The Retirement and Disability Research Consortium (RDRC) is an extramural research program that was established by the U.S. Social Security Administration (SSA) in 2018. Mathematica’s Center for Studying Disability Policy has partnered with the Center for Retirement Research (CRR) at Boston College in this effort, along with CRR’s other partner organizations: Syracuse University, the Urban Institute, and the Brookings Institution.
Prior to 2018, SSA funded separate extramural research programs on retirement and disability. Mathematica’s Center for Studying Disability Policy was one of the centers participating in SSA’s Disability Research Consortium (DRC); more information about our work under the DRC can be accessed here.
The mission of the RDRC is to:
- Research and evaluate a wide array of topics related to Social Security's Old-Age, Survivors, and Disability Insurance and Supplemental Security Income programs and related federal policies;
- Disseminate information on these topics to policymakers, researchers, stakeholder organizations, and the general public; and
- Provide training and education to scholars and practitioners in research areas relevant to these topics.
In the second year of the RDRC, our researchers were funded to conduct the following studies:
“Understanding the Local-Level Predictors of Disability Program Applications, Awards, and Beneficiary Work Activity”
Principal Investigators: Jody Schimmel Hyde and Dara Lee Luca, Mathematica; Jonathan Schwabish, Urban Institute, and Paul O’Leary, U.S. Social Security Administration
This project will document the local-area predictors of flows onto the DI and SSI programs, and flows out of those programs due to work using combined data from SSA’s Disability Analysis File (DAF) and other national sources. Numerous studies have documented that local-level factors contribute to the share of working-age adults reporting a disability, the employment rate of workers with disabilities, and disability benefit receipt (for example, Rupp 2012; Nichols et al. 2017; Sevak and Schmidt 2018; and Gettens, et al. 2018). Yet, that research to date has not comprehensively assessed how these factors predict flows into and out of DI and SSI. The results of this study will inform SSA’s understanding of trends in disability and the drivers of disability benefit application, benefit receipt prevalence, and beneficiary employment milestones. The project will also produce a public use file of local-area data to facilitate future research and policy analysis.
“Employment Outcomes for SSDI Applicants Who Use Opioids”
Principal Investigators: April Yanyuan Wu and Denise Hoffman, Mathematica and Paul O’Leary, U.S. Social Security Administration
This project will use SSA administrative data to examine the relationship between opioid use and employment outcomes among SSDI applicants. Although recent studies suggest that opioid use may have negative consequences for economic outcomes (Krueger 2017; Franklin et al. 2014; Savych et al. 2018), little is known about this relationship among SSDI applicants. This study will build on the results of an ongoing RDRC project that uses a supervised machine learning method to classify medication information and identify opioids among SSDI applicants. The study will provide SSA with information about the work capacity and post-adjudication economic well-being of SSDI applicants who use opioids, a notable share of the pool of SSDI applicants. The results of this study will be indicative of the extent to which this population would be receptive to services that could facilitate return to work after entry, as well as the extent to which services delivered to similar workers prior to SSDI entry would keep them in the labor force instead.
“Comparative Regression Discontinuity and Regression Discontinuity as Alternatives to Randomized Controlled Trials: Evidence from BOND”
Principal Investigators: Duncan Chaplin and Denise Hoffman, Mathematica and John Jones, U.S. Social Security Administration
This project will use data from the evaluation of the Benefit Offset National Demonstration (BOND) to assess the efficacy of comparative regression discontinuity (CRD) and regression discontinuity (RD) relative to each other and to randomized controlled trials (RCTs). RD is known as a relatively rigorous non-experimental method but produces imprecise results that apply to small populations. CRD addresses these issues. We will estimate CRD and RD models using simulated assignment to the BOND treatment group based on cut-points on duration of benefit receipt at the start of the BOND program. We will compare those estimates to each other and to RCT estimates for the treated group, for CRD, and to RCT estimates for those near to the cut-point, for RD. The findings will support interpreting CRD and RD studies on DI and related programs. They might also help SSA decide whether to consider these methods for testing impacts of proposed rules, using criteria such as duration of benefit receipt and beneficiary age to define those eligible and ineligible for the new rules. The CRD and RD methods are potentially attractive because they can avoid many of the challenges and costs of an RCT, but only if their validity is high.
In the first year of the RDRC, our researchers were funded to conduct the following studies:
“What Is the Extent of Opioid Use among Disability Insurance Applicants?”
Principal Investigators: April Yanyuan Wu and Denise Hoffman, Mathematica and Paul O’Leary, U.S. Social Security Administration
This study will use data from SSA’s Structured Data Repository (SDR) to document trends in opioid use among applicants to disability insurance (DI) and the extent to which opioid use is predictive of initial award, an award within five years, and mortality. The findings will inform SSA’s projections about future program entry based on aggregate trends in opioid use. The rise in opioid use nationwide – coupled with the numerous DI applicants who have conditions associated with opioid use, such as musculoskeletal problems – suggests that opioid use may be common and increasing among DI applicants. Beneficiaries cannot qualify for DI on the basis of drug addiction, but opioid use may exacerbate the effects of other conditions that meet DI qualifications. This study will address a major data issue that limits what is known about opioid use among DI applicants. Applicants are required to report their medications, but medications are recorded as a combination of coded and open-ended text fields. This study will capitalize on a previously developed supervised machine-learning algorithm to identify opioids recorded in freeform text and combine that information with opioids identified in populated medication codes.
"Trends in Opioid Use Among Social Security Disability Insurance Applicants," April Yanyuan Wu, Denise Hoffman, and Paul O’Leary, August 2019. (2019 RDRC Meeting Research Summary and Slides.
Principal Investigators: Purvi Sevak and Dara Lee Luca, Mathematica
This project will use data from the 1997 National Longitudinal Survey of Youth to examine the implications of the proliferation of gig employment and jobs with irregular scheduling on workers ages 18-36 who have disabilities. Trends in labor force participation for young adults with disabilities may signal future DI benefit claiming patterns. Changes in the labor market including alternative work arrangements in the gig economy and “just-in-time” scheduling” have introduced significant volatility in household incomes by creating more uncertainty in schedules, hours worked, and earnings, particularly for the working poor and young adults (Hannagan and Morduch 2015; Lambert, Fugiel, and Henly 2014). Although the flexibility of the gig economy and the challenges created by unpredictable schedules and earnings have been documented by the media, little is known about how these changes affect people with disabilities. For some young people with disabilities, such changes could make working less attractive than receiving DI benefits. But for others, the availability of a job with more flexible hours could have the opposite effect.
“Can Behavioral Insights Reduce Overpayments from DI?”
Principal Investigators: Denise Hoffman and Jonah Deutsch, Mathematica
This project will draw on insights from behavioral science to diagnose process problems and highlight best practices that may reduce overpayments from Social Security Disability Insurance (DI). Overpayments are prevalent among DI beneficiaries engaged in substantial gainful activity and are undesirable for both program administrators and beneficiaries. The Social Security Administration (SSA) has made several efforts to reduce work-related overpayments, but those efforts have not addressed a root cause of overpayments: beneficiaries’ failure to promptly and accurately report their earnings. This study will help SSA understand how communicating reporting requirements to beneficiaries can affect their compliance.
