SSA Disability Program Entry
The Efficacy of Supported Employment as an Early Intervention Option and Work Entry/Reentry Option for Individuals Served by State Vocational Rehabilitation Agencies
Investigators: Michael West (Virginia Commonwealth University) and Fong Chan (University of Wisconsin – Madison)
Inadequate employment is a serious obstacle to economic self-sufficiency for individuals with severe disabilities. When these individuals are unemployed or underemployed despite a motivation to work, they are forced to rely on Social Security disability benefits for income. To achieve gainful employment and reduce their reliance on federal disability benefits, individuals with severe disabilities often require interventions that minimize or mitigate the effect of their disabilities in the workplace. Supported employment (SE) is an evidence-based employment intervention for individuals with severe disabilities, and it is often used by vocational rehabilitation (VR) agencies. The extent to which SE can promote employment at a level that makes disability benefits unnecessary is unknown. An examination of the current and future levels of economic self-sufficiency of SE clients who are served by VR agencies will help us gauge the efficacy of SE as an early intervention strategy for non-beneficiaries and a stay-at-work/return-to-work option for beneficiaries.
In this study, we examine the efficacy of SE as an employment intervention in two important circumstances: (1) to prevent or delay receipt of disability benefits; and (2) to help both beneficiaries and non-beneficiaries achieve economic self-sufficiency, as measured by substantial gainful activity (SGA). The Rehabilitation Services Administration Case Service Report data files—for clients exiting services from FY 2000 to FY 2013 (RSA-911)—will be analyzed twice for this study. For the first analysis, we will extract a sample of VR clients who were non-beneficiaries at intake and compare the earnings and beneficiary status at case closure for those who received SE services and those who did not. In the second analysis, we will extract samples of SE and non-SE clients and use a classification and regression tree analysis, with economic self-sufficiency as the key outcome measure. Predictor variables will include client demographics, receipt of SE, and beneficiary status at intake. Comparisons of clients who received SE and those who did not will be made on the key outcome variable: self-sufficiency at closure, as measured by earnings either above or below SGA.
How the Course of Mental Illness and Multiple Health Conditions Predict Change in SSI/SSDI Beneficiary Status and Labor Force Participation for People Under Age 50
Investigators: Judith Cook and Jane Burke-Miller (University of Illinois-Chicago)
For this project, we will study (1) the extent to which deteriorating mental health conditions cause people under age 50 to exit the labor force prematurely and go onto the SSI/SSDI rolls, (2) whether co-occurring physical health conditions exacerbate this process, and (3) whether evidence-based supported employment services act together with the presence of these health co-morbidities and deteriorating mental illness to influence work outcomes. We previously showed that psychiatric symptoms and hospitalizations are associated with poorer work outcomes over time. We also reported that co-occurring medical conditions or physical disability are common among people with serious mental illness, and that participants with supported employment plus co-occurring conditions had lower earnings, worked fewer hours, and were less likely to work in competitive employment. We will extend these findings to examine the relationship between deteriorating mental health (defined by changes across symptom thresholds), co-occurring medical conditions, and labor force exit and enrollment in disability benefit programs, with additional focus on the timing and potential effectiveness of interventions.
The study will use data from the Employment Intervention Demonstration Program (EIDP), which follows a multisite cohort of SSI/SSDI beneficiaries and non-beneficiaries who receive evidence-based and supported employment. EIDP offers detailed information on the course of their mental illness and their co-occurring medical conditions, administrative data on clinical and vocational service utilization, and longitudinal data on their labor force participation and SSI/SSDI beneficiary status. We will use multivariable random effects logistic and linear regression models to analyze 24 months of longitudinal data from 1,455 EIDP participants.
Understanding what interventions will assist potential Social Security Disability Insurance (SSDI) applicants to remain in the labor force is an ongoing priority for the Social Security Administration (SSA). Early intervention after disability onset appears to be key; indeed, recent research highlights the promise of targeting disabled workers before they apply for SSDI (and if possible, while they are still employed). In addition, other research suggests that former workers from certain employers and industries are disproportionately likely to enroll in SSDI. But despite these findings and recommendations, relatively little work has been done to identify candidates for early intervention or to learn more about the characteristics and employer history of SSDI applicants, with an eye toward encouraging them to remain at work.
