Download:
pdf |
pdfATTACHMENT C
EXPLORATORY STUDY MEMORANDUM
This page is deliberately left blank.
Exploratory Study Memorandum
Final Submission
Authors:
Anne Paprocki
Marian Negoita
Prepared for:
Barbara Murphy and Kameron Burt
Evaluation of Alternatives to
Improve Elderly Access to SNAP
Contract AG-3198-C-16-0008
Social Policy Research Associates
March 14, 2017
This page is deliberately left blank.
Contents
Introduction .................................................................................................................................... 1
Recent Trends in Elder SNAP Enrollment and Participation ........................................................... 2
Barriers to Elder SNAP Access ......................................................................................................... 7
Interventions to Increase Elder SNAP Access ............................................................................... 14
Overall Themes, Challenges, and Areas for Further Study ........................................................... 28
Next Steps for the Study ............................................................................................................... 30
References ................................................................................................................................... R-1
Appendix A: SNAP Participation Rates for Eligible Elder Individuals, FY 2014 ............................ A-1
Appendix B: State Intervention Index .......................................................................................... B-1
Appendix C: Methodology Used to Calculate SNAP Participation Rates ..................................... C-1
i
This page is deliberately left blank.
ii
Introduction
This memorandum is intended to ground the Evaluation of Alternatives to Improve Elderly
Access to Supplemental Nutrition Assistance Program (SNAP). The United States Department of
Agriculture’s Food and Nutrition Service (FNS) commissioned the multi-year study to better
understand how to maximize elder (60+) access to SNAP. The evaluation will have four key
components: (1) an exploratory study to lay the groundwork and guide the rest of the project;
(2) a study of State interventions, based primarily on interviews with State SNAP administrators
from 10 States; (3) a study of elder participant perspectives, based on interviews and focus
groups with elder SNAP applicants, recipients, and eligible non-participants in the same States;
and (4) a quantitative analysis of the effect of various interventions to increase elder SNAP
access on enrollment and churn.
As the key deliverable for the exploratory study, this memorandum summarizes what is already
known about elder SNAP participation levels and caseload dynamics over time; factors
influencing elder participation in SNAP; and the scope, range, and effects of State interventions
for elders to date. It draws primarily on the following sources of data:
an analysis of SNAP Quality Control (QC) data from Fiscal Years 2010 to 2015 that
provides an update to known trends in elder SNAP participation; data from the Annual
Social and Economic Supplement (ASEC) of the Current Population Survey (CPS), the
American Community Survey, individual income tax data from the Census Bureau’s
Small Area Estimation Branch, and Census Bureau population estimates for 2010
through 2014 were used to estimate the number of elders who were eligible for the
program in years 2010 through 2014;
a literature review of just over 100 sources related to elder SNAP access, SNAP access
more generally, and elder access to other government benefit programs;
an updated index of State policy options, waivers, and demonstrations related to elder
SNAP access based on reports and communication with FNS;
interviews with the SNAP Directors from all seven of the FNS Regional Offices and with
individuals from five key national organizations focused on elder SNAP access1; and
a group interview with FNS National Office staff members about current challenges and
successes with elder SNAP access.
1
The five national organizations are the AARP Foundation, Benefits Data Trust, Center on Budget and Policy
Priorities, the Massachusetts Law Reform Institute, and the National Council on Aging.
1 | Elder SNAP Access
The overall goals of the memorandum are to (1) weave together findings from each of the
above data sources to provide the research team with the context needed to plan and carry out
subsequent project activities and (2) identify key areas for exploration in these forthcoming
activities. To accomplish these goals, the memorandum describes recent trends in elder SNAP
participation, identifies barriers that may hinder elder participation in SNAP, and discusses the
various interventions that have been used to date to improve elder access. The memorandum
ends with a discussion of some of the overall themes and challenges associated with elder SNAP
access and a review of States selected for further study. Appended to this memorandum are
two additional resources: a table that illustrates the Fiscal Year 2014 elder SNAP participation
rates by State (Appendix A) and a State intervention index that lays out the current relevant
interventions implemented by each State (Appendix B).
Recent Trends in Elder SNAP Enrollment and Participation
Low take-up rate for eligible individuals age 60 and older is one of SNAP’s most enduring
challenges. Estimates of the SNAP take-up rate suggest that, historically, only about a third of
eligible elders have participated in SNAP (Cunnyngham, 2010), while the national participation
rate among eligible individuals of all ages is typically twice as high (Cunnyngham, Castner, &
Sukasih, 2012; Eslami, Leftin, & Strayer, 2012). More recently, however, elder participation has
trended up. An analysis of the four most recently available years of participation and eligibility
data carried out by Mathematica Policy Research (Mathematica) indicates a rise in participation
rates2 that parallels a rise in the overall population of SNAP-eligible individuals (see Exhibit 1
below).
2
Appendix C describes the strategy used to calculate program participation rates.
2
Exhibit 1: Recent SNAP Participation Rates
90.0
80.0
National
70.0
Participation Rate
60.0
50.0
40.0
Elder
30.0
20.0
10.0
0.0
FY 2010
FY 2011
FY 2012
FY 2013
FY 2014
Fiscal Year
Sources: SNAP QC, Annual Social and Economic Supplement (ASEC) of the Current Population Survey (CPS),
American Community Survey, and individual income tax data and Census Bureau population estimates for 2010
through 2014.
As shown in Exhibit 1, participation rates have risen by approximately 10 percentage points for
both the general population and elder individuals, with the largest increase recorded between
2010 and 2012. The elder participation rate reached 40 percent in 2012, which can be
considered high by historical standards3.
The SNAP participation rate depends, of course, on two numbers: a numerator (the number of
people who enroll in SNAP) and a denominator (the number of people who are eligible to
enroll). The increase in participation rates has resulted from the number of participants
growing faster than the number of eligible individuals. As shown in Exhibit 2, this is particularly
true for elders—the number of eligible individuals grew 11 percent between Fiscal Year 2010
and Fiscal Year 2014, whereas the number of participants grew by a considerable 40 percent.
3
https://www.ers.usda.gov/topics/food-nutrition-assistance/supplemental-nutrition-assistance-programsnap/charts/snap-participation-rates/
3 | Elder SNAP Access
Exhibit 2: Number of SNAP Participants and Eligible Individuals
.
.
All eligible individuals
FY 2010
FY 2011
FY 2012
FY 2013
FY 2014
Elder eligible individuals
FY 2010
FY 2011
FY 2012
FY 2013
FY 2014
90% Confidence Intervals
Lower
Upper
Bound
Bound
Eligible
Individuals
(000s)
Participation
Rate
37,481,880
40,693,688
42,129,048
43,230,759
42,300,155
52,263,519
52,160,864
50,708,090
50,716,212
51,025,996
71.7
78.0
83.1
85.2
82.9
70.8
77.0
82.0
84.1
81.8
72.6
79.1
84.2
86.3
84.0
2,936,925
3,395,405
3,720,288
3,857,771
4,118,324
8,898,062
8,872,267
8,944,627
9,436,616
9,867,805
33.0
38.3
41.6
40.9
41.7
31.5
36.5
39.8
39.0
40.0
34.5
40.0
43.4
42.7
43.5
Participants
(000s)
Sources: SNAP QC (for the calculation of program participants); Annual Social and Economic Supplement (ASEC) of
the Current Population Survey (CPS), American Community Survey, and individual income tax data and Census
Bureau population estimates for 2010 through 2014 (for the calculation of numbers of SNAP eligible individuals).
Exhibit 3 below charts the evolution of elder SNAP participation against two important
measures of poverty—the number of individuals 60 and over whose income is below 125
percent of the Federal Poverty Line (FPL) and the number of individuals 60 and over whose
income is below 200 percent FPL4. We can glean two insights from these data. First, SNAP
appears to be an effective countercyclical policy tool because it responds to increases in
poverty. Although the overall economy has improved, the number of elders in poverty has
increased, leading to increased SNAP enrollment. These trends corroborate the evidence from
numerous other sources that SNAP acts as an effective countercyclical policy tool (Mabli,
Martin, & Castner, 2009; Klerman & Danielson, 2009; Ganong & Liebman, 2013; Mabli et al.,
2014).
Second, after a period of relative stagnation, it appears that the number of eligible elders has
been growing since Fiscal Year 2013. We can only speculate about what might have contributed
to this trend, but at least two explanations are possible. First, the increased number of eligible
elders might be the result of policies that relax program eligibility criteria, such as broad-based
categorical eligibility (BBCE), which decreases or eliminates the asset test and/or increases the
gross income limit for SNAP applicants, thus allowing more people to be eligible for the
program. Second, it could be the result of secular demographic trends, including the aging of
4
The reason to include both measures of poverty is that the SNAP gross income limit is 130 percent of the FPL for
most households, and some elder individuals may still qualify for SNAP once deductions are taken into account.
4
the Baby Boomer generation, general poverty trends, and overall population growth. This study
aims to shed light on the factors behind this growth.
Exhibit 3: Recent Trends in Elder SNAP Participation
20,000,000
18,000,000
Number of Individuals
16,000,000
14,000,000
12,000,000
10,000,000
8,000,000
6,000,000
4,000,000
2,000,000
0
FY 2010
FY 2011
FY 2012
FY 2013
FY 2014
Fiscal Year
Eligible Elders
Elder SNAP Participants
Individuals 60 and above with incomes below 125%
FPL
Sources: SNAP QC (for the calculation of program participants); Annual Social and Economic Supplement (ASEC) of
the Current Population Survey (CPS), American Community Survey, and individual income tax data and Census
Bureau population estimates for 2010 through 2014 (for the calculation of numbers of SNAP eligible individuals);
Aging Integrated Database (for poverty data).
In addition to the national trends in elder SNAP participation described above, for this study, an
understanding of State participation rates also is necessary. Although SNAP benefits are
federally funded, program costs and administration are shared by federal, State, and county
governments. State agencies (and in some cases, counties under State supervision) administer
the program on a day-to-day basis and have broad discretion in establishing State-level
program policies (including those regarding enrollment). Because of this flexibility, as well as
the demographic and economic characteristics of each State, elder SNAP participation rates
vary widely across States. For example, in Fiscal Year 2014, elder SNAP participation rates
ranged from a low of about 21 percent in Wyoming and California to a high of about 68 percent
in Vermont and 72 percent in New York. Appendix A provides more detail, illustrating the elder
SNAP participation rate for each State for Fiscal Year 2014. These participation rates will be an
important part of State selection for subsequent components of this study.
5 | Elder SNAP Access
QC data can also be used to track changes in the demographic makeup of elder SNAP
participants over time (see Exhibit 4). Of the total number of program participants, the
proportion of elders has been gradually increasing from eight percent in Fiscal Year 2010 to
almost 11 percent in Fiscal Year 2015. This finding confirms trends shown earlier in the report.
As shown in Exhibit 2, the number of elder SNAP participants has grown from 2.9 million to 4.1
million (40.2 percent) between FY 2010 and FY 2014, whereas the overall number of adult
participants has grown from 37.5 million to 42.3 million (12.9 percent). The faster rate of elder
caseload growth accounts for the increase in the proportion of elders among the total number
of program participants.
Another notable trend has been the increase in the percentage of elder participants who live in
a metropolitan area, from 78 percent in Fiscal Year 2010 to 82 percent in Fiscal Year 2015. A
majority of the elders who participate in SNAP live alone (almost three quarters); almost twothirds of the elders who participate in SNAP are women.5
Exhibit 4: Socio-demographic Characteristics of Elder SNAP Participants
Percentage of elders among all SNAP participants
Percentage of elder female SNAP participants
Percentage of elder participants in one-person
SNAP household
Percentage of elder participants in SNAP
household with only other elders
Percentage of elder SNAP participants living in a
metropolitan area
FY 2010
7.9
65.9
FY 2011
8.5
65.9
FY 2012
9.0
65.2
FY 2013
9.3
64.0
FY 2014
10.1
63.3
FY 2015
10.6
63.6
73.2
72.4
74.0
72.6
74.7
72.9
15.5
16.9
15.8
16.8
15.5
17.1
78.0
79.1
78.3
78.2
81.0
81.6
Sources: FY 2010 to FY 2015 SNAP QC data files
The QC database contains information about other programs in which SNAP participants enroll,
and these data constitute another useful way to describe participants and trends. Exhibit 5
below offers a glimpse at the shifting makeup of elder program participants between FY 2010
and FY 2015. In particular, the percentage of elder SNAP participants who earn income has
increased slightly (by two percentage points). In addition, the percentage of program
participants whose gross income is under the FPL fell by seven percentage points from Fiscal
Year 2010 to Fiscal Year 2015. In this same period, the percentage of elder SNAP participants
whose SNAP benefits are equal to the minimum SNAP benefit has almost doubled, from 10
percent to 19 percent. Thus, although all elder SNAP participants have low incomes, many of
the additional participants that the program gained during the observed period appear to have
5
Distributions by other socio-demographic characteristics such as age and race/ethnicity were not available by
year, so they are not presented here.
