II. Characteristics of Working-Age Beneficiaries and Use of Employment Services

Data from the 2004 NBS suggest that there could be substantial demand for employment-related services among working-age Social Security disability beneficiaries.¹ Clearly, many beneficiaries are unlikely to work. All have passed a rigorous determination process that found them unable to engage in substantial gainful activity (which SSA currently defines as working at a level to earn $900 a month). Furthermore, many have characteristics that would make finding employment difficult even in the absence of a disability: many are over 55 years old, have been on the rolls (and out of the labor force) for over a decade, and are in poor or deteriorating health. Nevertheless, a considerable share of these beneficiaries, representing millions of individuals, may be able to benefit from employment support services either because they are currently working or because they have expressed an interest in working at substantial levels.

This chapter first reviews the basic characteristics of working-age beneficiaries.² This descriptive information is helpful in interpreting the findings for TTW but has wider applicability since it comes from the first nationally representative cross-sectional survey of beneficiaries conducted in decades. After laying out characteristics of beneficiaries, the chapter discusses their self-reported work experience, attitudes, and expectations. It then addresses more concrete measures of the potential demand for TTW services by examining the reported use of and unmet needs for employment services.

A. Beneficiary Characteristics

1. Program Characteristics

From SSA administrative data we know that a large majority (69 percent) of working-age beneficiaries receive DI benefits (Exhibit II.1). Just over half (53 percent) are DI-only beneficiaries, and another 16 percent participate in both the DI and SSI programs. SSI-only recipients account for 31 percent of all beneficiaries.

The fact that the DI and SSI programs have different eligibility criteria means that beneficiaries in the two programs are likely to differ in their work histories and possibly in their work potential. While both the DI and SSI programs define disability in essentially the same way, the DI program provides disability benefits to people who have a substantial work history, regardless of income or assets, while the SSI program provides disability benefits only to people meeting the disability criteria with very low income and assets. As a result, most DI beneficiaries will have worked for 10 years (that is, a total of 40 quarters), while the work history of most SSI beneficiaries will be much more limited. Concurrent beneficiaries are likely to have a history of fairly low earnings and thus DI benefits that are low enough to make them eligible for SSI benefits.

Most (53 percent) working-age beneficiaries first entered the DI or SSI programs 10 or more years ago. Only a small share of beneficiaries (four percent) have been on the rolls for fewer than two years. Therefore, even while many beneficiaries have an extensive work history, the fact that they have been on the rolls and out of the labor force for a substantial amount of time is likely to make employment a challenge.³

When we interviewed beneficiaries in 2004, the average monthly disability benefit—including all federal and state supplement amounts—was $810.⁴ Most beneficiaries (63 percent) receive monthly benefits between $500 and $1000, about one-quarter (24 percent) receive monthly benefits in excess of $1000, and a relatively small share (13 percent) receive benefits that are less than $500 per month.

Relative to other types of benefits, it appears that DI and SSI benefits are an important source of income for beneficiaries. A large share of these individuals (65 percent) did not report receiving any other non-SSA cash or “near cash” benefits (such as food stamps) that could potentially be affected by earnings, and only a small share (17 percent) reported receiving $200 or more per month of these types of benefits.

Exhibit II.1. Program Characteristics of 2004 NBS Respondents at Interview

[D]

Source: Ticket Research File data about disability benefit eligibility and amounts matched to the 2004 National Beneficiary Survey data about receipt of other benefits. Sample size = 7,603. Note: Non-SSA cash and near cash benefits are defined to include only the following non-SSA benefits that could potentially be affected by earnings: food stamps; energy, housing, or other in-kind assistance; public assistance; workers' compensation; Veterans' benefits; private disability insurance; unemployment insurance; and pension income among those under age 59.

2. Sociodemographic Characteristics

Overall, most beneficiaries are over the age of 40 (78 percent) with almost 40 percent over the age of 55 (Exhibit II.2). Also, most are white (71 percent) and not married (67 percent).

