improve
Program Administration and Promote Access. GAO-02-409. Washington,
D.C.: February 22, 2002.

Page 52

GAO-12-670 Supplemental Nutrition Assistance Program

Related GAO Products

Means-Tested Programs: Determining Financial Eligibility Is
Cumbersome and Can Be Simplified. GAO-02-58. Washington, D.C.:
November 2, 2001.
Food Stamp Program: States Seek to Reduce Payment Errors and
Program Complexity. GAO-01-272. Washington, D.C.: January 19, 2001.

(131118)

Page 53

GAO-12-670 Supplemental Nutrition Assistance Program

GAO’s Mission

The Government Accountability Office, the audit, evaluation, and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and
policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding decisions.
GAO’s commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of
GAO Reports and
Testimony

The fastest and easiest way to obtain copies of GAO documents at no
cost is through GAO’s website (www.gao.gov). Each weekday afternoon,
GAO posts on its website newly released reports, testimony, and
correspondence. To have GAO e-mail you a list of newly posted products,
go to www.gao.gov and select “E-mail Updates.”

Order by Phone

The price of each GAO publication reflects GAO’s actual cost of
production and distribution and depends on the number of pages in the
publication and whether the publication is printed in color or black and
white. Pricing and ordering information is posted on GAO’s website,
http://www.gao.gov/ordering.htm.
Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537.
Orders may be paid for using American Express, Discover Card,
MasterCard, Visa, check, or money order. Call for additional information.

Connect with GAO

Connect with GAO on Facebook, Flickr, Twitter, and YouTube.
Subscribe to our RSS Feeds or E-mail Updates. Listen to our Podcasts.
Visit GAO on the web at www.gao.gov.

To Report Fraud,
Waste, and Abuse in
Federal Programs

Contact:
Website: www.gao.gov/fraudnet/fraudnet.htm
E-mail: fraudnet@gao.gov
Automated answering system: (800) 424-5454 or (202) 512-7470

Congressional
Relations

Katherine Siggerud, Managing Director, siggerudk@gao.gov, (202) 5124400, U.S. Government Accountability Office, 441 G Street NW, Room
7125, Washington, DC 20548

Public Affairs

Chuck Young, Managing Director, youngc1@gao.gov, (202) 512-4800
U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, DC 20548

Please Print on Recycled Paper.