Nonparticipation and problems of access in the food stamp program a review of the literature

              . nt*j m 001*4-A-o * AW2' V73
United States
Department of
Agriculture
Food and
Consumer
Service
Nonparticipation and
Problems of Access in the
Food Stamp Program:
A Review of the Literature
Office of Analysis and
Evaluation
«1-0^*
■ (!/
United States
Department of
Agriculture
Food and 3101 Park Center Drive
Consumer Second Floor
Service Alexandria, VA 22302
Nonparticipation and
Problems of Access in the
Food Stamp Program:
A Review of the Literature
Gretchen Schafit, PLD. and William Millsap, Ph.D.
A Product of
LTG Associates, Inc.
6930 Carroll Avenue 875 E. Canal
Suite 410 Suite 1
Takoma Park, MD 20912 Turlock, CA 95380
February 1996
£
ACKNOWLEDGMENTS
The authors wish to thank staff of the Food and Consumers Service for their
guidance in the production of this document. In particular, thanks to Gary Bickel, the
Government Project Officer, for his interest and guidance, in addition to Steven Carlson,
Robert Dalrymple, and Patricia Seward for their review. Staff members of LTG
Associates, Inc., including Cathleen Crain, Marilyn Madden, Beatrice de la Brosse, and
Jennifer Maahs, contributed to the editing and production ofthe document.
This study was conducted under contract number 53-3198-3-032 with the Food and Consumer Service,
U.S. Department of Agriculture. Points of view or opinions stated in the report do not necessarily
represent the official position of the Food and Consumer Service.
CONTENTS
Chapter Page
L INTRODUCTION AND BACKGROUND 1
II. ASSESSING NONPARTICIPATION IN THE FOOD STAMP PROGRAM 6
III. PROBLEMS OF ACCESS AND RELUCTANCE OF POTENTIALLY ELIGIBLE PERSONS
TO PARTICIPATE IN PUBLIC BENEFIT PROGRAMS 18
IV. EFFICACY OF CLIENT ASSISTANCE ACTIVITIES 25
V. SUMMARY AND CONCLUSIONS 35
VL BIBLIOGRAPHY ^38
ill
NONPARTICIPATION AND PROBLEMS OF ACCESS IN THE
FOOD STAMP PROGRAM: A REVIEW OF THE LITERATURE
I. INTRODUCTION AND BACKGROUND
This paper reviews literature about participation and nonparticipation in the Food Stamp
Program (FSP) and initiatives within it that have informed hard-to-reach client groups about food
stamps and/or assisted clients in the application process. The review encompasses five sections,
the first of which discusses the background of efforts to reach client groups who appear to be
underrepresented among food stamp participants. The second section reviews research literature
based on survey data that describes patterns of non-participation in the FSP. The third section of
the paper reviews literature based on direct interviewing or ethnographic observation of people
in financial need and their attitudes toward public benefit programs. These studies come from a
variety of programs, including those serving not only food stamp recipients but particular
populations of clients, such as people with AIDS, homeless persons, mothers and children, and
elderly individuals. The fourth section examines systematic efforts to provide outreach and client
assistance. The conclusion of the review, in the fifth section, discusses what the literature offers
to our understanding of patterns and problem of access to FSP participation.
Since the 1930's, publicly-funded programs have provided a safety net, so that a basic level
of health and welfare is available to all of the population. The Food Stamp Act of 1964,
establishing what has become one of the most significant of these benefit programs, was designed
to help distribute surplus farm produce as well as to improve the well-being of low-income families
and individuals in the United States (U.S. Senate 1985) Policy and program objectives within the
Food Stamp Program (FSP) have always focused on providing nutritional benefits to ensure a
healthy nation by enabling people who meet specific income, asset, and work requirements to
receive these benefits. Begun as a set of pilot projects initiated by President Kennedy through an
1
Executive Order in 1963, by 1973 the Agriculture and Consumer Protection Act changed the FSP to
a national entitlement program, providing benefits to all eligible persons certified within the
guidelines. The FSP is currently by far the largest food assistance program in the United States and
one of the largest Federal assistance programs generally, second only to the Medicaid program.
The FSP served 27 million people in fiscal year 1993 with a budget of $22 billion.
Distributing food stamps equitably to all those who are eligible and wish to participate,
even those hardest to reach, has been a central purpose of the FSP and other government benefit
programs. This goal has been monitored by a series of research studies which have analyzed
characteristics of participants and determinants of participation in the program. The program
objective, to provide the entitled benefits to all who qualify, has produced client-assistance
initiatives, including demonstration projects, designed to learn more about the barriers that exist to
enrolling and participating in the FSP among certain groups of potentially eligible people.
Public Law 91-671, passed in 1971, required State food stamp agencies to reach out to low-income
households to inform them of their potential eligibility for this benefit This outreach was
not tied to any nutrition education, but was meant to support participation by those most in need.
Specifically targeted through Federal and State support were the elderly, migrant farmworkers, and
named ethnic groups. In 1973, a Federal court order mandated that $278.5 million be spent on the
evaluation of these outreach plans and remedial action to "insure" participation of eligible
households. In 1975, regulations were passed that required staffing for outreach at the State level.
In the 1980's, these legislative mandates were weakened. Some State and local agencies
reported that they felt mandated outreach resulted in an unreasonable work load. As a result, in
1981, P.L. 97-35 left outreach to the discretion of the States and Federal funds were barred from
being used to share costs of outreach that States might wish to conduct In 1987, the prohibition on
outreach was lifted from the Federal FSP and money was made available to target the homeless
(Nichols 1989).
2
In 1988 Federal funding was once again authorized for food stamp outreach under the
Hunger Prevention Act. Under this act, State agencies could spend part of their Federal dollars on
outreach. States received 50-50 Federal matching funds for these client enrollment assistance
efforts. Also, the FSP began to require that bilingual services be provided to non-English-speaking
groups, including the use of translated certification materials.
Participation levels have become a commonly used criterion for evaluating Federal
programs. In order to inform the Federal government how well the FSP is serving the target
population, that is, those eligible for benefits, the participation rate is a useful marker, for it
measures the proportion of those persons eligible for food stamps who actually apply for and
receive food stamps.
While participation in the FSP has never encompassed all of the people in the population
estimated to be eligible, it has increased among both households and individuals in the past decade.
In the 1970's, more than half of the households eligible for food stamps did not participate in the
FSP. In 1983, a study reported that 55 percent of those eligible for food stamp benefits were still
not enrolled (Olson and Xiao 1991). A U.S. Department of Agriculture (USDA) study estimated,
however, that 60 percent of eligible households and 66 percent of eligible individuals were being
served by the FSP in 1984 (AARP 1991b, 18). Low enrollment among persons eligible for social
programs was not limited to the FSP, in the mid-1980's, but was found to be the case in the
Supplemental Security Income Program (SSI) and die Low Income Home Energy Assistance
Programs as well (Coe 1983).
The use of food stamps declined by one million participants between 1980 and 1987, with
one million additional persons classified as being below the poverty level but not using food stamps
(U.S. Conference of Mayors 1988). Currently, it is estimated that 69 percent of all eligible
households and 74 percent of all eligible individuals are participating in the FSP (Trippe and Sykes
1994). The current participation figures in the FSP may reflect efforts made by Regional, State, and
local food stamp offices to reach particular groups of recipients who were underrepresented in
enrollment figures, such as migrants, the elderly, Native Americans, linguistic minorities, and
others.
Many programs, including the FSP, have identified the elderly, disabled, homeless, rural
poor, low-income working poor, and non-English-speaking populations as target groups. These
groups are often the hardest to reach and may be found outside routine service provision networks.
The USDA has supported client assistance activities in the last few years, in part by funding
demonstration projects to determine effective means to reach specific populations such as homeless
individuals and families, elderly persons, low-income working people and their families, persons
living in rural areas, Native Americans, recent immigrants, and migrants. These projects,
conducted by non-profit organizations or governmental agencies, use specific techniques in efforts
to reach their target groups. The accesses and difficulties experienced by these projects are being
monitored to determine what might be effective and efficient ways of reaching the underserved
population groups.