6
slightly higher incomes (as reflected in the slightly higher probability to earn income, lower
percentage of those whose gross income is under 100 percent of FPL, and higher percentage
receiving the minimum benefit amount) than the traditional core of participants. This suggests
that at least some of the policies enacted to stimulate SNAP enrollment among elders are
functioning as intended; they are bringing participants to the program who perhaps would not
have traditionally enrolled. This study will provide some answers regarding the particular
policies that are associated with this trend.
Exhibit 5: Elder SNAP Participants’ Participation in Benefit Programs and Income
Percentage of elder SNAP participants receiving
Social Security income
Percentage of elder SNAP participants with
earned income
Percentage of elder SNAP participants with
benefits equal to the minimum benefit
Percentage of elder SNAP participants with gross
income under 100 percent of poverty
FY 2010
FY 2011
FY 2012
FY 2013
FY 2014
FY 2015
66.8
67.8
69.3
70.0
68.9
68.8
5.6
6.3
6.7
6.9
7.2
7.6
10.4
12.4
13.6
15.2
17.7
19.3
77.2
73.6
70.6
69.3
71.5
69.4
Sources: FY 2010 to FY 2015 SNAP QC data files
Barriers to Elder SNAP Access
As illustrated above, elder individuals have historically participated in SNAP at substantially
lower rates than the general population. Between 2010 and 2014, the participation rate among
the eligible elder population ranged from 33 to 42 percent, compared to 72 to 83 percent for
the eligible population as a whole (SNAP QC Data). While these numbers do show growth in the
elder SNAP participation rate, that rate remains half of the SNAP participation rate overall.
This low participation rate among low-income elder individuals is cause for public concern.
Without SNAP, elder individuals may be unable to meet their nutritional needs (Cody & Ohls,
2005), forgo medicine for food (Sattler & Lee, 2013), or be unable to pay utility bills or secure
safe or stable housing (O’Brien, Wu, & Baer, 2010). Some elder individuals without SNAP
benefits may have fewer resources to purchase food; elder diet insufficiency has been
connected to poorer mental and physical health outcomes as well as increased strain on
caregivers (Fuller-Thompson & Redmond, 2008). Previous research suggests that among all
SNAP households, SNAP not only increases food access and reduces food insecurity (Gundersen
& Ziliak, 2008), but also has significant positive effects on household incomes (LeBlanc, Lin, &
Smallwood, 2006), and raises millions of SNAP recipients out of poverty—over 10 million
individuals in 2012, according to the Center on Budget and Policy Priorities (2016). Thus,
increasing elder SNAP participation would appear to yield unambiguously positive benefits for
7 | Elder SNAP Access
society by improving the food security, financial security, and well-being for many elder
individuals living on limited incomes.
There is probably no single cause responsible for the persistently low elder SNAP participation
rate. Previous research and the interviews conducted for this memorandum suggest that a
variety of barriers make participation more difficult and combine to lower the participation
rate. These barriers can be divided into two broad categories: (1) barriers to applying for and
enrolling in SNAP and (2) barriers to remaining on SNAP.
Barriers to Applying for and Enrolling in SNAP
Two broad categories of barriers—societal and procedural—prevent and/or discourage elder
individuals from applying for and enrolling in SNAP.
Societal Barriers
Societal barriers to applying for SNAP include those that one interview respondent described as
“conceptual, personal, and psychological.” They involve the elder individual’s perception of the
program and who should take advantage of it. The following examples of societal barriers came
up frequently in the literature and in the interviews.
Stigma. Embarrassment, feelings of failure, hurt pride, dislike of government assistance, and
loss of independence are all reasons cited by elder individuals for not participating in SNAP
(AbuSabha et al., 2011; Bartlett & Burstein, 2004; Gabor et al., 2002; Kim & Frongillo, 2009;
Ponza & McConnell, 1996). In all these cases, the individual attaches some kind of stigma to
SNAP participation. Related to stigma, several interviewed respondents reported that some
elder individuals have the sense that SNAP is a “government handout” different from a benefit
program like Social Security for which they have “paid in.” However, while many respondents
identified the stigma of SNAP as a persistent barrier to elder SNAP access, some also said that
they believe it has lessened over time, especially with the advent of Electronic Benefit Transfer
(EBT). It is also worth noting that some research on government programs has shown that
stigma does not actually have a large effect on uptake rates (Blavin, Dorn, & Dev, 2014; Wu,
2009).
Perceptions of being undeserving of SNAP benefits. Research has documented that some
individuals feel they should not need SNAP, that others are needier, or that their participation
displaces other needy people (Bartlett & Burstein, 2004; Mack & Paprocki, 2016). Interview
respondents also reported that some eligible elders do not understand SNAP’s status as an
entitlement program. Some elder individuals also have said that they feel SNAP is “marketed”
towards younger families and is thus not designed for them (Geiger, Wilks & Livermore, 2014;
Ponza & McConnell, 1996).
8
Fear of fraud. Elder individuals may be reluctant to give out the personal information
(particularly social security numbers) required for a SNAP application due to fears about
identify theft (Gabor et al., 2002; Mack & Paprocki, 2016).
Perceived (or real) lack of need for SNAP. Some low-income elder individuals may choose to
get food assistance from other sources, such as through subsidized and congregate meals, food
banks, and senior centers (Fitzpatrick, Greenhalgh-Stanley, & Ver Ploeg, 2015; Gabor et al.,
2002, Oemichen & Smith, 2016; Wu 2009). Others may decline SNAP due to the challenges of
getting to a grocery store in a food desert (Fitzpatrick et al., 2015) or due to limited mobility
and lack of transportation.
Procedural Barriers
In addition to the societal barriers referenced above, there are also barriers to applying for and
enrolling in SNAP that one interview respondent described as “procedural.” These barriers are
more specifically connected to how SNAP is implemented and advertised by FNS and the States.
While respondents highlighted elders’ lack of knowledge about specific eligibility rules and
application procedures as barriers in this category, it is worth noting that many thought most
elder individuals are at least aware that SNAP exists. This general awareness about the
existence of SNAP as a program is also confirmed by the literature (Bartlett et al., 2004; Currie,
2004).
Lack of information about eligibility rules and/or application processes. While individuals are
usually aware that SNAP exists, they may not think that they qualify for the program (Ponza &
McConnell, 1996; Wu, 2009). The program’s complex eligibility calculations add to this
confusion, and those who are closer to the eligibility cut-off tend to be less sure that they
would qualify for SNAP than those who are well under it (Bartlett et al., 2004). In addition, elder
individuals may not realize that they fall under different eligibility criteria than the larger
population (Ponza et al., 1999). Many elder individuals also believe they are ineligible because
they have assets or they do not have dependent children living with them. Some do not know
how to apply for benefits or how to gain this knowledge (Bartlett & Burstein, 2004; Cody &
Ohls, 2005; Gabor et al., 2002; McConnell & Ponza, 1999).
Perceived or real burdens of applying. According to both the literature and the interviews
conducted for this memorandum, the burdens associated with applying for SNAP constitute a
major barrier to elder SNAP participation. Elder individuals may find it difficult to get to the
SNAP office because of lack of transportation, health issues, or physical limitations, and they
may not be aware of options to complete applications online or conduct interviews and “sign”
the application over the phone (Bartlett & Burstein, 2004; Bartlett et al., 2004; Cody & Ohls,
2005; Gabor et al., 2002). Research has documented that elder individuals often perceive
interactions with SNAP personnel as unpleasant, application requirements as difficult to
9 | Elder SNAP Access
understand, and documentation of income and assets, where needed, as burdensome and an
invasion of privacy (AbuSabha et al., 2011; Cody & Ohls, 2005; Gabor et al., 2002). The length of
the application form and the documentation required were also cited by several interviewed
respondents as specific barriers. Some noted that as States create more combined applications
that screen for several benefits programs at once, application forms necessarily become longer
and more complex. The form-filling burdens associated with applying for SNAP can be especially
intimidating or overwhelming for elder individuals with disabilities and those experiencing
cognitive decline (Herd, 2015).
Perception or reality of low benefit levels. Nearly every interviewed respondent said that the
“myth of the $16 [or minimum] benefit” plays a big role in preventing some eligible elder
individuals from applying. As indicated above, many individuals already perceive the SNAP
application process to be burdensome. If they believe that their efforts will only result in a small
monthly benefit, some will decide it is not worth the effort of applying (Bartlett & Burstein,
2004; Bartlett et al., 2004; Cody & Ohls, 2005; Gabor et al., 2002; Kim & Frongillo, 2009;
McConnell & Nixon, 1996; McConnell & Ponza, 1999). Some eligible elder individuals qualify (or
think they will qualify) for relatively low benefits because they live alone or with one other
person and have Social Security or Supplemental Security Income (SSI) that brings their income
to, or close to, the poverty level (Cody & Ohls, 2005). Elder individuals may not be aware that
the opportunity to claim medical expenses as deductions to household income could raise their
benefit amount (Jones, 2014). Even though the majority of elder individuals do get more than
$16 a month in SNAP benefits, the percentage of elder individuals who get the minimum
benefit amount has been rising. In Fiscal Year 2010, 10 percent of elder SNAP participants
received the minimum benefit, but this number had increased to 19 percent by Fiscal Year 2015
(SNAP QC Data). For some individuals, the perceived costs of applying for SNAP (particularly in
terms of time or loss of privacy) may outweigh the benefits (Cody & Ohls, 2005; Schanzenbach,
2009).
SNAP modernization and technology. As States have modernized their SNAP programs, there is
an increased use of call centers, online applications, and other automated technologies (Rowe
et al., 2010). While such updates may be more convenient for some elder individuals, others
struggle with these systems, which generally involve less one-on-one assistance from State
eligibility workers. In one study, most elder focus group participants said they preferred face-toface interviews to those conducted on the telephone (Gabor et al., 2002). In another small
study with participants of all ages, over half preferred traditional service delivery models to a
modernized system (Heflin, London, & Mueser, 2010). However, interviewed respondents also
pointed out that as new individuals age into the elder cohort, they may have more experience
with technology and more of a preference for these new ways of applying.
10
The literature and interviews for this memorandum also identified States’ learning curves
setting up their modernized systems as challenging to elder individuals. For example, as the
systems get set up, there may be long initial wait times when calling into a call center (Cody et
al. 2010). This challenge may also lessen over time.
Barriers to Remaining on SNAP
Some research reports that once they have enrolled, elder individuals are less likely to “churn
off” of SNAP than other individuals (Mills et al., 2014; Wu 2009). Typically, elder participants
have longer certification periods than do younger participants, and this results in increased
SNAP receipt (Ratcliffe, McKernan, & Finegold, 2007; Rutledge & Wu, 2013). Nevertheless, the
literature and interviewed respondents identified several specific barriers that elder individuals
face in staying enrolled once on SNAP. Several respondents said that while elder participants
may have fewer opportunities to churn, when these opportunities do come up, they are at
increased risk for falling off the benefit.