The goal of this project is to examine the characteristics of individuals who eventually leave enter SSDI, including the pathways by which they enter the program and the attributes of their employers. This information will help clarify which applicants would most likely benefit from early-intervention policies targeted at employers. It will also shed light on the employer characteristics most likely to be affected by targeted interventions. We will first conduct a systematic review of the existing research and then develop our findings using SSA and IRS administrative data. These data will come from an analytical database of information on employers developed by von Wachter and Song. We will investigate when and under what circumstances SSDI applicants left their employers; how applicant characteristics vary with these circumstances; and how employers differ by size, industry, and location.
"Firm-Level Early Intervention Incentives: Which Recent Employers of Disability Program Entrants Would Pay More?" Journal of Disability Policy Studies, David C. Stapleton, David R. Mann, Pragya Singh, and Jae Song, September 2017.
"Firm-Level Early Intervention Incentives: Which Recent Employers of Disability Program Entrants Would Pay More?" David C. Stapleton, David R. Mann, and Jae Song, March 2015.
"Increasing Employer Responsibility for Disability Benefits: Analysis of an Approach to Social Security Disability Insurance Reform," (brief) David R. Mann and David C. Stapleton, February 2015.
The paths that individuals take to Social Security Disability Insurance (SSDI) application are often characterized by disruptive changes in health, separation from employment, and increased program participation. Most DI applicants stop work for health-related reasons, such as the onset of a work limitation, rather than layoff or resignation. Those who leave because of illness or injury are more likely to apply for DI quickly, and they are less likely to seek other work. Understanding these paths may help the Social Security Administration (SSA) and other policymakers identify ways to intervene and promote employment supports for those at risk of qualifying for DI benefits.
The goal of this project is to use data from the Survey of Income and Program Participation matched to SSA administrative files to track the employment and program participation experiences of individuals up to 42 months before they apply for DI benefits. In addition, the investigators will examine the DI application rates of seven groups whose members may be at greater risk of qualifying for DI benefits. These groups include unemployment insurance recipients with a disability, workers with disabilities who are at risk of qualifying for unemployment insurance benefits, individuals with high health expenditures, workers’ compensation beneficiaries, private disability insurance beneficiaries, military veterans with a disability, and individuals with disabilities who received job training or education services within the past year. We will examine various demographic characteristics, employment, income, and program participation measures to describe the experiences of individuals before they apply for DI and to characterize individuals who do and do not apply for benefits.
"Patterns of Employment and Allowances for Disability Insurance Applicants," (brief) Kara Contreary, Todd Honeycutt, Michelle Bailey, and Joseph Mastrianni, June 2017.
The Role of Obesity in Social Security Disability Determinations
Investigators: Jody Schimmel (Mathematica Policy Research), Anne Stahl (Health Resources and Services Administration), and Jae Song (Social Security Administration)
In 1999, obesity, which had been designated a primary impairment in disability determinations for Social Security Disability Insurance (SSDI) and Supplemental Security Income (SSI) for 20 years, was delisted. It was moved to a secondary impairment category. After that, language was added to certain body system listings to ensure that adjudicators would continue to consider obesity’s effects on individuals’ ability to engage in substantial gainful activity (SGA).
The first purpose of this study is to assess the effect on the outcome of initial determinations of delisting obesity as a primary impairment, and to establish the impairment under which determinations were made. This effort will use the Disability Determination File (831) to document the compositional shift in the use of obesity as an impairment code (from primary to secondary) after delisting, and the extent to which obesity has remained prevalent among applicants and allowed applicants.
The second purpose is to assess how obesity contributed to applications and allowances from 2004 onward, using calculated Body Mass Index (BMI) from the Electronic Disability Claims System. Compared to the impairment code for obesity, this information is richer because obesity may contribute to the prevalence of other impairments even when it is below the threshold. We will produce annual statistics from 2004 forward on the distribution of BMI for applicants and allowed applicants by demographic characteristics, primary impairment, and state of residence.
"Trends in Obesity Among Social Security Disability Applicants, 2007–2013," (working paper) Jody Schimmel Hyde, Joseph Mastrianni, Yong Choi, and Jae Song, February 2016.