While the literature regarding food stamp participation is sizable and has been reviewed
regularly, this paper focuses on the reverse of participation, that is nonparticipation, an aspect
which is more difficult to understand. The paper reviews three distinct bodies of literature: first,
survey-based research on food stamp participants and participation; second, observational and
ethnographic studies of the problems of access to the FSP and other benefit programs; and third,
evaluation reports of efforts at outreach and client assistance in several benefit programs. The
review focuses on three issues:
• What information exists on nonparticipation rates and characteristics of those who do not
participate in the FSP?
• What information exists about problems people have in accessing public programs?
• What information exists on the efficacy of certain approaches to overcoming problems of
access?
Much more has been written about participation in the FSP than about nonparticipation. In
1991, the Center on Budget and Policy Priorities reported that there were only four studies on
nonparticipation in the survey literature, three of which were done in the early 1980's (AARP,
1991a). Most of the literature reviewed here centers on participation, with nonparticipation
addressed by extrapolation, for usually nonparticipation is treated as a reciprocal to an assumed,
optimal participation level. Exceptions are an early article by Coe which directly addressed
nonparticipation as a central issue in the FSP and some studies done by the USGAO. These studies,
however, are limited to demographic projections based on what is known about food stamp
participants' characteristics (Coe 1983) or telephone survey information (USGAO 1990).
The studies based on statistical analysis of sample survey data share several common
features:
• Eligibility estimates are drawn from census data or samples and are based on demographic
information;
• Potential participation in benefits programs such as food stamps is measured against the
estimated eligible population; and,
• The level of nonparticipation is assumed to be the difference between actual participation
and projected eligibility levels.
In short, statistical methods are employed to project estimated potential levels of
participation in the general populations. However, there is little research that uses specific case
studies, behavioral assessment, or other methods to draw conclusions about why populations
participate or do not participate in welfare benefit programs. The USGAO study, made in
collaboration with the University of Michigan (1990) described below, is an exception to this.
A second kind of literature and research is based on observation of the processes by which
people obtain or do not obtain government benefits. It also uses personal interviewing with clients
and others who would be eligible for benefits if they could, or chose to, apply in order to find out
what barriers may stand between potential clients and their participation. This research offers
description and insights into clients' decision-making processes. Examples, presented below,
include the work by Elliot Liebow and other researchers who gather data directly from the people
they describe.
Each kind of reporting has its advantages and limitations. Large-scale studies of
nonparticipation using national surveys reduce sampling bias, but cannot estimate the extent to
which processes are involved that respondents either cannot or will not describe. Projections of the
characteristics of nonparticipants from participation data do not attempt to assign reasons for the
patterns that are found, but merely describe the population groups who do or do not take part in
benefit programs. On the other hand, observational studies, based on in-depth knowledge of the
populations that are studied, do not sample and cannot project accurately to a national population.
Their description of problems of access are insightful, but are subject to researcher and sampling
bias.
DL ASSESSING NONPARTICIPATION IN THE FOOD STAMP PROGRAM
The survey research described in this section is based on a microsimulation model that
provides an estimate of the food stamp participation rate, "...where the numerator is the number of
persons or households participating in the program...and the denominator (is) the number of persons
or households eligible for the program...." (Trippe 1989, vii). Estimates of the denominator in this
ratio are, of course, what demand attention from researchers. In the past decade, the methods to
obtain these estimates have been improved to provide a closer approximation of those eligible for
theFSP.
In this section, literature from studies using two different data sets, the Michigan Panel
Study of Income Dynamics (PSID) and the Survey of Income and Program Participation (SIPP) are
examined, as well as several reviews of participation research conducted over the past years. PSID
survey data rely on yearly responses of a sample of people assumed to be eligible for food stamp?
based on their income and assets, while SIPP data are taken from quarterly surveys that include
more parameters, including monthly income and expenses, that better match food stamp eligibility
requirements. In addition, several reviews of participation studies are included in this section.
These household surveys may over-estimate the eligible population because they do not
take into account all exclusions, factor that are in the food stamp regulations. They may also
under-estimate participation rates when they depend upon self-reported food stamp participation,
because respondents may not wish to share that information. Because of their sampling of the
general population, however, they provide a broader-based view of participation measures than
ethnographic or observational studies are able to do.
Richard Coe (1983)
"Nonparticipation in Welfare Programs by Eligible Households: The Case of the Food Stamp
Program." Journal ofEconomic laauo- 17(4).
In this early article, Coe examined nonparticipation in public welfare programs and, in
particular, the FSP using the PS1D data set He looked at responses of those households to
questions asking why they did not participate in the FSP, and noted that it is an implicit assumption
that households which are eligible to participate in public transfer programs will do so. He found
evidence that nonparticipation in welfare programs by eligible households was far more common
than anticipated at that time. This was especially true of the elderly who qualify for SSI as well as
bousing assistance programs. The reasons why this is so are not well understood due to the lack of
adequate data. The author then examined this issue with respect to the FSP.
Based on these data, Coe estimated that only 45 percent of eligible households received
food stamps. This failure to participate in the program was attributed to five problem categories: 1)
informational concerns; 2) problems with program parameters, namely the amount ofthe benefit; 3)
administrative difficulties; 4) physical access considerations; and, 5) attitudinal issues.
Of the 55 percent nonparticipating households, the primary reasons given for not taking
part in the food stamp program were insufficient information about the program and uncertainty
about potential eligibility. Other reasons had to do with administrative difficulties with the
application process. Only one percent were discouraged by the small size of the benefit in their
own cases.
Coe found that participation in services for which one is eligible increases with the size of
the benefit that one can gain from application to a program and is negatively related to an employed
head of household. "Age-sex-marital" status is a configuration that affects participation. Married
couples appear more likely to be participants than single men. Unmarried men are more likely to
be told that they are ineligible for participation in welfare and food stamp programs regardless of
actual eligibility. This indicates that mere may be certain conditions, not barriers, that influence
one's perceived need for, and willingness to accept, assistance. Further demographic analyses, he
noted, are only proxies for understanding the behavioral reasons of groups not choosing to
participate in the FSP. Coe suggests in this article that policy ramifications of the findings are
clear-cut: more attention must be paid to informing potential clients about their eligibility to
participate in the FSP.
USGAO (January, 1990)
"Food Stamp Program: A Demographic Analysis of Participation and Nonparticipation."
Washington, D.C.: General Accounting Office, (GAO/PEMD-90-8).
The General Accounting Office (GAO) issued a series of reports on food stamp
nonparticipation based on data from 1979-1981 and, in this report, data from the 1987 PSID. This
study asked two questions:
1. What demographic characteristics or factors are associated with FSP participation?
2. What reasons do food-stamp-eligible households give for not participating in the program?
Researchers interviewed a sub-set of respondents to the PSID survey whom they believed
to be potentially eligible for food stamp benefits. They determined specific household
characteristics and correlated them with responses about food stamps.
The study found that 87 percent of households participating in other welfare programs also
participate in the FSP. Demographic characteristics associated with participation were persons
receiving SSI, households headed by a non-white single female, and those with children. Each
additional child in the household was associated with a 67 percent rise in the likelihood of
participation in the FSP. By contrast, with every $1,000 increase in annual income, households
were five percent less likely to participate in the FSP.
Nonparticipation was associated with the following factors:
• A reported lack of desire for benefits (382%);
• Lack of, or incorrect, information about the FSP (36.1%); and,
• Perceived or actual problems of program access (25%).
The GAO study, like Coe's work, identifies the reasons people give in interviews for not
participating in a benefit program. However, the content of the categories is not explored and
questions are left unanswered about the reasons why benefits are not wanted, what information is
incorrect or missing, and what '-'•oblems potential clients have with the program or access to it.
In a number of studies commissioned by the US Department of Agriculture (USDA), Food
and Nutrition Service (FNS), now Food and Consumer Services (FCSX attention was focused on
the levels of participation in the FSP and indirectly to nonparticipation in the program. The
following six studies are surveys with this focus.
9
Susan Allin and Harold Bccbout (November, 1989)
"Determinants of Participation in the Food Stamp Program: A Review of the Literature."
Alexandria, VA: USDA, FNS.
This extensive review of the existing literature on participation concluded that, according
to recent studies, only 60 percent of eligible households participate in the FSP. Three questions are
examined:
• What are the principle reasons eligible persons or households do not participate in die FSP?
