Summary
Policy remarks by Paul Van de Water (Center on Budget and Policy Priorities (CBPP)) at the Social Security Administration's (SSA) National Disability Forum, responding to SSA's 2015 Advance Notice of Proposed Rulemaking (ANPRM) on whether to revise the vocational factors (age, education, work experience) used in Disability Insurance (DI) eligibility decisions. The core argument: aggregate health improvements do not reach the low-education, manual-labor workers who actually rely on grid-based eligibility, so raising grid age thresholds would harm the wrong people and would not return them to work.
Key Claims
- Disability is not purely medical. The same back impairment can be career-ending for a manual laborer without a college degree but trivial for an economist. The Social Security Act explicitly requires considering age, education, and work experience alongside residual functional capacity (RFC) — this is statutory, not bureaucratic discretion.
- 2005 Notice of Proposed Rulemaking (NPRM) and 2009 withdrawal. SSA proposed in 2005 to add 2 years to all vocational grid age thresholds (e.g., the criteria for workers 50–54 would shift to 52–56). The proposal was withdrawn in May 2009 after substantial negative public comment. The 2015 ANPRM re-opened the question.
- SSA's 2005 rationale was flawed. SSA cited "average health improvements" and "rising life expectancy" to justify raising thresholds. But as Van de Water argues (citing colleagues Sherman and Sweeney 2006), studies showing that Americans on average are healthier at a given age are not relevant to those individuals who are not healthy — averages are misleading when health status varies sharply by socioeconomic group.
- Health improvements are concentrated at the top. Evidence since 2005 has confirmed that life expectancy gains are concentrated among high-socioeconomic status (SES) workers: Waldron (2007) shows the life expectancy gap between top and bottom half of earners at age 65 widened by 5.4 years between 1977 and 2006. College degree holders delay onset of poor health and activity limitations by ~20 years vs. high school (HS) diploma, ~30–40 years vs. HS dropouts (Zajacova, Montez, and Herd 2014).
- US DI criteria are already stringent by OECD standards. The Organisation for Economic Co-operation and Development (OECD) rates the U.S. as having among the most stringent eligibility criteria for disability benefits among advanced economies (OECD 2010). The US spends 1.3% of gross domestic product (GDP) on disability/sickness cash benefits vs. an OECD average of 1.9%. US beneficiaries are far more likely than European counterparts to be among their country's sickest citizens (Croda, Skinner, and Yasaitis 2013).
- Overall award rate ≈ 37–40%; allowance rate ≈ 56%. Over 2009–2011, SSA granted benefits to 37% of all disabled-worker applicants (award rate), and 56% of medical decisions (allowance rate, excluding technical denials).
- Rejected applicants have limited work capacity (von Wachter, Song, and Manchester 2011, American Economic Review (AER)): About 43–53% of rejected older males (ages 45–64) have any earnings 2 years after application, with median earnings ~$10,000. In contrast, comparable males who never applied for DI show ~79–82% employment with ~$35,000 median earnings. Even rejected applicants are substantially more work-limited than the non-applicant population, confirming that DI is not capturing people capable of substantial work.
- Labor market shift cuts both ways. The move toward "mind over muscle" work generally benefits older workers — but since the late 1990s, job opportunity gains for older workers have gone primarily to better-educated workers (Rutledge, Sass, and Ramos-Mercado 2015). This reinforces rather than undercuts the grid's rationale for less-educated workers.
- Literature review found no empirical basis for specific grid cutoffs (Mann, Stapleton, and de Richemond 2014, Mathematica), but also concluded it was unrealistic to expect such evidence — the grid sets bright-line rules for administrative consistency, not precise actuarial thresholds. This means criticism based on "the exact ages are arbitrary" misunderstands the grid's purpose.
- DI vs. ADA distinction. DI provides partial income replacement when impairment curtails earning capacity. The Americans with Disabilities Act (ADA) makes society more inclusive to reduce work barriers from impairments. These are complementary purposes; conflating the ADA's expanding concept of "disability" with DI's narrower definition muddles the reform debate.
Concepts Introduced or Extended
- Vocational Grid — adds the 2005 NPRM, 2009 withdrawal, 2015 ANPRM policy history; Van de Water's "averages are misleading" argument against raising thresholds
- Disability Insurance Program — adds OECD stringency comparison detail; DI vs. ADA conceptual distinction
- DI Denied Population — adds von Wachter/Song/Manchester rejected-applicant vs. non-applicant employment comparison
Entities Mentioned
Quotes
"Studies finding that Americans on average are healthier at a given age than Americans were a few decades ago are not relevant, however, to those individuals who are not healthy at these ages."
"Policymakers may decide that DI's eligibility criteria should be stricter, but should recognize that this would mean hardship for rejected applicants and would especially affect minorities and people with lower socioeconomic status."
"DI is designed to provide partial compensation to workers who develop a physical or mental impairment that, in light of their other circumstances, significantly curtails their capacity to earn a living. The ADA aims to make society more open and inclusive so that impairments are less likely to result in inability to work."
My Take
This is advocacy, not research — Van de Water marshals existing evidence to oppose a pending policy change rather than generating new findings. Its wiki value is primarily the policy-context layer it adds to Vocational Grid: the concrete 2005 NPRM/2009 withdrawal history, the structured normative argument against raising thresholds, and the introduction of the von Wachter/Song/Manchester (2011) AER data on rejected applicants' labor market outcomes (not otherwise present in the wiki). The core "averages are misleading" argument is logically sound and well-supported by Waldron (2007), though it does not engage with the harder question of how many truly able workers the current grid admits alongside the genuinely disabled.