Exhibit II.2. Sociodemographic Characteristics of Beneficiaries

[D]

Source: 2004 National Beneficiary Survey. Sample size = 7,603. Note: The applicable federal poverty level (FPL ) is determined by family size and the ages of family members. In 2004, the FPL for a family of one individual under age 65 was $9,827 per year.

Many beneficiaries have characteristics that suggest that they would have difficulty obtaining jobs even disregarding their disabilities. In particular, a substantial share of beneficiaries have less than a high school level of education (42 percent). ⁵ This rate is much higher than that of the general population of adults ages 25 and over (15 percent) (U.S. Census Bureau 2004). ⁶ A low level of education is likely to limit beneficiaries’ earnings, regardless of any disability-related challenges that must be overcome to return to work. Nearly half of all beneficiaries (49 percent) are living in families with annual incomes below the federal poverty level (FPL), and another 38 percent have incomes at or near the poverty level (100 to 299 percent of the FPL). The figures on family income are consistent with the figures reported in Exhibit II.1 that showed many beneficiaries receive cash and in-kind assistance from public programs other than SSA.

3. Disability and Health Characteristics

Given that a majority of working-age beneficiaries receive DI benefits and that the DI eligibility criteria require beneficiaries to have a substantial work history in order to qualify for benefits, it is not surprising that most disability beneficiaries first experienced the onset of their disabling health conditions during adulthood. Only 23 percent experienced childhood onset. A rather large share of all beneficiaries (42 percent) experienced onset of the disabling health condition after the age of 40 (Exhibit II.3). A later age of disability onset is much more common among DI-only beneficiaries than among SSI-only recipients. ⁷ A large body of literature examining return to work after illness or injury indicates that age at onset is an important predictor of return to work. ⁸ Most studies show that the likelihood of returning to work is significantly lower among those who are injured or become ill after age 50 relative to those whose injuries or illnesses occur at younger ages. While the literature points to numerous job-related, health, psychosocial, and economic factors that affect return to work (Krause et al. 2001), it provides little explanation about why older workers are less likely to return to work. Possible reasons might include decreased physiological ability to recover from illness or injury; age discrimination in ways that make access to job accommodations, retraining, or new employment opportunities more limited; access to higher wage-replacement benefits; and different preferences for work and leisure. The large share of beneficiaries experiencing onset after age 40, particularly among DI-only beneficiaries, suggests that many beneficiaries might face these barriers to work.

While a small percentage of beneficiaries (5 percent) reported having no conditions that limit their activities, most (62 percent) reported at least two health conditions causing limitations (Exhibit II.3).⁹ The most commonly reported limiting conditions were musculoskeletal (36 percent) and mental health (31 percent) conditions, followed closely by diseases of the circulatory system (24 percent).

The distribution of self-reported conditions generally mirrors the distribution of conditions recorded in SSA administrative records, with the conditions being most commonly reported in the survey also being the most prevalent in the administrative record (Exhibit II.4), although musculoskeletal conditions are much more frequently reported by respondents relative to their occurrence in the administrative data. Because any number of conditions could be reported in the survey data, while the administrative data typically list only one condition (or possibly two for concurrent beneficiaries), the percentages for self-reported conditions generally exceed the percentages reported in the administrative data. A notable exception is the mental retardation category, where a much smaller percentage of survey respondents reported mental retardation as a condition limiting their activities relative to its occurrence in the administrative data as the reason for qualifying for disability benefits. When we analyzed the extent to which a respondent’s primary or secondary diagnosis group in the administrative data concurred with any of his or her self-reported condition groups, we found that the rate of concurrence was 72 percent overall. The rates of concurrence are highest for circulatory and musculoskeletal conditions, and lowest for infectious/parasitic diseases.¹⁰

Approximately three-quarters of working-age beneficiaries reported that they are limited in their ability to conduct basic activities (Exhibit II.5). These include activities of daily living (ADLs), which involve such fundamental tasks as bathing or dressing, getting around the house, getting into or out of bed, and eating. They also include instrumental activities of daily living (IADLs), which include less fundamental but equally important activities such as getting around outside of the home, shopping for personal items, and preparing meals. Interestingly, beneficiaries seem to cluster somewhat at the extremes, with about equal shares reporting that they have no limitations in ADLs and IADLs (28 percent) or that they are limited in four or more ADLs or IADLs (27 percent).

