To view  the entire collection of publications available through ICI, go to:

Employment Services and Outcomes of People Receiving Welfare Benefits and Vocational Rehabilitation Services

Research to Practice 40


Originally published: 8/2005

This brief report is part of a series examining outcomes and service use of populations of people with disabilities that fit under the rubric of "emerging disability" developed by the National Institute on Disability Rehabilitation and Research. Emerging disability refers to the demographic pressures public systems face as they seek to provide supports for a diverse array of job seekers (Fujiura, 2000; Seelman & Sweeney, 1995).

State vocational rehabilitation (VR) agencies provide employment and independent living services to a range of people with significant disabilities. Many people served by VR agencies are also engaged in other public systems that may or may not be familiar with vocational services for people with disabilities. These other programs are recognizing that a portion of their population has a disability or health condition, and are looking to the VR system to provide additional employment supports. One such agency is the state welfare agency, which offers a variety of public programs to people in poverty.

Poverty programs have undergone substantial reform in the past decade, and there has been a heightened interest in discovering the prevalence of disability among people receiving welfare benefits. Multiple research studies have estimated that approximately one-half to two-thirds of single mothers receiving benefits under the Temporary Assistance for Needy Families (TANF) program have at least one disability (United States General Accounting Office, 2001). State welfare programs have increased their capacity to assist beneficiaries to apply for Social Security benefits and have developed collaborative relationships with state VR agencies (Foley, Marrone, & Simon, 2002; Marrone, Foley, & Selleck, in press).

Many states operate General Assistance (GA) programs that provide cash and in-kind assistance to people who meet income criteria and are either ineligible or waiting for TANF or Supplemental Security Income (SSI) benefits (Anderson, Halter, & Gryzlak, 2002). Income criteria are generally very strict and usually target the "severely poor" (Gallagher, Uccello, Pierce & Reidy, 1999). Disability rates among people receiving GA are high (Gallagher, Uccello, Pierce, & Reidy, 1999; Halter, 1996; Henly & Danziger, 1996) although descriptions of the population are difficult given the paucity of data collected on this population (Gallagher, et al., 1999). The GA population is much smaller than the TANF population in most states. It is also the only poverty program available to adults without children and those who may have health conditions that may not qualify for SSI or who may transfer to SSI.

This report profiles people with disabilities who had TANF, GA, or both at application to VR services and completed these services in the year 2003. We compare the following four groups:

  1. The general VR population, excluding people who receive either TANF or GA
  2. People who received TANF at application
  3. People who received GA at application
  4. People who received both TANF and GA at application

(The last three groups are not mutually exclusive.)

The data used for analysis came from the Rehabilitation Services Administration National Case Service Report (RSA-911) for fiscal year 2003.



About 4% of people using VR services were receiving TANF benefits at application, and about 4% were receiving GA at application. A very small percentage was enrolled in both programs. The TANF population was notably different from the general population in that 78% were women and the group had greater racial and ethnic diversity. Members of the GA population were more often men, and there was a notably higher frequency of Latino GA recipients than appeared in either the general population or the TANF population. All four populations had an average age in the middle to late thirties. Very few people who use VR services had a college degree; even so, people receiving TANF appeared to have fewer years of education than those in the general population or those receiving GA. About 4 out of 10 people receiving TANF had less than a high school education. While 20% of the general population were working at application, only 6% or less of the people receiving TANF, GA, or both were working at application.

Table 1
Demographics at Application (2003)
General VR population TANF GA Both TANF and GA
Race (N)569,01124,48225,9541,267
Non-Hispanic Black (%)21.434.033.933.7
Non-Hispanic White (%)67.351.248.451.9
Hispanic (%)8.911.714.611.8
Native American (%)
Asian/Pacific Islander (%)
Gender (N)571,89524,53926,0491,268
Female (%)44.277.743.879.6
Age (N)573,28424,53026,0481,268
Mean age37.035.939.935.5
Education (N)538,75723,50224,5531,222
Less than high school (%)31.840.833.238.1
High school degree or equivalency (%)
Some college (%)14.011.714.011.0
College degree or more (%)
Work status at application (N)538,75723,50224,5531,222
Not working (%)80.294.596.194.0
Working (%)