Recertification and reporting processes. All SNAP households must conduct periodic reporting
and undergo an occasional recertification process, though the frequency with which reporting
and recertification is done varies by State and household type. Elder individuals without earned
income often go through these steps less frequently and have to produce less documentation,
in part because their income and household situations are less likely to change (Bartlett et al.,
2004). In most States, households that have only elder or disabled members with no earned
income are either certified for 12 months (with no periodic reporting) or for 24 months (with a
12-month report) (U.S. Department of Agriculture, 2016). These processes can be burdensome
for all participants, with researchers noting that eligible households are more likely to leave the
program in recertification months (Bartlett et al., 2004).
Even though the reporting and recertification periods are less frequent for elder households
than for other households, these requirements can be barriers for elder individuals and can
lead to them dropping out of SNAP. Interviewed respondents from the FNS National and
Regional Offices and from key national organizations noted that elder individuals may have
trouble understanding what is required of them for recertification and reporting; they
explained that recertification notices are frequently in small print and use complicated legal
language. Because many States have abandoned the caseworker model in favor of the
transaction model, an individual may not have a specific caseworker to contact about these
notices. Elder participants may be unsure who to turn to with questions; as a result they may
get frustrated, not respond to notices, and eventually exit from SNAP. Recertifications can also
involve interviews, and respondents explained that elder individuals may have trouble
scheduling or attending interviews (even when they are done by phone). Elder individuals who
11 | Elder SNAP Access
have earned income may face greater burdens because they have more frequent reporting
requirements (Gabor et al., 2002).
Expunged benefits. While several studies have indicated that the adoption of EBT has led to an
improved program experience and reduced the stigma of SNAP participation (Danielson &
Klerman, 2006), two interviewed respondents indicated that the technology has caused
problems for some elder individuals. They explained that in some States, individuals who had
amassed several months of credit on their EBT cards had their benefits expunged after several
months of nonuse. According to these respondents, the States cleared these benefit build-ups
because they thought that the long period of nonuse indicated possible fraud. The
disappearance of their benefits caused some individuals to think they were out of the program
and to ignore subsequent notices. The problem with this expunging of benefits, according to
interviewed respondents, was that many elder SNAP participants have legitimate reasons for
not using their EBT cards for relatively long periods: some have smaller monthly benefits that
they purposely allow to grow for several months; others have mobility issues that make it
difficult to shop more frequently. One interviewed respondent also said that some elder
individuals may be unsure how to use EBT or forget their personal identification number, both
of which could lead to nonuse of the benefit or dropping out of the program. This latter type of
EBT challenge was reported in a focus group of elder SNAP participants in Washington State
(Gabor et al., 2002).
Change in household status. Interviewed respondents said that various changes in an elder
individual’s household status can lead to challenges staying on SNAP. For example, one
respondent noted that she has seen new widows who were not previously in charge of
household finances get confused about reporting and other requirements and eventually churn
off of SNAP as a result. Several respondents indicated that elder individuals who agree to take
care of their grandchildren may also be faced with new and confusing reporting requirements.
Barriers to Securing the Correct Benefit Amount
Interviewed respondents also highlighted barriers that make it more difficult for elder
applicants to be certified for the monthly allotment for which they are eligible. While not
directly associated with elder SNAP participation, barriers to larger benefit sizes may have a
secondary effect on participation, given the reports that some elder individuals do not apply for
SNAP because they do not think their allotment will be worth the effort involved. Interviewed
respondents explained that the documentation required to procure a deduction, including
shelter, medical and dependent care deductions, is often difficult for elder individuals to obtain.
Specifically, HIPPA regulations can make it challenging for Community Based Organizations
(CBOs) or State agencies to assist elder individuals with collecting the medical documentation
and receipts needed for an excess medical deduction. Some respondents also implied that
12
eligibility workers may sometimes not promote medical deductions or assess all elder
individuals for them due to the complexity of the process. In addition, respondents noted that
elder individuals who apply to SNAP through the Combined Application Project (CAP) or Elderly
Simplified Application Project (ESAP) may not have the opportunity to provide the extra
documentation that would make them eligible for deductions, even if such an opportunity is
technically supposed to occur, since these programs sometimes rely on automatic eligibility
determinations based on other data sources. One interviewed individual from a key national
organization said that he does not think most elder participants are aware of the opportunity to
use medical deductions.
How Barriers to Access are Experienced by Different Elder Subgroups
SPR’s research reveals that different subgroups within the broader elder population may have
an especially difficult time accessing or remaining on SNAP.
Rural elder individuals. In Fiscal Year 2015, 16 percent of elder SNAP participants were
considered to be living in a micropolitan or rural area, down from 20 percent in Fiscal
Year 2010 (SNAP QC Data). Interviewed respondents said that they thought rural elder
individuals were less likely to apply for and remain on SNAP than their urban peers. They
noted that there may be more of a philosophy of personal independence in rural areas
that makes stigma more prevalent. The possibility of stigma may also be increased by
the greater likelihood of rural individuals knowing the local SNAP eligibility workers or
grocery store employees. Respondents also said that transportation is often more
difficult in rural areas and there may be poorer broadband internet access and fewer
CBOs offering assistance.
Immigrant elder individuals. While certain legal immigrants qualify for SNAP,
respondents indicated that there is a lot of fear and confusion surrounding their
eligibility. Many worry that an application would have an impact on their legal status,
affect their household’s ability to naturalize, or even result in deportation. Rules about
whether and when an immigrant’s sponsor’s income needs to be counted for eligibility
may limit applications from these individuals, since they would require the potentially
sensitive step of an immigrant asking their sponsor for income information. A report
focused on California specifically noted that the State’s high number of immigrants
could be one reason for its low overall SNAP participation rate (Danielson & Klerman,
2011).
Elder individuals not fluent in English. SNAP material is technically required to be
translated into other languages, and some States have made doing so a priority. One
Regional SNAP Director explained that a State in his region provides its application in 13
languages. However, another respondent noted that materials other than the
application form (for example, notices) are often not translated into other languages.
According to both the literature and interviewed respondents, working with interpreters
13 | Elder SNAP Access
can be challenging. They may be difficult to access (Gabor et al., 2002), and some elder
individuals may not trust their translations.
“Older” elder individuals. Across Fiscal Years 2010 to 2015, about 16 percent of elder
SNAP participants were 80 or older (SNAP QC Data). Several respondents noted that
these “older” elder individuals may be less comfortable with using the internet or
automated phone systems to apply for SNAP. The literature also indicates that because
disabilities and cognitive impairment become more prevalent as individuals age, the
oldest members of the elder cohort may have greater challenges applying for SNAP and
maintaining their benefits than their younger peers (Herd, 2015; Cody & Ohls, 2005).
One study found that elder individuals in the age range of 65 to 74 used food stamps at
a rate nearly three times that of those 85 and older (Fuller-Thomson & Redmond, 2008).
However, two factors should be considered when interpreting statistics about the
participation of these “older” elders: the proportion of elderly who are older than 80 is
smaller than the proportion who are 60–80, and “older” elders are more likely to live in
community housing situations where they may not be eligible for SNAP.
Elder individuals who do not have internet access. Because most States (46) now have
SNAP applications online and some recommend that they be submitted this way,
individuals who do not have internet access (or who have very slow internet access)
may be at a disadvantage. Because broadband internet access is often less available in
rural areas, elder individuals who live in such areas may be especially likely to have slow
internet access.
Interventions to Increase Elder SNAP Access
Across all States, elder individuals are not subject to the same eligibility criteria as younger
SNAP applicants. For elder applicants, the following more generous criteria apply:
Households must only meet the net income test, rather than both the net and gross
tests.
Households can deduct elder medical expenses in excess of $35 a month when
calculating net income.
Households may have a higher amount of countable resources ($3,250, versus $2,250
for households that do not have elder or disabled members).
These different criteria recognize elder households’ specific needs. However, as the relatively
low rate of elder participation indicates, having a separate set of more generous eligibility rules
for elders has not been sufficient to bring the elder participation rate close to that of the
general population.
Over the years, FNS has developed a number of additional policy interventions that attempt to
increase the elder SNAP participation rate by directly addressing some of the barriers detailed
14
above. These include a series of demonstration projects, waivers, and policy options that States
may implement. Because demonstrations and waivers waive requirements of the Food and
Nutrition Act or existing SNAP regulations, they require approval from FNS and additional State
reporting responsibilities. In addition, demonstrations that impact household benefits must also
be deemed cost-neutral. Besides waivers and demonstrations, States have the flexibility to
adopt other policy options that do not require prior approval from FNS.
This section of the memorandum focuses specifically on a series of these interventions
identified by the FNS National Office to be of special interest for improving elder SNAP access. It
divides the interventions into those that may increase elder SNAP applications and enrollment,
those that increase benefit amounts, and those that may reduce churn and help elder
individuals stay enrolled. For each, the memorandum briefly describes the intervention,
summarizes any evidence from the literature about its effectiveness, and provides information
about how it is perceived by interviewed respondents. So far, there have been few impact
studies focused directly on these interventions, though the Elderly Nutrition Demonstration
(Cody & Ohls, 2005), conducted from 2002 to 2004, the Fiscal Year 2009 Pilots (Kauff et al.,
2014), and the Evaluation of the Effectiveness of Pilot Projects in Increasing SNAP Participation
among Medicare’s Extra Help Population in 2010 (Sama-Miller et al., 2014) examined some
similar models. In their group interview, FNS National Office staff members said that they are
unsure which current interventions have the most effect, in part because many States offer
multiple interventions and it can be difficult to disentangle their impacts. Despite this, the
interviewed respondents reported thoughts about the benefits, challenges, and possible
unintended consequences associated with each intervention. This information will help the
research team decide which interventions are especially important to include and know their
associated implementation challenges.
Other SNAP policy options and waivers not directly targeted to elder individuals may also have
an impact on elder access to the program. Among these, those identified by interview
respondents as especially likely to affect elder access, such as telephone interviews, online
applications, and call centers, are covered in this memo. Additional SNAP options, such as
Broad Based Categorical Eligibility (BBCE), which enables categorical eligibility for households
who are eligible for TANF non-cash benefits (U.S. Department of Agriculture, 2016), and vehicle
exemption in the calculation of household assets (U.S. Department of Agriculture, 2016), may
also influence elder access. However, these interventions are not a focus of this memo since
they have already been covered by prior research or are likely only applicable to a restricted
range of elder individuals.
Appendix B, the State intervention index, supplements this section of the memorandum. It
provides additional details about which States have each intervention and when they were first
implemented. It also provides a total count of interventions for each State, divided into those
15 | Elder SNAP Access
that target elder access specifically and those that aim to improve the application process for all
individuals. Some States offer only two total interventions, while others have as many as seven.
The number and type of intervention offered by each State will be a key consideration for study
State selection.
Policy Context for Interventions
In interviews with the FNS Regional Offices and the FNS National Office, respondents noted that
while they can and do encourage States to consider certain interventions, the choice to take
one on is ultimately up to the individual State. Respondents highlighted the following factors as
likely having an influence on a State’s uptake of an intervention:
The political climate and State priorities. Respondents explained that a State’s political
climate may affect whether the SNAP program prioritizes quality control or increasing
program access. Certain States are not interested in expanding their SNAP programs.
State budgets play a role in shaping priorities, as does the fact that gubernatorial or
commissioner approval may be required to take on a new intervention.
Management Information System (MIS) and data analysis capacity. Because States
that decide to implement a demonstration or waiver are required to conduct additional
reporting for FNS, some decide they do not have the capacity to take one on. This factor
is especially significant for States with older IT systems (legacy systems) or those that
are in the process of making a system transition.
Regional influence. Respondents said that when one State in a region has success with
an intervention, it sometimes makes others more likely to try it. Some FNS Regional
Office staff members indicated that sharing success stories to encourage emulation is
part of their role.
The advocate community. Several respondents said that the presence of a strong elderfocused advocate community could help drive a State’s decision to implement an
intervention.
Cost neutrality requirement. Opinions were mixed about whether the cost neutrality
requirement influences States’ decisions to implement an intervention; at least some
respondents indicated that it does play a role. Respondents noted that it is always a
difficult situation when adding an intervention may make some elder recipients better
off while making other participants worse off. According to a respondent from the FNS
National Office, it can be challenging to explain to clients who have their benefit amount
slightly reduced why this occurred.