Exhibit II.3. Age at Disability Onset and Most Prevalent Self-Reported Condition(s) Causing Activity Limitation

[D]

Source: 2004 National Beneficiary Survey. Sample size = 7,603. Note: Respondents were able to report multiple reasons for current activity limitations.

Exhibit II.5. Percent of Beneficiaries Who Have Difficulty with a Given Number of ADLs/IADLs

[D]

Source: 2004 National Beneficiary Survey. Sample size = 7,603.

Functional difficulties associated with the upper extremities (67 percent) and lower extremities (84 percent) were the limitations reported most often (Exhibit II.6). Impairments in sensory and communication functions such as seeing, hearing, or speaking were also very common (65 percent), as were difficulties coping with stress (59 percent) and concentrating (55 percent), two abilities that are required for most jobs. It is also noteworthy that 84 percent of beneficiaries reported being limited in their ability to get around (walking three blocks, climbing 10 steps, standing for an hour, and crouching), which might affect their ability to get to and perform a job.

Almost half (47 percent) of beneficiaries reported that their current health is either good or fair (Exhibit II.7). Nearly as many, however, said that their health is poor or very poor (43 percent). Only a small share of beneficiaries reported being in excellent or very good health (10 percent). When asked about how their current health compares to their health in the previous year, about equal shares indicated that their health is the same (43 percent) or worse (41 percent) than last year. Only a small share (16 percent) indicated that their health has improved since the previous year. These data suggest that health conditions may make it difficult for many beneficiaries to pursue employment, particularly if they often require treatment for poor or deteriorating health.

Exhibit II.6. Prevalence of Difficulty Performing Specific Activities

[D]

Source: 2004 National Beneficiary Survey. Sample size = 7,603.

Exhibit II.7. Current General Health and Current Health Compared to Last Year

[D]

Source: 2004 National Beneficiary Survey. Sample size = 7,603.

B. Employment, Reasons for Not Working, and Employment Expectations

While most working-age beneficiaries (87 percent) reported that they worked for pay at some time in their lives (Exhibit II.8), a nontrivial share had never worked at a job for pay (13 percent). At interview, 9 percent of all beneficiaries reported that they were working, and another 6 percent reported searching for a job during the previous four weeks. A somewhat higher share (13 percent) indicated that they had worked for pay for one month or longer some time during the previous year (2003). While only a relatively small percentage of beneficiaries have recently engaged in or actively sought employment (18 percent), as there are about 10 million working-age beneficiaries at any given time, this percentage translates into nearly two million beneficiaries.

We identified numerous demographic and experience characteristics that are at least somewhat predictive of participation (Appendix Table B.20). Among these characteristics, childhood disability onset stands out as a particularly important positive predictor, with employment likelihood six percentage points higher than for other beneficiaries after controlling for other characteristics. Beneficiaries with low to moderate levels of non-SSA benefits are much less likely to be employed than are other beneficiaries. Being only on SSDI benefits or on concurrent benefits, educated beyond high school, between 25-39, and having relatively good mental and physical health were other factors strongly predictive of employment. Strongly and negatively associated with employment were the following characteristics: Social Security benefits greater than $500 per month,¹¹ low to moderate levels of other cash and in-kind benefits ($1-$499), living with a spouse or relatives and having no children, and an ADL or an IADL requiring assistance.

Exhibit II.8. Beneficiary Employment

[D]

Source: 2004 National Beneficiary Survey. Sample size = 7,603. *Recent work activity defined as being employed at interview, having looked for work during the previous four weeks, or having worked during 2003.

When asked why they were not working, an overwhelming majority of beneficiaries (96 percent) reported that a physical or mental health condition prevents them from working (Exhibit II.9). These reports are consistent with the DI and SSI eligibility criteria. Other reasons for not working reflect the challenges many people with disabilities have in the labor market: being discouraged by previous work attempts (30 percent); inaccessibility of workplaces (29 percent); inability to find a job for which they are qualified (28 percent); and the perception by others that they cannot work (28 percent). Only 11 percent of all beneficiaries indicated that a potential loss of cash or health insurance benefits was a reason for not working, although many more might cite this potential loss as a work disincentive if they felt their other barriers could be resolved.