"The Effect of a 1999 Rule Change on Obesity as a Factor in Social Security Disability Determinations," (working paper) Anne Stahl, Jody Schimmel Hyde, and Harnam Singh, February 2016.
"The Prevalence of Obesity Among Recent Applicants to Federal Disability Programs," (brief) Jody Schimmel Hyde, January 2016.
"The Role of Obesity in Federal Disability Determinations," (brief) Jody Schimmel Hyde, January 2016.
Employment and Program Participation Characteristics of Successful Social Security Disability Insurance (SSDI) Applicants
Investigators: Todd Honeycutt and Allison Thompkins (Mathematica Policy Research) and Michelle Bailey (Social Security Administration)
Early intervention and return-to-work programs can be informed by better knowledge on (1) which SSDI applicants are more likely to be successful in obtaining benefits and (2) the types of supports that may be needed to improve the outcomes of individuals involved in diversion programs. Prior research has uncovered differences in the employment and program characteristics of those who do and do not apply for SSDI, as well as differences among various groups in the likelihood of applying for SSDI benefits. For example, new private disability insurance beneficiaries had the highest rate of SSDI application, and the applicant employment begins to decline more than two years before their SSDI application, although a large proportion of applicants work in the six months before they apply.
This study builds on the above research to identify earnings and program participation characteristics of successful SSDI applicants. We will also consider differences in pre-application employment patterns that might be used to differentiate successful and unsuccessful applicants. Using Survey of Income and Program Participation panels matched to Social Security Administration data, we will identify SSDI application and receipt for two distinct analyses. First, we will compare the characteristics of successful and unsuccessful SSDI applicants to determine whether there are distinctions in the employment and program participation patterns of these two groups. Second, we will focus on employment patterns of successful SSDI applicants, such as the month-to-month timing of employment, along with potential differences in the characteristics of those with different types of employment preceding their application.
Between 1980 and 2012, SSDI applications more than doubled, increasing from 1,262,300 to 2,820,800 (Social Security Administration 2013). Although changes in labor force demographics can explain much of this growth, the causes of the remaining growth are not well understood. One possible explanation for this growth is the job mismatch hypothesis: potential workers with impairments are unable, or at least increasingly less likely, to satisfy the demands of many jobs, which limits the set of jobs available to the individual and makes disability application more attractive.
In this study, we will investigate the extent to which changing job demands have contributed to the growth in the SSDI rolls. Using detailed occupational characteristics from the Occupational Information Network linked to the occupations of individual workers in the March Current Population Survey, as well as State Agency Monthly Workload Data, we will first describe the current and past distribution of job demands among workers. We will then use an Oaxaca-Blinder decomposition to investigate the extent to which growth in SSDI applications can be accounted for by changes in job demands. Furthermore, by making use of the occupational information from the electronic disability claims system (Electronic Disability Claiming System, 3368 and 3369 forms), we will explore the relationship between the distribution of pre-application job demands among SSDI applicants and that of nationally representative workers. The findings will contribute to understanding the causes of SSDI growth and will address SSA’s research priority areas of demographics and program growth.
Receipt of SSI among children under age 18 has increased 325 percent since 1990. Understanding the growth in child SSI participation has important fiscal implications for the Social Security Administration because many children on SSI become adult SSI recipients and, as such, are likely to have extremely long spells of SSI receipt. It also has implications for child and family well-being. However, existing research has not been able to explain this growth, even when examining differential growth rates across states. One possible explanation for the lack of comprehensive findings at the state level is that local economic and policy conditions may play a more important role than state conditions. Substantial heterogeneity in child SSI receipt at the sub-state level suggests this may be the case.
The goal of this project is to examine county-level variation in child SSI caseloads and caseload growth and to examine the factors that may underlie this variation. These factors include local economic conditions, fiscal incentives, social networks, and incidence of health conditions.
"Child Participation in Supplemental Security Income: Cross- and Within-State Determinants of Caseload Growth," Journal of Disability Policy Studies, Lucie Schmidt and Purvi Sevak, June 2017.
"What Can We Learn from County-Level Variation in Child SSI Participation Rates?" (brief) Purvi Sevak and Lucie Schmidt, June 2016.