Cause of Primary Impairment

RSA-911 includes a list of causes of impairment and then the impairments themselves. For example, a cause may be accident/injury and the impairment may be spinal cord injury. The top five leading causes of the primary impairment for each population are shown in Table 2. For both of the populations in poverty programs, the causes included depression and mood disorders, and accident or injury. The GA population appeared to be largely people with addictions and depression or mood disorders. The high rate of accident and injury may be related to drug and alcohol abuse, although this could also include violence, vehicular accidents, and other trauma. The TANF population appeared to be people with depression, mood disorders, learning disabilities, and cognitive disabilities. Many researchers have discussed the high rate of domestic violence among people receiving TANF, and one might question if the high rate of accident/injury could be an indicator of this. Alcohol and drug abuse did not make it into the top fivecauses of primary impairment for the TANF population using VR services. A surprising 8% of people receiving TANF had mental retardation listed as the cause of their primary impairment.

Table 2
Cause of Primary Impairment in Rank Order (2003)
Rank General VR population N=546,424 TANF N=23,489 GA N=25,232 Both TANF and GA N = 1,228
Impairment % Impairment % Impairment % Impairment %
1 Accident/injury (other than TBI or SCI) 10.9 Depressive and other mood disorders 16.6 Drug abuse or dependence (other than alcohol) 21.5 Depressive and other mood disorders 20.6
2 Specific learning disabilities 10.2 Accident/injury (other than TBI or SCI) 12.2 Depressive and other mood disorders 14.3 Accident/injury (other than TBI or SCI) 13.0
3 Cause unknown 9.8 Cause unknown 10.1 Alcohol abuse or dependence 9.8 Specific learning disabilities 8.1
4 Mental retardation 8.6 Mental retardation 8.1 Accident/injury (other than TBI or SCI) 9.1 Drug abuse or dependence (other than alcohol) 6.6
5 Schizophrenia and other psychotic disorders 5.8 Specific learning disabilities 8.0 Cause unknown 5.9 Mental retardation 5.9

Primary Impairment

The top five primary impairments across all groups were cognitive, psychosocial, other mental impairments, and other physical or mobility impairments. The predominant type of disability across all four groups was psychiatric disability. About 30% of the general population, 37% of the TANF population, 56% of the GA population, and 42% of the population in both poverty programs had psychiatric disabilities. About one-quarter to one-fifth of the general and TANF populations had a cognitive impairment. The other physical and orthopedic impairments in all four groups may be related to accidents and injuries sustained.

Table 3
Cause of Primary Impairment in Rank Order (2003)
Rank General VR population N=546,424 TANF N=23,489 GA N=25,232 Both TANF and GA N = 1,228
Impairment % Impairment % Impairment % Impairment %
1 Cognitive impairments 23.3 Psychosocial impairments 25.9 Psychosocial impairments 28.9 Psychosocial impairments 28.3
2 Psychosocial impairments 19.7 Cognitive impairments 20.0 Other mental impairments 27.6 Cognitive impairments 18.5
3 Other mental impairments 10.4 Other mental impairments 11.9 Cognitive impairments 11.0 Other mental impairments 14.3
4 Other physical impairments 8.6 Other physical impairments 8.3 Other physical impairments 7.0 Mobility orthopedic/ neurological impairment 7.2
5 Mobility orthopedic/neurological impairments 6.4 Other orthopedic impairments 7.9 Other orthopedic impairments 5.4 Other orthopedic impairments 5.9

Economic Indicators

The TANF and GA populations received SSI or Social Security Disability Insurance (SSDI) less frequently than the general population at both application and closure. However, about 12% of the GA/TANF population received SSI at application, which indicates a substantial amount of public income support for this group. Although people receiving both GA and TANF were a very small population, they appeared to be involved with multiple income support programs. One might suppose that this group included families in extreme poverty and who possibly were homeless. About 10% of the TANF population received SSI at either application or closure. These were likely families in which the adult was enrolled in SSI and the children received TANF; however, RSA-911 data provides no information about family size, marital status, or the presence or age of children. At the most, only one-tenth of families receiving TANF who used VR services could be listed as "child only" families. The low rate of SSDI receipt among the three poverty groups suggested limited work history for each population.

