Abstract
A comprehensive analysis of the 2009 Current Population Survey–Annual Social and Economic Supplement (CPS-ASEC) provides the first snapshot of the involvement of people with disabilities in a broad array of government programs. Focusing on the working-age population, this analysis reveals that people with disabilities represent one third of the persons who participate in safety-net programs. Among persons with disabilities, 65% participate in a safety-net program, compared with 17% of persons without disabilities. Among Social Security Disability Insurance (DI) beneficiaries, 38% participate in safety-net programs other than DI, Supplemental Security Income (SSI), Medicaid, and Medicare. Furthermore, results suggest that only 3% and 8% of low-income nonworking safety-net participants with and without disabilities, respectively, utilize employment services. Improved data collection on the use of employment services is needed to further evaluate the interaction of safety-net and employment services programs.
Introduction
There are far-reaching and persistent disparities between Americans with and without disabilities—Americans with disabilities are more likely to drop out of school, less likely to be employed, more likely to live in poverty, more likely to have unmet health care needs, and more likely to participate in government programs than Americans without disabilities (Houtenville, Stapleton, Weathers, & Burkhauser, 2009).
The goal of U.S. disability policy is to address these disparities by (a) supporting the adjustments that enable people with disabilities to be independent, self-directed, and self-sufficient, while (b) supporting their health, economic security, and well-being. Safety-net programs (e.g., antipoverty and public insurance programs) primarily provide financial and material support, while education and employment services programs mainly provide services and opportunities to obtain skills related to self-sufficiency and entry into the labor market.
Unfortunately, dire predictions abound about the ability of federal and state programs to continue to support Americans in need of public support and services. Many of the safety-net programs that touch the lives of persons with disabilities have faced exponential and, some would say, unsustainable growth over the past decade, in terms of the numbers of persons participating and the amount of funds expended. A recent report, for example, projects that the federal disability insurance trust fund will be exhausted by 2018 (Social Security Administration [SSA], 2011). The growth in expenditures for the two largest public health insurance programs, Medicare and Medicaid, has been identified as a critical public policy concern by key members of both prominent political parties. Other social programs will surely be under consideration for reduction given recent federal and state budgetary pressures.
One way to reduce the numbers of persons participating in such programs is to provide employment opportunities and services that foster economic self-sufficiency. According to the Government Accountability Office (GAO; the independent, nonpartisan investigative arm of the U.S. legislative branch), however, the lack of a unified and well-integrated disability policy in the United States has resulted in the development and perpetuity of a patchwork of public programs that do not share common goals (GAO, 2008, 2010, 2012). In addition, while expenditures for many safety-net programs have risen in recent years in the United States, particularly for persons with disabilities, the amount of federal government spending on programs that are designed to foster self-sufficiency through employment among working-age people with disabilities has remained relatively stagnant (Livermore, O’Toole, & Stapleton, 2011).
Information on program participation among persons with disabilities is rare because disability status is not routinely and consistently collected within safety-net programs. The purpose of this article is to examine the participation of working-age people with disabilities across a wide variety of government programs and to investigate whether the aggregate imbalance between participation in safety-net programs and employment service activities alluded to above holds equally for all safety-net programs. The section titled “Literature Review and Overview of Programs” reviews existing literature in this area. The “Method” section describes the data and methods used for this analysis. Results are presented in the “Results” section. The section titled “Discussion” discusses results, limitations, and policy implications. The concluding section summarizes the findings and suggests directions for future work.
Literature Review and Overview of Programs
Challenges Facing Employment Policy
There are several key challenges facing employment policy in the United States: (a) the large and long-standing disparity of the economic status between people with and without disabilities, (b) the lack of integration between employment policy and social insurance/public assistance (safety-net) policies, (c) an imbalance between expenditures on safety-net programs and expenditures on employment services, and (d) the interaction of programs that inadvertently create disincentives to work.
Economic disparities
The employment rate and income of people with disabilities are substantially and persistently lower than the employment rate and income levels of people without disabilities (Houtenville & Ruiz, 2011; Weathers & Wittenburg, 2009; Yelin & Trupin, 2003). In March 2009, 31.6% of working-age (age 25–61) people with disabilities were employed, compared with 77.7% of their counterparts without disabilities (authors’ calculations using the 2009 Current Population Survey–Annual Social and Economic Supplement [CPS-ASEC], the most recent data that are available on the utilization of employment services and activities). Empirical evidence from Trupin and Yelin (2003) suggested that employment disparities date back at least to the early 1970s.
In 2008, 22.3% of working-age people with disabilities worked full-time year-round, earning an average annual wage/salary of US$42,600, compared with 63.0% of working-age people without disabilities, earning an average annual wage/salary of US$52,900. With regard to family income, 25.9% of working-age people with disabilities lived in families with annual family incomes below the poverty line, compared with 9.3% for working-age people without disabilities (authors’ calculations using the 2009 CPS-ASEC). At the individual level, research has suggested that poverty rates within the United States are generally lower for the working-age population (ages 25–61) than for persons at either end of the age spectrum. Working-age persons with disabilities have been found to have substantially higher rates of poverty than those without disabilities, using standard poverty measures (Burkhauser, Rovba, & Weathers, 2009; Houtenville & Ruiz, 2011).
