Feb 15, 2018

They Call This The Waterfall Chart

Click to view full size


Anonymous said...

I think some Congressional investigations of the Appeals Council would be in order. Also, eliminate recon.

Anonymous said...

Splitting Allow and remand, and dismissal and denial makes it appear the reversal rates are not particularly significant upon judicial review. As I see it, it is 1/3 award at initial, 1/10 award at recon, 1/2 award at hearing, 1/10 award at AC, and 1/2 reversal at Court. Recon and AC are obviously not necessary levels.

Anonymous said...


"Recon and AC are obviously not necessary levels."

I think most, including the agency, would agree with you on recon, but the district courts would almost certainly disagree with you about the AC. If you dumped over 100k more appeals in their laps on an annual basis, they would scream.

Anonymous said...

Here in region 6 we haven't had Recon for years, and it works just fine.

The real scandal is that too many cases are being denied at the initial claim level. If initial was doing their job correctly, then very few claims would ever be approved at the other stages.

Anonymous said...

It would be nice if the chart also showed the number of cases and not just the percentages.

I would be interested in how many cases are dropped rather than appealed to the next stage.

Anonymous said...

You're wrong about the agency being on board with nixing recon, at least OHO management. Having recon slows down the deluge at the hearing level; OHO (and other top SSA brass) love recon because it slows their workloads. Any procedural hurdle, process, etc. between filing and a request for hearing decreases OHO's, and by extension SSA's, workload.

Anonymous said...

That is true for medical and non-medical reconsideration requests. Reconsideration has not been a priority workload for the back-end operations (i.e. program centers, earnings operation) in a while for a few regions. It may be different for reconsideration requests with the front-end (i.e. field).

Anonymous said...

In 2012, remanded 18 percent
But their get up and go
Got up and went

Now the AC, its got a lot colder
Turning more down
Turn the cold shoulder

Now 9 percent is all they remand
And they start to resemble
A big rubber stamp

Federal Court is your only fair chance
But you first you must do
The Appeals Council dance

So why did it change
In this steep downward flow?
18 to 9 is a lot you know

So many more lost, so many more fail
Makes me wonder what finger
Pushed on the scale

Tim said...

Fifty percent of all cases brought to Federal Court are approved/remanded! I wonder what percentage of those denied by ALJ/AC/FC end up applying again? Is it cynical of me to think they just want to push you past the date of last insured so they can limit you to SSI? Remember, just because you're paranoid doesn't mean they're not out to get you! For some of us, continuing the process is our only real option.

Anonymous said...

What are the number that actually go to fed court after being denied at the Initial, Recon, Hearing, and AC? How many Fed Court cases allow a case or kick it back with an amended onset date after DLI because severity or GRID is now met? Seems pretty reasonable after all, because the interpretation is much wider. An algorithm that just looks for key words and test values would deny a bunch more cases than humans.

Anonymous said...


I have yet to determine a claim lacks merit for judicial review based on an Appeals Council's determination. If we file an appeal with the AC, it will be appealed to Court absent extraordinary circumstances (Claimant got better, change in law, etc.).

It may vary by circuit, but in my experience a change in medical severity or gridrule is not a particular relevant consideration for the courts. If you want to make those arguments, file a new claim, as the ALJ's decision, not the claimant's present circumstances is being reviewed.

As to interpretation being wider, the interpretation is intended to be the same. AC should be reversing every case reversed by the Courts as lack of substantial evidence, or the decision being predicated on harmful legal error, are the same standards. The AC just just has an extremely limited interpretation of what does not constitute substantial evidence and they ignore legal errors entirely without explanation.

Anonymous said...

Sharing my experience, I have noted about a 40% remand rate in cases in which an extremely blatant and obvious material error is succinctly pointed out in a AC brief. Of those turned down by the AC, the Federal Courts are sending a majority of those back. Of cases that should be remanded but which aren't so obviously blatant, very few are remanded by the AC but the Federal Courts are sending back a a bit less than half.

Anonymous said...

this really isn't a waterfall chart - and SSA does NOT call it that.

true waterfall charts would follow the results of decisions on applications through the final decision - which may fall over several fiscal years.

this is a workload data chart that reports on the results of decisions at each decision level during that fiscal year.

in past years - the SSA budget document had the actual number of cases at each level.


curiously the case data is NOT in this year's budget data.

do some simple math - and you'll see that 80% of all allowances are completed at the initial and recon level.

and - on the issue of recon - there is only one prototype state (no recon) in each region......or 10 states total.

expect that Single Decision Maker to be phased out and the reconsideration step reinstated in the next few years according to SSA's strategic plan.

Anonymous said...

In recent years, there have been around 1 million initial allowances a year.

Anonymous said...

9:53, the "true" waterfall data is at https://www.ssa.gov/policy/docs/statcomps/ssi_asr/2016/sect10.html#table69 for SSI and https://www.ssa.gov/policy/docs/statcomps/di_asr/2016/sect04.html#chart11 for DI. It only goes to 2015 and even then so many cases are still pending. It will be many years until we're able to discern trends from these data.