Economic Impacts of the DRS Program on Consumers with Significant Disabilities in West Virginia

Transcription

1 Economic Impacts of the DRS Program on Consumers with Significant Disabilities in West Virginia A Realistic Return-on-Investment Model for State-Federal VR Programs Department of Education and the Arts West Virginia Division of Rehabilitation Services January 2011

2 DRS-ROI: Page 1 Economic Impacts of the DRS Program on Consumers with Significant Disabilities in West Virginia: A Realistic Return-on-Investment Model for State-Federal VR Programs Introduction West Virginia Division of Rehabilitation Services 1 The West Virginia Division of Rehabilitation Services (DRS) is a state-federal vocational rehabilitation (VR) program with the mission of promoting quality employment outcomes and independence for persons with significant disabilities in the state. The primary goal of DRS is to optimally utilize its fiscal and human resources to vocationally rehabilitate consumers with significant disabilities through effective and efficient management practices. DRS achieves this by providing qualified consumers the necessary VR and related services through our counselors, staff members, and service providers. This investment in the provision of services enables DRS consumers to ultimately achieve, return to, or maintain competitive employment and become more independent. The agency expects the investment in services to be administered over time to match the needs of consumers. The return on the investment is to be reflected in current and potential earnings of DRS consumers working in an integrated setting after receiving services under the individualized plan(s) for employment and exiting the VR system. In addition to the impact of DRS services on the current and potential earnings of its consumers, the post-vr earnings generated by consumers also produce tax revenues for the federal and state government and reduce Social Security benefit payments (e.g., SSI and SSDI) for the society. The gains in taxes, the savings in social benefit payments, and the current and potential earnings of DRS consumers represent benefits to the society. These benefits are incorporated in the return-oninvestment (ROI) model to measure the broader DRS impact on the societal optimization of fiscal and human resources. DRS uses the ROI model to empirically validate a strong and continuous return on investment for its VR program that provides VR and related services to persons with significant disabilities in West Virginia. DRS also uses the ROI statistics to annually measure the efficiency and accountability of the program. Literature Review Several prior studies have examined the costs and benefits of vocational rehabilitation and each can approach the subject from a unique angle. Table 1 demonstrates the various attributes of several of these projects: 1 This report was prepared by Pisnu Bua-Iam and Dr. Thomas K. Bias of the State Plan and Program Evaluation Unit, West Virginia Division of Rehabilitation Services. Special thanks to Director Deborah Lovely and members of the DRS Executive Management Group for their strong support of this project and to Basil White Jr., E. Susan Newhouse, and Beverly Shamblin for outstanding computer programming and data entry supports.

3 DRS-ROI: Page 2 Table 1: Attributes of the Literature Author (Year) Description Greenblum (1975) Used a linked database between SSA and RSA and presented wage and public benefit changes for approximately two years after closure. Rogers et al. (1995) Used a self-reporting sample of those with psychiatric disabilities to present wages and social security benefit changes over time. Estimated taxes using 23% of gross income. The cost-benefit ratio compared all this information to the total costs of a supported employment program. Gibbs (1991) Discussed different methodological approaches to outcome measures within vocational rehabilitation. Noted that using pre-program data is inadequate, as those who seek vocational rehabilitation may have a sudden need for services not reflected in wage or benefit data prior to application. Kenyon et al. (2003) This research utilized a survey instrument to determine wages for clients who closed from the Australian vocational rehabilitation program. It included taxes, public assistance, and wages in benefits for the cost-benefit analysis. The study also approached the issue of other non-quantifiable social benefits such as self-esteem. Cimera (2009) Used the RSA-911 application and closure data to examine VR clients with mental illness and their performance at closure using national data. Cimera made the important note that findings for one state cannot be generalized to other states because of economic and administrative differences. Hemenway & Rohani These studies (for the states of Florida and Massachusetts) used U/I (1999) and Uvin et al. data for 8 years and compared wages, social security (from RSA-911 (2004) and carried forward in time), taxes (estimated as 23% of gross income) between those who received services (Status 26 and 28) and those who did not (Status 30). Using the difference between the two groups as the benefit, they projected a 30 year work-life of benefits Hollenbeck (2008) and Uvin et al. (2004) Dean & Rowe (2006 and 2010) Dean (1991) Wilhelm & Robinson (2010) and savings to the public and compared to costs of service. Assumed a constant growth model of wages from the U/I data for 10 to 20 years and compared to costs of services. This study also compared Status 30 to Status 26 and 28. Taxes were estimated (6.2% for social security, 1.45% Medicare, and 25% for state and federal taxes in the case of Massachusetts). It also included fringe benefits of 25%. Noted that Status 28 should be used for costs and benefits because a large percentage of Status 28s actually generate wages after closure. Using U/I data, this looks at all application and closures back to Social Security was briefly discussed, but neither it nor taxes were included in the ROI. Compared Status 26 and 28 to Status 30 using U/I data. Used U/I data to compare Status 26 and 28 to Status 30 in order to estimate the differences and calculate economic impact similar to Hemenway and Rohani (1999). This study excluded all federal contributions to create a cost-benefit ratio for the state of Utah. It also assumed each client would work 24 years after closure (using a calculation between average client age and the age of retirement.)