Despite the numerous employment challenges implied by the beneficiary characteristics discussed above, a substantial share of all beneficiaries (30 percent) indicated that their personal goals include getting a job (if not currently working), moving up in a job, or learning new job skills (Exhibit II.10). Overall, 20 percent of beneficiaries see themselves working for pay in the next year. A somewhat higher share (26 percent) see themselves working for pay in the next five years. Overall, only 7 percent of all beneficiaries see themselves as able to earn enough to stop receiving benefits in the next year, but a larger share (15 percent) see themselves able to do it within the next five years. While the percentages of those who see themselves working in the next five years and working enough to leave the rolls in five years are small, they represent very large numbers of individuals – about 2.5 million and 1.5 million, respectively. Even if the reported expectations tend to be overly optimistic, the findings imply that there are a large number of beneficiaries who might benefit from services and policies designed to promote employment. If we focus only on beneficiaries who indicate having future work expectations, among those expecting to work in the next year, 37 percent see themselves earning enough to leave the rolls in the next year, and among those who see themselves working in the next five years, 58 percent see themselves earning enough to leave the rolls in that time frame.¹²

Exhibit II.10. Expectations About Future Employment

[D]

Source: 2004 National Beneficiary Survey. Sample size = 7,603.

C. Use of Health, Employment, and Education Services

1. Service Use in 2003 and Characteristics of Users

Among all working-age beneficiaries, 31 percent reported using services in 2003 for purposes of improving their ability to work or to live independently (Appendix Table C.6).¹³ It is interesting to note that beneficiaries who report being TTW participants represented a only a small fraction (1.4 percent) of all beneficiaries using services in 2003 who resided in Phase 1 states where TTW had been fully implemented for just over one year by the end of 2003. Clearly, a large percentage of beneficiaries use services both in the absence of TTW, and outside of the sponsorship of TTW where TTW is available.

To examine the characteristics of those who used services, we estimated a multivariate (logit) model of the likelihood of using services in 2003 among all beneficiaries, controlling for various sociodemographic, programmatic, and health characteristics (Appendix Table C.21). The model indicates that, all else constant, beneficiaries who used services in 2003 were significantly more likely to be under age 55; have experienced adult onset of disability; have been on the disability rolls for between two and five years; have a high school education or higher; report a mental health or musculoskeletal condition or mental retardation as a main reason for being limited; and have at least one ADL or IADL limitation requiring assistance. They were also significantly less likely to be either black or married with children. The characteristics that have the strongest association with the likelihood of service use are reporting a mental illness as a main reason for limitation (all else constant, this increases the likelihood of using services by 26 percentage points) and having education beyond high school (all else constant, this increases the likelihood of using services by 16 percentage points).

Only a few of the characteristics significantly associated with service use were also significantly associated with the likelihood of employment: age; education; living with a spouse or relatives and having no children; and having an ADL or an IADL requiring assistance. In the case of ADLs/IADLs requiring assistance, the direction of the association is reversed: those needing assistance were significantly less likely to be employed, but were significantly more likely to be using services. The differences in the findings of the analyses of employment and service use likely reflect the fact that a large share of beneficiaries use services mainly for purposes of improving health and functioning, rather than for purposes of employment. We discuss this further in the next section.

2. Reasons for Using Services and Types of Services Among Users

Most beneficiaries who used services in the previous year (2003) report using them for purposes of improving their health (70 percent) or to improve their ability to do daily activities (25 percent) (Exhibit II.11). Only a small share of all beneficiaries report using services for purposes of finding a job (9 percent) or to increase their income (1 percent).

