Experiences of Selected Federal Social Welfare Programs and State LIHEAP Programs in Targeting Vulnerable Elderly and Young Child Households
Analysis of Social Programs – Findings and Recommendations
This section of the report presents information on the findings and recommendations from the research on targeting in other Federal social welfare programs. It describes the procedures used for identifying literature on targeting of elderly households and young children households, and summarizes the literature for those social programs that have addressed the issues of targeting. It looks separately at general population programs, programs that specifically serve the elderly, and programs that specifically serve young children. It finds that only a few programs have focused significant research efforts on program targeting. However, the information available from those programs that have studied targeting furnishes valuable insights for the LIHEAP program.
Overview of Method
The literature review included a variety of sources, including relevant reports available online from Federal and State program websites, evaluator websites, and other complementary sources. This search yielded a great deal of information on program targeting activities, as well as bibliographic references for other research. The search also included accessible databases for academic peer-reviewed literature. The academic articles were less helpful due to their focus on finding general associations between demographics and participation rates rather than exploring details of targeting strategy success and failure.
This search looked both for subject coverage across sources and lines of study within groups of sources. For example, a wide variety of sources and programs touched upon stigma as a barrier to enrollment, but a narrower set of literature discussed the relatively new Medicare Savings Programs in order to study their immediate effects or SCHIP in order to address the possibility of continued funding. Complying with standard procedure for a literature review, the search continued until a particular inquiry was exhausted or repeated themes in the information gathered were found.
Research focused on the major programs serving elderly households and households with young children. The general population programs included: Food Stamps (FSP), Medicaid, Housing Vouchers, and the Weatherization Assistance Program (WAP). The programs specifically serving elderly households included: the Medicare Savings Programs, SSI, and the Nutrition Programs for the Elderly (NPE). The programs specifically serving households with children included: EITC, WIC, Head Start, and SCHIP. There was a wide range in the levels of information available on these different programs. There is very little written about some programs’ targeting activities, but a great deal on others. This initial finding is in concert with the findings of GAO Report discussed in Section III. The literature review found:
General Population Programs
- FSP estimates a participation rate, uses it as a performance measure, includes it in its performance reports, and has a substantial body of literature exploring all the different facets of program participation and targeting;
- Medicaid and SCHIP taken together as jointly administered programs by CMS have recently been the subject of a series of evaluations and reports.
- The Housing Choice Voucher Program does not estimate a participation rate, but does monitor the “utilization rate,” or the rate at which voucher recipients are able to redeem benefits. It therefore does not have an accompanying literature on participation barriers and strategies.
- Searches for information on WAP did not yield very much information.
- Programs Serving Elderly Households
- The Medicare Savings Programs, namely the Qualified Medicare Beneficiary (QMB) and Specified Low income Medicare Beneficiary (SLMB) programs, have a somewhat moderate set of studies exploring the response of the elderly population to these relatively new programs.
- SSI has a much smaller set of literature, likely owing to the fact that the SSA does not regularly estimate participation rates.
- The Elderly Nutrition Program, like the Weatherization Assistance Program, was not included in the GAO analysis and generates few studies, likely resulting from a small budget.
- Information on the EITC participation rate was available, but there was no literature that examined the reasons for differences in participation by demographic groups.
- WIC has limited targeting literature, likely owing to the fact that participation rates are not used as a performance measure.
- For Head Start, very little has been written on participation rates.
- Programs Serving Households with Children
Given these assets and limitations, the comparative analysis is focused on two programs for the general population: Food Stamps, with analysis for both populations, and Medicaid, with analysis for children only. The analysis also focuses on three programs targeted to households with elderly members or children, i.e., the Medicare Savings Programs, SCHIP, and WIC.