More than half of those receiving TANF were enrolled in Medicaid at application, and about one-third of those enrolled in GA were receiving Medicaid at application. About half of those receiving TANF at application were still receiving TANF at closure. Similarly, about half of those receiving GA at application were receiving GA at closure. People in poverty programs enrolled in Medicaid likely entered through the welfare programs rather than disability programs. In fact, one might suspect that the numbers should be higher for Medicaid receipt among these three populations than for the general VR population.

Of interest is that 40% of the TANF population was not receiving TANF at closure of VR services. This also appeared true of the GA population as well. TANF services are time-limited, and it is unclear from this data whether receiving VR services influenced TANF receipt. The question merits further investigation.

Table 4
Economic Indicators at Application and Closure (2003)
General VR population TANF GA Both TANF and GA
Social Security (N)573,33024,53026,0491,268
SSI at application (%)15.810.57.911.4
SSI at closure (%)67.351.248.451.9
SSDI at application (%)
SSDI at closure (%)
Insurance (N)547,56123,01624,7421,266
Medicare at application (%)
Medicare at closure (%)
Medicard at application (%)21.464.143.564.2
Medicard at closure (%)20.255.541.647.9
Other Public Assistance (N)561,99322,72823,0731,133
TANF at application (%)01005.0100
TANF at closure (%)0.359.34.555.5
GA at application (%)05.6100100
GA at closure (%)0.44.454.347.2

Use of VR Services

Table 5 provides service utilization frequencies by service type. For the purposes of this report, service utilization is displayed for everyone who applied to VR. This is a very conservative estimate, as many people do not make it to the service delivery stage (i.e., their case is closed prior to instituting an individual plan for employment [IPE]). The purpose of this table is to show what percentage of people who approached VR (through state agency referral or any other referral type) received a particular service.

The average (mean) cost of purchased VR services was lower for the TANF population (by about $1,000) and the GA population (by about $300) than for the general population. The VR agency may represent one of several sources for employment services available to the TANF and GA populations. The majority of individuals in all groups received assessment and a substantial percentage in each group received rehabilitation counseling. The general population received diagnosis and treatment more frequently than those on TANF and GA. About one-third of the GA population received transportation. The TANF and GA populations received job placement services, job search services, and on-the-job training with less frequency than the general population.

Table 5
Services Received (2003)
General VR population (N=573,330) TANF (N=24,530) GA (N=26,049) Both TANF and GA (N=1,268)
Mean cost of purchased services$2,180$1,239$1,790$1,191
Services provided (%)
Rehab counseling & guidance47.539.036.048.0
Diagnosis and treatment30.519.019.627.5
Job placement service17.413.415.810.2
Job search service17.012.414.015.5
Other services15.
Occupational/vocational training8.57.910.19.7
Miscellaneous training7.
Job readiness training7.
On-the-job supports9.
Disability-related training2.
On-the-job training2.


The TANF population and the GA population achieved an employment outcome less frequently than the general population. About one-third of people in the general VR population were closed out of VR prior to receiving services, whereas about one-half of people receiving TANF, GA, or both were closed out of VR prior to receiving services. Note that although a higher percentage of people receiving TANF left VR services prior to service delivery, their reasons for closure were similar to those of the general VR population. Of those who disengaged:

  1. 12% left because the disability or condition was too significant to benefit from VR services, a disabling condition did not exist, or the condition did not merit VR services.
  2. 19% were closed because the VR caseworker could not locate or contact the person.
  3. 25% refused services or further services.
  4. 25% were listed as failing to cooperate
Table 6
Distribution of Closure (2003)
General VR population (N=573,330) TANF (N=24,530) GA (N=26,049) Both TANF and GA (N=1,268)
Type of closure (%)
Employment outcome35.520.425.222.7
Services, no employment outcome24.326.228.727.4
After eligibility, before IPE21.626.828.223.8
Before eligibility determined15.523.415.024.0

Successful Closures and Earnings

Of those people on either TANF, GA, or both who closed into an employment outcome, a slightly higher percentage closed into integrated employment as compared to the general population. This may have to do with a comparatively lower utilization of supported employment for these groups. Weekly earnings were lower for each of the poverty populations than for the general population. Translated into yearly earnings and assuming a 52-week year, the TANF population averaged about $14,500 per year; the GA/TANF population about $14,700; the GA population about $16,600; and the general population about $17,400. In 2003, the year of the data, the federal poverty level was $15,260 for a family of three (Federal Register, 2003).

