Mr. Tremblay and Mr. Smith are affiliated with the Division of Vocational Rehabilitation, State of Vermont, 103 South Main Street, Weeks Building 1A, Waterbury, Vermont 05671-2303 (e-mail, firstname.lastname@example.org). Dr. Xie and Dr. Drake are affiliated with the Psychiatric Research Center, Dartmouth College, Lebanon, New Hampshire.
A major reason for the high unemployment rate among Social Security Administration (SSA) disability beneficiaries is fear and misunderstanding about the impact of employment on their benefits (1,2). Beneficiaries often do not work or are underemployed because they fear a loss of benefits, especially health care coverage, or they are unaware of or misunderstand available work incentives built into the SSA disability programs (1,2). This situation is a particular problem for beneficiaries with mental illness, whose disabilities may be less visible and less predictable than those of other persons with disabilities and for whom substantial work activity may be more likely to lead to a cessation of public benefits, health coverage, or both (1). Fear of losing benefits also may be a contributing factor in the finding that vocational service consumers with psychiatric disabilities have shown smaller earnings gains over time than the general population of people with disabilities (3).
One response to this problem has been to help SSA disability beneficiaries to understand their benefits and make optimal decisions about employment by providing specialized benefits counseling. A benefits counselor is a staff person who is expert in benefits issues and focuses exclusively in this area (4). SSA rules for the Supplemental Security Income (SSI) and Social Security Disability Insurance (SSDI) programs are extremely complex and difficult to understand. The benefits counselor position was conceived to provide an expert consultant on state and federal benefits for both the consumer and the supporting professional.
Recently, SSA's State Partnership Initiatives, the Ticket to Work and Work Incentives Improvement Act, benefits planning and outreach grants, and other federal-state vocational rehabilitation programs (5) have promoted specialized benefits counseling services nationwide. Despite the extensive implementation of benefits counseling services, however, quantitative outcome evaluation efforts have until recently been minimal, as measured by the current lack of published studies. Even less evidence is available for the effectiveness of those programs for individuals with mental illness. Vermont has used its SSA-funded State Partnership Initiative to create a specialized benefits counseling service for vocational rehabilitation. The purpose of this article is to report employment outcomes for participants in the Vermont program who have been identified as having a psychiatric disability.
To evaluate the effects of Vermont's benefits counseling program for persons with psychiatric disabilities, we used a quasi-experimental study. Benefits counseling services were provided as part of the Vermont State Partnership Initiative project funded by the SSA. Division of Vocational Rehabilitation consumers who were identified by the division as having a primary disability of mental illness and who received specialized benefits counseling were compared with two matched groups comprising concurrent and past participants in the vocational rehabilitation system who did not receive benefits counseling. Earned income was compared for each group over four years.
Benefits counseling intervention
The Work Incentive Initiative program of the Vermont Division of Vocational Rehabilitation began providing specialized benefits counseling services in 1999. Eleven benefits counselors, located in Division of Vocational Rehabilitation offices and community mental health centers throughout Vermont, provided benefits counseling services to persons receiving disability benefits from the SSA under the SSDI program or the SSI program. SSDI is a federal insurance program for workers with disabilities, whereas SSI is a program for low-income people with disabilities who do not have a sufficient work history to be eligible for SSDI. Referrals for services primarily came from vocational rehabilitation and community mental health agency counselors and were made for individuals who were actively seeking employment that could potentially affect their public benefits. In Vermont there is an unusually high level of collaboration between vocational rehabilitation and the state community mental health centers. The benefits counseling intervention was integrated with existing vocational rehabilitation services and supported employment services for the intervention group. As such it is best conceptualized as a supplemental service designed to enhance vocational rehabilitation and supported employment services.