16
Interventions to Increase the Number of Elders Applying for and Enrolling in
SNAP
The following interventions were identified as possibly increasing the number of elder
individuals who apply for and become enrolled in SNAP. Most of these interventions make the
application process itself easier; the last one to be discussed involves targeting individuals who
might not otherwise know about SNAP or have applied. The first two interventions described
are most important for this study given that they target elder individuals (or elder and disabled
individuals) specifically, while the rest aim to improve the process for all individuals.
Elderly Simplified Application Project (ESAP)
Currently, eight States have an Elderly Simplified Application Project (ESAP). Overall,
interviewed respondents believe this demonstration project, which was first implemented in
2009, has the most promise for increasing elder SNAP access. It combines several strategies
that may both increase the ease of applying for SNAP and reduce churn. The ESAP includes a
streamlined SNAP application form (for example, in Alabama the ESAP form is three pages), the
option to use data matching to verify applicant information, a waiver of the recertification
interview, and a 36-month certification period (U.S. Department of Agriculture, 2016). In the
past, it also involved a waiver of the certification interview, but the FNS National Office has
reinstated this requirement as States apply for the waiver for the first time or get approval to
reinstate it. One interviewed respondent from a national organization focused on SNAP noted
that the ESAP “removes [the] heavy burden [of applying] from applicants.”
In addition to implementing the ESAP package as defined above, some ESAP States have taken
additional steps within the ESAP framework to simplify the SNAP application process for elder
individuals. For example, Alabama and South Carolina have units within their State agencies
that work only with elder applicants. Alabama also has a waiver of interim reporting for ESAP
participants, which FNS National Office staff members indicated that they may add to the ESAP
model going forward. Other States that do not have the full ESAP have implemented some
components of it. For example, several States, including Massachusetts and Minnesota, have a
simplified application form for elder applicants and others have 36-month certification or
recertification interview waivers.
While there has been no research that evaluates the effectiveness of the ESAP in its current
form, some information that points to its potential is available. There have been two
evaluations that examined the impacts of similar simplified application strategies in several
States. The Evaluation of the USDA Elderly Nutrition Demonstrations included an evaluation of
a simplified application model used in two counties in Florida. In this model, the SNAP
application form was only one page and applicants did not have to submit documentation of
income and expenses or do a certification interview. Researchers found that it increased elder
17 | Elder SNAP Access
SNAP participation by more than 20 percent in 21 months compared to similar counties without
this simplified process (Cody & Ohls, 2005). Another evaluation that assessed demonstration
projects that targeted elder individuals and used a simplified application process in Michigan
and Pennsylvania also found statistically significant increases in SNAP participation (Kauff et al.,
2014). While not identical to the ESAP, these simplifications reduced required contact with the
State agency. However, the demonstrations also involved additional components (such as
application assistance from CBOs in Michigan and a waiver of the initial certification interview
in Pennsylvania) that make a one-to-one link between the positive findings from these
demonstrations and the current ESAP model impossible. Nevertheless, these positive impacts
for similar interventions indicate that the ESAP has promise and is worthy of further study.
While they were very positive about this intervention overall, interviewed individuals from key
national organizations also expressed some concern about the reinstatement of the
certification interview for the ESAP. FNS National Office staff members explained that its
reinstatement is intended to give elder applicants one-on-one contact with the program and a
chance to get questions answered; some individuals from national organizations fear the
additional step will keep some applicants from successfully applying. A prior study that
evaluated Oregon’s and Utah’s waiver of both the certification and recertification interviews
(not just for elder applicants) found mixed results, with no effect on the number of applications
approved and a decrease in application approval times (Rowe et al., 2015). Interviewed
respondents were interested in examining the merits of the certification interview more closely
for elder applicants specifically, though, with one interviewed respondent recommending that
FNS allow at least one ESAP State to keep the certification interview waiver through this study
so that it could be part of the evaluation. In addition, several respondents recommended
including Alabama in the study because it once had a waiver of the certification interview and
no longer does.
As for the issue of unintended consequences, the FNS National Office said they worried that
without a certification interview, some ESAP applicants were not given an opportunity to
explore medical deductions that might ultimately increase their benefit amount. This was also a
finding in the evaluation of the certification and recertification wavier (Rowe et al., 2015) and
part of the reason why FNS has reinstated this interview.
Combined Application Project (CAP)
The Combined Application Project (CAP) is a demonstration implemented for the first time in
1995 that is currently implemented in 17 States. The intervention involves a partnership
between the State SNAP agency and the Social Security Administration (SSA); its purpose is to
connect those receiving Supplemental Security Income (SSI) to SNAP (U.S. Department of
Agriculture, 2016; Rowe et al., 2010). By Federal law, all SSI applicants are required to have the
18
chance to apply for SNAP while applying for SSI. However, without CAP, this joint application
process does not always happen or work smoothly (Dorn, Minton, & Huber, 2014).
There are two types of CAPs—a “standard CAP,” which involves a simplified joint application for
both SSI and SNAP that is processed by the SSA and waives a certification interview with the
SNAP administering agency, and a “modified CAP,” wherein SSA participation data is used to
target potential SNAP participants (U.S. Department of Agriculture, 2016). In the latter case, the
SSA does not process or submit the SNAP application. Both the standard and modified CAP use
either standard benefit or standard shelter deduction amounts (U.S. Department of Agriculture,
2016) to make the application process easier. Many interviewed respondents were supportive
of CAP in theory, but they identified implementation challenges, such maintaining a
relationship with the SSA, that they believe have weakened the effects of the intervention. In
part because of these challenges, which are described in more detail below, FNS is not currently
approving new CAP demonstrations.
Because CAP has been an option for over two decades, prior research has examined its role in
the SNAP application process. Initial findings were promising. A survey of the States that had a
CAP in 2008 had positive findings, with all States and 65 percent of local offices believing that
CAP increased participation among elder individuals (Rowe et al., 2010). Another study
reported that from 2000 to 2008, “CAP states experienced a 48% increase in SNAP participation
levels among 1-person SSI households, at a time when such households’ enrollment in other
states saw little change” (Dorn, Minton, & Huber 2014, p. ii). However, another report from
2015 was more mixed about CAP’s influence over time. Based on participation data, this report
states that the percentage of elder SNAP participants who were eligible for SNAP through CAP
fell from 14 percent in 2009 to eight percent in 2013 (Eslami, 2015), but notes that this
reduction may have been a result of changes in SSI eligibility. The same report also reveals
significant variation across States. Five CAP States enrolled less than 10 percent of elder SNAP
participants through CAP, while five others enrolled over 20 percent through the
demonstration (Eslami, 2015).
Interviewed respondents from both FNS Regional Offices and key national organizations
identified challenges working with the SSA as a reason why CAP varies by State and does not
always work as intended. Respondents reported that the relationship between the State SNAP
agency and SSA can be very different in different States, even within the same region. One
Regional SNAP Director explained that one State in his region has a modified CAP that seems to
work well, with SSA providing the needed data to target SSI recipients for SNAP. In contrast, he
noted that another State in his region is struggling with a standard CAP, and that the SSA there
had helped fewer people apply to SNAP this year than it had in years prior. One individual from
a national organization focused on SNAP surmised that SSI administering offices in some States
might be dealing with their own technology changes or modernization processes, which could
19 | Elder SNAP Access
affect their ability to be strong partners. During the group interview with the National FNS
Office, one respondent said she had heard that some State SNAP agencies had not
communicated with the State agency administering SSI in over a decade.
Another potential drawback to CAP noted by the literature and identified in interviews is that it
can make it harder for applicants with high medical or shelter costs to get connected to a
benefit amount that would take into account all of the deductions for which they are eligible.
This problem occurs because the standard CAP model uses standardized benefit amounts in
order to meet the cost neutrality requirement. Applicants are supposed to be notified that they
can contact the SNAP administering agency to follow up about potential deductions that could
increase their benefit, but this may not always happen, or participants may not understand the
notices (Dorn, Minton, & Huber, 2014). Several interviewed respondents agreed with this
assessment, with one individual from a national organization focused on SNAP saying she does
not think CAP does a “robust” enough job looking for deductions. Nevertheless, another
individual from a different national organization focused on SNAP added that even if an elder
individual who applied to SNAP through CAP is not getting all the deductions for which he or
she is eligible, at least this person is receiving the benefit.
Community Partnership Interview Demonstrations
Currently, four States have a Community Partnership Interview Demonstration, and this
intervention does not specifically target elder individuals. Nevertheless, the demonstration,
which enables local CBOs to help individuals with the SNAP application process and conduct
certification interviews, may be especially helpful for those 60 and over. Respondents noted
that many elders express a preference for one-on-one, in-person assistance, which this
intervention provides. Because CBOs may be more trusted than government agencies in certain
communities and may have more culturally competent staff members, respondents also
suggested that this intervention could be especially effective at reaching elders in immigrant
communities or those who speak languages other than English.
Three studies provide mixed evidence about the effectiveness of community partner
interviews. One examined six Community Partner Interviewer demonstrations from 2009-2010,
including one that focused specifically on reaching the elder population. In two States,
households that were interviewed by a CBO were more likely than households that were
interviewed for SNAP through the regular channels to include elder individuals (Wilson et al.,
2015). In a survey, applicants who applied through a CBO were also more likely to be satisfied
with the length of time it took to get their SNAP benefit and with the customer service
experience (Wilson et al., 2015). However, the study also found that in two States, denials for
SNAP applications for procedural reasons were higher for applications from CBOs as opposed to
those done through the regular channels, and in one State, the length of time it took to receive
20
the benefit was longer for applications from CBOs than through the regular channels (Wilson et
al., 2015). Another study examined the impact of application assistance strategies (offered by
both CBOs and individuals hired specifically for this purpose by State SNAP agencies) provided
in three States. While the models varied by State and county (and were not identical to the
Community Partnership Interview Demonstration), there were some positive findings. In two
counties in two different States, the application assistance led to greater than a 30-percent
positive impact on elder SNAP participation (Cody & Ohls, 2005). However, in one county within
one of these same States, the impact on elder SNAP participation was slightly negative, and in
another State it was only slightly positive (Cody & Ohls, 2005). The researchers involved in the
study identified implementation challenges as a possible reason for these differences. A third
study evaluated the Fiscal Year 2009 Pilots, one of which targeted elder applicants with a
strategy similar to the Community Partnership Interview Demonstration. In several areas of
Michigan, CBOs helped elder applicants fill out applications over the phone, and this phone call
was then counted as the certification interview. Researchers found that elder participation in
SNAP increased more in the sites with these strategies than in the comparison sites that used
the regular SNAP application process (Kauff et al., 2014).
Our interviewed respondents had mixed thoughts about the value of involving CBOs in the
SNAP certification interview process. Several supported the idea, emphasizing the ability of
such organizations to reach out to new groups and explaining that state eligibility workers could
also make procedural errors like those made by some CBOs in the studies. However, others
thought that CBOs should “complement, not replace” State agency work. In general,
respondents thought that CBOs should play a role in outreach and application assistance even if
they did not take on the certification interviews themselves.
Telephone Interviews
There are several policies that States can use to allow SNAP applicants or recipients to conduct
certification or recertification interviews by phone instead of in person. As of 2001, FNS
declared that all applicants with “hardships” must be able to have a telephone or at-home
interview if they request it (Gabor et al., 2002). Since 2001, States have also applied for an
additional waiver that allows all interviews to be conducted over the phone (Rowe et al., 2015),
though households can still request in-person interviews if that is their preference. Currently,
all States except North Dakota and Connecticut have this waiver to offer telephone interviews
in lieu of face-to-face ones.
All respondents supported telephone interviews as an option, though some emphasized that it
is important to make clear that in-person interviews are also available. While advocates believe
that telephone interviews reduce the burden on applicants (Cody et al., 2010), some elder
individuals may prefer to talk in person or have trouble with hearing or understanding over the
21 | Elder SNAP Access
phone. In a focus group conducted in Washington State with elder individuals, most said they
would prefer a face-to-face interview (Gabor et al., 2002). The authors noted that participants
who spoke Spanish or Korean were more likely to feel they could better communicate in person
(Gabor et al., 2002). One individual from a national organization focused on SNAP identified
several best practices in phone interviews: allowing general scheduling flexibility (including
letting applicants reschedule interviews as needed), ensuring that calls happen on time, and
allowing benefits to be approved without a phone interview if it is the only missing part of an
application. Despite some acknowledged challenges, multiple respondents emphasized how
helpful it is for elder individuals to be able to conduct interviews over the phone as opposed to
needing to go into an office.
