What this covers

David Autor of MIT examines the explosive growth of SSDI—now costing roughly $200 billion annually—in conversation with EconTalk host Russ Roberts. The core puzzle is counterintuitive: workplace safety has improved, medical care has advanced, and population health indicators have strengthened, yet disability rolls among working-age adults have doubled since 1981. Autor argues that the surge reflects not a sickening population but the collision of two forces: congressional liberalization of what counts as disability, and a labor market that has grown harsher for low-education workers. The elastic definition of disability—centered on inability to work rather than any specific medical diagnosis—became far more permissive after 1984, when Congress restructured the program to prioritize claimant-supplied evidence and weight hard-to-verify conditions like back pain and mental illness. Simultaneously, disappearing jobs and stagnant wages for workers without college degrees made SSDI a rational last resort.

The conversation traces how this mechanism operates. Disability applications spike when unemployment rises, as demonstrated by Appalachian coal regions where awards surged when prices fell. Once awarded—yielding roughly $1,100 per month plus Medicare, with a present value exceeding $275,000 for someone in their mid-40s—recipients almost never leave: fewer than 1% exit voluntarily each year. Those who do work engage in income targeting, deliberately staying beneath the Substantial Gainful Activity threshold to preserve their benefits. The program also created a for-profit ecosystem: attorneys and law firms now collect over a billion dollars annually in fees from successful appeals before Administrative Law Judges. Autor discusses how age and education shape approval odds—older, less-educated workers are effectively "totaled" rather than retrained—and how the program masks labor-market deterioration by absorbing people who would otherwise register as unemployed, distorting economic statistics about workforce participation.

Sharpest takeaway

Autor argues the dramatic growth of SSDI is driven not by worsening population health but by the program's increasingly elastic, employment-based definition of disability colliding with a deteriorating labor market for low-education workers, making disability a rational last resort that people rarely leave.

  • Jobs have gotten safer and population health has improved, yet disability rolls have doubled as a share of working-age adults
  • Congress's 1984 liberalizations made the definition more elastic, weighted hard-to-verify mental and pain claims, and made removal nearly impossible
  • Disability applications spike when unemployment rises because displaced low-skill workers exhaust other options

The argument · threads5 threads · 20 claims
0.85

Rising disability awards reflect labor-market collapse and program liberalization, not worsening health or worker conditions.

5 pointscentrality 5/5
  • The major factor behind rising disability is the system itself—a program made more valuable by rising medical-benefit value and more accessible by congressional liberalization of the elastic definition—colliding with a harsher labor market for low-education workers (lower real wages, fewer opportunities, less employer health insurance), making disability an attractive last resort; obesity and diabetes are minor countervailing factors, not the main driver.

    “The major factor is the disability system itself. That it's become more accessible; the definition of disability has actually been made considerably more elastic over time.”

  • SSDI has masked the depth of labor-market problems by absorbing people who would otherwise appear as long-term unemployed, contributing to the illusion of a lower unemployment rate and to declining labor force participation—especially among low-education men—but it did not cause the jobless recoveries, since a binding labor shortage would have produced wage increases that have not appeared.

    “it has actually masked the depth of the job market problems; that basically it has absorbed a lot of individuals who would otherwise appear on the rolls as long-term unemployed.”

  • There was no big surge in disability awards in the late 1990s because the economy was booming; rapid sustained growth began with the early-1990s jobless recovery and has not reversed since, ratcheting upward with each recession because people remain on the rolls for long periods.

    “It wasn't until the early 1990s, that first jobless recovery that you mentioned, that Disability started growing extremely rapidly and has not looked back since.”

  • The disability epidemic is counterintuitive because jobs have become safer and more sedentary, medical care and workplace accommodations have improved, and population health has improved (lifespans up, disability among over-65s down ~25% in 30 years), so the rise cannot be explained by worsening worker health—few SSDI entrants even come from workplace injury or Workers' Comp.