Lack of program integration
The economic disparities described above exist despite two major goals of government policies and programs: (a) ensure that people with disabilities are not in poverty, having adequate and appropriate food, shelter, education, and health care access and (b) increase the employment of people with disabilities so that individuals may be self-sufficient. Achieving these two goals simultaneously is difficult and requires a high level of coordination among diverse government programs. However, the programs providing employment services and safety-net supports for working-age people with disabilities are fragmented across numerous agencies and lack coordination (GAO, 2012). For instance, antipoverty income support is provided by the SSA Supplemental Security Income (SSI) program, which also pays for employment services for its participants. Earnings replacement benefits are provided by the SSA Disability Insurance (DI) program and state/private workers’ compensation (WC) programs, all of which also pay for employment services to help beneficiaries return to work. However, the U.S. Department of Education (USDOE) provides employment services through its vocational rehabilitation program for people with disabilities, including SSI participants and DI beneficiaries.
Imbalance in expenditures
The vast majority of disability-related federal expenditures are for safety-net programs related to income and health (GAO, 2008). The GAO estimates that only 2% of federal expenditures on all disability-related programs were for programs providing employment services. A more recent analysis by Livermore et al. (2011) echoed this finding, suggesting that government expenditures on disability-related programs are imbalanced—only a small proportion of federal expenditures are devoted to assisting individuals with disabilities to become employed, while a much larger proportion supports income and in-kind assistance. In addition, Livermore et al. found that total federal expenditures related to working-age people with disabilities increased from 11.4% to 12.0% of total federal expenditures from 2002 to 2008.
When employment services are provided, they are most often provided as “late interventions.” SSA, for example, focuses its employment support efforts on its existing DI and SSI beneficiaries, with very limited success in moving beneficiaries back to work (Liu & Stapleton, 2011). Other research has suggested that few people with disabilities without jobs are actively looking for work (Erickson & Lee, 2008) and that only a small number participate in public vocational rehabilitation services that are designed to increase participation in the labor market (Stapleton, Honeycutt, & Schechter, 2008).
Inadvertent disincentives to work
General and disability-related safety-net programs may inadvertently expose participants to strong work disincentives, thus impeding efforts to increase the employment of people with disabilities so that individuals may be self-sufficient (see Aaron, 1973, Blank, 1997, and Moffitt, 2003, with respect to welfare policy, and Bound & Burkhauser, 1999, with respect to disability policy). Economic theory posits that the supply of labor depends, in part, on the availability of nonlabor income, such as financial support from government programs. More specifically, as nonlabor income increases, the quantity of labor supplied decreases, and vice versa. Work disincentives have been identified as an important reason that state vocational rehabilitation clients who receive SSA disability benefits have worse employment outcomes than other state vocational rehabilitation clients (Stapleton & Erickson, 2004). As a result, as employment and education programs support individuals to increase their self-sufficiency through employment, antipoverty and public health insurance may provide individuals with disincentives to being employed. In recognition of these issues, advocates, policy makers, and researchers are continually developing and exploring ways to increase and improve employment services, remove work disincentives, and integrate programs.
Participation in Safety-Net Programs
For purposes of this article, safety-net programs include energy assistance, Medicaid, Medicare, military or veterans’ health insurance, the Indian Health Service (IHS), public housing, rent subsidies, DI, Supplemental Nutrition Assistance Program (SNAP), SSI, Temporary Assistance for Needy Families (TANF), other welfare programs (including general assistance), unemployment insurance (UI), veterans’ disability compensation, veterans’ pensions, and WC. These programs (shown in Table 1) were chosen as those representing the largest share of overall safety-net program participation in terms of both numbers of participants and cost. These programs provide income, other supports, or some combination of income and other supports.
List of Programs Identified in the CPS-ASEC.
Note. CPS-ASEC = Current Population Survey–Annual Social and Economic Supplement; DI = Social Security Disability Insurance; SSI = Supplement and Supplemental Security Income; WC = workers’ compensation; SNAP = Supplemental Nutrition Assistance Program; TANF = temporary assistance for needy families; UI = unemployment insurance.
Energy assistance
Energy assistance offers financial support to low-income families struggling to pay home heating or cooling costs. Administered by the U.S. Department of Health and Human Services (USDHHS), the low-income home energy–assistance program (LIHEAP) is a block grant program that provides funding to states. States use the funds to offer temporary financial assistance designed to cover immediate home energy needs for low-income families (USDHHS, 2010). During 2008, 6.1 million households, including persons of all ages, received either heating or cooling assistance (USDHHS, 2012a). Data from the 2009 CPS-ASEC suggest that 3.5 million working-age persons participated in energy-assistance programs during 2008 (authors’ calculations).
IHS
IHS is a health care program administered by USDHHS. IHS provides medical assistance to eligible American Indians at IHS facilities and pays a portion of the cost of selected health care services provided at non-IHS facilities (U.S. Census, 2010). The program offered services to 1.9 million American Indians or Alaskan Natives during 2008 (IHS, 2012). Approximately 357,000 working-age persons received services through IHS during 2008 (authors’ calculations using the 2009 CPS-ASEC).