Benefits counseling may be understood as a specialized form of financial planning and consultation for people with public cash and health care benefits. Benefits counseling services are individualized, longitudinal, and of variable intensity. These services include general education of beneficiaries regarding SSA disability programs; the various work incentives available under those programs, such as SSI Earned Income Offset, impairment-related work expenses, and trial work periods; and other federal and state public benefits. Such benefits include Medicaid, housing assistance, food stamps, and other state health care programs. Benefits counseling also includes individualized research and counseling regarding enrollees' current benefits packages and how those benefits would be affected by work activity and changes in earnings; assistance to the enrollee in managing his or her benefits through the transition to employment; and provision of information to supporting professionals, such as vocational rehabilitation counselors and mental health case managers. The goals of benefits counseling were to maximize beneficiaries' income earned from work and to reduce dependency on public benefits. The duration and extent of benefits counseling varied by individual according to the beneficiary's needs and preferences, but counseling was available throughout the duration of the project. Follow-up information was gathered from enrollees at least quarterly, and enrollees commonly sought and received follow-up benefits counseling services as new benefits-related questions arose or as their personal situations changed—for example, when receiving pay raises or getting married. The amount of counseling varied by individual but averaged approximately eight hours per consumer.
Sampling pool. The three study groups for this evaluation were drawn from a common sampling pool of individuals who were current and former Division of Vocational Rehabilitation consumers (approximately 95 percent of benefits counseling enrollees in Vermont are current or former consumers of vocational rehabilitation services), had received SSA disability benefits, and had been identified by the Division of Vocational Rehabilitation as having a primary disability of mental illness. This sampling pool was used to obtain comparison groups drawn from populations as similar as possible to project enrollees, given that all vocational rehabilitation consumers have actively sought employment supports. All benefits counseling participants gave informed consent for their involvement in the demonstration project, all outcome data were obtained from preexisting state administrative records, and the analysis was conducted as a component of ongoing program evaluation of public benefit and service programs by the State of Vermont under the review of the SSA.
Matching procedures. The two matched comparison groups consisted of vocational rehabilitation consumers who had not participated in the benefits counseling program: a "contemporaneous" group and a "historical" group. Selection of pairwise matches was based on the start date of each person's vocational rehabilitation services to control for the period in which employment support was received and thus control for statewide macroeconomic and historical effects. The purpose of the contemporaneous group was to minimize historical confounds in examining outcomes, and the purpose of the historical group was to provide a replication of the comparison but with a temporally distinct sample.
For the contemporaneous comparison group, each match to a benefits counseling participant was drawn from the pool by selecting a noncounseled consumer of vocational rehabilitation services whose vocational rehabilitation start date was closest to that of the project participant. For the historical comparison group, each match was drawn from the pool by selecting a noncounseled vocational rehabilitation consumer whose start date was most similar to that of the participant's but minus 2.5 years, which was the earliest time point for which outcome data were available.
Because duration of exposure to vocational rehabilitation services is related to job readiness among vocational rehabilitation consumers, pseudoenrollment dates were assigned to all contemporaneous or historical comparison group members in a manner to parallel the length of time between the benefits counseling participant's vocational rehabilitation start date and that participant's benefits counseling enrollment date. All outcomes were compared on a time scale relative to an individual's enrollment or pseudoenrollment date. The goal of this matching process was to control for both time of entry into vocational rehabilitation services and maturation in the vocational rehabilitation service system.
Resulting study participants. As of December 31, 2002, the benefits counseling program had 364 enrollees who were SSA disability beneficiaries (SSDI, SSI, or dual eligibility), were between the qualifying ages of 18 and 60 years, had received vocational rehabilitation services, and had a vocational rehabilitation primary disability of mental illness. As a result of the pairwise matching procedures, the contemporaneous and historical comparison groups each included 364 persons. Table 1 presents demographic characteristics of the three groups, and, as shown, the three study groups were similar on the basic demographic variables of minority status, sex, and age.
Prior work history, one of the strongest predictors of future earnings for individuals, was controlled for directly in the regression model, rather than through matching. This was accomplished through statistical adjustments for group differences in baseline earnings and in proportions of SSA beneficiary types, that is, SSDI, SSI, or both.