Table 7
Employment Outcomes at Closure (2003)
Employment outcomes General VR population (N=203,859) TANF (N=5,018) GA (N=6,574) Both TANF and GA (N=288)
Integrated employment%84.488.387.890.6
Mean weekly earnings$334.57$279.32$318.47$282.97
Mean weekly hours34333533
Self-employment (except BEP)%
Mean weekly earnings$300.76$252.51$245.50$318.00
Mean weekly hours28262428
Supported employment%
Mean weekly earnings$168.73$207.39$211.98$230.25
Mean weekly hours24292827
Homemaker and unpaid family worker%4.44.0 3.33.4
Mean weekly earningsN/AN/AN/AN/A
Mean weekly hoursN/AN/AN/AN/A


The RSA-911 data give a profile of the people using VR services who are also enrolled in TANF and/or GA.

The TANF Population

The TANF population was largely comprised of women in their mid-thirties with a high school education or less who were not working at application. About one-tenth also received SSI, and most had depression, accident/injury, mental retardation, or learning disabilities as the cause of their primary impairment. The primary impairments were generally cognitive, psychiatric, and other physical disabilities. About one-half of those who received TANF at application did not receive TANF at closure. Only 20% of the population closed into an employment outcome and, of those who did, most had weekly earnings that put them below the federal poverty line for a family of three. Those closed into integrated employment worked an average of 33 hours per week. About one-half of the people who applied to VR and received TANF eventually received services from the state VR agency. Reasons for closure suggested that people left VR prior to receiving services for reasons similar to those given by the general VR population.

Policy Implications

  • In general, VR agencies should look at what it takes to keep this population engaged in services. Although the reasons for disengagement appeared similar to the general VR population, the proportion of people leaving was higher for those receiving TANF than for those who did not. VR services are voluntary and based upon the stated choices of individuals served. It is a very different cultural norm than the welfare system, which has work requirements, sanction policies, and time limits. Improved engagement may lead to improved cross-agency service delivery, and it may also lead to an increase in the number of people entering competitive employment.
  • A significant percentage of people receiving TANF who used VR services had mental retardation listed as their primary impairment. To what degree are state welfare agencies and VR agencies working with the state mental retardation/developmental disabilities (MR/DD) agency? People with intellectual disabilities may have access to additional employment services and supports through the state MR/DD agency.
  • The welfare populations also received fewer employment services than the general VR population, at least from VR. They had access to multiple employment programs, which makes it difficult to determine the total package of employment supports received. Research that can look across programs would be useful to provide a picture of the constellation of services. For instance, most systems do not have comprehensive employment services for single mothers with disabilities, indicating that many agencies will be involved in case management.
  • Welfare receipt declined between application and closure. This suggests that VR services may assist people to leave TANF. It is also true that TANF is a time-limited benefit and that people may actually leave TANF prior to completing their VR services and acquiring an outcome. Further investigation of this situation would be useful to determine if VR is helping people with disabilities leave welfare.
  • Of those that left VR with an integrated employment outcome, 88% worked an average of 33 hours per week. Current policy debates in welfare question whether or not people with disabilities can work at all, and some question whether they can work even 20 hours per week. With this in mind, it is important to note that a substantial percentage of TANF recipients using VR services (who were likely to have more significant disabilities than those that do not qualify for VR services) worked more than 20 hours per week.
  • Earnings are a concern. Efforts to improve the weekly wages of this population should be investigated. The RSA-911 data does not provide information about household income, only individual income. Therefore, it is difficult to know the financial means of the population. However, by virtue of receiving TANF, this population includes parents with disabilities and likely single mothers. Some might be combining earnings with SSI income. The average weekly earnings of those receiving TANF were below the federal poverty line for a family of three.

