What the ALJ is actually deciding
Four of the five steps in a disability claim can only deny. The fifth is where almost every contested case is won or lost — and most hearing briefs spend their best pages on the wrong one.
Social Security disability is the practice area most often described as paperwork, which is a shame, because the decision underneath it has a very particular structure — and once you can see the structure, it is obvious where a brief should spend its pages. Most briefs spend them somewhere else.
The claim moves through five questions in a fixed order. They are not weighted equally and they do not behave the same way.
The five steps, and which ones matter
- Step one — is the claimant working above substantial gainful activity? A threshold question with a number attached. It denies or it passes.
- Step two — is there a severe medically determinable impairment? A low bar, and one that turns on whether the impairment is established by acceptable medical evidence rather than on how much it hurts.
- Step three — does it meet or equal a listing? If yes, the claim is allowed outright. Most do not, and a great deal of brief-writing energy is spent here for a low hit rate.
- Step four — can the claimant do their past relevant work? Decided on residual functional capacity against the demands of jobs they actually held.
- Step five — can they do any other work in the national economy? The burden shifts to the agency. This is where contested cases live.
Note the asymmetry. Steps one, two and four can only end the claim or pass it along. Step three can allow it. Step five is the only place where the agency has to prove something — and the only place the grid rules, the vocational expert, and the precise wording of the RFC all collide.
Why the RFC is the whole case
Residual functional capacity is what the claimant can still do, sustained, in a work setting. Everything at steps four and five is an argument about that one finding, which means a hearing brief is, structurally, a brief about the RFC whatever else it says it is about.
Two things decide it, and neither is how severe the condition sounds:
- Specific functional limits, sourced to the record. Not "significant limitations in standing" but what the file actually establishes about how long, how often, and with what consequence.
- Sustainability. The single most under-argued concept in this practice. The question is not whether someone can lift twenty pounds. It is whether they can do it repeatedly, on a schedule, five days a week, with normal breaks and acceptable absences.
A claimant who can do a task once and then needs two days to recover does not have that capacity in any sense an employer would recognise. The record usually says so — in treatment frequency, in the pattern of missed appointments, in what the claimant reported between visits. It rarely says so in a single quotable sentence, which is why it gets missed.
Where the record quietly decides it
Four things carry disproportionate weight and are easy to lose in a file of any size:
- Treatment frequency over time. A sustained pattern argues sustainability far better than any single opinion, and it is in the file already.
- Gaps, and their reasons. An unexplained gap is read as improvement. A gap with a documented reason — lost coverage, no transport, a hospitalisation for something else — is a different fact entirely, and the explanation is usually somewhere in the notes.
- Side effects. Frequently documented, almost never argued. Medication that causes drowsiness bears directly on attention and absence, which are exactly what the vocational expert is asked about.
- The claimant’s own reports, over time. Consistency across years of notes is corroboration. Inconsistency will be found by someone, so it is better found by you first.
The onset date is a separate argument
Established onset gets treated as a detail and is often worth more than any other single issue in the case — months of benefits, and sometimes the difference between a claim inside the insured period and one outside it.
It is also where the record is thinnest, because the beginning of a condition is rarely a dated event. It is the point where a pattern starts, which has to be argued from the file rather than quoted from it. Briefs that assert an onset date without building it usually get the date the agency prefers.
A note on language
Small thing, consistently done wrong by firms that also run personal injury or employment work: in this practice the person is the claimant, not the plaintiff. There is no defendant. Nobody is being sued. An ALJ reading "the plaintiff" in the first paragraph has learned something about how much of this brief was adapted from somewhere else, and that is not the first impression worth making.
What this means for preparation
If steps four and five decide most contested claims, and both turn on functional capacity sustained over time, then hearing preparation is mostly an exercise in building one timeline out of a long record: treatment, reported function, work attempts, side effects, gaps and their reasons.
That is slow work and it does not feel like lawyering. It is also the part that decides the case, which is a bad combination — the highest-value task in the matter is the one most likely to be compressed when the hearing is next week.
Luma builds that chronology from the record with each entry citing its page, which is the shape of the work whoever does it. The argument about what it means is still yours to make.