The primary outcome for this analysis was earnings from competitive employment, based on administrative records of Vermont's unemployment insurance program. This information is submitted by employers to the Vermont Department of Employment and Training as quarterly tax reports, which are subject to state unemployment insurance laws and the federal employees program. Both public- and private-sector workers are included in this system. Omissions include earnings from self-employment or from out-of-state work and the following employee groups: elected officials, religious nonprofit organizations, charitable and educational organizations, unpaid family members, farm workers (with some exceptions), and some railroad employees. On the basis of self-reports of project enrollees who worked, approximately 5 percent were self-employed and 6 percent were working out of state, constituting the main categories of noncovered earnings. Although this data set does not include all earnings, it covers an overwhelming majority of wage earnings in Vermont and was used in this study as an economic indicator variable for group comparisons. The earnings obtained from state unemployment insurance records are in quarterly increments, and the data were obtained as of December 2003.
Before the analysis, two conversions were applied to the earnings data. First, all monetary amounts were converted to 1997 (baseline) dollars to adjust for inflation effects across time. Second, the calendar dates associated with each earnings record were converted on a person-by-person basis to time relative to the individual's date of enrollment in the benefits counseling program or, for comparison group members, relative to the date of pseudoenrollment. Thus, for an individual with an enrollment or pseudoenrollment date of May 15, 2000, earnings reports for the first, second, and third calendar quarters of 2000 would be translated into reports for the first quarter before the quarter of enrollment, the quarter of enrollment, and the first quarter after the quarter of enrollment, respectively. For the group comparisons, all records for the first quarter after the quarter of enrollment were compared with other records for the first quarter after the quarter of enrollment, and so on. These two temporal conversions allowed group comparisons of intervention effects over time for a program with rolling enrollments and for an intervention with multiple baselines by individual.
To analyze the differences in group earnings outcomes, we used mixed-effect linear models (6,7,8,9). Such models are an extension of traditional regression techniques and are especially suitable for the examination of longitudinal data and multilevel data, given that individuals may have begun and ended their participation at different points in time. Autocorrelation is assumed for repeated-measures observations. The outcome variable for the model was earnings. Group (intervention, contemporaneous, or historical) and time (eight baseline quarters and eight postenrollment quarters) were treated as explanatory variables, and group-by-time interaction effects (for three groups and 16 quarters) were included in the model. Age, sex, and benefit type (SSI, SSDI, and dual eligibility) were treated as covariates. Preexisting differences between the earnings levels of participants and eligible non-participants were taken into account by examining changes in the earnings gaps among groups.
Figure 1 shows the longitudinal plots of quarterly earnings over four years, two years before and two years after the initiation of benefits counseling, and before and after the pseudoenrollment date for comparison group participants. The earnings of the benefits counseling intervention group clearly increased after the initiation of benefits counseling. The average quarterly earnings for this group ranged between $464 and $612 during the preintervention period but then increased to a nonoverlapping range of $765 to $928 in the postintervention period. In contrast, average earnings for the contemporaneous and historical comparison groups remained at roughly the same levels as they had been at baseline. Earnings for the contemporaneous group ranged from $382 to $575 in the preenrollment period and from $353 to $602 in the postenrollment period. Earnings for the historical group ranged from $284 to $511 in the preenrollment period and from $290 to $384 in the postenrollment period.