4 DRS-ROI: Page 3 DRS Methodology for ROI in FY 2011 Return on investment is always a prominent factor in discussions about vocational rehabilitation services and their impact on society. In order to comprehend the return on public investment for financial costs of DRS, it was decided to examine cases which received services and closed in Federal Fiscal Year (FY) 2007 (commonly referred to as Status 26 Rehabilitated and Status 28 Closed after Services without Employment). Step 1: Sampling Methodology Because of time and other resource limitations, it became clear that looking at all closed cases who received services in 2007 would be difficult, especially if this methodology was adopted to be used over multiple years or on a wider scale. It was decided to extract a random sample of cases from the population of 2,521 cases closed after services in cases were selected for a margin of error of +/- 4.71% at a 95% confidence interval. Table 2 presents a comparison of the population and sample groups. Table 2 Comparison of Population and Sample Group: Population Sample Total Cases Closed Rehabilitated 1587 (63.0%) 236 (63.8%) Mean Age at Application Female 47.6% 50.8% Minority 6.4% 8.1% Less than High School Graduate at Application 50.5% 53.0% High School Graduate and Above at Application 49.5% 47.0% Mean Hours Worked per Week at Application Mean Weekly Earnings at Application $54.42 $62.65 Transitional Youth (Age <25 at application) 58.3% 63.5% Primary Disability Sensory/Communicative 11.0% 14.1% Primary Disability Physical 34.9% 28.9% Primary Disability Mental 54.1% 57.0% Primary Source of Support Personal Income 15.4% 18.1% No Medical Insurance at Application 27.8% 29.7% Counties in WV represented 55 of of 55 Step 2: Data Collection and Sources The next step in the DRS return on investment study was to collect relevant data. The goal was to find, using the best accessible and already available information, what the cost for going through rehabilitation services was for each closure in FY 2007 and how many quarterly wages they generated after closure. Costs included administrative and actual services costs for each client. Table 3 demonstrates the data sources used in this study:

5 DRS-ROI: Page 4 Table 3 Cost and Benefit Data Elements: Data Source Measurement Source Costs Administrative Costs $ per Quarter client was served at DRS. 2 Value for RSA-2 each client was found by taking the average cost for all clients served in a given year and summing them for each year client was in the DRS electronic case iecm at DRS management system (iecm). This included all preceding and subsequent cases the client had on record. Cost of Services $ for total services in each case (including preceding RSA-911 and subsequent) the client had with DRS 3. The actual Benefits Wages amount for each individual client was reported. Gross wages in $ per quarter for 3 years (12 quarters) including the quarter in which the case was closed 4. When U/I data was unavailable, this was supplemented by reported wages at closure data to determine the wages received in the closure quarter RSA-911 (but not in any subsequent quarters). Social Security Savings $ in SSI or SSDI benefit payments received by client at RSA-911 Closure or Application if the client stopped receiving social security benefits during services by achieving the SGA level. Social Security Taxes 6.2% of Gross wages up to $106,800 annually U/I Data Medicare Taxes 1.45% of Gross wages annually U/I Data Federal Taxes $ annually calculated using the federal employer withholding tables State Taxes $ annually calculated using the WV state employer withholding tables Unemployment Insurance (U/I) Data U/I Data 2009 Federal Withholding Tables U/I Data 2009 State Withholding Tables 2 Administration costs were directly from the RSA-2 administration costs and total # of individuals. Costs for years prior to 2001 were estimated at 2001 values ($ per client served). This was thought to be a conservative method, as it would be expected for administration costs to, in general, rise slightly over time. 3 The study included all prior and subsequent cases on record as well as the costs for the case closed in FY In some rare cases, closures prior to 2001 did not have accurate information on cost of services (n=5). These costs could not be estimated accurately and were excluded from the model, but expected to have a minimal impact on results. 4 This method is expected to underestimate actual wages received by clients because it assumes when there is no data available in the U/I database that the client received zero wages. In some cases, no data would not realistically indicate no wages (I.E.: A client who moves to a different state and receives wages there).