Call Centers
Thirty-five States currently use call centers in some or all of their counties. Considered part of
the SNAP modernization process, call centers help local SNAP offices by answering questions
about the program, providing benefit information, conducting certification interviews, and
sometimes directly certifying and recertifying applicants (U.S. Department of Agriculture, 2016).
As an intervention, call centers can reduce wait time and make information about SNAP more
accessible. Respondents indicated that while the convenience of call centers can be a benefit to
elder individuals, the quality of call centers varies greatly across States. When they are not well
implemented, there can be long waits to reach someone and calls may be dropped (Cody et al.,
2010; Rowe et al., 2010). Both reports and respondents also noted that some elder individuals
prefer one-on-one, in-person communication with an eligibility worker dedicated to their case
rather than speaking to someone new each time they call the center (Cody et al., 2010). Several
respondents also wondered how well elder individuals are able to understand and navigate
automated phone prompts, which are a part of some call center systems.
Online Applications
Online applications are now very common, with 46 States providing them as an option. While
applicants have the option to submit an application in another format, some States encourage
online submittal. On the one hand, online applications can be faster to complete (Cody et al.,
2010) and can be beneficial for elder applicants with mobility issues because they can be done
at home. However, respondents also indicated that there is a wide range of technological savvy
and internet access within the elder category. Respondents explained that elder individuals,
especially those in rural areas, may not have a fast internet connection at home. Going to the
library or a CBO for internet access or help with the application can be challenging due to
transportation or mobility issues. One report noted that online systems may be especially
difficult for those with limited English language skills or disabilities (Cody et al., 2010). A
22
respondent added that drop-down menu items specifically, common in online applications, can
be confusing for those who have less internet experience.
Several reports have found that significant numbers of participants prefer traditional service
delivery methods to an online SNAP application process (Cody et al., 2010; Heflin, London, &
Mueser, 2010; Kauff et al., 2014). For example, some elder individuals prefer paper applications
because they do not feel comfortable submitting personal information electronically (Kauff et
al., 2014; Rowe et al., 2010). Interviewed respondents noted that older elder individuals likely
experience online applications differently than younger ones, and that resistance to online
methods of applying for SNAP may therefore lessen over time. On the positive side, online
applications can be programmed to prompt for medical expenses in a systematic way,
something that may not be done consistently by eligibility workers (Jones, 2014).
Benefits Data Trust and Data Matching
Five States currently work with the nonprofit organization Benefits Data Trust to conduct what
they call “proactive” SNAP outreach and application assistance. While these partnerships with
Benefits Data Trust are not a demonstration project or waiver and do not exclusively target
elder individuals, the data matching done by the organization may still play an important role in
elder SNAP access. Benefits Data Trust collects administrative data from States, then uses this
data with its own management information system, call center, and web-based telephone
system to conduct targeted SNAP outreach and provide application assistance (Kauff et al.,
2014). Staff members from the organization explained in an interview that they might, for
example, match a State’s Medicaid and SNAP records to find individuals who are only enrolled
in Medicaid, then send each of these individuals a SNAP application that is already partially
filled out with the information they have collected. Benefits Data Trust staff members then
follow up with each individual to help them complete the application by phone.
Benefits Data Trust work providing this type of outreach to elder individuals in Philadelphia was
evaluated as part of the Fiscal Year 2009 Pilots, and researchers found positive impacts. The
strategy that Benefits Data Trust used in this pilot also had some additional components, such
as a waiver that allowed elder individuals to self-declare medical expenses and another that let
Benefits Data Trust staff members conduct certification interviews, which went beyond the
current Benefits Data Trust strategy that focuses on targeted outreach through data matching.
Compared to a comparison county, significantly more elder individuals were participating in
SNAP in Philadelphia 17 months after the demonstration began (Kauff et al., 2014). However,
the evaluation did not find significant differences in the number of applications processed
(Kauff et al., 2014).
Similar data matching strategies were also conducted and evaluated as part of the three Fiscal
Year 2010 pilots aimed at increasing elder SNAP participation in New Mexico, Pennsylvania, and
23 | Elder SNAP Access
Washington. Each targeted SNAP application outreach and assistance to individuals who were
also applying for public assistance for medical costs, such as through the Medicare Low Income
Subsidy (LIS). The States (or, in Pennsylvania’s case, Benefits Data Trust) linked SNAP caseload
data with medical assistance program data to find elder individuals likely to be eligible for
SNAP. These individuals were then mailed simplified applications or information about SNAP. In
one State, part of the application was already filled out based on preexisting data from the
medical program. Using a quasi-experimental design, researchers found that the pilots had
positive effects on applications in all three States, though the overall number of individuals who
applied was limited in some areas (Sama-Miller et al., 2014).
Overall, interviewed respondents viewed the work done by Benefits Data Trust positively, and
judged data matching in general to be very promising. Interviewed individuals from key national
organizations noted that such a process removes much of the application burden from the
potential participant and makes use of information the State already has. This type of strategy
is growing, with Benefits Data Trust planning work with several additional States in the future.
Interventions to Increase Elder SNAP Benefit Amounts
As noted above, low monthly benefit allotments were identified by the literature and nearly
every respondent as a major barrier to elder SNAP access (Bartlett & Burstein, 2004; Bartlett et
al., 2004; Cody & Ohls, 2005; Gabor et al., 2002; Kim & Frongillo, 2009; McConnell & Nixon,
1996; McConnell & Ponza, 1996). Given this finding, it is reasonable to suppose that raising
benefit amounts directly or indirectly might be interventions that help expand SNAP
participation.
Increased Benefit Amount
Several States have simply increased the minimum benefit amount. According to an
interviewed individual from a national organization focused on SNAP, in 2015 both D.C. and
Maryland chose to use State or local funds to provide a higher monthly SNAP allotment for
those technically eligible for only $16 a month, and New Mexico has also experimented with a
similar increase. In Maryland, this increase is only for those in the elder category. While there
has not yet been time to assess the outcomes of these benefit increases, other studies have
shown that increased SNAP allotments (for example, during the American Recovery and
Reinvestment Act) generally raise participation (Besharov, 2016; Coe & Wu, 2014). For the most
part, interviewed respondents were very positive about the idea of raising the minimum SNAP
allotment for elder individuals, though they also noted that most qualify for more than this
amount anyway. One individual from a national organization focused on SNAP explained that
higher benefit amounts lead to increased access to healthy food, which is correlated with
increased health and frees up other resources for the participant to spend on medical expenses
and other needs.
24
Standard Medical Deduction (SMD)
The Standard Medical Deduction (SMD), a demonstration currently active in 19 States, can have
the effect of indirectly increasing the benefit amount by lowering net income. Under regular
SNAP rules, all elder SNAP recipients are entitled to deduct out-of-pocket medical expenses
greater than $35 per month from their income. Claiming a medical expense can add significant
amounts to a SNAP allotment, with one report indicating that claiming $50–$200 in medical
expenses can lead to a SNAP allotment that is $7–$69 higher each month (Jones, 2014).
Expenses can include costs for transportation to medical appointments and over-the-counter
medication as well as insurance copays (Jones, 2014).
The SMD makes this process simpler by setting a standard medical deduction amount for those
applicants who can prove their medical expenses are above $35 a month but might struggle to
document the exact amount (U.S. Department of Agriculture, 2016). Applicants still have the
option of a medical expense deduction equal to their actual expenses if they are higher than
the SMD threshold (U.S. Department of Agriculture, 2016). Medical deductions in general, and
the SMD specifically, are considered underused by elders (Jones, 2014). Unlike the other
interventions profiled in this memorandum, the SMD may not directly increase overall elder
participation rates. However, by potentially increasing benefit allotments, it may draw more
elder individuals to the program.
Overall, elder participant uptake of the excess medical deduction has risen only slightly since
the SMD became a State waiver option. The average size of the excess medical deduction has
also fallen when inflation is taken into account. In Fiscal Year 2007 (prior to the implementation
of the SMD in any State), about 13 percent of elder SNAP households received an excess
medical deduction with an average monthly size of $163 in 2007 dollars (SNAP QC Data). In
Fiscal Year 2015, about 16 percent of elder SNAP households received an excess medical
deduction, with an average monthly size of $170 in 2015 dollars (SNAP QC Data), the equivalent
of $149 2007 dollars6. Compared to 2007, there were more elder individuals on SNAP in 2015,
and among them, a greater proportion received a medical deduction. If the additional program
participants have lower medical expenses, this could be one reason why the average medical
deduction shrunk. Wide variations in these numbers by State (Eslami, 2015) may also reflect
SMD implementation challenges.
The SMD is less complicated to administer than the Excess Medical Deduction for both SNAP
participants and eligibility workers because only $35.01 worth of medical expenses needs to be
verified. In fact, one Regional SNAP Director indicated that a State in his region decided to
implement the SMD not because it particularly wanted to increase elder SNAP participants’
6
Inflation adjustment made using the United States Department of Labor Consumer Price Index Inflation
Calculator https://www.bls.gov/data/inflation_calculator.htm
25 | Elder SNAP Access
benefit amounts, but because the eligibility worker error rate in calculating the Excess Medical
Deduction was too high. A report focused on Washington State confirmed the complexity of the
Excess Medical Deduction for both caseworkers and elder participants there (Gabor et al.,
2002), and a survey of States from 2008 found that many indicated a desire for a SMD due to
the complexity of the alternative (Rowe et al., 2010).
Despite being a simplification of the Excess Medical Deduction, respondents indicated that the
SMD could still be challenging for elder participants and eligibility workers. Several respondents
explained that verifying even $35.01 of medical expenses can be difficult, especially when
HIPPA laws prevent eligibility workers from helping an applicant collect the necessary receipts.
In addition, according to interviewed respondents, some elder individuals who first access SNAP
through an alternative method, like the ESAP or CAP, may not be given details about the
availability of any kind of medical deduction.
Interviewed respondents also noted that the cost neutrality requirement associated with the
SMD can be a barrier to its adoption. Because elder individuals may receive a higher medical
deduction with the SMD than they would have without it, overall SNAP costs can go up.
According to respondents, many States deal with this dilemma and maintain cost neutrality by
lowering the Standard Utility Allowance (SUA) slightly. Several regions noted that, in a sense,
States are taking money from one group to give more to another, which can be politically
sensitive. At least two respondents recommended that the SMD become standard as opposed
to a demonstration, because this would remove the cost neutrality barrier and might make it
more administratively feasible for States.
Overall, interviewed respondents had mixed feelings about the SMD in its current form. One
individual from a national organization focused on SNAP said that the data they have collected
leads them to believe that use of the SMD causes overall medical deduction amounts to
decrease slightly but increases overall take-up of the medical deduction. Because of this effect,
they think that it is beneficial to the elder population on a broad scale. Another approved of the
intervention on principle, but feels that States often do not do the training and outreach
necessary for it to work well. Multiple respondents asked that the SMD and its effect on elder
SNAP access be included in the study.
Interventions to Help Elders Remain on SNAP and Reduce Churn
In addition to the two categories of interventions discussed above, FNS has created
interventions that attempt to make it easier for elder individuals to remain on SNAP. These
interventions address “churn” by minimizing the effort that elder participants must make once
on SNAP.
26
Elderly and Disabled Recertification Interview Waiver
Eight States have an Elderly and Disabled Recertification Interview Waiver that enables
households made up entirely of elder and disabled participants with no earned income to
bypass the recertification interview. This waiver was first established in 2009. For the most part,
interviewed respondents were very supportive of this intervention and any efforts to reduce
the contact participants must make with the State SNAP agency after enrolling. Several
respondents said that States should be able to get much of the information they would get in
recertification interviews through the data they already have. Note that this waiver is also one
component of the broader ESAP.