General Population Programs
Food Stamp Program
The FSP is a Federally-funded entitlement program providing low income households the means to buy food at retail stores through the use of electronic benefit cards. The USDA administers the program and is responsible for program design, including levels of availability, eligibility and benefits. As a means-tested program, FSP requires participants to meet certain income and resource thresholds. Participants must also be U.S. citizens or eligible noncitizens, and all “able-bodied adults without dependents” must be employed or take part in an employment and training program. In the late 1990s, the number of households participating in the FSP and the percent of eligible households participating declined significantly (Bartlett, 2004).8 As a result, there has been considerable research on FSP participation in the last decade.
a. FSP Participation
In 2001, the Urban Institute estimated the overall FSP participation rate to be between 46 and 48 percent. They estimated the participation for households with elderly members to be between 27 and 28 percent and for households with children to be between 55 and 57 percent (GAO, 2005). A recent study by Mathematica Policy Research (Wolkwitz, 2007) gives a similar overall estimate of the participation rate at about 48 percent in 2001. The Wolkwitz paper also produced participation rates for FY 2005, estimating overall participation rates at 59 percent for households and 65 percent for individuals.
Studies agree that FSP participation rates vary by group. Rates are highest for households on public assistance (i.e., TANF and SSI) and for households with incomes below poverty. Rates are lowest for elderly households, eligible noncitizens, and households with earnings. (Wolkwitz, 2007) Households applying to the program are also more likely than non-applicants to be younger, single-parent families in worse financial condition (Bartlett et al, 2004). Wolkwitz (2007) finds that the program reaches the neediest eligible individuals because 65 percent of all eligible individuals are being served with 80 percent of the potential benefits distributed. That means that the households who are eligible for the smallest benefits (i.e., those with the least need) are the least likely to participate.
Most households participate in the FSP as a result of either a change in income, such as a change in employment status, or a change in family composition, such as the death of a spouse. Households will first rely on their personal networks and then use more informal social programs, such as food pantries, before turning to the FSP. Family and friends can be instrumental in either encouraging or discouraging individuals’ or households’ participation in the program (McConnell and Ponza, 1999).
Overall, elderly households see many benefits to the program, particularly as a critical form of food and economic assistance. By providing funds for food, the program frees up other money to pay for medical, utility, and housing expenses. The program also allows elderly households to buy healthy food such as fruits and vegetables or ethnic items unavailable from food pantries. Elderly individuals also appreciate the added independence that comes from being able to shop and cook for themselves rather than relying on their families (Gabor et al., 2002).
b. Barriers to Enrollment
Researchers describe five sets of barriers that eligible populations face when trying to enroll in the FSP. Most of the findings in the literature point to barriers faced by all groups, but some specifically point to barriers faced by households with elderly members or children.9 The five main categories are: 1) eligibility rules and confusion about those rules; 2) application procedures; 3) office procedures; 4) personal feelings about the program; and 5) lack of awareness.
i. Eligibility Rules
Many potential applicants are confused about eligibility rules (Dion and Pavetti, 2000; Gabor et al., 2002; Bartlett et al., 2004). Eligible individuals might believe they are ineligible if they have certain assets like a bank account or car, are employed, have been denied in the past, or are not U.S. citizens. Elderly households in particular sometimes believe that in order to receive Food stamps an applicant needs to have a child or be on TANF. They also might believe that they have to sell their house or car or not live with their children in order to become eligible for the program. Some of this confusion is likely due to the association of FSP with other programs through coordinated outreach efforts. One model estimated that coordinating outreach efforts with Medicaid/SCHIP reduced the likelihood that an apparently eligible household believed it was eligible by 16 percentage points (Bartlett et al., 2004).10
Elderly households who are immigrants face an additional set of worries surrounding their own or their families’ immigration status. Some think there is a three to five year waiting period to obtain benefits. They might also think participation in the program will affect their eligibility or their children’s eligibility to become naturalized. Perhaps most frightening is the prospect that if they join the program, their sponsor will have to pay all the money back or “will be damaged or hurt” For all elderly households, most information about the program circulates through family and friends, so without active intervention, there is ample opportunity for misinformation (Gabor et al., 2002).