Understanding Subsequent Social Security Disability Insurance Award Among First-Time Supplemental Security Income Awardees
Investigators: Priyanka Anand and Yonatan Ben-Shalom (Mathematica Policy Research)
Actuaries have documented a long-term positive trend in the incidence of Social Security Disability Insurance (SSDI) awards among adults. This study will examine one pathway to SSDI that is taken by a substantial number of adults, but frequently ignored in the literature on SSDI growth: entry into the Supplemental Security Income (SSI) program only, followed by entry into SSDI after earning a sufficient number of quarters of coverage to become disability insured.
We will use the 2014 Disability Analysis File to create a database of eight annual cohorts of first-time SSI awardees from 2001 to 2008. We will identify, by cohort, how many of these awardees then received SSDI benefits in the next five years and whether they were first entitled to SSDI benefits because they became disability insured after being awarded SSI benefits versus whether they became eligible for other reasons, such as qualifying as a disabled adult child or a disabled widow(er) beneficiary. Using both descriptive and regression analyses, we will then explore the characteristics that are associated with a subsequent SSDI award for individuals who were initially awarded SSI. Finally, we will examine whether the work activity that led first-time SSI-only awardees to acquire subsequent SSDI benefits is sustained after the SSDI award. This analysis will provide not only a better understanding of the events that led up to SSDI award among first-time SSI-only awardees but also insight into the timing of work-related behavior before and after the receipt of SSDI benefits.
Pathway to SSA Disability Program Entry Among Medicaid Enrollees, 1999–2013: The Role of Serious Mental Illness, Multiple Impairments, and Recent Health Care Utilization
Investigators: Judith Cook and Jane Burke-Miller (University of Illinois at Chicago)
One of the Social Security Administration’s (SSA’s) priorities is to identify pathways to disability programs, opportunities for early intervention, and the intersection of state benefit programs and SSA disability programs. This project will address this priority by examining how the presence of chronic mental health conditions in low-income populations who are eligible for state Medicaid benefits leads to enrollment in SSA’s disability programs. The study builds on research that shows (1) the high prevalence of multiple medical impairments among people with severe mental illness, (2) the prevalence of untreated psychiatric disorders among low-income women on welfare who could be eligible for disability benefits, and (3) the relationship between poverty and psychiatric disability.
The study will use the Centers for Medicare & Medicaid Services’ Medicaid Analytic eXtract data system to analyze Medicaid enrollment, service use, and expenditures from 1999 to 2013 in adults who were 21 to 64 years old. Severe mental illness and other impairments will be identified from claims data by using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) Major Diagnostic Categories and Diagnostic Related Groups. We will describe individuals in terms of the number and type of impairments, demographic characteristics, pathways to Medicaid eligibility, state/region, recent health care utilization, and changes in all of these characteristics over time. Using summary and longitudinal analyses, we will examine three outcomes: (1) the change from nondisabled to disabled status among working-age Medicaid enrollees, (2) the change to dual Medicare-Medicaid status among working-age disabled Medicaid enrollees, and (3) whether and how Medicaid-eligible individuals move from disabled to nondisabled status.
Using Survey of Income and Program Participation data linked to Social Security administrative files, we examine the preapplication employment patterns of Social Security Disability Insurance (DI) applicants for periods of varying lengths up to 24 months before application. Based on their employment histories, we identify two types of applicants. Type 1 applicants are characterized by stable employment in well-paying jobs; most proposals related to workforce retention or DI diversion target this type of worker. Type 2 applicants have either intermittent or no work experience in the preapplication period. Proposals that focus on DI applicants who have recent or long-term attachments to the workforce are therefore likely to miss about half of those who eventually apply. Future proposals should consider outreach to people who lack a strong labor force attachment and who might need a broader array of supports to remain in or return to the workforce.
"Employment Patterns Before Applying for Disability Insurance,"Social Security Bulletin, Kara Contreary, Todd Honeycutt, MIchelle Stegman Bailey, and Joseph Mastrianni, November 2017.