While the Elderly and Disabled Recertification Interview Waiver has not been evaluated as a
unique intervention, researchers did evaluate a 2011 waiver that Oregon and Utah received to
waive both the certification and recertification interviews. This waiver did not focus on elder
participants, but such individuals were included in the study. While overall findings were mixed,
especially concerning the certification interview waiver, researchers did determine that the
recertification interview waiver reduced churn in Utah (Rowe et al., 2015).
36-Month Certification Demonstration
While only two States currently operate 36-Month Certification Demonstrations, which were
established in 2010 and 2011, longer certification periods for elder individuals were supported
by respondents as one of the key ways to reduce churn. The ESAP incorporates a 36-month
certification, so the eight States with that demonstration effectively implement this policy as
well. While they do not look specifically at 36-month certifications, several studies have found
that extending certification periods leads to increases in SNAP participation (Besharov, 2016;
Peterson et al., 2014; Ratcliffe, McKernan, & Finegold, 2007; Rutledge & Wu, 2013). Multiple
interviewed Regional SNAP Directors felt that extended certification periods had lessened elder
churn in their areas. Some also supported simplified recertification processes generally, with
one saying, “if you don’t have to take an action to recertify, then there would be no reason to
churn.” One concern with longer certifications is that a participant might not have their benefit
appropriately adjusted should their income or household situation change. However,
interviewed individuals from key national organizations also added that even with extended
certification, there are still annual reporting requirements, so elder participants would have the
opportunity to provide the information needed for a new deduction, for example, should their
medical expenses increase.
Simplified Reporting
SNAP households must submit a periodic report at least once every six months (U.S.
Department of Agriculture, 2016). However, States may choose to adopt simplified reporting,
27 | Elder SNAP Access
which enables households made up exclusively of elder or disabled members with no earned
income to be certified for 12 months without a periodic report or certified for 24 months with a
12-month periodic report (U.S. Department of Agriculture, 2016). During their group interview,
respondents from the FNS National Office indicated that they believe elder individuals who
churn off of SNAP often do so at the time of the 12-month report. They are interested in
reducing or eliminating periodic reports for this group, possibly by replacing the reports with
data matches to verify needed information. According to an interviewed individual from a
national organization focused on SNAP, Alabama uses this technique already and does not
require interim reporting for those on the ESAP.
Overall Themes, Challenges, and Areas for Further Study
Overall, this exploratory research has revealed that while there are past studies that identify
impacts of some of the interventions created to increase elder SNAP access, little is known
about the current combination of interventions and how they work together. Future
components of this study will play an important role in clarifying what works best to increase
elder participation in SNAP.
Several major elder SNAP access themes emerged from the exploratory research. These themes
will be important to consider for the next steps of the project.
Overall Themes
According to research conducted by FNS, the elder SNAP participation rate is
trending up, though overall it is still about half that of the national participation
rate.
These data also show that there remains wide variation in elder SNAP
participation rates across States. Because each State has a different political
context, elder population, and combination of demonstrations, waivers, and other
SNAP policy options, this variation is not unexpected. The research team will
attempt to exploit these differences in State participation rates and intervention
use to determine the impact of the various SNAP policy options.
According to respondents, the most promising interventions—and therefore the
ones they are most interested in learning more about—are the ESAP, SMD,
longer recertification periods, and the “proactive” data matching outreach and
application assistance conducted by Benefits Data Trust. In addition, several
respondents would like to learn more about how the various interventions
interact.
28
There is a delicate balance between simplifying the elder SNAP application
process and ensuring that applicants are fully familiar with the program and the
deductions for which they may be eligible. On the one hand, most respondents
indicated that long application forms, required certification and recertification
interviews, and complex reporting processes could be very burdensome for elder
applicants and create major barriers for stable enrollment in SNAP. On the other
hand, some respondents explained that they did not want the application process
to become so streamlined that elder applicants never get clear and detailed
information about the program, particularly about excess medical deductions,
what the recertification process entails, and how to use EBT. Currently,
respondents appear to be thinking through this challenge particularly as it relates
to the certification interview. Some intermediaries felt that requiring such an
interview for elder SNAP recipients prevents some elder individuals from ever
enrolling in SNAP, while others felt that waiving it removes an important point of
introduction to the program. In addition, some respondents thought that the
interview could be useful in theory, but in practice it often was too rushed to
provide applicants with valuable information. Because Alabama’s ESAP previously
waived the certification interview and now does not, data from that State may be
a way to evaluate the effects of this interview.
Respondents felt that the program must weigh the benefits of enrolling
applicants at all with the costs of applicants not receiving all possible deductions
for which they are eligible. When an applicant enrolls in SNAP through certain
simplified avenues, such as CAP, they may not receive information about medical
and other deductions, even if this information is supposed to be provided.
However, some of these individuals would not otherwise be on SNAP at all, and so
their enrollment with a slightly lower monthly benefit allotment might still be
considered a success. One respondent suggested that in future components of this
study, elder applicants themselves should be asked what they thought about the
idea of trading ease of enrollment for a slight reduction in their benefit.
Respondents believe that SNAP modernization is likely difficult for some elder
applicants even though it can bring improvements to the SNAP application and
enrollment process. In theory, modernization practices such as call centers and
transaction models can speed up the application process and minimize the effect
of applicants’ mobility issues. In practice, however, some elder individuals prefer
the one-on-one, in-person attention that is no longer common. Respondents also
reported that some States have struggled with the implementation of these new
practices to the point where long wait times and dropped calls become common.
Respondents felt that the oldest elder applicants might be most affected by some
aspects of modernization, because they may be less comfortable with automated
phone prompts, online applications, and other technologies included in this
model.
29 | Elder SNAP Access
Many respondents were supportive of using data matching technology to simplify
the application and recertification processes (examples are using Medicaid data to
target potential SNAP applicants and using SSI records to update income or address
information without burdening the elder participant).
Overall, respondents emphasized that there is sometimes a disconnect between
the ideal implementation of an intervention versus what actually happens on the
ground. For example, respondents indicated that States do not always translate
recertification notices, eligibility workers do not always explain medical deductions,
and program material does not always inform applicants that they can choose to
do an in-person interview if they prefer. While Appendix B, the State intervention
index, provides details about which States have taken on which demonstrations,
waivers, and other options, it will be important to nuance this data with the
information collected about States’ implementations of these interventions during
the Study of State Interventions. Understanding a State’s local context, such as its
political climate, will also be a necessary part of this data collection.
Next Steps for the Study
To determine which States to include in the subsequent components of the study, the research
team is currently weighing various relevant criteria. These criteria provide the opportunity to
compare States across a number of important areas identified in this memo, including the 2014
elder SNAP participation rate, the growth in elder SNAP participation rate since 2010, the
number and type(s) of interventions to address elder access, the type of SNAP administration
(county or State), and the State region. Additional criteria, such as State data quality, will also
be considered. The State Selection Memo (provided to FNS on February 1, 2017) goes over the
State selection process in more detail. The proposal for States is listed below. These States have
a diverse array of elder SNAP-access experiences, including States with both high and low elder
SNAP participation rates, States from every region, States with SNAP programs that are
administered by the county and the State, and States with a variety of interventions.
List of Proposed States for Inclusion in the Study
The research team proposes including the following 10 States in the study:
Alabama
Arkansas
Florida
Idaho
Massachusetts
Minnesota
Nebraska
New York
Pennsylvania
Washington
The research team proposes the following as alternate States:
Arizona
Colorado
Maryland
North Dakota
South Carolina
Texas
30
References
AbuSabha, R., Shackman, G., Bonk, B., & Samuels, S. (2011). Food security and senior
participation in the Commodity Supplemental Food Program. Journal of Hunger &
Environmental Nutrition, 6(1), 1–9. Retrieved from
https://doi.org/10.1080/19320248.2011.549358
Bartlett, S., & Burstein, N. (2004). Food Stamp Program access study: Eligible nonparticipants.
Abt Associates Inc. Retrieved from http://abtassociates.com/reports/efan03013-2.pdf
Bartlett, S., Burstein, N., Hamilton, W., Kling, R., & Andrews, M. (2004). Food Stamp Program
access study. U.S. Department of Agriculture, Economic Research Service. Retrieved
from https://www.ers.usda.gov/webdocs/publications/efan030133/30283_efan030133_002.pdf?v=41515
Besharov, D. J. (2016). Reconceiving SNAP: Is nutritional assistance really income support?
Research on Social Work Practice, 26(3), 316–327. Retrieved from
http://journals.sagepub.com/doi/pdf/10.1177/1049731515589373
Blavin, F., Dorn, S., & Dev, J. (2014). Using behavioral economics to inform the integration of
human services and health programs under the Affordable Care Act. Urban Institute.
Retrieved from http://tpcprod.urban.org/UploadedPDF/413230-Using-BehavioralEconomics-to-Inform-the-Integration-of-Human-Services-and-Health-Programs-underthe-ACA.pdf
Center on Budget and Policy Priorities. (2016). Policy basics: Introduction to SNAP. Retrieved
from http://www.cbpp.org/research/policy-basics-introduction-to-the-supplementalnutrition-assistance-program-snap
Cody, S., & Ohls, J. (2005). Evaluation of the USDA elderly nutrition demonstrations: Volume I,
Evaluation findings. Mathematica Policy Research. Retrieved from
http://www.mathematica-mpr.com/~/media/publications/PDFs/reachingoutI.pdf
Cody, S., Reed, D., Basson, D., Pedraza, J., Sama Martin, E., Santos, B., & Smith, E. (2010).
Simplification of health and social services enrollment and eligibility: Lessons for
California from interviews in four states. Washington, DC: Mathematica Policy Research.
Coe, N. B., & Wu, A. Y. (2014). What impact does Social Security have on the use of public
assistance programs among the elderly? (CRR WP 2014-5). Center for Retirement
Research at Boston College. Retrieved from
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2440759
Cunnyngham, K. (2010). Estimated effects on the Supplemental Nutrition Assistance Program of
eliminating California’s SSI cash-out policy: Final report. Washington, D.C.: Mathematica
Policy Research.
Cunnyngham, K, Castner, L., & Sukasih, A. (2012). Empirical Bayes shrinkage estimates of state
Supplemental Nutrition Assistance Program participation rates in 2007-2009 for all
eligible people and the working poor. Washington, D.C.: Mathematica Policy Research.
R-1 | Elder SNAP Access
Currie, J. (2004). The take up of social benefits. National Bureau of Economic Research.
Retrieved from http://www.nber.org/papers/w10488
Danielson, C., & Klerman, J. A. (2006). Why did the Food Stamp caseload decline (and rise?):
Effects of policies and the economy. University of Wisconsin-Madison, Institute for
Research on Poverty. Retrieved from
http://www.ssc.wisc.edu/irpweb/publications/dps/pdfs/dp131606.pdf
Danielson, C., & Klerman, J. A. (2011). California’s Food Stamp Program: Participation and cost
challenges for the state. Public Policy Institute of California. Retrieved from
http://www.ppic.org/content/pubs/report/R_911CDR.pdf
Dorn, S., Minton, S., & Huber, E. (2014). Examples of promising practices for integrating and
coordinating eligibility, enrollment and retention: Human services and health programs
under the Affordable Care Act. Urban Institute. Retrieved from
https://aspe.hhs.gov/sites/default/files/pdf/77016/rpt_PromissingPractices.pdf
Eslami, E. (2015). State trends in Supplemental Nutrition Assistance Program eligibility and
participation among elderly individuals, fiscal year 2008 to fiscal year 2013. Washington,
D.C.: Mathematica Policy Research.
Eslami, E., Leftin, J. & Strayer, M. (2012). Supplemental Nutrition Assistance Program
participation rates: Fiscal year 2010. U.S. Department of Agriculture. Retrieved from
https://www.fns.usda.gov/sites/default/files/Trends2010.pdf
Fitzpatrick, K., Greenhalgh-Stanley, N., & Ver Ploeg, M. (2015). The impact of food deserts on
food insufficiency and SNAP participation among the elderly. American Journal of
Agricultural Economics, 98(1), 19–40. Retrieved from
https://doi.org/10.1093/ajae/aav044
Fuller-Thomson, E., & Redmond, M. (2008). Falling through the social safety net: Food stamp
use and nonuse among older impoverished Americans. The Gerontologist, 48(2), 235–
244. Retrieved from https://doi.org/10.1093/geront/48.2.235
Gabor, V., Williams, S. S., Bellamy, H., & Hardison, B. L. (2002). Seniors’ views of the food stamp
program and ways to improve participation–Focus group findings in Washington State.