Medicaid
Medicaid is a federal–state program that provides health insurance to needy individuals. Income and asset eligibility levels are set by states, within certain guidelines provided by the federal government. In December 2008, 59,523,600 persons were enrolled in Medicaid, nationwide (Kaiser Family Foundation, 2012). An estimated 12 million working-age persons received Medicaid during 2008 (authors’ calculations).
Medicare
Medicare is a federally administered program that provides health insurance coverage to persons who are disabled or are age 65 or older. Eligibility does not depend upon income. In 2008, 44,831,390 persons were covered by Medicare. Only a portion (16%, or 7.2 million) was between the ages of 19 and 64 (Kaiser Family Foundation, 2011). Approximately 5.6 million Medicare beneficiaries were working-age in 2008 (authors’ calculations).
Military health
Military health insurance includes the Civilian Health and Medical Program of the Department of Veterans Affairs (CHAMPVA), Veteran Affairs (VA), or other military health insurance. During 2008, more than 9 million persons were covered within the military health system (U.S. Department of Defense, 2009). An estimated 4.3 million working-age persons were covered by some form of military health insurance during 2008 (authors’ calculations).
Public housing
Housing assistance is available to ensure that adequate standards of living are available for low-income families. Public housing provides housing to eligible low-income families. The federal Department of Housing and Urban Development provides funding to local housing agencies. The housing agencies then determine eligibility based on income, demographics (elderly, person with a disability, family), and citizenship status (U.S. citizen or eligible immigration status). For purposes of eligibility, an elderly household is defined as a household where the head of household or the spouse is age 62 and older. A household with a disability includes households where either the head of household or the spouse is less than 62 years old and has one of the six functional limitations listed elsewhere in this article (see Table 2; U.S. Department of Housing and Urban Development [USHUD], 2011b). In 2008, 4.2 million persons were housed in approximately 3.8 million public housing and/or Section 8 units (USHUD, 2012). Data from the 2009 CPS suggest that more than 2.5 million working-age persons resided in public housing during 2008 (authors’ calculations).
Program Participation Rates for Individuals, Ages 25 to 61, 2009 CPS.
Note. CPS = Current Population Survey; DI = Social Security Disability Insurance; SSI = Supplemental Security Income; WC = workers’ compensation; SNAP = Supplemental Nutrition Assistance Program; UI = unemployment insurance.
Rent subsidies
The USHUD offers a housing choice voucher program to assist persons in securing housing in the private market. The vouchers are administered by local housing agencies and are paid directly to landlords (USHUD, 2011a). In 2008, more than 5 million persons residing in 2.2 million housing units received support from housing choice voucher programs (USHUD, 2012). Approximately 1.4 million rent subsidy program participants were working-age in 2008 (authors’ calculations).
DI
DI pays monthly cash benefits to persons who have sufficiently participated in jobs covered by Social Security benefits and who are deemed unable to work for a year or more due to a disability (as defined by the SSA). DI is an earnings replacement/insurance program that is administered by SSA. More than 8 million persons received DI benefits during 2008 (Houtenville & Ruiz, 2011). The DI program provided benefits to more than 4.1 million working-age persons in 2008 (authors’ calculations).
SNAP
Nutrition assistance is a shared federal–state program that aims to allow low-income households to purchase a minimal level of nutritionally adequate food. While the federal government pays the benefit costs and sets the rules for program participation, states administer the program. The program provides a monthly amount of funding based on household size. Funds may only be used to purchase food. The presence of a person with a disability in the household is not necessary for participation in this program. During fiscal year 2009, an average of 33.5 million persons participated in SNAP each month. Per person average monthly benefit receipt was US$125.31 (U.S. Department of Agriculture, Food, and Nutrition Service, 2012). According to data from the CPS-ASEC, approximately 11 million working-age persons participated in SNAP during 2008 (authors’ calculations).
SSI
SSI is administered by SSA. A means-tested program, it provides monthly cash payments to low-income aged, blind, and disabled individuals. Recipients may also participate in the DI program if dual eligibility requirements are met. SSI provided income support to 7.5 million persons during 2008 (Houtenville & Ruiz, 2011). During 2008, 2.6 million working-age persons participated in the SSI program (authors’ calculations).
TANF
TANF is a federal–state program that provides income assistance to needy families. Monthly income is provided, variable by household size. Participation in employment or employment-related activities is a requirement for this program. Approximately 3.8 million persons received TANF during 2008 (USDHHS, 2012). Nearly 2.8 million working-age persons participated in TANF during 2008 (authors’ calculations).
Other welfare programs
Public assistance programs such as general assistance programs offered in 30 states are categorized as other welfare programs, to distinguish them from the primary welfare program (TANF) described above. Such programs generally provide financial assistance to low-income persons who do not have minor children, do not meet Social Security criteria for disability benefits, and are not elderly (Schott & Cho, 2011). Approximately 375,000 working-age persons received such supports during 2008 (authors’ calculations).
UI
UI programs are overseen by the U.S. Department of Labor (USDOL) but each state administers its own program. Eligibility conditions, maximum benefit levels, and possible duration of benefit receipt vary by state. All states do require that UI recipients be able to and available for work. The amount of weekly UI payments varies by state. UI benefit expenditures, for approximately 9.1 million people, equaled US$120 billion in the fiscal year 2009 (Congressional Budget Office, 2010; Wolf, 2010). Data from the 2009 CPS suggest that 8 million working-age persons participated in UI programs during 2008 (authors’ calculations).