• How do participation rates vary across different types of households? What are the reasons
for that variation?
• On what basis do eligible households make their decisions regarding participation? How
might changes in program structure or operations influence those decisions?
These questions were then answered by reviews of studies that use survey data to elicit
reasons for nonparticipation, correlate participation levels with household characteristics, and use
multivariate analysis to identify the best predictors of participation and/or nonparticipation.
Nonparticipation was attributed in these studies to faulty information about eligibility,
administrative hassles, stigma, and distance to the food stamp offices, which some respondents
considered to be too far in relation to the size ofthe benefit they were likely to be granted.
The authors noted that the findings in the literature varied substantially due to the data
used, methods employed, and assumptions made in conducting the studies. Indeed, there were
variations in participation rates across different eligible populations. In order to determine what
factors are associated with participation or nonparticipation, it was argued, one must compare
households participating in FSP with those not participating in the program.
However, the authors also pointed out that no survey data set has all of the information
needed to make such comparisons. Measurement error is a real problem for making any reliable
projections about eligibility, participation, and, by extrapolation, nonparticipation in FSP. Having
stated the limitations ofmany studies cited in their literature review, they presented comparisons of
different analytic approaches to examining FSP populations.
Aliin and Beebout described the limitations of methodologies that have been used to
understand variations in participation rates across eligible populations and identified characteristics
that have been associated with participation and nonparticipation. Large-scale survey research
produce responses that are too general to allow much interpretation about actual decision-making
processes that individuals and families engage in when deciding to apply for or not to apply for
food stamps, but they cannot explain the actual behavior of households. This lends little guidance
to those interested in identifying ways to influence participation behavior among eligible
households. To gain further understanding about how low-income households acquire information
and make program participation decisions, the authors suggested, observational and ethnographic
studies may be needed. All studies must be up-to-date if they are to provide infomation that can be
used to develop policy, for older studies may not be predictive over periods oftime. SIPP data may
be best used to update results of older data, for it best replicates household characteristics relevant
to FSP participation. The authors urge a better understanding ofthe motivation behind the behavior
of nonparticipants in order to make policy decisions.
Carole Trippe (1989)
"Estimating Rates of Participation in the Food Stamp Program: A Review of the
Literature." Alexandria, VA: USDA, FNS.
In a complementary report to that of Allin and Beebout (1989), Trippe reviewed literature
on FSP participation rates stressing that it was unreasonable to expect universal participation in any
voluntary social welfare program, but that participation should be representative of eligible
populations. As with the Allin and Beebout study, the estimated participation rates varied
depending upon measurement procedures, data sources, and the study methodology employed.
Researchers have used two different types of data sources to estimate the number of participants in
FSP; administrative data which provide actual counts of participants but do not directly address
eligibility and, therefore, nonparticipation levels; and, household surveys such as the Current
11
Population Survey (CPS), the Panel Study of Income Dynamics (PSID), and more recently, the
Survey of Income and Program Participation (SIPP). Trippe discussed how to account for the
disparities in the figures, why they vary, and their changes over time. The report reviewed the
latest data source for calculating participation levels of the FSP, the SIPP, developed to estimate
with greater precision the number of eligible clients who actually participate in public benefit
programs. This survey collects data on a monthly basis from a statistically drawn sample of
households. Of interest to any serious reviewer of food stamp participation and nonparticipation
levels is the discussion on the inability to directly measure food stamp eligibility, the lack of data
on this topic, and how different data sources are used to get at this issue.
The author's conclusions are that monthly SIPP data allow greater precision in estimating
eligibility, FSP participation levels, and lastly, nonparticipation rates in the program. It is argued
that the use of SIPP data, even employing differing analytic methods, can arrive at a more precise
understanding of food stamp population dynamics.
Harold Beebout, et ah, eds. (1990)
"Food Stamp Policy Issues: Results from Recent Research." Washington, D.C.:
Mathematics Policy Research, Inc.
Taking a snapshot on the "state of the art" in food stamp policy, the Food and Nutrition
Service brought together a number of recognized scholars on food stamps to assess where the
program is and where it might be going-problems, issues, and policy ramifications for welfare.
Noting that the FSP has been in place for decades, it has understandably undergone considerable
evolution to meet competing program objectives in a changing national environment
Despite significant growth in the FSP, the authors stressed that hunger and problems of
welfare of low-income households appeared to be increasing as the FSP entered the nineties. The
papers in the volume cover such topics on FSP participation as:
• Flow well does FSP reach the target population of low-income households?
12
• Does the FSP improve diet quality for low-income households?
• How well do FSP and other food assistance programs meet the needs of vulnerable groups
such as the homeless? and,
• How should FSP employment and training policies be structured to best serve participants?
As with earlier reports, the editors noted that like other welfare programs, not everyone
who is eligible for food stamps participates in the program, yet the level of participation is used as
one indicator of FSP success. Further, the report acknowledged that in most areas, the FSP is
administered jointly with the Aid to Families with Dependent Children (AFDC) and Medicaid
programs, with food stamps providing about 33 to SO percent of the family's total purchasing
power. Some limited discussion addresses the question of why some eligible households do not
participate in the FSP. The author suggests from the data that they:
• Did not know they were eligible;
• Did not want or believe they needed food stamps; or,
• Found the cost of participation too high in relation to the level ofbenefits received.
One conference participant concluded that SO percent of FSP dynamics (changes in
participation levels and composition) may be due to changes in household composition (e.g.,
income changes, number of persons, etc.) and reflects the changing family in U.S. society.
In order to understand nonparticipation by seemingly eligible households more adequately,
the author suggested that data collection methods need to go beyond national household surveys
(citing focus groups, detailed interviews, and other market research methods) to arrive at a more
lucid picture ofFSP nonparticipation patterns.
In addition to articles on FSP participation (nonparticipation, only inferentially), this report
also described the impact of the FSP in meeting the dietary needs of FSP participants, FSP and the
homeless, the elderly, and a comparison of the FSP with the Food Distribution Program on Indian
Reservations. Lastly, the authors discuss the FSP's relationship to employment and training
13
programs for able-bodied food stamp recipients and the role ofwork in reducing the need for public
assistance. Nonparticipation is not directly addressed in this compendium of articles.
Carole Trippe and Pat Doyle (1992)
"Food Stamp Participation Rates: January 1989." Alexandria, VA: USDA, FNS.
While the primary focus of this study by Mathematica Policy Research was to determine
FSP participation rates, nonparticipation was addressed far more thoroughly in this study than in
many other works. This was the fourth in a series of reports seeking to provide estimates of rates of
participation from the total eligible population and among selected subgroups within the overall
population.
Comparisons were made with the earlier estimates of participation rates from the August
1985 and January 1989 studies which derived the estimated participation rates from a consistent
data base (the SIPP) and a common methodology for generating the estimates. Based on these data,
the report indicated that in the SO States (and the District of Columbia):
• Fifty-nine percent of the eligible individuals participated in the FSP;
• Fifty-six percent ofthe eligible households participated in the program; and,
• Participating households received 66 percent of the benefits payable had all the eligible
households participated.
The authors noted that even using a common data base and methodology, h was still quite
difficult to calculate the number of eligible households with certainty, given problems of income
under-reporting and various sampling issues.
Nevertheless, the authors concluded that the SIPP data enable a much improved simulation
of the actual dynamics of FSP participants and nonparticipants within the eligible FSP population.
It concluded that FSP participation stabilized at 59 percent among eligible persons between 1988-
1989 after declining five percentage points between the years January 1985-January 1988. The
14
stable participation level of 1988-89 is attributed to no changes being made in the program in these
years.
This study found that higher benefits correlated directly with higher participation levels.
Indeed, participation rates were found to increase as potential benefits rose; participation rates were
much higher among persons in households whose income was below the poverty level (74%); and
households with earnings participated at a lower-than-average rate (32%). In other words, the
higher the level of poverty, the more need there is, and the higher the benefit level, the more
attractive program participation becomes to the eligible person in need.
