There is a number printed on the fee agreement you signed this week, and it is $9,200. It is worth about as much as the RRP sticker on a car.

In 2022, the last year SSA has published, the agency made 311,047 direct payments to claimant representatives. They came to $923,992,943.10. That is an average payment of $2,970.59, and the cap that year was $6,000, so this is not a story about the ceiling being too low. Almost nothing was touching the ceiling.

The reason is in the statute and it is not subtle. You get 25 percent of past-due benefits or the cap, whichever is smaller, so the cap only becomes the operative number once a quarter of past-due clears $9,200. That takes $36,800 in past-due benefits. Run the average backwards and the typical case is carrying something like twelve grand. It would need to roughly triple before the number on your fee agreement did a single thing.

So when NOSSCR fights for a higher cap, and it should, understand what is actually being won. It is a better outcome on the minority of files with a long enough wait and a big enough monthly benefit to get there. On the ordinary case it changes nothing at all, because the fee was never being set by the cap. It was being set by somebody's earnings record and by how long SSA sat on the claim.

The number nobody watches

The average payment peaked in 2007 at $3,576.87. By 2022 it was $2,970.59. That is seventeen percent lower before you touch it, and once you deflate it with CPI-U it is 41 percent lower in real terms. Fifteen years of the fee quietly getting smaller.

Now put the cap next to it over exactly the same window. In 2007 it was $5,300. Today it is $9,200, which is up about 74 percent. One line went up three quarters and the other fell by more than a third, and they are sitting in the same agency's files pointing in opposite directions. The entire public argument in this practice area has been about the line that went up.

I keep coming back to how boring the mechanism is. The fee is a fixed slice of past-due benefits. Past-due benefits are the monthly benefit times the months SSA took, and the monthly benefit comes off a work history nobody in your office can influence. You are being paid a percentage of a number that is set somewhere else entirely, and the only reason it moved at all was inflation eating it from underneath.

SSA also stopped publishing

The table on that statistics page runs monthly all the way from January 2000, and then it just stops. The last row is February 2023, 30,951 payments and $82,003,050.12, and there has been nothing since.

I went back three times looking for a continuation because it seemed like an oversight. It is not. The agency that decided in May 2025 that publishing a yearly notice about the cap would be an unnecessary administrative burden had, by then, already gone quiet on the underlying numbers for more than two years. The OIG had flagged a version of this in 2023, pointing out that SSA reports the Title II side publicly and has never published anything equivalent for SSI direct payments. Now neither one is visible.

Which means that if you want to know what representatives are being paid nationally right now, in 2026, you cannot. Nobody can. The most recent complete year on the public record is 2022 and that is not a choice I made.

There is one oddity in it I still cannot explain. November and December of 2022 carried 78,568 payments between them, well above the pace of any other month that year, at averages around $2,173 and $1,876. Through October the year was running at $3,298. Something happened in that year-end push and the public data does not say what. If you would rather use the pre-surge figure, use $3,298, and everything above holds with more room rather than less.

What this leaves you with

Two numbers run an SSD practice and only one of them is yours.

Revenue per approved case is not. It is a quarter of past-due benefits, it has been sliding in real terms since before the last recession, and no amount of lobbying about a ceiling most of your files never reach is going to move it. That fight is worth having on principle. It is not a plan.

Cost per case is yours, entirely, and it is the whole remaining lever. If the average fee is somewhere around three thousand dollars gross of every single thing it took to earn, then every hour a file eats is being spent against a fixed and slowly shrinking amount, which is a fairly grim sentence to write but is also just what the arithmetic says. The firms that stay comfortable through the next ten years of this are going to be the ones whose cost per case falls faster than their fee does.

That is the actual reason to care about any of it. Not the rescission, which was a paperwork decision about a paperwork commitment. The rescission is just the thing that made me go and add up the columns.

Check it yourself

The monthly data is at ssa.gov/representation/statistics.htm and it takes about twenty minutes. Sum the twelve 2022 rows and you should land on 311,047 and $923,992,943.10.

The rescission is 90 FR 19241, May 6, 2025. The promise it withdrew is at 89 FR 40523. Every cap amount and effective date since 1991 is in POMS GN 03920.006, and the statute is 42 U.S.C. 406(a)(2)(A). A couple of the older rows in the archive have typos in them, an Oct-13 amount missing a comma and a Feb-06 amount with two decimal points, so treat 2006 and 2013 with suspicion. The 2022 rows are clean.

One thing the data cannot tell you: the column counts payments, not cases and not representatives. A concurrent Title II and SSI claim throws off a Title II payment that shows up here and an SSI payment that has never been published anywhere, so the fee per case is higher than the fee per payment. By how much, I genuinely do not know.

We build the operational side of this for disability firms, so cost per case is the number I spend my working life on.

If you want the other half of the picture, we pulled every ALJ decision for a year to work out how much of an outcome comes down to which judge picks up the file. More than you would like.