Veteran disability compensation
The U.S. Department of Veterans Affairs (USDVA) provided monthly disability compensation checks to approximately 2.9 million disabled veterans in 2008. Benefit amounts vary by disability rating (Associated Press, 2009; USDVA, 2012a). More than one million working-age persons received veteran disability compensation during 2008 (authors’ calculations).
Veteran pension
Veteran pension payments are available for working-age persons who have limited or no income, are deemed permanently and totally disabled, a patient in a nursing home, or are receiving DI payments (USDVA, 2012b). USDVA reports that 315,763 people received veterans’ pensions at the end of fiscal year 2008 (USDVA, 2009). According to 2009 CPS data, approximately one quarter of a million working-age people received veteran pension payments during 2008 (authors’ calculations).
WC
WC is a social insurance program that provides cash payments, medical payments, and, in most states, rehabilitation and training to support workers who are injured on the job. Costs are borne by employers and states administer the program. During 2008, the program covered approximately 130.6 million workers in the United States (Sengupta, Reno, & Burton, 2011). Nearly one million working-age people received payments from WC programs during 2008 (authors’ calculations).
Measures of poverty play an important role in eligibility determination for many of the safety-net programs described above. The federal government publishes two prominent measures of poverty: poverty guidelines and poverty thresholds. Poverty guidelines are used for administrative purposes to establish eligibility for certain government programs and are provided by the USDHHS on an annual basis, based on family size. Separate guidelines are provided for the 48 contiguous states and Washington, D.C., Alaska, and Hawaii. Poverty guidelines are used to determine eligibility for the LIHEAP, parts of Medicaid, the subsidized portion of Medicare (the prescription drug coverage), and the SNAP. TANF, SSI, general assistance programs, Section 8 low-income housing assistance, and low-rent housing assistance do not use these poverty guidelines to establish program eligibility (USDHHS, 2011). Poverty thresholds are used to calculate the number of persons in poverty and are provided on an annual basis by the U.S. Census Bureau, providing the same figures for all 50 states and the District of Columbia (USDHHS, 2011).
Employment Services
The review of safety-net programs provided above demonstrates that a large number of individuals are participating in publicly funded safety-net programs in the United States. Participation in employment services is one option that may help to improve the economic situation of safety-net program participants and reduce rates of program participation. Research has suggested that persons with disabilities who participate in employment-related services can improve both their overall employment rates and earnings potential (Wittenburg, Rangarajan, & Honeycutt, 2008).
Some safety-net programs explicitly include a work activity participation requirement (TANF, UI), whereas others do not. Even where employment service participation is not a requirement, however, program administrators have begun to experiment with ways to increase levels of employment for persons with disabilities. The USDOL One-Stop program expanded and integrated many employment-related services. The SSA Ticket to Work program promotes expanded choices when SSA beneficiaries choose among providers of employment services, and a new feature of the program is designed to promote collaboration with vocational rehabilitation agencies. The Centers for Medicare and Medicaid, through Medicaid Infrastructure Grants, are providing funding to states to expand and coordinate employment services. Medicaid Buy-In programs, adopted by many states, are designed to remove a strong work disincentive: the loss of Medicaid due to increased income. SSA is currently conducting demonstration projects to rigorously test alternative employment services and benefit designs (Wittenburg, Mann, & Stapleton, 2012).
Individuals can access employment services through private or public sources. Public employment services programs include USDOL-funded One-Stop programs, and, for persons with disabilities, USDOE-supported state vocational rehabilitation programs. In Federal Fiscal Year 2009, state vocational rehabilitation agencies processed 618,375 applicants and closed 177,984 cases (Houtenville & Ruiz, 2011). Data on participation in specific employment services programs, such as state vocational rehabilitation agency programs, are not available in the CPS.
The range of employment services offered at either private or public sources might include job training, resume writing, interviewing guidance, and potential employer identification. Participation in work programs such as community service might also serve to increase employment opportunities.
Research Questions
Participation of people with disabilities in individual safety-net programs, multiple safety-net programs, and employment services are key areas of interest for the disability policy field. The research conducted here will explore these issues. Levels of safety-net participation are expected to differ between persons with and without disabilities. Levels of participation in employment services are expected to differ between persons with and without disability and also among programs. Of particular interest is the extent to which working-age persons with disabilities who are participating in the safety-net programs described above are also engaging in employment-related services and activities.
Method
Program participation may be measured in various ways and with varying detail. Other researchers have examined current and past applications for program benefits, basis for eligibility, current and past receipt of benefits, nature of benefits (such as cash, insurance, and in-kind material supports and services), amount of benefits (dollars received and the dollar value of in-kind benefits to participants), and overall program expenditures (Burkhauser, Daly, & Houtenville, 2001; Burkhauser & Simon, 2010; Goodman & Stapleton, 2007; Livermore et al., 2011). Braddock and Hemp (2008) analyzed program data, including number of participants and annual program expenditures, related to the population with developmental disabilities. Using a similar approach for all disability types, Goodman and Stapleton (2007) and Stapleton, Wittenburg, and Thornton (2009) examined program participation statistics using administrative data.