The Impact of Parental Health Insurance Coverage Availability on Disability Benefit Applications for Young Adults
Investigators: Heinrich Hock and Michael Levere (Mathematica Policy Research), and Nancy Early (Social Security Administration)
Gaining health insurance coverage may play a role in the decision to apply for Supplemental Security Income (SSI) or Social Security Disability Insurance (SSDI) benefits. By qualifying for Medicaid or Medicare through SSI or SSDI, people with disabilities may be better positioned to pay for their medical expenses. For some, this can mean avoiding financial hardship. The motivation might be particularly strong for youth facing complex challenges in their transition to adulthood. Since the passage of the Affordable Care Act in 2010, children can stay on their parents’ health insurance plan until age 26, and this could impact their tendency to apply for other benefits.
We will estimate the impacts of access to parental health insurance on SSI/SSDI applications, awards, and benefit receipt, focusing on youth near the age limit for dependent coverage. We will use a regression discontinuity design to assess how SSI/SSDI applications, awards, and benefit receipt change with age. Our impact estimates will be based on any sharp differences in outcomes at age 26. Finding a jump in claims rates for SSI/SSDI at age 26 would help us understand the role of health insurance coverage in applying for disability benefits. Given the long time period over which young adults may continue receiving benefits, understanding how significant this reason for entering the SSI/SSDI rolls could be may yield insights that can help youth achieve self-sufficiency and save taxpayers money.
Trends in Co-Occurring Serious Mental Illness and Multiple Chronic Conditions and Their Association with Work Disability and SSI/DI: Implications for Early Intervention Policy and Practice
Investigators: Judith Cook and Jane Burke-Miller (University of Illinois at Chicago)
Working-age adults who are Supplemental Security Income or Social Security Disability Income (SSI/DI) beneficiaries are usually eligible for benefits because they have an impairment or chronic condition that limits their ability to work. However, many people with disabilities experience multiple chronic conditions (MCCs) that can lead to earlier onset of functional limitations, extensive impairment, and high medical expenses. People with serious mental illness (SMI) have significantly higher rates of many chronic conditions than the general population does, and most of these conditions are amenable to self-care interventions that have the potential to delay the onset or exacerbation of disability. Moreover, compared to people with no chronic conditions, people who have both psychiatric disorders and chronic physical conditions have higher odds of work disability (as measured in total or partial disability days and extra effort days), after controlling for their sociodemographic characteristics, occupation, and region.
This study has two objectives. The first is to describe trends in the U.S. population since the year 2000 with regard to the prevalence of depression and other types of SMI co-occurring with metabolic syndrome-related and other chronic conditions (SMI-MCCs), including the relationship of SMI-MCCs, age, and status over time. The second objective is to examine employment and disability outcomes associated with SMI-MCCs over two years of longitudinal follow-up (2013–2014) on a nationally representative sample. These outcomes include employment type, hours worked, earnings, income sources other than a job, days off for disability, physical and mental health functioning, public health insurance coverage, SSI/DI status, and health care utilization. The findings will be used to describe the significance of SMI-MCC’s for SSI/DI disability policy and the availability of effective interventions for this population in the areas of weight loss, smoking cessation, increased physical activity, and self-management of chronic illness.
Migration Among Persons with Disabilities: Does Migration Explain Geographic Variation in Disability Prevalence?
Investigators: Alexis Henry and Jack Gettens (University of Massachusetts Medical School)
States and counties across the nation have widely varying participation rates in Social Security Disability Insurance (SSDI) and Supplemental Security Income. In a previous project for the Disability Research Consortium we examined the variation and found that much of the variation, particularly for SSDI, was explained by the correspondingly wide variation in the prevalence of disability across these geographic areas. There are two possible explanations for the latter variation: geographic variation in the types of impairment that are prevalent, and differences in the net migration patterns of persons with and without disabilities. This study will focus on explaining any variations in the latter, yielding information that is critical to understanding the determinants of disability prevalence and disability program participation. Different migration patterns might be important in explaining observed patterns in participation rates. Understanding of the proximate causes of geographical patterns of prevalence is important in order to support policy efforts intended to address program participation in areas where it is exceptionally high.
The study has three components. In the first, we will produce statistics that describe the migration patterns of persons with and without disabilities, using thematic maps, bivariate comparisons of means, and regression analysis. For the second component, we will use regression methods to determine the characteristics of areas that could be associated with the migration patterns of persons with disabilities. In the final component, we will determine the extent to which migration accounts for the geographic variation in disability prevalence.