The GA Population

The GA population was the most racially and ethnically diverse of the four groups. About one-third was African-American and about 15% Latino. The majority were men with an average age of 40. Their educational level was similar to the general VR population but with fewer college graduates. Only 4% were working at the time of application. Most appeared to have drug and alcohol addictions, depression and other mood disorders, and impairments due to accidents and injuries. This demographic profile suggests that many may be homeless. This group had a lower utilization of SSI or SSDI than any other group, and very few received Medicare. About half received Medicaid. A little more than one-quarter received transportation services through VR, which was much higher than any other group. One-quarter closed to an employment outcome, but about two-thirds left VR after eligibility determination but before reaching an outcome. The GA population earned about $40 more per week than the TANF population that closed to integrated employment.

Policy Implications

  • The GA population is likely transient, and long-term contact with a public employment program may be difficult. The racial and ethnic diversity may suggest that this population is likely to be urban. It may also be reasonable to conclude that the group receives services from multiple public programs, including veteran's services, the department of corrections, the state mental health agency, and alcohol and substance abuse programs. This population is just as large as the TANF population in the VR program and should not be overlooked despite the emphasis on TANF programs in most states.
  • Not every state offers general assistance. VR agencies may want to consider examining employment outcomes for people with a similar profile, including those with substance abuse, psychiatric disabilities, low rates of SSI or SSDI receipt, and limited work history.

The GA/TANF population

This population represented a very small group using VR services. However, given the demographic profile, it is likely that a significant percentage was comprised of homeless families with children. Most had psychiatric or cognitive impairments, and there was some prevalence of substance abuse. About 1 in 10 received SSI. Of those that acquired an employment outcome, 90% closed to integrated employment, and they had higher wages than the TANF-only group. Despite the small size of the population, this is likely to be a highly disadvantaged group that receives intensive services from multiple programs other than VR and welfare, including homeless shelters, corrections, mental health agencies, substance abuse providers, child protective services, and domestic violence shelters. TANF receipt indicates that these are families with young children. An important social policy goal is to understand and improve the cross-system constellation of services received by these families.


Anderson, S.G., Halter, A.P., & Gryzlak, B.M. (2002). Changing the safety net of last resort: Downsizing general assistance for employable adults. Social Work, 47(3), 249-258.

Federal Register, Vol. 68, No. 26, February 7, 2003, pp. 6456-6458.

Foley, S., Marrone, J., & Simon, M. (2002). Cruise ships and kayaks: Welfare and rehabilitation approaches for women with disabilities in poverty. Behavioral Sciences and the Law, 20(6), 659-680.

Foley, S.M. (2002, April). Preliminary findings from the national survey of state systems and employment. Paper presented at the meeting of the Council of State Administrators of Vocational Rehabilitation, Bethesda, MD.

Fujiura, G. (2000). The implications of emerging demographics: A commentary on the meaning of race and income inequity to disability policy. Journal of Disability Policy Studies, 11(2), 66-75.

Gallagher, L.J., Uccello, C.E., Pierce, A.B., & Reidy, E.B. (1999). State general assistance programs 1998. Washington, DC: Urban Institute.

Halter, A. (1996). State welfare reform for the employable poor: The facts behind the assumptions. Social Work, 41(2), 106-110.

Henly, J.R., Danziger, S.K. (1996). Confronting welfare stereotypes: Characteristics of general assistance recipients and post-assistance employment. Social Work Research, 20(4), 217-228.

Marrone, J., Foley, S.M., & Selleck, V. (in press). How mental health and welfare reform interact in the context of welfare-to-work: The need for hope, hassling, systems, and support. American Journal of Psychiatric Rehabilitation.

Seelman, K, & Sweeney, S. (1995). The changing universe of disability. American Rehabilitation, 21(3/4), 2-18.

United States General Accounting Office. (2001). Welfare reform: Moving hard to employ recipients into the workforce (GAO-01-368). Washington, D.C: Author.

For further information on this publication, please contact the lead author at 617/287-4317 or

This is a publication of the Emerging Disabilities, Employment Outcomes, and Systems Change Project (H133A021503) funded by the National Institute on Disability Rehabilitation and Research at the U.S. Department of Education.