Burkhauser et al. (2001) and Burkhauser, Feng, Jenkins, and Larrimore (2008) used changes in the share of household income coming from various sources (including public programs) as a means of understanding the relationship between program participation and employment outcomes. Their findings suggest that declines in wage/salary income due to the reduced employment of people with work limitations were replaced by the wage/salary incomes of other household members and program income (mostly DI and SSI income, but also income from WC, TANF/Aid to Families With Dependent Children [AFDC], and UI compensation). However, the values of employer-provided health insurance programs (a major form of compensation), public health insurance programs, and in-kind programs (e.g., public/subsidized housing) were not included in household income.
For the analysis conducted here, population survey data are used and program participation will be defined as a participation rate (the percentage of persons with and without a disability who participate in a particular program). While the use of survey data to measure program participation has been found to result in significant underestimation (Klerman, Ringel, & Roth, 2005; Lynch et al., 2008; Meyer & Goerge, 2010; Meyer, Mok, & Sullivan, 2009; Wheaton, 2008), survey data provide the best opportunity to examine participation among and across multiple safety-net and employment services programs. In addition, there is no reason to expect that underestimation that occurs is dissimilar between persons with and without disabilities, making comparisons between these two groups still relevant.
While certain data limitations exist, the CPS-ASEC can be used to estimate the numbers of persons with disabilities who are participating in social safety-net programs and who are participating in certain employment services and activities. This article will use data from the 2009 administration of the CPS-ASEC, weighted at the person level. Basic demographic information as well as labor force data, income, and noncash benefit receipt are included. The CPS uses a multistage-stratified nationally representative sample of the U.S. population, sampling households, and noninstitutional group quarters (e.g., dormitories) and collecting survey information on all individuals in these units. Individuals living in institutions are not sampled (U.S. Census, 2010). Given that the reporting period covered the past year, it is possible that some individuals reported joint participation when in fact participation occurred sequentially. For example, an individual cannot jointly receive TANF and SSI, but might first receive TANF before later qualifying for SSI. Results are described as the year of program participation, rather than the year of survey administration.
Analysis Approach
Program participation rates were first calculated for each program, among persons with and without disabilities. To explore levels of cross-participation, programs were collapsed into four categories—health insurance programs (i.e., Indian Health, Medicaid, Medicare, military health), disability programs (i.e., DI, SSI, DI&SSI, veteran disability compensation, veteran pension), housing programs (i.e., energy assistance, public housing, rent subsidies), and other income support programs (i.e., TANF, other welfare, UI). The percentage of persons participating in any such program, one program or more than one program within each group was also calculated. Difference of proportion tests were used to assess differences between persons with and without disabilities, with p < .10 as the chosen level of significance. In essence, the difference in proportion test uses the adjusted standard error of the difference in the percentage of people with and without disabilities participating in a given program. The standard error of the difference is used to calculate a confidence interval of the difference between the two compared percentages. Confidence intervals that include 0 indicate that no significant difference exists. Sample weights were used in the calculation of all estimates to adjust for the complex design of the CPS-ASEC sample and to produce nationally representative estimates. And furthermore, design factors were used in the calculation of standard errors to adjust for the stratified design of the CPS-ASEC sample in accordance with U.S. Census Bureau guidelines (U.S. Census Bureau, 2010).
Key Variables
To conduct the planned analysis, several key constructs must first be defined. The remainder of this section describes how disability, safety-net program participation, employment service and activity participation, and low-income, and working-age were defined.
Disability
Beginning in 2008, the CPS adopted a set of disability questions:
Is this person deaf or does he or she have serious difficulty hearing?
Is this person blind or does he or she have serious difficulty seeing even when wearing glasses?
Because of a physical, mental, or emotional condition, does this person have serious difficulty concentrating, remembering, or making decisions?
Does this person have serious difficulty walking or climbing stairs?
Does this person have difficulty dressing or bathing?
Because of a physical, mental, or emotional condition, does this person have difficulty doing errands alone such as visiting a doctor’s office or shopping?
Recent research has suggested that the six-question sequence fails to adequately capture the population of persons that are participating in DI and SSI, successfully identifying only 63% of participants (Burkhauser, Houtenville, & Tennant, 2012). As an alternative, Burkhauser et al. (2012) suggested adding a work limitation question (which has been available in the CPS-ASEC since 1981) to the six sequential disability questions, because doing so was found to increase the ability to identify the population with disabilities as reflected by the receipt of DI and SSI benefits. This strategy will be used in the primary analysis. The work limitation question(s) reads as follows: “[d]oes anyone in this household have a health problem or disability which prevents them from working or which limits the kind or amount of work they can do? If so, who? (Anyone else?)”
In addition, given the complexity of disability, it is wise to consider alternative definitions of disability. Estimates are therefore also produced using only the six disability questions and only the work limitation question. The results from these two alternative definitions are briefly mentioned below and are available upon request. (Overall, the results when using these two definitions are very similar to the results found when using the primary definition.)