ii. Application Procedures
The second set of barriers relates to application procedures. Several studies show the use of identifying technology, such as fingerprinting, significantly decreases the likelihood that applicants will complete the process (Bartlett et al., 2004; Ratcliffe et al., 2007). In fact, Bartlett (2004) estimate that fingerprinting applicants causes a 23 percentage point decrease in the number of applicants that complete the process, while Ratcliffe et al. (2007) estimate that the use of fingerprint imaging reduces food stamp receipt by 15 percent for households below 175 percent of the poverty threshold. Studies suggest elderly individuals tend to deeply distrust this process due to a perceived loss of privacy and dignity (Gabor et al., 2002; Zedlewski et al., 2005). Even without this technology, elderly individuals feel very uncomfortable giving out their personal information such as their Social Security number, and are particularly frustrated if this information can be obtained by office staff through other means, such as, through their participation in other programs (Gabor et al., 2002).
Others point to the difficulty of obtaining multiple verification documents and the generally overwhelming amount of paperwork (Dion and Pavetti, 2000). Elderly participants have a particularly difficult time getting documents from third parties, such as landlords or bank officials. The complexity of the FSP medical deduction, often relevant to elderly applicants, means case workers are reluctant to process that portion of the application, leading to lower benefits. Those elderly households with limited English proficiency face the additional barrier of poorly-done or overly-difficult translations (Dion and Pavetti, 2000; Gabor et al., 2002).
Many applicants are also dissuaded by overly personal questions on the application form and a perception that the benefit will not compensate for these difficulties. Some elderly households argue the questions on the application for the FSP are irrelevant and designed to make the applicant feel ashamed for applying. Interaction with office staff might add to this loss of dignity, particularly if staff are rude or acting “like it’s coming out of their own pocket” (Gabor et al., 2002, p 47). Studies argue this kind of behavior on the part of office staff can make applicants feel like they are being judged as undeserving or even criminal for receiving benefits (Gabor et al., 2002). One study estimated that an additional positive response by a supervisor on a three response index designed to measure the positivity of food stamp office staff is associated with a 10 percentage point increase in the likelihood of application completion (Bartlett et al., 2004). This theme carries over into the section below on stigma.
For elderly households, the primary concern is the low expected benefit coming from the program. Many elderly nonparticipants assume they will only receive $10 a month if they already receive Social Security. Though the actual benefit is greater than this for the average elderly applicant, this perception is a significant barrier when combined with other obstacles to enrollment. To these households, the benefit just does not seem worth the hassle (Gabor et al., 2002; Zedlewski et al., 2005). This barrier is reflected in the very low participation rate for individuals receiving the minimum benefit (about 15 percent), as shown in Table 2 below.
Table 4-1 - Participation Rates for Individuals by Household Benefit as a Percentage of Maximum Benefit, FY 2005
Percent of Maximum Benefit
Minimum benefit (≤ $10)
1 – 25%
26 – 50%
51 – 75%
76 – 99%
SOURCE: Table 4 (Wolkwitz, 2007)
* Participation rates over 100 percent are due to reporting errors in the CPS.
The participation in non-entitlement programs is sometimes called a “coverage rate” (GAO, 2005) in order to acknowledge that the goal of these programs is not to reach all eligible households, but only those who can be served within given funding limits.
In an entitlement program, every household that meets the program eligibility requirements must be served by the program, no matter what the total cost. SSI is an example of an entitlement program. In a non-entitlement program, there is a limit to program funding. So, households that meet the eligibility requirements can receive benefits only as long as funding is available.
A respondent to the CPS-ASEC survey is interviewed in February, March, or April and furnishes information on annual income for the prior calendar year. A respondent to the SIPP survey is interviewed three times a year and furnishes information on monthly income for each of the previous four months. Since participation in many social welfare programs is based on retrospective income for the last month or three months, or on prospective income for the next month or three months, the SIPP does a better job of estimating program eligibility.
The decline in both the number of participants and the participation rate was of concern to program managers. If the number of participants had declined because fewer households were eligible for the program, program managers might have considered that a positive outcome. However, with participation rates falling, program managers were concerned that the program was not reaching households that needed assistance.
