For every SSDI and SSI decision that reaches the vocational stage, the Social Security Administration still measures a claimant's ability to work against job categories that predate the personal computer. Despite an eight-year effort and more than $300 million spent updating the occupational data behind those decisions, the system that Disability Determination Services actually uses was effectively frozen in the late 1970s. That gap is now showing up in the way Residual Functional Capacity gets translated into denial letters across every state.
The role of occupational data in a disability decision
Most SSDI and SSI claims turn on five steps. The first two are nonmedical: whether the claimant meets the insured status or means-tested financial rules. Steps three and four are the medical listings: if the condition is severe enough, on paper, to meet or equal a listing, the claim is allowed on medical grounds. Step five is the one most denied applicants actually face. Social Security asks whether there is work the claimant can still do, given their residual functional capacity, age, education, and work history.
That last step is powered by occupational data. A vocational expert or a state agency analyst maps the claimant's RFC, expressed as a set of exertional and non-exertional limits, against a dictionary of jobs to see whether anything exists in significant numbers in the national economy that the claimant can still perform. The dictionary is the part that has not been updated.
Where the data comes from
For decades, the underlying numbers came from a Department of Labor file maintained as the Dictionary of Occupational Titles. When the Labor Department stopped publishing it, Social Security began relying on a privately maintained snapshot whose entries are tied to occupations as they existed around 1977, with adjustments made in the 1990s. The occupations an analyst finds in that file include roles and skill requirements that bear a passing resemblance to the actual labor market of the 2020s.
During the 2010s, Congress and SSA began a multiyear effort to replace that dictionary with a modernized system built on the Occupational Information Network, known as O*NET. The plan was to redesign the regulations that govern how RFC is scored, retrain the agency staff who use the data, and roll out a validated system that reflects the work people actually do today. The expected cost, by the time work slowed, had climbed above $300 million.
Why the update stalled
In late 2025, Nextgov reported that the regulatory overhaul needed to put the new occupational system into effect had been quietly withdrawn. Reporting from a taxpayer watchdog group in early 2026 confirmed that the agency is still applying the older job file to disability decisions that will affect claimants through 2026 and into 2027.
Internal reviews cited the cost and complexity of validating the new dataset across thousands of job titles. Critics have noted that the older dictionary codes many current roles poorly and may misclassify the physical and mental demands of modern jobs. Without an updated regulatory framework, the new data cannot simply be substituted into the step-five analysis, even where it is technically available.
What this means for SSDI and SSI claimants
The practical effect is that vocational experts are still being asked to defend findings that an applicant could perform jobs that look little like the ones available today. A claimant whose RFC limits them to light work may be told that a sufficient number of light, unskilled jobs exist somewhere in the national economy, even if the jobs listed no longer exist in that form. The mismatch has been a point of contention at hearings for years and shows no sign of disappearing in the current cycle.
For claimants and representatives, the takeaway is to address the data problem directly in the hearing record. A vocational expert's testimony should be challenged where it relies on occupations that no longer reflect the labor market or that misstate the demands of the work. Counsel can introduce current Department of Labor statistics, employer surveys, or state labor data to show that the older job descriptions are not a fair measure of what work actually exists today.
What to watch next
Members of Congress on both sides of the aisle have continued to raise the issue. A Congressional Research Service report published in December 2025 reviews the long history of the agency's occupational data and lays out the legal and operational questions that any successor system would have to answer.
Until a new framework is in place, claimants and representatives should expect step-five arguments to rely on the same dictionary SSA has used for decades. Pushing back on the realism of those findings is one of the more useful levers in a denied appeal, and it is likely to remain relevant through the rest of 2026.