Of the projected 5.7 million nonparticioating households among an estimated 12.7 million
eligible in 1989, more than half (56%) had incomes above the poverty line, and 39 percent were
eligible for only a small monthly benefit (less than one-fourth die maximum allotment). Overall,
more than half of the nonparticipants (56%) were households with elderly persons or wage-earners
(45%); about half of each of these groups fell below die poverty line. Finally, Trippe noted that
most of the eligible nonparticipating households above the poverty line would have been eligible
for relatively small monthly food stamp benefits (S23 on average), which may explain their relative
lack of participation.
No data were provided on reasons for nonparticipation in this report Conclusions were
inferred from the demographic profiles of the general population and a review ofthe subgroupings.
Alberto Martini and Susan Allin (December, 1993)
"Did the Characteristics of Food Stamp Program Entrants Change During the 1990-91
Recession?" Alexandria, VA: USDA, FNS.
This research attempted to determine how the characteristics of food stamp entrants
changed during the 1990-1991 recession. Indeed, the authors noted that dramatic changes occurred
in the number of persons participating in the FSP, increasing from 18.6 million persons in July
1989 to 27.3 million persons in April 1993. From July 1990 (the date given for the start of the
15
recession) to July 1991, the FSP experienced an increase of 3 million participants or a quarter
million persons a month.
The authors were interested in: 1) investigating whether the new FSP participants were
similar to participants who were in the program prior to the recession, and 2) finding possible
linkages between different rates of FSP participation and the slow recovery from the recession. The
data for this study were obtained from administrative reporting of participating FSP households by
the Integrated Quality Control System (IQCS). Household information was gathered by use of
SIPP data. The authors identified four groups of participants in the FSP based on a range of factors,
such as income, employment skills, assets, etc. These groups were labeled "secure," "precarious,"
"distressed," and "on the edge." Additionally, the groups were also classified as "long-term" and
"short term" participants in FSP.
The report concluded that FSP participation rates were higher during the recession than
prior to it, although former middle class households participated at much lower rates than predicted
by their potential eligibility levels. It would appear that the dramatic increases are derived from a
higher rate of participation of persons eligible for the FSP who prior to the recession had not
participated. The additional economic stress apparently resulted in changing behavior patterns
regarding this benefit The authors did not attribute these decisions to participate to the state of the
economy, nor did they address the issue ofbow such decisions might be made.
Carole Trippe and Julia Sykes (1994)
"Food Stamp Program Participation Rates: January 1992." Alexandria, VA: USDA, FNS.
The authors reported that by January of 1992, the FSP was providing benefits to 74 percent,
or 24 million, of die 33 million persons estimated to be eligible for food stamps—representing 69
percent of eligible households and receiving 82 percent of the benefits. By way of evaluating the
participation levels among subgroups, it was found that:
16
• Almost all eligible children participated (under 5, 95%; under 18,86%);
• One in three eligible elderly persons participated (33%); the majority of
nonparticipating eligible elderly lived alone; and,
• Single parent households participated more than any other household, that is, if
children lived with a single adult, their households were more likely to participate
(! 00%) than if they lived with two or more adults (78%).
Confirming conclusions of the Martini and Allin (1993) study, this study also reported that
the dramatic increase between January 1989 and January 1992 was driven largely by a higher
participation rate among those already eligible rather than an increase in the number of eligible
households. In short, the nonparticipation rate among eligible households went down, with 26
percent of eligible persons reported as not participating in the FSP. Nonparticipants were more
likely to be elderly persons and households headed by a white non-Hispanic, ethnic groups other
than African Americans, those with earnings or those eligible for only small benefit amounts. It
was estimated that 92% of eligible African Americans participated, but only 61% of eligible
Hispanics and 59% of eligible white non-Hispanics. Those with the least resources participate at
higher rates than other low-income people, with 86% of those assumed eligible who live below the
poverty line participating versus 21% of those above the poverty line. A combination of earned
income and low benefits is related to nonparticipation, as could be expected.
In this study, excellent demographic characteristics were provided in several tables on
these nonparticipants, but no discussion was devoted to reasons why they chose not to participate in
FSP.
To summarize this section of the paper, survey data have shown that nonparticipation has
been associated with a lack of information about the FSP, problems with physical access to food
stamp offices, negative attitudes on the part of potential clients toward the procedures required to
apply and participate, and less financial need than others who do participate. Nonparticipants are
often older people, white non-Hispanics, those with incomes, or those eligible for small benefits.
17
Survey data cannot provide further information beyond the self-reporting provided by this format
Understanding possible underlying causes for nonparticipation requires a different approach to
respondents.
III. PROBLEMS OF ACCESS AND RELUCTANCE OF POTENTIALLY
ELIGIBLE PERSONS TO PARTICIPATE IN PUBLIC BENEFIT PROGRAMS
Problems of access and the reasons why many potentially eligible clients are hesitant to
enter into the public benefit systems are not easily studied through systematic survey research
(Allin and Beebout 1989). Either observational research and ethnography or multi-methodological
studies are needed to reach beyond the answers people will give to an anonymous interviewer in
short-answer, survey formats. Going beyond the first response of a respondent, which may be
given to save face or to protect oneself by giving simplified and perhaps inaccurate answers to
personal questions, is a job requiring a different approach from that of the statistician.
Ethnographic methodologies include participant observation in which the researcher spends
a significant amount of time with the subjects of the study in order to obtain their view of the
subject matter that is under investigation. Traditional ethnographies often take more than a year to
complete, but this method has been increasingly used in short-term studies that incorporate careful
observation and recording with goal-oriented interviewing and other data gathering techniques.
In most ethnographic or observational studies, observational data are used in conjunction
with numerical data from demographic reports, records, surveys and the like to describe certain
parameters of the research question. The numerical data are usually secondary to the narrative
analyses. Researchers use a qualitative approach in order to learn about and to describe complex
social relationships and to arrive at conclusions that may enlighten decisions for action and/or
policy. Researchers usi j this approach are not as interested in tightly circumscribed questions or
in testing propositions that have been developed outside of the social scene itself, nor do they
18
expect that social inquiry will lead to ultimate understanding ofthe nature of large-scale systems of
social interaction. They are more likely to look for operating systems and negotiated rules for
behavior and beliefs around specific domains of activity, events, and behavior. The results of their
work point to areas in which survey research may be particularly fruitful.
Marie Rank (1994)
Life on the Edge. New York: Columbia University Press.
In this book, the subject matter is the welfare system, and recipients of food stamps play a
large role. Using the state of Wisconsin as his focus in the years 1980-1988, Rank examined trends
in participation in a variety of programs, including the FSP, Medicaid, and AFDC, through
interviews with fifty families receiving assistance, through a statewide longitudinal caseload
sample of welfare recipients, and observations of the welfare system and its operations. His
observations on the welfare "climate" over a period of years included selected quotations from the
press and speeches given by government leaders. In personal observations and interviews with
clients during their application process, Rank found that they experience real fear in doing
something new that requires significant literacy skills. Further, the loss of independence and the
required reliance on government largesse is frightening for many. These are components of stigma
in America with a long tradition in public opinion, according to Rank. The recipients of welfare
benefits whom Rank interviewed believed that they were treated differently than others in the
population. Two-thirds of his respondents believed that they were subject to others' disapproval as
. .elfarc recipients. As one respondent stated (Rank 1994,137-138):
You really do have to be a strong person to be able to use
Food Stamps and not get intimidated by how people treat you
when you use them. And even then rfs soil bard. You feel
people's vibes, you know, in the line. And the checkout people
are almost without exception rude, unless you really get to know
them. And I always feel like, "God, I'll be glad when I don't have
to use these." So every time you check out, they always have to
go up to the office to get change, so you got all these people
waitin' in line - it's like, you know, "These Food Stamp people."
19
Once about six weeks ago I turned to the woman behind
me and said, "I don't know, I have not once come up here and
bought something with Food Stamps where they didn't have to go
and get change for, like a five or something, that they had in the
drawer." She says, "Well, I guess it's just one of those ways that
they're not making it easy for you."
Rank found that food stamp recipients participate in the negative attitudes toward persons
receiving public benefits by complaining about their lack of initiative and their willingness to
accept charity.
Elliot Liebow (1993)
Tell Them Who I Am- New York: The Free Press.