U.S. Department of Agriculture, Economic Research Service. Retrieved from
http://162.79.45.195/media/1772673/efan02012.pdf
Ganong, P., & Liebman, J. B. (2013). Explaining trends in SNAP enrollment. Retrieved from
https://www.hks.harvard.edu/jeffreyliebman/ExplainingTrendsInSNAPEnrollment.pdf
Geiger, J. R., Wilks, S. E., & Livermore, M. M. (2014). Predicting SNAP participation in older
adults: Do age categorizations matter? Educational Gerontology, 40(12), 932–946.
Retrieved from https://doi.org/10.1080/03601277.2014.912837
Gray, K. F., & Cunnyngham, K. (2016). Trends in Supplemental Nutrition Assistance Program
Participation Rates: Fiscal Year 2010 to Fiscal Year 2014. Washington, D.C.: Mathematica
Policy Research.
R-2
Gundersen, C., & Ziliak, J. P. (2008). The age gradient in Food Stamp Program participation:
Does income volatility matter? In Jolliffee, D., & Ziliak, J. P. (Eds.) Income Volatility and
Food Assistance in the United States (pp. 171-214). Kalamazoo, MI: W.E. Upjohn
Institute. Retrieved from http://dx.doi.org/10.17848/9781435684126.ch7
Heflin, C., London, A.S., & Mueser, P. (2012). Clients’ Perspective on a Technology-Based Food
Assistance Application System. American Review of Public Administration, 43(6), 656673. Retrieved from http://journals.sagepub.com/doi/pdf/10.1177/0275074012455454
Herd, P. (2015). How administrative burdens are preventing access to critical income supports
for older adults: The case of the Supplemental Nutrition Assistance Program. Public Policy
& Aging Report, 25(2), 52–55. Retrieved from https://doi.org/10.1093/ppar/prv007
Jones, T. (2014). SNAP’s excess medical expense deduction: Targeting food assistance to lowincome seniors and individuals with disabilities. Center on Budget and Policy Priorities.
Retrieved from http://www.cbpp.org/sites/default/files/atoms/files/8-20-14fa.pdf
Kauff, J., Dragoset, L., Clary, E., Laird, E., Makowsky, L., & Sama-Miller, E. (2014). Reaching the
underserved elderly and working poor in SNAP: Evaluation findings from the fiscal year
2009 pilots. Washington, D.C.: Mathematica Policy Research.
Kim, K., & Frongillo, E. A. (2009). Patterns of food insecurity and participation in food assistance
programmes over time in the elderly. Public Health Nutrition, 12(11), 2113-2119.
Retrieved from https://doi.org/10.1017/S1368980009005357
Klerman, J. A., & Danielson, C. (2009). Determinants of the Food Stamp Program caseload. U.S.
Department of Agriculture, Economic Research Service and RAND Corporation.
Retrieved from https://naldc.nal.usda.gov/download/32849/PDF
LeBlanc, M., Lin, B.-H., & Smallwood, D. (2006). Food assistance: How strong is the safety net?
Amber Waves, 4(4), 11-15. Retrieved from: http://purl.umn.edu/125754
Mabli, J., Godfrey, T., Wemmerus, N., Leftin, J., & Tordella, S. (2014). Determinants of
Supplemental Nutrition Assistance Program (SNAP) participation from 2008 to 2012 final
report. Arlington, VA: Decision Demographics.
Mabli, J., Martin, E. S., & Castner, L. (2009). Effects of economic conditions and program policy
on state Food Stamp Program caseloads 2000 to 2006. Washington, D.C.: Mathematica
Policy Research.
Mack, M., & Paprocki, A. (2016). NCOA Senior SNAP Enrollment: Promising Practices Brief.
Oakland, CA: Social Policy Research Associates.
McConnell, S., & Nixon, L. (1996). Reaching the working poor and the poor elderly: Report on
literature review and data analyses. Mathematica Policy Research. Retrieved from
https://www.mathematica-mpr.com//media/publications/pdfs/nutrition/rwppe_litreview_dataanalyses.pdf
McConnell, S., & Ponza, M. (1999). The reaching the working poor and poor elderly study: What
we learned and recommendations for future research. Washington, D.C.: Mathematica
Policy Research.
R-3 | Elder SNAP Access
Mills, G. B., Vericker, T., Koball, H., Lippold, K., Wheaton, L., & Elkin, S. (2014). Understanding
the rates, causes, and costs of churning in the Supplemental Nutrition Assistance
Program (SNAP): Final report. Washington, D.C.: Urban Institute.
O’Brien, E., Wu, K. B., & Baer, D. (2010). Older Americans in poverty: A snapshot. AARP Public
Policy Institute. Retrieved from http://assets.aarp.org/rgcenter/ppi/econ-sec/2010-03poverty.pdf
Oemichen, M., & Smith, C. (2016). Investigation of the food choice, promoters and barriers to
food access issues, and food insecurity among low-income, free-living Minnesotan
seniors. Journal of Nutrition Education and Behavior, 48(6), 397–404.e1. Retrieved from
https://doi.org/10.1016/j.jneb.2016.02.010
Peterson, A., McGill, B., Thorn, B., Suchman, A., & Ribar, D. C. (2014). Examining the growth of
zero-income SNAP caseload: Characteristics, circumstances, and dynamics of zeroincome SNAP participants: Volume I. Arlington, VA: Insight Policy Research, Inc.
Ponza, M., & McConnell, S. (1996). Reaching the working poor and poor elderly: Interim report
on focus groups. Princeton, NJ: Mathematica Policy Research, Inc.
Ponza, M., Ohls, J. C., Moreno, L., Zambrowski, A., & Cohen, R. (1999). Customer services in the
Food Stamp Program: Final report. Princeton, NJ: Mathematica Policy Research, Inc.
Ratcliffe, C., McKernan, S.-M., & Finegold, K. (2007). The effect of state Food Stamp and TANF
policies on Food Stamp Program participation. Urban Institute. Retrieved from
http://www.urban.org/sites/default/files/alfresco/publication-pdfs/411438-The-Effectof-State-Food-Stamp-and-TANF-Policies-on-Food-Stamp-Program-Participation.PDF
Rowe, G., Gothro, A., Brown, E., Dragoset, L., & Eguchi, M. (2015). Assessment of the
contributions of an interview to SNAP eligibility and benefit determinations: Final report.
Washington, DC: Mathematica Policy Research, Inc.
Rowe, G., O’Brien, C. T., Hall, S., Pindus, N. M., Eyster, L., Koralek, R., & Stanczyk, A. (2010).
Enhancing Supplemental Nutrition Assistance Program (SNAP) certification: SNAP
modernization efforts: Final report – volumes I & II. Washington, D.C.: Urban Institute.
Rutledge, M. S., & Wu, A. Y. (2014). Why do SSI and SNAP enrollments rise in good economic
times and in bad? (CRR WP 2014-10). Center for Retirement Research at Boston College.
Retrieved from http://crr.bc.edu/wp-content/uploads/2014/06/wp_2014-10.pdf
Sama-Miller, E., Makowsky, L., Rowe, G., Brown, E., Clary, E., Castner, L., & Satake, M. (2014).
Evaluation of the effectiveness of pilot projects in increasing Supplemental Nutrition
Assistance Program (SNAP) participation among Medicare’s Extra Help population.
Washington, DC: Mathematica Policy Research.
Sattler, E.L. & Lee, J.S. (2013). Persistent food insecurity is associated with higher levels of costrelated medication non-adherence in low-income older adults. Journal of Nutrition in
Gerontology and Geriatrics, 32(1), 41-58. Rerieved from
http://dx.doi.org/10.1080/21551197.2012.722888
R-4
Schanzenbach, D. (2009). Experimental estimates of the barriers to Food Stamp enrollment.
Institute for Research on Poverty, University of Wisconsin-Madison. Retrieved from
http://irp.wisc.edu/publications/dps/pdfs/dp136709.pdf
U.S. Department of Agriculture, Food and Nutrition Service. (2015). Fact sheet: USDA support
for older Americans. Retrieved from
https://www.fns.usda.gov/pressrelease/2015/020215
U.S. Department of Agriculture, Food and Nutrition Service. (2016). State options report:
Supplemental Nutrition Assistance Program. Retrieved from
https://www.fns.usda.gov/sites/default/files/snap/12-State_Options.pdf
Wilson, C., McGill, B., Trippe, C., Gaddes, R., Estes, B., Tucker, M., & Tadler, C. (2015). An
assessment of the roles and effectiveness of community-based organizations in the
Supplemental Nutrition Assistance Program. Arlington, VA: Insight Policy Research, Inc.
Wu, A. Y. (2009). Why do so few elderly use food stamps? (WP 10.01). Harris School of Public
Policy, University of Chicago. Retrieved from
http://harris.uchicago.edu/sites/default/files/working-papers/wp_10_01.pdf
R-5 | Elder SNAP Access
This page is deliberately left blank.
Elder SNAP Access
Appendix A: SNAP Participation Rates for Eligible Elder Individuals, FY 2014
.
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Participants
Eligible
Individuals
Participation
Rate
65,382
7,385
61,405
33,391
161,023
43,838
44,829
7,208
11,748
469,562
126,184
23,852
14,596
178,271
56,374
24,137
23,512
62,188
65,897
24,176
66,630
135,512
133,631
41,578
52,686
80,536
10,054
11,977
28,714
10,688
109,943
29,527
538,695
128,984
3,851
137,858
42,351
81,165
156,265
20,833
69,206
8,003
126,394
312,066
13,298
12,848
66,832
90,361
32,745
57,779
2,355
215,426
19,670
248,861
141,926
755,163
116,479
89,927
23,241
27,309
843,498
349,671
43,600
41,771
369,830
170,255
65,889
72,610
177,600
221,402
46,537
154,856
215,533
296,203
101,100
160,378
206,523
28,303
36,963
82,820
25,737
223,009
80,193
752,124
375,235
11,563
351,095
151,160
129,049
374,520
35,469
221,803
21,525
270,142
837,340
51,973
18,947
215,971
164,371
95,686
126,107
11,441
30.4
37.5
24.7
23.5
21.3
37.6
49.9
31.0
43.0
55.7
36.1
54.7
34.9
48.2
33.1
36.6
32.4
35.0
29.8
52.0
43.0
62.9
45.1
41.1
32.9
39.0
35.5
32.4
34.7
41.5
49.3
36.8
71.6
34.4
33.3
39.3
28.0
62.9
41.7
58.7
31.2
37.2
46.8
37.3
25.6
67.8
30.9
55.0
34.2
45.8
20.6
90% Confidence Intervals
Lower Bound Upper Bound
27.3
34.4
21.5
20.4
18.7
34.3
45.1
27.6
36.9
51.6
32.7
50.1
31.3
44.8
29.9
32.8
29.2
31.9
26.7
47.8
39.4
58.0
41.4
37.5
29.0
34.9
31.5
28.4
31.2
37.5
45.1
33.4
66.7
31.7
29.0
36.4
24.5
57.4
38.9
54.7
28.3
32.7
43.1
34.1
22.1
63.2
27.8
50.5
30.4
41.6
16.6
33.4
40.6
27.9
26.7
23.9
40.9
54.6
34.4
49.2
59.7
39.5
59.3
38.6
51.6
36.3
40.5
35.6
38.1
32.9
56.1
46.7
67.8
48.8
44.8
36.7
43.1
39.5
36.4
38.1
45.5
53.5
40.2
76.5
37.0
37.7
42.1
31.6
68.4
44.5
62.8
34.1
41.6
50.4
40.5
29.1
72.5
34.1
59.5
38.0
50.1
24.6
Rank by
Rate
44
24
48
49
50
23
10
42
17
6
29
8
32
12
37
28
40
31
45
9
16
4
15
20
38
22
30
39
33
19
11
27
1
34
36
21
46
3
18
5
41
26
13
25
47
2
43
7
35
14
51
Sources: SNAP QC, Annual Social and Economic Supplement (ASEC) of the Current Population Survey (CPS), American
Community Survey, and individual income tax data and Census Bureau population estimates for 2010 through 2014.