Program participation
Program participation is defined as a participation rate (the percentage of persons with and without a disability that participate in a particular program). The CPS collects income and program participation data from the head of household for all household members of age 15 or older. To calculate participation rates, the percentage of individuals with or without a disability designated as participants in a specific program were counted in the numerator of the calculation. The denominator included all persons with or without disability.
Employment services/activities
As mentioned above, the CPS does not contain data on the utilization of specific employment services programs. However, a number of questions designed to measure participation in employment search activities and programs related to welfare reform were added to the CPS-ASEC, beginning in March 2001 (Minnesota Population Center, 2011). The questions, asking about participation over the past calendar year and ascertained for each individual in the household, are as follows:
[h]ave you (or others in your household) . . .
attended a job search program or job club, or used a job resource center to get lists of jobs and employers, to schedule job interviews, or to fill out job applications? Who did that? Anyone else?
attended a training program to learn a specific job skill, such as computer word processing, auto mechanics, nursing, providing child care, or a skill for some other job or vocation? Who did that? Anyone else?
attended job readiness training to learn about resume writing, job interviewing, or building self-esteem? Who did that? Anyone else?
participated in work programs such as community service to receive cash assistance? Who did that? Anyone else?
Unfortunately, these questions were only asked of individuals in low-income households and were discontinued after the 2009 administration of the CPS-ASEC.
Low income
The employment services and activities questions described above were only asked of survey participants in households that met low-income criterion, as defined by the Census, and that included at least one person who was between the ages of 15 and 64 (Julie Walker [U.S. Census], April 12, 2012, personal communication).1 Given this limitation, employment service and activity participation can only be calculated for persons with and without disabilities who meet this low-income criterion.
Working-age
The working-age population will be defined as persons aged 25 to 61. This narrower definition attempts to address the large number of persons aged 18 to 24 engaged in educational pursuits as well as the large number of people of age 62 to 64, who are retired.
Results
Sample Description
The sample of persons with and without disabilities did not differ greatly on either gender or race. Fifty-one percent of each sample was female. Eighty-one percent of the sample without disabilities and 77% of the sample with disabilities was White. Working-age persons with disabilities were slightly older (M = 47.38, s = 10.093) than working-age persons without disabilities (M = 42.27, s = 10.389). 2
Share of Program Participants With Disabilities
Considered from the program perspective, persons with disabilities comprise more than one third of working-age program participants in safety-net programs, based on 2009 CPS data. Results vary by program. For example, an estimated 43% of the 3.5 million working-age persons who participated in energy-assistance programs during 2008 had disabilities. Nineteen percent of IHS and 23% of military health insurance program participants were persons with disabilities. Forty-seven percent of the more than 12 million working-age Medicaid recipients and 83% of the 5.6 million working-age Medicare beneficiaries were persons with disabilities. Also, 41% of the more than 2.5 million working-age public-housing participants, at the individual level, and 45% of the 1.4 million working-age rent subsidy program participants had disabilities. More than one third (35%) of the more than 11 million working-age SNAP participants were persons with disabilities. In the TANF program, 47% of the nearly 2.8 million working-age program participants were persons with disabilities. In addition, 18% of the 8 million working-age program participants in UI insurance and 42% of the nearly 1 million WC program participants met the CPS definition of disability (authors’ calculations). Fifty-one percent of veteran disability compensation and 37% of veteran pension program working-age participants were persons with disabilities.
Program Participation Rates by Disability Status
Considered from the population perspective, working-age people with disabilities are more likely to participate in safety-net programs than people without disabilities. Table 2 includes detailed program participation rates for the entire working-age sample, comparing participation rates for persons with and without disabilities as measured using the seven disability questions. Nearly two thirds (65%) of working-age people with disabilities participated in one or more such programs during 2008, compared with 17% of their counterparts without disabilities. 3 In nearly every program, larger proportions of working-age persons with disabilities participated. Thirty-two percent of working-age persons with disabilities participated in Medicaid and 26% participated in Medicare. More than one in five working-age persons with disabilities participated in DI (22%). More working-age people with disabilities participated in SNAP (22%) than SSI (17%). Among persons without disabilities, the highest proportions of participation hovered around five percent, in the SNAP, UI, and Medicaid programs. Interestingly, similar percentages of people with and without disabilities participated in UI (4% and 5%, respectively), even though people with disabilities are less likely to participate in the labor force. 4
Cross-Program Participation
Table 3 shows levels of cross-participation within broad categories of programs. In keeping with the individual-level results, participation in safety-net programs was significantly higher for working-age persons with disabilities than for working-age persons without disabilities. Approximately 52% of persons with disabilities participated in at least one public health insurance program, whereas only 8% of persons without disabilities participated in a public health insurance program. More than 15% of persons with disabilities and 3% of people without disabilities participated in a housing-related program such as energy assistance, public housing, or rent subsidies. Forty-three percent of persons with disabilities and 1% of persons without disabilities participated in a disability program. More than one quarter of persons with disabilities and nearly 10% of people without disabilities participated in another support program such as SNAP, TANF, other welfare, or UI.
Cross-Program Participation for Individuals, Ages 25 to 61, 2009 CPS.
Note. CPS = Current Population Survey.