In this ethnographic account of homeless women living in a shelter in a suburb of
Washington, D.C., Liebow described the difficulties in accessing appropriate help, including food
stamps. Liebow volunteered within the shelter providing a number of services in the routine of
daily life over a period of several yean in the 1980*3. At the same time, he observed and
interviewed dozens of women (with their understanding of the research and their cooperation)
about their lives and perceptions of public services, among other topics.
Liebow discussed issues faced by homeless women in depth. His approach to public
benefits and the possible nonparticipation among his respondents is couched in both philosophical
and observational terms. For example, he discussed the problem many homeless women have in
answering questions. Among the homeless people Liebow interviewed, this was a major issue and
a deterrent to entering public programs. Screening procedures of all kinds were abhorrent to the
women be interviewed, and acceptance of services often hinged on the ability to retain a degree of
privacy while being granted a benefit. Liebow saw the requirement to answer questions about their
lives a "price they paid for being powerless."
Liebow reflected on the perceptions of service providers in the midst ofeligible clients who
do not respond as expected (Liebow 1993,139):
20
To the rebuffed service provider or would-be friend, these
are simply more instances of people who cannot act in their own
self-interest, more instances of behavior unconnected to reality,
more instances of craziness among homeless women. There
seems to be little awareness that the failure of some women to
cooperate may be, at bottom, a prideful refusal to capitulate to
what they see as an oppressive system, and therefore a kind of
victory for them.
This dissonance between service providers and clients was cited by Cohen (1990) as well,
who reported that service providers, in this case mental health clinicians, could respond quite
positively to patients whom others had defined as difficult and unable to be helped when additional
training was offered. The health care workers' feelings of helplessness in this case were treated
effectively
Mary Belenky and her co-authors (1986) support some of Liebow's observations. They
examined women in formal educational settings and in informal service provision settings. Using
observational and interview methods, the authors identified characteristics of women's ability to
obtain information and process it One pattern they observed was common among women who
were silently obedient to authority. "While they feel passive, reactive, and dependent, they see
authorities es being all-powerful, if not overpowering...there is little evidence that the silent women
could imagine themselves actively listening to the authorities' ideas, understanding what they were
saying, and then choosing to obey" (Belenky, et al., 1986,28). This inability to process information
could lead to a reluctance to enter into situations in which interactive question and answer sessions
are required and would tend to support Liebow's observations.
Other authors observing clients' response or non-response to participation in services have
referred to the difference in needs between clients and providers. Nonpardcipation as an expressive
act of defiance is described in more academic terms by James C. Scott (1990). Passive aggression,
failure to produce documents, to keep appointments, and to behave in expected ways may reflect,
for some clients, a legitimate response to feelings of being on the fringe of society. Issues of
21
privacy are discussed in a 1994 report that examined the electronic transfer of food stamp benefits
(Casey, et al 1994). In roundtable discussion with FSP and WIC Program recipients, privacy issues
emerged as major concerns about participation in these programs. Curtis (1994a) also refers to the
expectation on the part of providers that clients should be "grateful" recipients of food services at
food pantries and within organizations that distribute food. Volunteers and staff expect that their
reward will be, in large part, found in the recognition of their efforts through interactions with
beneficiaries This plays into the powerlessness of the poor, underscored by the lack of choice in
food that is put before them. Proclaiming their status as "needy" through documentation required to
receive the benefits, those who would like to receive benefits are viewed suspiciously until they
prove their rightful position in the ranks of the certified.
Geraldine Olson and Jing-jian Xiao (Fall, 1991)
"Information-Seeking Characteristics of Hungry People." Journal nf Home Economies. 11-
16.
There is readily available information about consumers' information-seeking behavior, but
little information about how people on welfare get information when they seek emergency food. In
1988, the Hunger Factors Assessment Project of the Oregon Food Share Program randomly
selected 17 towns or cities in Oregon to collect this iformation. Surveys were sent to 944
respondents and 610 responded to the question: "Where do you hear about getting food help?"
Twenty-seven percent of the respondents used the food stamp office as a source of information
about emergency food. The authors reported that people who had used food stamps were more
likely than others to use additional sources of information about where to get emergency food. This
study indicated that the food stamp office itself may be an outreach mechanism through wjrd of
mouth, ofeligibility worker to client, or through posted information.
Gretchen Schafft, et al. (April, 1988)
"WIC Benefit Target Study: Phase I Report" Alexandria, VA: USDA, FNS.
In an observational study of benefit targeting among 18 WIC programs, beneficiaries
reported in focus group discussions reasons for nonparticipation among people they knew. The
barriers that they reported or that the researchers observed were subsumed under three categories:
informational barriers; logistical and management barriers; and, affective and attitudinal barriers.
Many potential clients did not know about the program or believed that they were not eligible.
Many thought they were not poor enough to use the WIC program. This rinding reflects that of
Coe in 1983. Others did not understand the age limitations of the program or understand that
women, in addition to children, were eligible, particularly during pregnancy. Citing Belenky's
work, these authors observed that many women were unable to access programs on their own
without waiting for an authority figure to reach out to them with information and with an invitation
to participate. Information also was not communicated because of language differences and
illiteracy in their native language, as well as in English. Simply translating materials was not
enough when this double literacy handicap was present
Logistics provided a problem for women with many children and no access to an
automobile. WIC offices that adhered to a standard schedule were not able to reach people whose
transportation and other obligations prevented them from coming into the office during particular
times of day. Some pregnant women or women with small children had to wait to be seen for long
periods of time. This was difficult for them and a discouragement to their participation. They
found that applicants spend, on average, almost five hours completing the application process.
Many chose not to complete the application because of the "hassle" involved: length of the process
or problems getting to the food stamp office, confusion about the process, long waits at the office,
and "unobtainable documents" required in the application process.
23
The WIC study also showed that clients and eligibility workers often do not understand one
another's perspectives (a further confirmation of the Liebow study). Adversarial relationships
occasionally developed as a symptom of power relationships, as well as differing values and needs.
At times, staff members were anxious about physical interactions with clients due to possible
infectious diseases, and in the health care setting, this caused changes in physical examination
schedules and procedures. As in the Liebow study, this research showed that privacy was an issue
for clients in the application and reapplication interviews which were often conducted in rooms .hat
were not soundproof.
Similar to Rank (1994) these authors found that dealing with the visibility of being a
recipient of a social welfare benefit was difficult for many WIC recipients. C shing their vouchers
at stores exposed them to the attitudes of the community that were sometimes negative. Other
issues that concerned them included disclosing one's personal status regarding immigration,
incarceration, or other information that could lead to drastic action, such as losing one's children,
being deported, fined and/or jailed.
To summarize the literature in this section of the paper, major findings of ethnographic and
observational research indicate that many potential clients of public benefit programs have a fear of
losing self-esteem by participating in the application process. This comes about through lack of
real skills, such as literacy and counting, but also as clients and/or applicants to public benefit
programs are treated in ways they perceive to be demeaning. Many also fear the label that will be
attached to them as recipients oftaxpayers' dollars. Also, some potential participants refuse to take
part in benefit programs because they do not want to give up iirforrriatjon they hold privately. Some
authors found that potential clients often waited for information to be brought to them rather than to
seek out programs that could help them and their families. This pattern of waiting passively for an
authority figure to tell them what to do may be more prevalent in women who also have problems
with literacy and feel excluded from routine social interactions in their communities.
24
IV. EFFICACY OF CLIENT ASSISTANCE ACTIVITIES
Deciding on vhat methods best reach potential clients and can assist them in applying for
the FSP or other benefits requires a two-step process. The first step is a needs assessment of the
population groups that he program aims to affect The results of such needs assessments indicate
what barriers to participation exist and suggest which activities might provide potential clients with
sufficient information about the program and its eligibility requirements. The secouu step in the
process of determining what methods should be used to reach clients and potential clients is an
evaluation of the efficacy of these methods, including their effectiveness and their costs. The
results of these studies can help the FSP or other programs decide where to place effort and use
resources.
Rigorous evaluations of client assistance activities are difficult to perform for a number of
reasons, but can be extremely useful to program planners if sufficient care is given to their design.
Studies that combine an overview of either single or extended programs and combine data
gathering from a number of sources and critical respondents offer a strong chance of producing
material that can be used for program improvements and policy initiatives.