The studies found do not distinguish in their discussion of barriers between households with older children and those with younger children.
The model parameters represent the change in the number of percentage points. For example, in this case the change in the percentage of households that believed they were eligible for the program might have dropped from 66 percent to 50 percent, a change of 16 percentage points.
iii. Office Procedures
Elderly individuals face many of the same barriers in office procedures as other applicants, but often to a greater extent. While most applicants have difficulty finding blocks of time to go through the extensive process, elderly participants must cope with the physical difficulty of waiting in sometimes excessively long lines. Again, those elderly households with limited English proficiency face an additional barrier of being unable to find someone who speaks their primary language (Gabor et al., 2002; Zedlewski et al., 2005).
Many households experience difficulties getting themselves to the food stamp office, particularly during restricted weekday-only hours when many are at work (McConnell and Ponza, 1999; Bartlett et al., 2004). Families with children face an additional barrier when requested to not bring children into the office, an office practice reduces the number of applicants who finish the application process by an estimated 21 percentage points (Bartlett et al., 2004). This adds the additional expense and hassle of finding child care while attempting to access basic services. The attitudes and behaviors of office staff also come into play in encouraging or discouraging enrollment. Some studies discuss the importance of a household’s prior negative experiences with the program staff as dissuading it from returning to the office (Dion and Pavetti, 2000; Bartlett et al., 2004).
iv. Personal Perceptions
Households also carry with them strong feelings about what it means to themselves and to others to use the FSP. Some feel that using the program means they relinquish their personal independence, a feeling only strengthened by the overly intrusive application procedures discussed above (Bartlett et al., 2004). Others feel that using a program as charity reflects poorly upon them; they do not want to look as if they need help. They may feel embarrassed to be seen shopping with the benefits card or want to hide their participation from friends and family (Dion and Pavetti, 2000; Zedlewski et al., 2005).
For many seniors, the perceived loss of independence they feel that comes with enrolling in the program is again not worth the small benefit they think they would receive. Seniors also do not want to look as if they need help after having worked and paid taxes their whole lives. In some situations, stigma was a family concern. If an elderly person was seen asking for assistance, the family might look bad for not being able to keep their need hidden to avoid subjecting their families to embarrassment (Gabor et al., 2002; Zedlewski et al., 2005). However, though many studies agree stigma exists, some downplay its importance in dissuading households from actually participating in the program (McConnell and Ponza, 1999).
A final set of barriers involves a lack of awareness (McConnell and Ponza, 1999; Zedlewski et al., 2005). Some households do not know how to apply for benefits, while some have never heard of the program. This was only mentioned by a few studies as an important factor in participation; however, many programs spent a good amount of time on outreach. It is unclear whether the small effect of awareness on determining participation was a result of these outreach efforts or whether the two are unrelated, though it has been shown that employing a larger number of outreach methods increases the likelihood that a household thinks it might be eligible (Bartlett et al., 2004).
c. Strategies for Enrollment
This study identified four other key studies that seem to give the LIHEAP program useful information on how to increase the participation of elderly and young child households in LIHEAP.
- Food Stamp Program Access Study (Bartlett, 2004) – This study worked directly with 109 local FSP offices to collect information on office practices, collect information on applicant and recipients households, and conduct surveys with eligible nonparticipants.
- Food Stamp Outreach Grant Study (Zedlewski, 2005) – This study examined the experiences of 19 FSP offices that received grants to conduct outreach to low income households. The project durations varied by site from 10 months to 24 months.
- Elderly Nutrition Demonstrations Study (Cody and Ohls, 2005) – This study evaluated six elderly nutrition demonstrations that employed three different demonstration models: simplified eligibility, application assistance, and commodity alternative benefit.11
- Senior Views of the Food Stamp Program (Gator, 2002) – This study conducted focus groups with eligible and recipient seniors.
The Bartlett study found that program awareness, program understanding, program stigma, and program application barriers reduce FSP participation. The first three issues can be addressed by program outreach, while the last is affected by policies implemented by the State and local FSP offices.