Results of the linear mixed-effects regression model presented in Table 2 show that virtually all the group comparisons favored the intervention group. Pre- versus postintervention changes in group means, equivalent to group-by-time interaction effects, showed a significant increase in earnings over time for the intervention group but no significant change for either of the two comparison groups. Even taking into account the differences in baseline earnings among the three groups, the intervention group improved significantly more than either the contemporaneous or historical comparison groups. That is, even though the three groups started at different levels of earnings and each varied over the four years, the improvements for the benefits counseling group were significantly greater than any (nonsignificant) improvements demonstrated by the two comparison groups. The increase in the earnings gap between the intervention group and the contemporaneous comparison group was an adjusted average of about $1,256 per person per year (or $314 per person per quarter), and the increase in the earnings gap between the intervention group and the historical comparison group was an adjusted average of a little more than $1,272 per person per year (or $318 per person per quarter). The same pattern of results was obtained in more conservative tests of applying the model separately to the comparison of the intervention and contemporaneous group alone and then to the intervention and historical group alone: $1,260 per person per year difference for the contemporaneous comparison (p<.001), and $1,268 per person per year difference for the historical comparison (p<.001). The adjusted averages include zero earnings for roughly half of the intervention recipients who were unemployed, so adjusted average increases for only the employed individuals are roughly double those amounts.
The model in Table 2 also shows the effects of some significant covariates. Younger participants and men showed greater earnings, as did those with SSDI rather than SSI or dual-eligibility status.
The main finding of this analysis was that significant increases in mean earnings for Social Security disability beneficiaries with psychiatric disabilities were associated with the receipt of benefits counseling services, even after the analysis controlled for preexisting earnings advantages and for key demographic predictors of earnings such as age, sex, and Social Security beneficiary type. This finding suggests that benefits counseling as an employment intervention may improve earnings outcomes for participants, above and beyond whatever increments might be attributed to services from vocational rehabilitation and the local community mental health agency.
As noted, benefits counseling is best conceptualized as a supplemental service designed to enhance vocational rehabilitation and supported employment interventions for people who receive SSI and SSDI benefits. All three groups—the intervention group and the contemporaneous and historical comparison groups—received vocational rehabilitation services, and many received supported employment services. However, the greater increase in earnings of the intervention group suggests that the benefits counseling added supplemental value to the vocational rehabilitation and supported employment interventions. Benefits counseling consumers were better informed about the potential impact of work on their benefits and therefore were better able to take advantage of available work incentives. As a result they chose to work at a higher level on average than the comparison groups.
Among other predictors, the earnings of women and older persons lagged behind those of other individuals, and SSDI beneficiaries consistently showed greater earnings than either SSI or dual SSDI-SSI beneficiaries.
Several limitations warrant mention. Most important, the lack of random assignment leaves open the possibilities of selection bias and other uncontrolled effects of group non-equivalency. Although this study attempted to control for those problems through a combination of careful matching and statistical adjustments, such threats to internal validity are best handled by random assignment. As a next-best approach to control for self-selection and motivational biases, however, only recent vocational rehabilitation participants were included in the study.
Second, the SSA introduced a number of new work incentives for disability program recipients during the course of the benefits counseling intervention. Although the influence of most of these changes was controlled for in the contemporaneous comparison, one exception was an SSI waiver program that was introduced in Vermont in April of 2001, which was available only to benefits counseling enrollees. Any possible confound between the benefits counseling program and the SSI waiver would not account for the results of this study, however, because the greatest earnings increases observed here were obtained by SSDI-only participants.
Third, the limited participation of individuals from ethnic minority groups, although representative of Vermont's population, must be considered in assessing the generalizability of these results. A final limitation lies in the nature of the earnings data used in this analysis. Vermont unemployment insurance records do not include several categories of earnings. Thus the subset of earnings examined in this analysis, while encompassing a large majority of all wage earnings, must be considered a strong economic indicator rather than a precise or complete tally of total wage earnings.
This study provides empirical evidence that individualized benefits counseling services are associated with significant increases in earnings for SSA disability beneficiaries with psychiatric disabilities. It suggests that benefits counseling should be a fundamental component of evidence-based employment services for persons with psychiatric disabilities if participants are to maximize earnings. Future research on the impact of benefits counseling should include larger samples, random assignment, and alternative outcome variables such as utilization of public benefits.
This study was supported by grant 12-D-70346-1-05 from the Social Security Administration and the Substance Abuse and Mental Health Services Administration, by grant 11-P-91237 from the Centers for Medicare and Medicaid Services, and by grant 035436 from the Robert Wood Johnson Foundation.