    “jobs have gotten safer. They are more sedentary. Medical care has greatly improved... Even if population health were holding constant, there ought to be fewer and fewer people who are effectively disabled”

  • Disability applications rise rapidly when unemployment rises and jobs disappear—as shown by Appalachian coal workers whose disability awards spiked when coal prices plummeted—not because people become disabled but because job loss plus underlying health conditions makes it easier to claim inability to work; applications more than doubled from 1.2 million in 1999 to 2.9 million in 2011.

    “There's a classic study by Kermit Daniel, Seth Sanders, and Dan Black that looks at coal workers in Appalachia and essentially shows that when coal prices plummet, disability applications and disability awards from the Appalachia region rise very rapidly. And it's not because people become disabled when the price of coal... declines. It's because they lose jobs.”

0.79

SSDI's benefit structure and program rules directly discourage recipients from working and engaging in formal employment.

4 pointscentrality 5/5
  • Because earning above the Substantial Gainful Activity threshold (about $1000/month) can trigger benefit review and loss of the lifetime annuity-like award, the program creates a powerful incentive against meaningful formal labor-market participation, pushing capable recipients toward under-the-table cash work to avoid jeopardizing benefits.

    “The thing is not by intention but the program creates a very strong incentive against meaningfully participating in the formal labor market.”

  • A typical SSDI recipient gets about $1100/month indexed to inflation plus Medicare after two years, and because this is guaranteed income for life backed by the government, its true value far exceeds the cash figure—the present value of an award for someone in their mid-to-late 40s is roughly $275,000, equivalent to a large lump sum.

    “the kind of present value of someone of a disability award for someone in their mid-to-late-40s is something like $275,000. So, it's the equivalent of someone giving you $275,000 in cash all at once”

  • Fewer than 1% of recipients voluntarily leave SSDI per year, and while 10-15% of claimants do some registered work-for-pay, they engage in 'income targeting'—deliberately staying below the Substantial Gainful Activity threshold to avoid forfeiting their benefits.

    “Oh, it's tiny. It's less than 1%... Per year.”

  • SSDI benefits are calculated like Social Security retirement using a progressive formula (replacing 90%, then 32%, then 15% of earnings tiers), so low earners get a higher replacement rate, while the payroll taxation that funds it is regressive.

    “the first small amount of income is replaced as 90%, then at 32%, then at 15%. So the benefits can get reasonably high; for someone with very high earnings you might have a Social Security Disability benefit as high as $2500-$3000 a month.”

0.65

SSDI's fundamental purpose and design contain an irreconcilable contradiction with disability-rights and employment principles.

4 pointscentrality 4/5
  • The ADA (1990) and SSDI embody contradictory conceptions: the ADA aims for equality of opportunity, full participation, and economic self-sufficiency—treating disability as impairments that need support to enable work—while SSDI defines disability as the inability to work and only helps after a person proves they are not in the labor force, and ironically the employment rate of the disabled has fallen since the ADA passed.

    “the disability program views work and disability as offsets... the ADA, in 1990, says... 'The nation's proper goals regarding individuals with disabilities are to ensure equality of opportunity, full participation, independent living, and economic self-sufficiency.'”

  • Disability under SSDI is defined by inability to engage in substantial gainful activity for health reasons, not by any medical diagnosis—there is no medical condition called 'disability'; it is an elastic societal category about who is deemed unable to work, which expands when jobs are scarce.

    “Disability is not a medical term; there is no medical diagnosis called disability. There are medical diagnoses for a variety of illnesses, but disability is just a broad category that is a societal definition: What do we want to deem as unable to work?”

  • SSDI, started in 1956, insures workers against loss of income and health care access from a work-limiting disability before retirement age; it is funded by about 1.6% of payroll-taxed earnings, and roughly five years of near-full-time work makes a worker insured.

    “The Social Security Disability Insurance (SSDI) program was started in 1956. Its intention is to insure workers against loss of income and access to health care in the event of a work-limiting disability that ends their careers prior to their reaching full retirement age.”