Reflecting the predominance of health-related reasons for using services, the types of services used during the previous year were most frequently health-related (Exhibit II.12). Personal counseling or group therapy was most common (69 percent), followed by occupational, physical, or speech therapy (39 percent). (Although only a small share of service users indicated that finding a job or increasing income was a reason they sought services (Exhibit II.11), much larger shares actually used services specifically geared towards employment: 22 percent of service users received on-the-job training and/or advice about modifying a job; and 20 percent received work assessments and/or assistance finding a job (Exhibit II.12).

Exhibit II.11. Selected Reasons for Using Services Among All Beneficiaries Who Used Services in 2003

[D]

Source: 2004 National Beneficiary Survey. Sample size = 2,775.

Exhibit II.12. Types of Services Used in 2003 Among Service Users

[D]

Source: 2004 National Beneficiary Survey. Sample size = 2,775.

3. School Enrollment and Degree-Seeking Behavior

Only a very small proportion (3 percent) of beneficiaries were enrolled in school at the time they were interviewed (Exhibit II.13). Among those who were enrolled, most (65 percent) were enrolled for purposes of obtaining a degree or professional license.¹⁴ Of those seeking degrees, the largest share (50 percent) were working towards Associate or Undergraduate degrees, followed by 16 percent who were working towards a high school equivalence credential through the General Education Development (GED) testing process or other high school equivalency certification.

D. Unmet Service Needs

All respondents, whether or not they had used any services, were asked whether there were any services, equipment, or supports that they needed in 2003 to improve their ability to work, but did not receive. About 10 percent of all beneficiaries indicated that they did not receive services they thought they needed during the previous year (Exhibit II.14). Again, though the percentage is small, it represents roughly one million individuals. Among those indicating an unmet need for services, the most common reasons for not obtaining services were being ineligible for or refused services (23 percent), inability to afford services (18 percent), and a lack of information about where to get services (16 percent).

E. Summary and Conclusions

The survey data suggest that there is potential demand for employment and employment-related services among Social Security disability beneficiaries. Although at any given time, only a small share of beneficiaries are employed or actively seeking employment, rather substantial proportions have goals that include work and see themselves working in the future. Many even see themselves earning enough to leave the rolls in future. In addition, a good share of beneficiaries used services during the previous year to improve their ability to work and live independently. While most indicated that they used the services primarily to improve their health and functioning, a considerable number received services intended specifically to address employment. Many beneficiaries indicated that they were unable to get needed services for reasons related to a lack of information, inability to afford services, and being ineligible for services. While the percentages of beneficiaries indicating an interest in employment—either through their actions or their expectations—represent a minority of all beneficiaries, they translate into millions of individuals given the size of the federal disability rolls, and thus a potentially large pool of beneficiaries who might benefit from a program like TTW.

The survey data also clearly indicate that a large share of beneficiaries are likely to have difficulty pursuing employment. A very large proportion of beneficiaries are age 55 and older. Even larger shares report having poor or deteriorating health, and experience difficulty performing activities that are essential to most forms of employment, such as getting around outside of the home, concentrating, and coping with stress. In addition, over half of all beneficiaries have been on the rolls for 10 years or longer, and therefore may have lost, or never established, a significant attachment to the labor force.

Finally, the survey data suggest that even if beneficiaries have employment aspirations and attempt to work, many potential challenges to successful employment may need to be addressed. In addition to the activity limitations and poor health associated with their disabling conditions, most beneficiaries have low levels of education that may limit their employment opportunities; most are living at or near poverty, suggesting that they and their families may rely on public programs for which eligibility could be jeopardized by earnings; and rather substantial shares have experienced work-related obstacles such as a lack of reliable transportation¸ inaccessible workplaces, and discouragement from work either by others or through their own experiences.

In summary, while many beneficiaries seem unlikely to use TTW-funded services, our overview of beneficiary characteristics and use of services indicates that there is indeed some potential demand for a program like TTW. A substantial share of beneficiaries indicate an interest in employment, and many of these beneficiaries have needs and challenges that a program like TTW could address. Whether there is sufficient demand for services to support the TTW market will depend on beneficiaries’ decisions to assign their Tickets. We turn to those decisions in the following chapters.