Evaluations of client assistance activities often depend upon the efforts of program staff,
who may be untrained in evaluation methodologies, to assess their own activities or to provide
useable information for outside evaluators (Schafft and Randolph 1994; View 1994). Technical
assistance in evaluation is then an important part of the evaluation effort. This both provides skill-building
and promotes an effective outcome of evaluation that leads to future decision-making.
Many evaluations do not provide this extra support for program staff. Under these
circumstances, the efficacy of the client-reaching and assisting methods is often estimated rather
than analyzed in detail (AARP 1991b). Multi-site studies that evaluate such activities often show
25
great variation among sites, mitigating meaningful generalizations about the efficacy of client
assistance methods JKLSS (Schafft and Randolph 1994, View 1994, Peterson 1993, Silverstein and
Puma 1992).
The use of comparative figures, such as those produced in the 1993 Social Security Study
(Peterson), is very helpful in judging the value of activities designed to assist clients. Comparing
the results in participation rates assumed to be due to the intervention of particular methods of
client assistance with other matched demographic areas not using such methods or with the same
area in another time frame approaches an experimental design. Such designs are not always
available, however, because the researcher does not have access to comparative sites that can be
used for analysis nor comparative data from other time periods. This is often a result of geographic
boundaries that are not similar between the demonstration site and others or aggregate data that is
collected for different purposes and from different catchment areas (Schafft 199S). Researchers
also cannot control for other outreach activities that may take place independently and
simultaneously to those they are following (Siverstein and Puma 1992).
Although the literature reviewed here on effective methods to improve client access and
assistance offers only one example of a quast-exprrimental research design, many studies are able
to identify the unmet needs of potential clients as they attempt to gain access to various benefit
programs and suggest ways in which programs can become potentially more accessible.
James L. Peterson, et al. (1993)
"Phase I & II Analysis for the Social Security Administration's Outreach Demonstration
Projects." Baltimore, MD: Social Security Administration
In this study of Supplemental Security Income (SSI) outreach activities, researchers studied
and reviewed 22 demonstration projects conducted by nonprofit organizations to increase access to
the SSI Program for potentially eligible clients. Researchers used a combination of research
methods and incorporated into their findings statistical data on:
• Client characteristics and attitudes from intake interviews conducted within the 22 projects
and extracted from the files;
• Social Security data on the history of payments made to SSI recipients and to individuals;
• Information on the demographics, staffing, and workload of the Social Security offices in
specific locations; and,
• Monthly awards and their amounts for specific Social Security offices.
Textual data sources were more limited and consisted of materials in project files,
including financial, quarterly, and final reports, observations made by government project officers
in the form of site visit reports, and evaluations produced by the Social Security Administration
(SSA) personnel. All of the textual analysis was done on secondary data, not first-hand
observations or interviews Information on individual clients was extracted from standardized
intake interviews conducted at each of the projects by their own staffs and then aggregated into a
database.
At the time of the projects' inception, about 60 percent of the estimated eligible aged
individuals participated in the SSI Program. The evaluation of these projects included individual
evaluations of each she and analyses of aggregate statistics across all sites. The researchers were
able to compare applications and enrollments during the project period to those in the same period
prior to the project; they compared applications and enrollments with projected figures that would
have followed a trend line from previous years; and, they compared applications and enrollments to
matched sites over the same period oftime.
This research particularly looked at the procedure of doing client assistance and made
recommendations for improving the efficiency and effectiveness of this process. An interesting
comparison made of the barriers grantees expected to find, but did not, and those they did find
which they had not anticipated, indicated that an examination of client assistance activities over a
period of time can reveal important information.
27
The researchers found that "some outreach effort was expended on people who would have
found their way to SSA on their own, rather than on finding difficult-to-reach individuals who
needed extra attention and assistance in applying." This phenomena of shifting outreach targets to
those easier to reach was documented by Schafft and Randolph (1994) as well.
The authors were able to identify barriers by characteristics of the population such as race,
sex, age, and size of community. Data showed that lack of information about the program and its
eligibility requirements and a lack of knowledge of the existence of the program were two major
barriers, but that stigma about receiving SSI was not as strong a disincentive as had been expected.
Gender played virtually no role in enrollees' decisions to use SSI. Age and disability were
significant barriers, as was the lack of transportation for some rural minorities, particularly Native
Americans. Many groups, including those with specific illnesses, the homeless, and substance
abusers often needed payee representative services and did not always understand this option.
The variability among projects in their ability to reach potentially eligible clients would
have been "washed out" in any report which aggregated figures from all demonstration project
sites. In this study, these individual outcomes and variations in the quantity and quality of data
gathered is clear from detailed case study materials from each she appended to the body of the
document Cost analysis of the projects and some estimates of cost-effectiveness are included. The
fact mat outreach is an activity whose costs are often hidden within agency overhead also was noted
byBateman(1991).
This study found that the most significant barrier to SSI participation was lack of
information about how to apply for benefits. This was the only barrier mentioned by people in
every project On the other hand, expected barriers, such as fear or distrust of the government
program, did not occur in the way they had been predicted by project staff in their proposals. No
barriers were found to be related to gender, but "disability-related barriers, isolation-related
barriers, and information-related barriers rise with age."
Projects contributed approximately three percent of new applicants to the SSI Program in
their local areas, but the authors believed that indirect effects of the projects may account for a
larger influx of clients over the coming years. They also found mat outreach is a more difficult
process than the projects had anticipated, and that to conduct client assistance required strong
management, particularly when projects were composed of components that were spread
throughout a large area. Case management techniques, rather than simply attempting to enroll
clients immediately in a benefit program, appeared to be necessary to the success of reaching the
end goal of participation. Using client lists from another program to conduct outreach is an
effective tool, as is disseminating information and using a toll-free telephone number followed by
individualized client assistance. The public-private cooperation was both critical to the success of
projects and more difficult to achieve than had been anticipated due to "organizational imperatives
and philosophies that impede the development of such a relationship."
USGAO (October, 1990)
"Social Security: District Managers' Views on Outreach for Supplemental Security
Income Program." GAO/HRD-01-19FS.
This is a report of a telephone survey in which district office managers in 640 offices
nationwide were questioned about outreach activities in the SSI program. The motivation for the
study was a belief that potentially eligible recipients of this benefit did not know oftheir eligibility.
Findings of this study indicated that there was no clear understanding of the intended
outcome of the outreach efforts that had been put in place in SSI offices, although managers felt
that outreach was important They also believed that they were doing enough outreach, although
they spent only ten hours a month doing it and had no measures of their success in reaching
potential enrollees. This study indicated that outreach was considered necessary, but there was no
effort expended at the local level to determine its effectiveness.
29
American Association of Retired Persons (AARP) and Center on Budget and Policy
Priorities (1991a)
"Supplemental Security Income (SSI): A Review of Participation Rates and Outreach
Activities." Washington, D.C.: AARP.
AARP and Center on Budget and Policy Priorities (1991b)
"Participation Rates by the Elderly in the Medkaid Program: A Review of Participation
Rates and Outreach Activities." Washington, D.C.: AARP.
A series of demonstration projects and reports emanated from the American Association
for Retired Persons (AARP) in the early 1990's on the subject of outreach. AARP staff, in
conjunction with program personnel, undertook projects and studies of outreach in SSI, Medicaid,
the FSP, and the Low Income Home Energy Assistance Program from which they produced
manuals to help AARP volunteers participate in outreach activities to improve access to potential
clients. Authors of this series reviewed program statistics, studied the literature on participation
and problems of access, and interviewed State officials.
The AARP developed a model of outreach to older people that stressed producing and
disseminating information about benefits of these public programs and the eligibility criteria for
them. They used a multi-media approach and made presentations in community centers. Rather
than concentrating on a single benefit program, AARP volunteers presented the benefits of each of
the four programs and their individual requirements for participation. Overall, the AARP found
that this multi-program approach to outreach received a positive response from older people.
Outreach strategies that worked for one program, they believed, could be transferred to another.
They recommended that outreach be made a mandated aspect of all public benefit programs.