  • Whether an injured worker qualifies depends not just on the impairment but on age and education: a younger, more educated claimant is expected to retrain into sedentary work and denied, while an older worker with low education is effectively deemed unable to do any job—analogous to an insurer 'totaling' an old car rather than repairing it.

    “typically, the criteria actually change as a function of your age and education... if you were 55 and you had no more than a high school education... if you can't do your old job, you can't do any job, and you would receive Disability.”

0.59

Congressional decisions since 1984 made SSDI far more accessible by weakening verification standards and empowering claimant-favorable evidence.

3 pointscentrality 4/5
  • Since a 1984 Congressional decision, the SSA must give first weight to evidence from the claimant's own medical counsel over its own consultants and can only disregard it if it can show fraud or distortion, creating a powerful for-profit advocacy industry of law firms (e.g., Binder and Binder, which received over $20 million in 2010) that build cases to get people approved.

    “the SSA is bound by law--again, this is Congress's decision in 1984--to give sort of first weight to the evidence provided by the disability claimant's own medical counsel... only in cases where it has reason to believe the applicant is distorting the information can it have the right to ignore it.”

  • In the early 1990s more than half of new SSDI claims were for cancers and heart disease, but at present more than half of awards are for mental disorders and musculoskeletal (largely back) disorders—conditions that are difficult or impossible to objectively verify, enabling motivated claimants to establish eligibility.

    “25-30 years ago the predominant source of new claims for people going onto the SSDI program were cancers and heart disease... At present more than half of all awards are for mental disorders and musculoskeletal disorders.”

  • A revolt against the Reagan-era continuing-disability-review clampdown (which terminated ~40% of reviewed cases during the deep early-1980s recession and triggered state-court orders to stop complying) led Congress in 1984 to redefine disability around ability to function in a work-like setting, require greater weight to hard-to-verify pain and mental illness, prioritize claimant-supplied evidence, and require proof of medical recovery to terminate benefits—making the standard far more elastic for the following 25 years.

    “about 40% of the cases that were reviewed by the SSDI process were summarily terminated... this kind of created almost a civil war. And I believe it was 21 U.S. State Supreme Courts ordered their Social Security Field Offices to stop complying with the continuing disability review process.”

0.52

The program has grown to massive scale with profound fiscal and demographic consequences, absorbing millions who would otherwise be counted as unemployed.

4 pointscentrality 3/5
  • As of 2011, about 8.6 million disabled adults receive SSDI—roughly 4.7% (1 in 20) of adults aged 25-64—at a cost of about $130 billion in cash plus $70 billion in Medicare, totaling roughly $200 billion a year, or over $1500 per US household; the share of working-age adults receiving SSDI has doubled from about 2.2-2.3% in 1981.

    “currently there are 8.6 million disabled adults receiving SSDI benefits, about 4.7% of all adults aged 25-64 in the United States... approximately $130 billion a year in cash payments... the Medicare component adds another $70 billion to that. So you are talking about basically a $200 billion dollar program”

  • The disability program is now almost half women, a major change from 30 years ago when women often failed to qualify due to insufficient labor force history; growth has been more rapid among women though substantial for both sexes.

    “it is the case now that the disability program is almost half women. That wasn't true 30 years ago. And an important reason why that wasn't true was that women wouldn't qualify for SSDI because they didn't have enough labor force history.”

  • Because awards increasingly go to early-onset, low-mortality conditions like back pain and mental illness, recipients may stay on SSDI for 15-20 years until retirement age, unlike earlier cohorts with high-mortality diseases like cancer; this means lifetime net benefits per recipient are far greater and the program's effective size is understated at any point in time.

    “those are diseases with relatively early onset. They could easily start in your 40s. And they have very low mortality. Which means that people will be on the program for many years.”

  • When a denied applicant appeals and wins before an Administrative Law Judge, their counsel collects 25% of back benefits up to about $6000 (a cap set by law to protect claimants), and the SSA pays more than a billion dollars a year directly to attorneys who prevail against it on behalf of claimants.

    “If you are represented by counsel, your counsel gets 25% of those back benefits, up to about $6000.”