¹ More detailed tabulations of the survey results are shown in Appendix C. Many of the findings presented in this chapter were examined in greater detail in our previous evaluation report (Thornton et al. 2006). In some cases, however, the statistics presented in this chapter differ slightly from those presented in the previous report. This is due to the fact that when the analyses for the previous report were conducted, final survey weights and imputations for selected missing values had not been completed. The statistics presented in this chapter are derived using the final survey weights, and imputed variables where applicable. (back)

² Additional information from the 2004 NBS about the health, work-related goals and activities, service use, and income sources of various subgroups of beneficiaries (SSI-only, DI-only, and concurrent beneficiaries; TTW participants; and working beneficiaries) is presented in Appendix B. (back)

³ Now that the initial rollout has been completed, Tickets are being mailed only to new beneficiaries, whereas during the initial rollout Tickets were mailed to virtually all working-age beneficiaries. As a result, as the TTW program moves into the future the characteristics of beneficiaries with new Tickets will differ substantially from the characteristics of the current population of Ticket holders, which includes all current working-age beneficiaries. (back)

⁴ Data on benefits were obtained from SSA administrative data and appended to the survey data. The reported statistics for the combined monthly federal and state supplement amounts include only federally administered state supplement amounts. State-administered supplements received by a small share of disability beneficiaries are not included. In December 2004, approximately 340,000 blind and disabled SSI recipients were receiving state-administered supplements that averaged about $123 per month (SSA 2005, Table 14). Appendix Table C.2 provides additional information about program eligibility and benefits. (back)

⁵ The extremely high rate of low levels of education corresponds with the findings of a previous study, which found that, in 1999, 34 percent of DI beneficiaries and 54 percent of SSI recipients reported having less than a high school level of education (Martin and Davies 2004). We found similar percentages when beneficiaries were disaggregated by program status (Thornton et al. 2006). (back)

⁶ The percentage of disability beneficiaries with less than a high school level of education (42 percent) is much higher even when compared to the rate in the general population of those ages 25 and older who are not in the labor force (26 percent). Those in the labor force are much less likely to report having less than a high school level of education (10 percent) (U.S. Census Bureau 2004). (back)

⁷ In our previous evaluation report (Thornton et al. 2006) we showed that 65 percent of DI-only beneficiaries and 25 percent of SSI-only recipients report an age at disability onset of 40 or older. (back)

⁸ See, for example, Fox, Borba, and Liu (2005); Blackwell et al. (2003); Yasuda et al. (2002); and Krause et al. (2001). (back)

⁹ Administrative data indicate that among those reporting no limiting conditions at interview, most were awarded disability benefits on the basis of a mental illness (33 percent) or mental retardation (31 percent). (back)

¹⁰ It should be noted that the rates of concurrence between the self-reported conditions and those recorded in the administrative record will be affected both by the degree to which respondents were able to describe their health conditions accurately, and the degree to which the survey interviewers were able to interpret and code the responses appropriately. Also, the administrative data may differ from the survey data because the SSA records list the condition for which establishing eligibility is most straightforward, even if the beneficiary does not see that condition as the most significant barrier to employment or functioning. (back)

¹¹ In the logit models, the variables representing Social Security benefit amounts are defined as the benefits that would be received in the absence of earnings. This is calculated based on the benefit amounts due and countable earnings information obtained from administrative data. As most beneficiaries are not employed at levels that affect benefits, the value of benefits in the absence of earnings is equal to the amount of benefits due. (back)

¹² Author’s calculations based on data shown in Exhibit II.10. (back)

¹³ The NBS solicited information about a broadly defined set of services that beneficiaries saw as helping them to work or to live independently. These included job-search services; medical services; therapy or counseling; and the education or other training needed to secure a new job or to advance in a career. This broad definition was used to reflect the very broad latitude given to ENs and SVRAs to provide services that would help beneficiaries earn their way off the rolls. Throughout this section, the use of the term “services” reflects the many varied supports reported being used by beneficiaries to improve their ability to work or to live independently. (back)

¹⁴ The other 35 percent of beneficiaries enrolled in school at interview indicated that they were “just taking classes.” (back)