6 DRS-ROI: Page % of all cases (successfully rehabilitated and those who received services but closed without unemployment) in the sample had U/I data in the system. Return on Investment = the ratio of Costs to Benefits where: Costs = Administration Costs + Cost of Services Benefits = Gross wages for 12 quarters including the quarter in which the case was closed Step 3: Modeling Return on Investment In order to produce the most realistic return on investment possible without having to estimate other possible benefits, a streamlined model was formulated to determine an actual three-year figure. Other potential benefits that rely on estimating techniques will be discussed in the subsequent sections of this report. As presented in Table 4, the total cost of all 2007 Status 26 and 28 closures included in the sample (n=370) was estimated to be $1,954, Wages were estimated to be $10,780, for the sample over three years. This can be broken down into yearly figures to produce the following return on investment estimates, as well as extrapolated figures for the entire FY 2007 at DRS: Table 4: 1-year, 2-year, and 3-year Return on Investment Year Costs 5 Cumulative Gross Wages ROI Sample 1 st Year $1,954, $3,633, $1:$ nd Year $1,954, $7,337, $1:$ rd Year $1,954, $10,780, $1:$5.51 Extrapolation 1 st Year $13,312, $24,758, $1:$ nd Year $13,312, $49,995, $1:$ rd Year $13,312, $73,445, $1:$ See Table 3 Costs include cost of services and administration costs for FY2007 closures and for all cases they had preceding or subsequent to that one.

7 DRS-ROI: Page 6 Step 4: Estimating Social Security Savings In order to estimate the amount of savings in discontinued social security payments (SSI and SSDI), each case was broken down into either a blind or non-blind group since there are different substantial gainful activity (SGA) limits for each. Using the U/I wage data, each client who received SSI or SSDI at closure was examined to see how many quarters they exceeded SGA. In those quarters, it was assumed that the client no longer received SSI or SSDI benefits. In some cases, clients received SSI or SSDI at application, but had already stopped receiving the payments by closure. This was assumed to be an effect of DRS services, so in these cases the application amount was used to determine social security savings each quarter after closure that SGA was achieved. As noted in Table 5, summing all these quarterly savings results in the following estimate for the sample and extrapolated to the entire 2007 closure year: Table 5: Social Security Savings Estimates Year Estimated Cumulative Savings to Social Security Sample 1 st Year Sample $169,491 2 nd Year Sample $327,162 3 rd Year Sample $465,867 Extrapolation 1 st Year Population $1,154, nd Year Population $2,229, rd Year Population $3,174, Step 5: Estimating Taxes In order to get a reasonable (although estimated) idea of the taxes generated by DRS clients, it was decided to use the U/I wage data in conjunction with the federal and state income tax withholding tables. This assumes that the federal and state governments have a reasonable method of estimating how much the average individual will pay in taxes at the end of the year. Clients were also assumed to have one tax exemption and to be single for purposes of parsimony. Finally, social security taxes and Medicare taxes are calculated as a percentage of gross wages (See Table 3 for further discussion and explanation of the data elements). Table 6 shows estimated taxes for clients in the sample and an extrapolation to the closure population as a whole:

8 DRS-ROI: Page 7 Table 6: 1-year, 2-year, and 3-year Tax Estimates Year Tax Cumulative Estimate Sample 1 st Year Federal Income Taxes $232, nd Year Federal Income Taxes $489, rd Year Federal Income Taxes $750, st Year State Income Taxes $112, nd Year State Income Taxes $230, rd Year State Income Taxes $342, st Year Social Security $225, nd Year Social Security $454, rd Year Social Security $668, st Year Medicare $52, nd Year Medicare $106, rd Year Medicare $156, Extrapolation 1 st Year Federal Income Taxes $1,584, nd Year Federal Income Taxes $3,334, rd Year Federal Income Taxes $5,112, st Year State Income Taxes $764, nd Year State Income Taxes $1,569, rd Year State Income Taxes $2,336, st Year Social Security $1,535, nd Year Social Security $3,099, rd Year Social Security $4,554, st Year Medicare $358, nd Year Medicare $724, rd Year Medicare $1,065, Step 6: Estimating ROI including Taxes and Social Security Savings This step reports the net income of the client (rather than gross income), plus tax benefits and social security savings and compares the sum to the total costs. This can estimate an inclusive ROI that includes and helps present these societal benefits. Results are in Table 7:

9 DRS-ROI: Page 8 Table 7: 3-year Return on Investment including Estimated Savings Year Costs Total Benefit ROI Sample 1 st Year $1,954, $3,802, $1:$ nd Year $1,954, $7,664, $1:$ rd Year $1,954, $11,245, $1:$5.75 Extrapolation 1 st Year $13,312, $25,898, $1:$ nd Year $13,312, $52,194, $1:$ rd Year $13,312, $76,584, $1:$5.75 Conclusions Based on the inclusive ROI data presented above, the estimated cost of human capital investment on DRS consumers with significant disabilities ($13,312,864.73) is more than fully recovered in less than one-year of estimated earnings ($25,898,304.90) generated by DRS consumers after exiting the VR system. The $1:$1.95 ROI after the first year clearly validates DRS program as being efficient and accountable state-federal VR program. An average DRS rehabilitant spent two to three years in the VR system. It is estimated that the investment made by the DRS program on its consumers with significant disabilities has resulted in a $1:$5.75 ROI for the short-term period of three years. The estimated $13,312,864 investment in FY 2007 DRS case closures has resulted in an estimated $76,584, in the short-term of DRS consumers work-life. Additionally, DRS has proven its ability to move people with disabilities toward quality employment outcome with competitive earnings and more independence in integrated settings. Based on the collected UI data of the 186 sampled DRS rehabilitants with reported quarterly earnings in the threeyear period, Table 8 presents the short-term work pattern after exiting the VR program for these individuals. Table 8: Working Longevity of Rehabilitants Range Still Working More $ Less $ Not Working 1 st Year to 2 nd Year % 59.1% 29.6% 11.3% 1 st Year to 3 rd Year 79.6% 50.5% 29.0% 20.4% DRS plans to monitor the quarterly earnings of the sampled consumers for the next few years to measure the continuing work-life and the medium-term economic impacts of the DRS program. DRS will also examine the estimated lifetime earnings of DRS consumers by incorporating a realistic work-life 6 Each year is a combination of four quarters. Clients showing wages for one or more quarters were considered working for that year.

10 DRS-ROI: Page 9 table and discount rates to measure the long-term economic impacts of the program on consumers with significant disabilities, especially the long-term economic impacts of transition youth with disabilities in West Virginia. Replicating DRS-ROI Model This report on ROI at DRS presents a conservative and sophisticated methodology that is consistent with the latest literature and/or applications of ROI for public VR programs. The DRS-ROI methodology also enhances the current evaluative approach by using realistic and precise definitions, data elements, and measurements for computing benefits and costs. By following the step-by-step instructions presented in this report, other state-federal VR agencies can clearly replicate the DRS-ROI model to realistically measure and validate the efficiency and accountability of their VR programs.

11 DRS-ROI: Page 10 References Cimera, R.E. (2009). The outcomes and costs of public vocational rehabilitation consumers with mental illnesses. Journal of Applied Rehabilitation Counseling, 40(2), Dean, D.H. (Spring, 1999). Comparing employment related outcomes of the vocational rehabilitation program using longitudinal earnings. American Rehabilitation, 5-9. Dean, D.H., Ashley, J.H., Schmidt, R.M., & Rowe K.L. (Fall, 2006). The long-term impact of comprehensive vocational assessment for youth with disabilities in transition: Evaluation of Virginia s post-secondary education/rehabilitation transition (PERT) program. National VECAP Journal, Dean, D.H., & Rowe K.L. (September, 2010). Using readily available administrative data to examine VR s long-term impact on consumers employment and earnings. Presented at the Pathway to Performance Program Evaluation Summit, Vancouver, WA. Gibbs, W. E. (1991). Vocational rehabilitation outcome measures: The probability of employment and the duration of periods of employment. American Rehabilitation, 17(1), Greenblum, J. (1975). Evaluating rehabilitation programs for the disabled: National long-term followup study. Social Security Bulletin, 38, Hemenway, D.E., & Rohani, F. (1999). A cost-benefit analysis of the employment of people with disabilities in Florida: Final report. The Educational Services Program at Florida State University. Tallahassee, FL. Hollenbeck, K. (2008). State use of workforce system net impact estimates and rates of return. Paper presented at the Association for Public Policy Analysis and Management Conference, Los Angeles, CA. Kenyon, P., Koshy, P., & Wills-Johnson, N. (2005). A cost benefit analysis of vocational rehabilitation services provided by CRS Australia. Institute for Research into International Competitiveness, Curtin University of Technology, Perth Western, Australia. Rogers, E. S., Sciarappa, K., MacDonald-Wilson, K., & Danley, K. (1995). A benefit-cost analysis of a supported employment model for persons with psychiatric disabilities. Evaluation and Program Planning, 18(2), Uvin, J., Karaaslanli, D., & White, G. (2001). Evaluation of Massachusetts public vocational rehabilitation program final report. Commonwealth Corporation, Boston, MA. Wilhelm, S., & Robinson, J. (2010). Utah state office of rehabilitation economic impact study. The Center for Public Policy and Administration, Salt Lake City, UT.