Health insurance programs include Indian Health, Medicaid, Medicare, and military health. bHousing programs include energy assistance, public housing, and rent subsidy programs. cDisability programs include Social Security Disability Insurance (DI), Supplemental Security Income (SSI), DI&SSI, veteran disability compensation, veteran pension, and workers’ compensation (WC). dOther support programs include Supplemental Nutrition Assistance Program (SNAP), Temporary Assistance for Needy Families (TANF), other welfare, and unemployment insurance (UI).
Persons with disabilities were more likely to participate in more than one public health insurance program (12%) than people without disabilities (less than 1%). Most housing program participants (90% of persons without disabilities and 85% of persons with disabilities) and most other support participants (90% and 88% of persons without and with disabilities) were participating in only one program, regardless of disability status. Similarly, most participants in disability programs were only participating in one program (97% of persons without disabilities and 96% of persons with disabilities). 5
Table 4 shows safety-net participation rates among the participants of the two largest disability income support programs (DI and SSI). Among DI beneficiaries, nearly 38% were concurrently participating in programs other than DI, SSI, Medicaid, and Medicare. Higher rates of cross-participation were found for SSI (54%) and DI&SSI concurrent beneficiaries (61%). (DI&SSI concurrent beneficiaries are typically people with work experience and poverty-level family income.) SNAP participation was, by far, highest across disability program participants. Results also suggest some overlap in public health insurance support to persons receiving DI and SSI, although DI beneficiaries are typically tied to Medicare and SSI recipients typically receive Medicaid. Thirty-two percent of DI beneficiaries received Medicaid during the past year and 27% of SSI recipients received Medicare. The use of cross-sectional CPS data does not allow for a further exploration of the timing of health insurance receipt.
Program Participation Rates for SSA Disability Program Participants, Ages 25 to 61, 2009 CPS.
Note. SSA = Social Security Administration; CPS = Current Population Survey; DI = Social Security Disability Insurance; SSI = Supplemental Security Income; WC = workers’ compensation; SNAP = supplemental nutrition assistance program; TANF = temporary assistance for needy families; UI = unemployment insurance.
Program Participation and Employment Service/Activities
Table 5 summarizes safety-net participation by the presence of disability and employment status (for the overall working-age and for the low-income working-age populations) as well as employment service and activity participation by the presence of disability and employment status (for the low-income working-age sample) of the 2009 CPS. Working-age people with disabilities are more than twice as likely to participate in safety-net programs as persons without disabilities, regardless of employment status. Participation in any employment service or activity could only be calculated for the low-income sample, as the relevant questions were restricted to this population. Sixty-two percent of the working-age sample with disabilities met the low-income criteria, while only 35% of the working-age sample without disabilities was identified as low income. Nineteen percent of the low-income sample had a disability. Participation in employment services and activities was low for both low-income groups, ranging from 3% to 8%.
Summary Table Comparing the Total and Low-Income Populations, Individuals, Ages 25 to 61, 2009 CPS.
Note. CPS = Current Population Survey. Bold denotes significant difference (p < .10) between low-income persons with and without disabilities.
Overall, 76% of the total low-income working-age population with disabilities participated in some form of a safety-net program during 2008. Only an estimated 4% of these persons, however, participated in employment activities. Safety-net program participation was significantly lower for low-income persons without disabilities (27%), while participation in employment services and activities was similar (3%). Among low-income persons who were employed, significantly higher (p < .10) proportions of persons with disabilities participated in employment services than did persons without disabilities. Six percent of persons with disabilities who were employed used employment services, compared with less than 2% of persons without disabilities. Low-income persons without disabilities were significantly more likely (p < .10) to use employment services than persons with disabilities who were not employed. 6
As Table 6 shows, rates of participation in any work activities were low for both persons with and without disabilities who were not employed, ranging from a low of less than 1% to a high of 15%. The percentages included in columns indicate the percentage of persons who participated in any employment service or activity as a proportion of those who participated in the safety-net program noted in the relevant row. People without disabilities were significantly more likely (p < .10) to participate in employment services for many programs (shown in bold text in Table 6). Perhaps of more interest are those programs where employment service participation rates were similar for both persons with and without disabilities. Rates were similarly low in the Medicare, military health, and DI programs (2%–5%). Rates were similarly high in the TANF and UI programs (14%–15%). For people without disabilities, those who were participating in TANF, UI, or energy-assistance programs had the highest participation in any type of employment activity. 7 Fifteen percent of persons with disabilities who were participating in UI participated in some type of employment activity, as did 14% of persons with disabilities who were participating in TANF.
Employment Services/Activities Participation Rates for Low-Income Safety-Net Participants Who Are Not Employed, Ages 25 to 61, 2009 CPS.
Note. CPS = Current Population Survey; DI = Social Security Disability Insurance; SSI = Supplemental Security Income; WC = workers’ compensation; SNAP = Supplemental Nutrition Assistance Program; TANF = temporary assistance for needy families; UI = unemployment insurance. Bold denotes significant difference (p < .10) between low-income persons with and without disabilities.
The lowest rates of participation in employment services and activities for low-income, nonworking persons without disabilities (as identified in the CPS-ASEC) were found in the DI and Medicare programs. Similarly, rates of participation in employment services and activities hovered around 2% for low-income, nonworking persons with disabilities who were participating in DI, DI&SSI concurrently, Medicare, and SSI.