AARP found that outreach has three components: dissemination of program information,
client assistance with applications, and reduction of negative attitudes toward receiving public
benefits. Hotlines, establishing referral networks, and reducing the literacy level of outreach
materials to a fifth to seventh grade level and in several languages are helpful tools in reaching
potential clients for public benefit programs. With regard to the FSP, AARP suggested that States
30
should target elderly people with their outreach materials and inform them through their SSI
mailings of their potentially eligible status in the FSP. Mailings should also be targeted to rural
areas. The authors also suggested that policy changes within the FSP be inaugurated, such as
indexing the minimum benefit for inflation, paying small benefits on a quarterly basis, and
simplifying the medical deduction so that more older people could use it Person-to-person
outreach using community leaders to disseminate information might also be important. They
suggested that a single point of entry be established between SSI programs and the FSP. Moving
application sites to agencies close to those in need as outreach locations was effective, as well as
providing assistance to people making these applications.
Gary Sihverstein and Michael Puma (1992)
"Evaluation of the AARP Outreach Demonstration Projects. Bethesda, MD: Abt
Associates.
This outside evaluation of the AARP demonstration projects, which were begun in 1989 to
increase participation among the elderly in four assistance programs, Medicaid, Low Income Home
Energy Assistance Programs, SSL and the FSP, found that the 20 demonstration sites were able to
conduct outreach, but were unable to show statistically significant results. In a two-year effort, an
impact analysis of 16 of the sites produced a great deal of anecdotal evidence that outreach
activities attracted elderly persons in need of public assistance, but failed to reach isolated, elderly
persons who are perhaps most in need.
The outreach was conducted through broad-based media efforts. The demonstration sites
tracked calls to their toll-free numbers following public service announcements or mailings of
information. Telephone calls requesting appointments and information were more numerous
following periods of outreach activity in communities. The messages helped to eliminate incorrect
information and reduced negative attitudes toward participation.
31
The authors reported that broad-based campaigns were often targeted to people who were
ineligible, already receiving benefits, or were unwilling or unable to follow up on the application.
Many may have been previously enrolled. The universe of those to whom the information would
be particularly useful was much smaller than the outreach targeted. As a result of this observation,
the authors recommended that cost-effectiveness of outreach could be enhanced by targeting certain
communities that have large numbers of low income residents and where no other outreach is being
undertaken. The authors also found that local demonstrations require flexibility in developing site-specific
outreach and establishing appropriate activities and contacts with other agencies, as well as
engaging the caseworkers in public assistance offices. They urged that client assistance with
applications be a part of any outreach program and pre-screening be used as a targeting device in
on-going outreach. Finally, applications should be available at a neutral site to reduce the effects of
stigma among the elderly.
Wendee Wechsberg, et al., (1993)
"A Comparison of Injecting Drug Users Reached Through Outreach and Methadone
Treatment" Journal ofDrug Issues. Inc. 23(41.667.687.
In this study, researchers compared certain characteristics of injecting drug users (IDUs)
contacted either through outreach or through their participation in methadone treatments programs
in four cities. Both the outreach and methadone treatment programs were part of the National
AIDS Demonstration Research (NADR) program. The characteristics of both groups were also
compared with a national probability sampling derived from the 1990 National Household Survey
on Drug Abuse.
The authors reviewed the literature about outreach to patients with AIDS and identified
common methods, such as sending ex-addicts as outreach workers into areas where drug use is
known to be prevalent, using peer education programs among teenagers, and providing
motivational messages for bringing people into treatment
32
Both the outreach IDU population and the IDU population in methadone treatment were
interviewed using structured, standardized protocols. The authors found that outreach accessed a
population with different characteristics from those who were in methadone treatment programs.
The outreach clients were older when they first injected drugs, had used drugs more extensively and
for a longer period, and were less likely to have been in treatment previously. The authors conclude
that outreach presents "attractive and appropriate alternatives to those groups that are underserved
and most vulnerable." Recognizing differences within drug user populations provides an "impetus
for continued innovation and refinement of new and current strategies." This is one of the few
studies that compares clients reached through outreach with populations that come into programs
through other means. This might be a useful approach in other areas of interest Many studies of
AIDS and drug prevention outreach programs report on a single outreach method: person-to-person,
street-based approaches. Generally, they provide more description of the activities and
settings for the purpose of understanding the context of exposure and usage than evaluation of
outcomes (Weeks, ed. 1991).
Louise Warwick, et aL (1992)
"Evaluation of a Peer Health Worker Prenatal Outreach and Education Program for
Hispanic Farmworker Families." Journal of Community Hearth. 17(1).
Hispanic migrant farmworker families received prenatal education from peer health
workers in this study. This approach, similar to that used in drug and AIDS client assistance
programs, stems from programs initiated during the Great Society in the 1960's. High health care
costs continue to make the approach attractive to health care planners.
The project was undertaken in three migrant farmworker communities in the border area of
Arizona near Mexico. The project recruited local women to further recruit and train other women
in 12-week courses on prenatal health. The evaluation of this ten-month demonstration project for
33
low-income Mexican and Mexican-American women used both descriptive and survey methods to
estimate the effects ofthe intervention on the clients and the peer health workers.
The evaluation found that the women studied shared many of their views on childbirth and
subsequent care of newborns and young children with the peer health workers, allowing medical
staff to understand the cultural context of prenatal and neonatal indigenous practices. Women
attended classes and reported that they had incorporated the material presented to them into their
plans for child care. The health care peer workers fulfilled their obligations under the
demonstration, and they reported that the women were very pleased to take part in the outreach
activities. The staffwere satisfied, therefore, with the results ofthe demonstration.
The study is more successful as a description of cultural expectations of researchers and
clients than a careful study of the effectiveness of outreach using convincing data. It does
underscore the need to carefully record clients' perspectives on service provision and how these
perspectives support or interfere with the goals and outcomes of programs planned for them.
In summary, this section which reviews client assistance methods shows that lack of
information about programs is problematic, but that providing information alone, without case-management
assistance, may not be sufficient to make benefit programs available to some potential
clients. The required attention to hard-to-reach cases is not inexpensive and must be carefully
targeted to be efficient Again, changing target groups to reach those easier to serve makes client
assistance appear to be effective, but may bring clients who would have enrolled OP their own into
benefit programs. Public-private partnerships are important in reducing negative attitudes ofclients
toward programs and reducing the costs of client assistance.
V. SUMMARY AND CONCLUSIONS
The literature reviewed in this piper addresses nonparticipation in government benefit
programs using:
• Analyses of answers people have given in household surveys;
• Extrapolations from FSP participation and estimated eligible population data; and,
• Observations of and interviews with people living in poverty.
Nonparticipation in some of the studies has been treated as the inverse of participation, in
others as an aside to other themes, and in still others as the subject of secondary analyses of
household surveys. From this selection of literature, one can better understand the dynamics of
participation in the FSP, reasons people give for their nonparticipation, and the disjunction between
the requirements of government benefit programs and the lives of people who live on the margins
oftheir own society.
Identifying reasons for nonparticipation is an important way to provide guidance to the FSP
about equitable distribution of food stamps. This information can also provide a better
understanding of specific client assistance methods that could be helpful in certain local food stamp
offices. Survey literature, based on a microsimulation model, indicate that nonparticipation is
associated with people who are older, have fewer children, less need, and more financial resources.
On the other hand, people with greater need, those who are African American and other non-white
minorities, those with more children, and those who are eligible for larger benefits are more likely
to participate in the FSP.
Potential clients have offered as reasons for not applying for benefit programs, their lack of
information about their eligibility and how to apply, the amount of the benefit not being worth the
hassle entailed in the application process, and not wanting to assume the stigma ofbeing a recipient
of public benefits at the taxpayers' expense. Many of ttie reasons for nonparticipation, such as
illiteracy and cognitive confusion, are not issues that will be discussed openly with a survey
35
researcher, but can be documented through observational studies and ethnographic fieldwork. On
the other hand, when ethnographic and observational studies address nonparticipation in benefit
programs, it is usually an aside or an addendum to another agenda (Rank 1994; Liebow 1993;
Belenky 1987).