A-1 | Elder SNAP Access
This page is deliberately left blank.
Elder SNAP Access
Appendix B: State Intervention Index
Exhibit B1: State Interventions Targeted towards Elders
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of
Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Elderly Simplified
Application Project
(ESAP)
In Place
Start Date
10/2008
Combined Application
Project (CAP)
In Place Start Date
Standardized Medical
Deduction (SMD)
In Place
Start Date
10/2014
Elderly and Disabled
Recertification
Interview Waiver
In Place
Start Date
36-Month Certification
Demonstration
In Place Start Date
―
11/2011
10/2016
10/2006
12/2012
1/2005
10/2015
11/2013
10/2010
10/2007
1/2011
3/2007
11/2016
10/2013
11/2005
4/2008
7/2013
9/2011
9/2009
B-1 | Elder SNAP Access
6/1/2012
Total State
Interventions
2
0
1
2
0
1
1
0
0
2
2
0
1
0
0
1
2
1
2
0
2
3
1
0
State
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Total
Elderly Simplified
Application Project
(ESAP)
In Place
Start Date
10/2012
Combined Application
Project (CAP)
In Place Start Date
10/2001
Standardized Medical
Deduction (SMD)
In Place
Start Date
9/2011
7/2016
7/2004
6/2013
4/2013
―
10/2015
1/2014
9/2013
6/2011
10/2010
10/1995
1/2010
10/2012
10/2015
10/2007
9/2002
9/2002
12/2008
10/2011
1/2007
1/2007
12/2001
7/2013
8
17
36-Month Certification
Demonstration
In Place Start Date
10/2003
5/2009
12/2009
7/2003
7/2005
Elderly and Disabled
Recertification
Interview Waiver
In Place
Start Date
4/2010
3/1/2005
1/2006
19
8
2
Sources: SNAP Policy Database, SNAP State Options Report, SNAP Current and Historical Waiver Databases,
USDA Support for Older Americans Fact Sheet, and communication with FNS
B-2
Total State
Interventions
2
2
0
1
0
1
1
1
1
1
1
0
1
1
2
1
3
2
0
3
0
1
2
2
0
0
1
Exhibit B2: Other State Interventions
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of
Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Community Partner
Interviewer
Demonstration
In Place Start Date
6/2009
8/2009
Telephone Interviews
in Lieu of Face-to-Face
In Place
Start Date
7/2009
6/2009
4/2004
6/2010
7/2009
6/2010
8/2005
3/2007
10/2008
10/2005
6/2009
10/2009
5/2009
10/2010
3/2007
3/2010
7/2009
12/2010
10/2010
5/2009
12/2006
9/2009
9/2009
7/2012
Call Centers
In Place
Start Date
―
8/2010
7/2006
6/2007
12/2008
11/2009
10/2008
9/2004
8/2004
7/1998
12/2008
11/2001
10/2007
10/2006
―
―
―
1/2002
7/2007
3/2008
11/2009
Online Applications
In Place
Start Date
―
―
12/2008
―
10/2007
10/2011
―
8/2005
4/2005
9/2008
―
―
1/2009
11/2007
5/2007
9/2003
―
5/2010
10/2011
12/2006
7/2006
8/2009
―
B-3 | Elder SNAP Access
Benefits Data Trust State
BDT Center Launch Year
2015
2012
Total State
Interventions
3
2
3
2
3
4
2
3
2
4
3
2
3
3
3
3
2
3
3
3
4
3
2
3
1
State
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Total
Community Partner
Interviewer
Demonstration
In Place Start Date
8/2009
2/2010
4
Telephone Interviews
in Lieu of Face-to-Face
In Place
Start Date
9/2015
10/2013
9/2008
5/2009
3/2009
10/2007
2/2008
7/2006
7/2005
8/2008
9/2008
2/2008
6/2009
3/2009
6/2002
4/2010
10/2006
5/2005
6/2003
1/2007
3/2009
12/2003
7/2009
10/2007
9/2010
49
Call Centers
In Place
Start Date
―
―
9/2004
12/2008
12/2008
―
7/2000
12/2008
1/2000
6/2005
―
6/2007
11/2004
1/2006
7/1998
11/2009
7/2001
9/2000
7/2000
7/2003
35
Online Applications
In Place
Start Date
3/2011
9/2008
10/2011
10/2011
6/2004
―
9/2004
―
3/2011
10/2010
―
―
4/2002
10/2007
12/2009
―
10/2007
1/2006
10/2007
10/2010
4/2005
1/2002
6/2003
6/2006
46
Benefits Data Trust State
BDT Center Launch Year
2014
2008
2015
5
Sources: SNAP Policy Database, SNAP State Options Report, SNAP Current and Historical Waiver Databases,
and communication with FNS
B-4
Total State
Interventions
2
2
3
4
3
2
3
3
3
1
2
3
3
4
3
4
2
2
4
3
2
2
3
3
3
1
Exhibit B3: Total Interventions, by State
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Total Elder-Targeted
Interventions
Total Non-ElderTargeted Interventions
Total State
Interventions
2
0
1
2
0
1
1
0
0
2
2
0
1
0
0
1
2
1
2
0
2
3
1
0
2
2
0
1
0
1
1
1
1
1
1
0
1
1
2
1
3
2
0
3
0
1
2
2
0
0
1
3
2
3
2
3
4
2
3
2
4
3
2
3
3
3
3
2
3
3
3
4
3
2
3
1
2
2
3
4
3
2
3
3
3
1
2
3
3
4
3
4
2
2
4
3
2
2
3
3
3
1
5
2
4
4
3
5
3
3
2
6
5
2
4
3
3
4
4
4
5
3
6
6
3
3
3
4
2
4
4
4
3
4
4
4
2
2
4
4
6
4
7
4
2
7
3
3
4
5
3
3
2
B-5 | Elder SNAP Access
This page is deliberately left blank.
Elder SNAP Access
Appendix C: Methodology Used to Calculate SNAP Participation Rates
We derived estimates for elder individuals and all eligible people for each State in each of the
five fiscal years FY 2010 to FY 2014 using empirical Bayes shrinkage estimation methods.
Specifically, we used a shrinkage estimator that optimally averaged direct estimates of SNAP
participation rates with predictions from a regression model. We obtained the direct estimates
by applying SNAP eligibility rules to households in the Current Population Survey Annual Social
and Economic Supplement (CPS ASEC) to estimate numbers of eligible people and used SNAP
Quality Control (QC) data to estimate numbers of participating people. The regression
predictions drew on data from the American Community Survey (ACS), Internal Revenue Service
individual income tax data obtained from the Census Bureau’s Small Area Estimation Branch,
population estimates, and administrative records.
This procedure, summarized by the flow chart in Figure 1, has the following four steps:
1.
From CPS ASEC data, SNAP administrative data, and population estimates, derive direct
estimates of State SNAP participation rates.
2.
Using a regression model and the direct estimates derived in Step 1, predict State SNAP
participation rates based on SNAP administrative, individual income tax, and ACS data
and population estimates.
3.
Using a shrinkage estimator, average the direct estimates from Step 1 and the
regression predictions from Step 2 to obtain preliminary shrinkage estimates of State
SNAP participation rates.
4.
Adjust the preliminary shrinkage estimates from Step 3 using national estimates of
eligible people derived from the CPS ASEC to obtain final shrinkage estimates of State
SNAP participation rates.
Each step is described in the remainder of this section.
1.
From CPS ASEC data and SNAP administrative data, derive direct estimates of State SNAP
participation rates.
The first step is to directly estimate SNAP participation rates by dividing an estimate of the
number of people participating in SNAP by an estimate of the number of people eligible for
SNAP, with the resulting ratio expressed as a percentage. To derive a participation rate for
elder individuals, we divided the number of elder participants by the number of eligible
elder individuals. We used SNAP QC data to estimate numbers of participants in an average
month in the
C-1 | Elder SNAP Access
Figure 1. The Estimation Procedure
fiscal year and CPS ASEC data to estimate numbers of eligible people in an average month.
The CPS ASEC collects income data by calendar year, so we used two years of CPS ASEC
data for each set of fiscal year estimates. For example, we obtained estimates of eligible
people in FY 2014 (October 2013 through September 2014) from the 2014 and 2015 CPS
ASEC. For details on how we derived the direct estimates, see Farson Gray and
Cunnyngham (2014), who used the same methodology to estimate national SNAP
participation rates.
2.
Using a regression model, predict State SNAP participation rates based on
administrative, ACS, and other data.
The second step was to use data from outside the CPS ASEC to estimate a regression model
and formulate a prediction for each group (elder individuals and all eligible people) in each
C-2
State in each year. Our regression model consisted of ten equations, with five predicting
SNAP participation rates for eligible elder individuals in fiscal years 2010, 2011, 2012, 2013,
and 2014, and five predicting SNAP participation rates for all eligible people in fiscal years
2010, 2011, 2012, 2013, and 2014. The ten equations were estimated jointly, and the
values of the regression coefficients could vary from equation to equation. The seven
predictors used were (in addition to an intercept):
Percentage of the population receiving SNAP benefits according to administrative
data and population estimates
Percentage of SNAP participants who are elderly according to SNAP QC data
Percentage of renter occupied housing units that spent 30 percent or more of
household income on rent and utilities according to ACS one-year estimates
Percentage of households with a female householder, no husband present, and
related children under age 18 according to ACS one-year estimates
Percentage of occupied housing units that are owner occupied according to ACS
one-year estimates
Median adjusted gross income according to individual income tax data
Percentage of all people not claimed on tax returns according to individual income
tax data and population estimates
In addition to the predictors that we selected for our model, we considered many other
potential predictors. All of the predictors considered had three characteristics: (1) it is
plausible that they are good indicators of differences among States in SNAP participation
rates; (2) they could be defined and measured uniformly across states; and (3) they could
be obtained from nonsample or highly precise sample data—such as the ACS or
administrative records data—and, thus, measured with little or no sampling error.
3.
Using shrinkage methods, average the direct estimates and regression predictions to
obtain preliminary shrinkage estimates of State SNAP participation rates.
To derive preliminary estimates of state SNAP participation rates, we averaged the direct
estimates calculated in Step 1 and the regression predictions from Step 2 using an
empirical Bayes shrinkage estimator. We call the estimates from this step “preliminary”
because we make some fairly small adjustments to them in the next step.
C-3 | Elder SNAP Access
4.
Adjust the preliminary shrinkage estimates to obtain final shrinkage estimates of State
SNAP participation rates and numbers of eligible people.
We adjusted the preliminary shrinkage estimates of participation rates so that the counts
of eligible people implied by the rates sum to the national count of eligible people
estimated directly from the CPS ASEC. This adjustment was carried out separately for each
year and for the two groups (elder individuals and all eligible people). To implement the
adjustment, we calculated preliminary estimates of the numbers of eligible people from
the preliminary estimates of participation rates derived in Step 3 and the administrative
estimates of the numbers of SNAP participants obtained in Step 1. Using the FY 2014
estimates for eligible elder individuals as an example, the State estimates summed to
10,071,353 and the national total estimated directly from the CPS ASEC was 9,867,805. To
obtain estimated numbers of eligible elder individuals for States that sum (aside from
rounding error) to the direct estimate of the national total, we multiplied each of the State
preliminary estimates of eligible people by 9,867,805/10,071,353 (0.9798). Such
benchmarking of estimates for smaller areas to a relatively precise estimated total for a
larger area is common practice. Applying this adjustment, we obtained our final shrinkage
estimates of the numbers of people eligible for SNAP. From those estimates and our
administrative estimates of the numbers of SNAP participants, we derived final shrinkage
estimates of participation rates.
C-4
| File Type | application/pdf |
| File Title | Report Cover |
| Author | Rachel Lindy |
| File Modified | 2017-06-30 |
| File Created | 2017-03-14 |