Discussion
People with disabilities comprise one third of working-age safety-net program participants in the United States and not just within disability-related programs. In addition, at least 65% of all working-age persons with disabilities participate in one or more of the safety-net programs highlighted here. Participation in more than one health insurance program is more common for people with disabilities, emphasizing the importance of health care for this population.
Rates of program participation are even higher among low-income persons, with 76% of working-age persons with disabilities participating in one or more safety-net programs during 2008. Safety-net participation is particularly high for those low-income persons who are not employed, ranging from 39% for persons without disabilities to 81% for persons with disabilities. Even for the 44% of low-income persons who are employed, past year safety-net participation remains high at 49% for persons with disabilities and 22% for persons without disabilities. Half of low-income persons with disabilities who work still utilize safety-net programs, suggesting that the income received from employment does not fulfill basic material needs for this population.
Reported participation in job search, job clubs, job training, or other work programs is low among all low-income persons. Low-income persons with disabilities who are employed are significantly more likely to report participation in employment services over the past year, however, than low-income persons without disabilities who are employed. The utilization of employment services may provide the support necessary for some persons with disabilities to be successful in the workplace. For low-income safety-net participants who are not working, rates of participation in employment services are significantly (p < .10) higher for people without disabilities (5%) than for people with disabilities (3%), although low overall. Further work can explore why participation in employment services is so modest for low-income people who are not working.
The intersection of public programs with working-age persons with disabilities on such a large scale provides a unique opportunity for persons with disabilities to access needed services and supports. While it is clear that such programs are contributing important services that improve the health, safety, and well-being of persons with disabilities, better coordination among programs may be needed. Programs are authorized under a variety of regulations, administered by multiple federal, state, and local agencies, and have a correspondingly complex web of eligibility and program continuation policies. The concerns of people with disabilities may therefore not get the dedicated attention they need. The high level of cross-program participation for persons participating in the two largest disability income support programs, for example, suggests that persons with disabilities need access to an array of additional services and supports to supplement the income support received from DI and SSI. Future work, using longitudinal data, can perhaps better explore the timing, duration, and impact of these additional supports for disability beneficiaries.
The imbalance between levels of participation in safety-net programs and levels of participation in employment services and activities alluded to at the start of this article appears to hold true for low-income persons. Among low-income working-age persons with and without disabilities who were not employed, the highest rates of participation in employment services and activities were found among those participating in UI and TANF. The high rates of participation for those in UI may not be surprising, given that participants have a recent work history. The higher rates of participation in the TANF program, a program that does not typically include persons with high levels of employment or educational experience, suggest that program policies may have an influence on job preparation and seeking behavior. Both UI and TANF share a time limit and a requirement for participation in employment-related activities.
Future research can begin to untangle the effect of these two general policy programs (UI and TANF) in encouraging participation in employment services among persons with disabilities. Analysis of any disincentives to participation in employment-related activities should consider not only program-level variables (e.g., the interactions among program eligibility, built-in program disincentives for employment) but also the individual characteristics of and constraints faced by people who choose to forego employment services.
Limitations
People with disabilities may face particular challenges in obtaining or maintaining employment. While a number of these safety-net programs offer linkages to employment services, detailed and consistent cross-program participation in available employment services by persons with disabilities is lacking. The employment services and activities data used here were limited in scope. Results, for example, could not distinguish between publicly and privately funded employment services and activities. In addition, individuals who undertook job search activities, for example, on their own, without using any outside resources, would be included as employment activity participants. Although the results cannot validate the imbalance in public investments for safety-net programs and employment services for persons with disabilities alluded to earlier in this study, the results can suggest that participation in employment-related activities is far lower than participation in safety-net programs. The ability to collect more detail about the types of employment-related activities and services in which people are participating would help policy makers better target scarce employment-related funding. The development of a comprehensive taxonomy of employment services that could be used across program and national data collection efforts might help to facilitate the collection of better data.
Further limitations of the current analysis, including the use of cross-sectional and self-reported data, suggest that future work in this area might benefit from the integration of workforce development and/or vocational rehabilitation administrative data into analytical plans. The inclusion of qualitative research can help to understand how and why program participants access employment services.
Conclusion
Safety-net programs provide an important source of support to millions of working-age Americans with disabilities, many of whom participate in more than one program. Efforts to understand program interactions must continue, as fiscal resources continue to tighten and the need for improved coordination among programs becomes more critical. Collecting more detailed data about employment-related behavior, at both the population-based data and program-based data, can facilitate such analyses. Results can then be used by disability policy makers as they seek to craft policies that will remove and/or overcome disincentives and facilitate an increase in employment-related activities and services participation for safety-net program participants who have disabilities.
Footnotes
Authors’ Note
The contents do not necessarily represent the policy of the Department of Education and you should not assume endorsement by the federal government (Edgar, 75.620 (b)). The authors retain sole responsibility for any errors or omissions.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding of this study was made possible by the Employment Policy and Measurement Rehabilitation Research and Training Center, which is funded by the U.S. Department of Education, National Institute for Disability and Rehabilitation Research, under cooperative agreement H133B100030.