Issues identified as reasons for nonparticipation are often large categories and lack specific
content Why do people not know about the FSP? Why are they confused about their own
potential eligibility? When they identify administrative problems as being disincentives to
participation, are they speaking from their own experience or from hearsay? If the food stamp
office is too far from where they 'ivc, do they travel equal distances for other services? A careful
delineation of the conicm of the reasons given for nonparticipation is not available in the literature
reviewed here.
Identifying reasons for nonparticipation by client group is helpful and can provide a better
understanding of specific client assistance methodologies that can inform the practices of certain
local food stamp offices. On the other hand, the development of this literature carries an inherent
danger of stereotyping "the homeless" or "ihfi elderly" in ways that ignore the immense
heterogeneity within groups. Factors that keep people from taking part in benefit programs in
which they are eligible may be found in quite separate target groups.
There are few studies of the interaction of clients and programs in which programmatic
changes are suggested that might enhance access. A study of eligibility workers as gatekeepers
would be useful in this regard as well as an analysis of food stamp rules and regulations and their
effect on enrollment and participation. The locus of barriers to access is still difficult to identify
from existing literature. For instance, when respondents indicate that they do not wtnt the "hassle"
of the application process, one does not know if the problem is within themselves or in the
procedures and interactions in the food stamp office. Addressing such issues might provide policy
and program guidance.
Evaluation of client assistance programs produces various indications of methods that are
effective. While no study to date has produced statistically validated measures of the effectiveness
of specific methods of conducting client assistance, many have provided valuable information
about these activities. It is in the descriptive narratives of clients in service provision settings that
an understanding of their problems of access will become more clear. A variety of studies using
observational and ethnographic approaches would complement the more readily available reviews
of statistical data on participation.
The literature is fruitful in discussions of nonparticipation in the FSP. Studies based on
household surveys have provided broad descriptions of those most and least likely to participate in
public benefit programs. Studies of actual demonstrations have depended upon the review of
secondary data for narrative accounts of client interviews, but have enhanced their findings with
statistical program data. Ethnographic and observational studies have provided primary data from
clients through both self-reporting and researcher observation, but have not been able to develop
alternative measures of effectiveness of methodologies for client assistance. With the limitations of
each approach, the combination of literature from a variety of studies and reviews helps the reader
understand the complexity of the issue of nonparticipation and use a combination of methods to
understand the phenomenon better.
37
BIBLIOGRAPHY
Allin, Susan and Harold Beebout. "Determinants of Participation in the Food Stamp Program: A
Review ofthe Literature." Alexandria, VA USDA FNS, November 19S9.
American Association of Retired Persons (AARP) and Center on Budget and Policy Priorities.
"Supplemental Security Income (SSI): A Review of Participation Rates and Outreach
Activities." Washington, D.C.: AARP. 1991a.
AARP and Center on Budget and Policy Priorities. "Participation Rates by the Elderly in the
Medicaid Program: A Review of Participation Rates and Outreach Activities." Washington,
D.C.: AARP, 1991b.
AARP and the Center on Budget and Policy Priorities. "Food Stamp Program: A Review of
Participation Rates and Outreach Activities." Washington, D.C.: AARP, 1991c.
AARP. "Public Benefits Outreach Conference Report" Washington, D.C.: AARP, 1992.
Bateman, Peter. "The Economic Aspects of Outreach." Practicing Anthropology- 13(2), 1991:
10-13.
Becbout, Harold, et al. (eds.). "Food Stamp Policy Issues: Results from Recent Research."
Washington, DC: Mathematica Policy Research, Inc., 1990.
Belenky, Mary Field, et al. Women's Wavs of Knowing. New York: Bask Books, Inc., 1986.
Casey, J.T., et al. "Data Privacy Issues for Food Benefit Programs; Final Report." Washington,
D.C.: Price Waterhouse, 1994.
Coe, Richard. "Nonparticipation in Welfare Programs by Eligible Households: The Case of the
Food Stamp Program." Journal of Economic Issues. 17(4) 1983.
Cohen, Neal L. "Stigma is in the Eye of the Beholder A Hospital Outreach Program for
Treating Homeless Mentally 111 People." Rochester, MN: The Menninger Foundation, 1990.
Curtis, Karen. "May I See Your Identification, Please?: The Changing Face ofEmergency Food
Assistance in Delaware." Paper presented at the Society for Urban Anthropology meetings,
Atlanta, GA, December 2,1994a.
Curtis, Karen. "Food Stamp Participation by Emergency Food Recipients." Practicing
Anthropology 16(4), 31-34, 1994b.
Danziger, Sheldon H., Gary D. Sandefur and Daniel H. Weinberg. Confronting Poverty.
Boston: the President and Fellows of Harvard College, 1994.
DeVauh, Margorie L. and James P. Pitts. "Surplus and Scarcity: Hunger and the Origins of the
Food Stamp Program." Social Problems, 31(5) 1984: 545-557.
Liebow, Elliot Tell Them Who 1 Am New York: The Free Press, 1993.
Martini, Alberto and Susan Allin. "Did the Characteristics of Food Stamp Program Entrants
Change During the 1990-91 Recession'7" Alexandria, VA: USDA, FNS, December 1993.
Nichols, Abigail C. Internal Memorandum to Dustin J. Van Vleet, Midwest Regional Director
USDA, FNS, 9/26/89.
Olson, GeraJdine 1. and Jing-jian Xiao. "Information-Seeking Characteristics of Hungry
People." Journal of Home Economics- Fall, 1991:11-16.
Peterson, James L., et al. "Phase I & II Analysis for the Social Security Administration's
Outreach Demonstration Projects." Baltimore, MD: Social Security Administration, 1993.
Rank, Mark Robert Life on the Edge. New York: Columbia University Press, '994.
Scott James C. Domination and the Am of Resistance New Haven: Yale University Press,
1990.
Schafft, Gretchen, et al. "WIC Benefit Target Study: Phase I Report." Alexandria, VA: USDA,
FNS, April 1988.
Schafft, Gretchen and Frances I. Randolph. Innovative Community Rased Services for Older
Persons with Mental Illness Rockv.lle, Maryland: NIMH, November 1994.
Schafft, Gretchen. "Barriers to Participation in the Food Stamp Program Identified through
Research Demonstration Project Grants." (unpublished) 199S.
Sirverstein, Gary and Michael Puma. "Evaluation of the AARP Outreach Demonstration
Projects." Bethesda, MD: Abt Associates, 1992.
Trippe, Carole. "Estimating Rates of Participation in the Food Stamp Program: A Review of
the Literature." Alexandria, VA: USDA, FNS, 1989.
Ti ipr-., Carole and Pat Doyle. "Food Stamp Participation Rates: January 1989." Alexandria, VA:
USDA, FNS, 1992.
Trippe, Carole and Julia Sykes. "Food Stamp Program Participation Rates: January 1992."
Alexandria, VA: USDA, FNS, 1994.
U.S. Conference of Mayors. Barriers to Participation in Benefit Pmy.-inn Washington, D.C.:
U.S. Conference of Mayors, 1988.
USGAO. "Social Security: District Managers' Views on Outreach for Supplemental Security
Income Program." GAO/HRD-01-19FS, October 1990.
U.S. Senate, Committee on Agriculture, Nutrition, and Forestry. The Food Stamp Program:
Hi«~y n*«<TfTlfini linn fflWl Optifflll" Washington, D.C.: U.S. Government Printing
Office, April 198S.
39
View, Virginia and Kim J. Amos. Living and Testing the Collaborative Process: A Case Study
Of Community-Baaed Services Integration- Arlington, Virginia: National Center for Clinical
Infant Programs, 1994.
Warwick, Louise H., et al. "Evaluation of a Peer Health Worker Prenatal Outreach and
Education Program for Hispanic Farmworker Families." Journal of Community Health. 17(1),
1992.
Weeks, Margaret R. "Community Outreach Prevention Effort" Hartford, CT: Institute for
Community Research, 1991.
Wiecha, Jean L., Johanna T. Dwyer, and Martha Dunn-Strohecker. "Nutrition and Health
Service Needs Among the Homeless." Public Health Reports 106(4), 1991:364-374.
Wechsberg, Wendee M., et al. "A Comparison of Injecting Drug Users Reached Through
Outreach and Methadone Treatment" Journal ofDrug Issues. Inc. 23(4), 1993:667-687.
40