827 Veterans had a legacy hearing before the Board of Veterans’ Appeals in 2024. As of the end of September, a hundred and sixty-eight of them still did not have a decision, and the Board told me it does not track the one number that would explain why. Then, the same VA announced that no more than 20% of its covered employees are allowed to be rated “outstanding”. VA is not bad at counting. VA has decided what to count.
I hope none of you poor souls reading this still have a Legacy appeal pending. Truth is, though, there are lots of you out there. Hell, I have found two this year on my own. Both times it has been guys that hired me, got a decision, filed an appeal, and VA just did nottin’ with it. Life went in, Vet. let it go. Hired me, and I see the file. Uhhh, hey VA, how about that Statement of the Case you owe us from 2003? Red rover red rover can you mail it right over?
I filed a FOIA request with the Board on July 31. The answer finally came yesterday. Also this week, another little birdie forwarded me two VA all-hands messages. The two things landed within a fortnight of each other, and they are the same story told from opposite sides of the counter. This story is both halves.
Askin' is Free, Gettin' is What Cost
Before the data, a word about the price of admission, because it tells you something. Not just something, everything you need to know.
I do a lot of these FOIA requests. Every single one I publish, free of charge, for you, the public to see and draw your own conclusions. I always ask for a fee wavier due to that reason. Denied. Then the FOIA office put me in the commercial use category, which is the most expensive of the three categories the statute allows. Here is the stated reason:
So, The Spectator is not news. The Spectator is advertising. That is the Board’s position, and it came with a bill for $477.10, which I paid on August 28. Oh well, cost of reporting, I guess.
The Board then blew its own deadline. Because no response issued within thirty working days, VA refunded the search fees, $207.99 of the $477.10 came back. The irony there. The agency charged me for the search, missed the statutory deadline, and the penalty for missing the clock was that I only paid about half. Not bad, I guess?
$269.11 for 202 pages. I’d pay it again. But a Veteran who wants to check the Board’s math on his own appeal is not going to. But I digress.
What the Board Says It Does Not Keep
I asked for eleven data fields on every legacy hearing appeal. Most of them came. Two did not, and the two that did not are the interesting ones.
First, the Board told me it does not maintain any docket-management code beyond the Advance on Docket designation. No priority codes. No queue designations. Nothing that would let anybody see how a case is being routed.
Second, and this is the one:
Read that again. The Board does not keep, in any reportable form, the date a case was assigned to the Veterans Law Judge who is supposed to decide it. Soooo, what if a Judge has just had your appeal for 6 months with no decision? Guess you are just SOL.
The Board called this a partial denial. I would call it a confession. You cannot measure how long an appeal sits on a particular judge’s desk, because the Board does not record when it got there.
But hold that thought.
The Rule That Explains the Whole Thing
Along with the data, the FOIA office released two pages of the Board of Veterans’ Appeals Handbook from August 2021. Sections 6.1 through 6.1.2. Two pages out of 202, and they are the most important two pages in the stack.
Section 6.1 explains that cases are normally handed out by an automatic distribution algorithm. Everything in the working range of the docket goes into a pool, and the pool feeds the judges. That is the system most people picture when they imagine a docket. An automated algorithm, you say?
Then comes 6.1.2, titled "Other Than General Population." Item (1):
"VU" is the Board’s in-house shorthand for its Veterans Law Judges, as best I could make out. "Pre-assigned" means your appeal is not in the pool. It never goes back in. The moment you testified, your case stopped being the Board’s problem and became one specific judge’s problem.
Now I am not going to pretend this rule came from nowhere. It comes from 38 C.F.R. § 20.604, and the idea behind it is decent: the judge who looked you in the eye and heard you describe your knee ought to be the judge who decides whether to believe you. There is a fairness argument for it and I am not against it.
But look at what it does when you put it next to the other fact. Your case is locked to one human being, and the Board does not track the date it got locked to him. If that judge is slow, or buried, or out on leave, or working a caseload that would stagger a mule, there is no report anywhere that shows your case aging on his particular shelf. There is no alarm. There is nobody whose job it is to notice.
The docket does not save you, because you are not on the docket anymore. You are in a drawer, and the drawer is controlled by a Board judge, and no one is tracking to see how long the Judge takes once it gets in the drawer. Smoke and mirrors, it is all in how you shift the numbers and the data, and the VA is a mastermind behind it.
Now the Numbers
Here is what the Board’s own report shows for legacy hearings held in calendar year 2024.
- 827 legacy hearings held.
- 659 had a Board decision by September 25, 2026.
- 168 did not.
For the ones that got decided, the Board was quick. The median was 125 days from hearing to decision, and 68.7% were decided within six months. That is a real number and it deserves saying out loud: when a legacy hearing case gets worked, it usually gets worked fast. Key word: worked.
Which makes the other number worse, not better.
The 168 still waiting had been waiting a median of 807 days since their hearing. 150 of them were past the two-year mark. The longest had been sitting 974 days.
And remember, the hearing is not the start of the clock. For these same 168 Veterans, the median time from the VA Form 9 (what starts the appeal) to the hearing was about four years. Counting from the Form 9, the median Veteran in that group has now been at the Board 6.3 years. The longest has been there just shy of ten.
One more. Of the 168 still waiting, exactly one was advanced on the docket. So this is not a story about AOD cases cutting in front. These are ordinary appeals, sitting.
Comparin’
I also asked the Board to build a comparison set. Show me every legacy hearing appeal with a Form 9 filed within six months of one particular Veteran’s, before or after. A peer group. People who got in line at the same time.
The Board found 8,524 of them.
8,301 have a Board decision. 223 do not.
For the ones decided, the median run from Form 9 to decision was 3.6 years. The 223 still waiting have been at it a median of 6.4 years, and they have been waiting a median of 1,360 days since their own hearings — nearly four years sitting after testifying.
97.4% of that group is done. The remaining 2.5% are not a backlog. They are a residue of something that ain’t moving.
What I Think Is Happening
I am not going to accuse anybody of anything I cannot prove, so here is the careful version.
The Board has a rule that pulls legacy hearing appeals out of automatic distribution and attaches them to one judge. The Board does not keep reportable data on when that attachment happened. And roughly one in five legacy hearing cases from 2024 is still sitting two years later while two-thirds of its classmates were decided inside six months.
Those three facts fit together in exactly one obvious way, and she ain’t purtty.
What would settle it is the number the Board says it does not track. So that is the next FOIA, and I will say plainly what I am asking for: assignment dates, by case, and pending counts by judge. If the Board is right that those do not exist in reportable form, that is itself an answer, and Congress ought to hear it.
The Other Side of the Counter
Alright. That is the Board’s half. Here is VA’s.
A little birdie forwarded me the all-hands. Two of them, actually — the one 810 Vermoint Ave., N.W., sent to every employee in the VBA on October 13, and the one a regional office director sent down to his own people a couple of hours later trying to explain it.
The birdie’s own commentary was four lines long and about as good a summary as anybody is going to write.
This is the month VA built an elaborate machine for ranking its own employees against each other. The agency that cannot tell you where your appeal is can now tell you, to the percentage point, how many of its people are allowed to be good at their jobs.
Let us take it in order: what the government did, what VA chose to do on top of it, and who it lands on.
What OPM Actually Did
On July 7, 2026, the Office of Personnel Management published a final rule rewriting how federal employees get performance ratings. It took effect August 6, 2026. Compliance with the distribution piece is required beginning January 1, 2027.
The headline is one sentence. The rule, in OPM’s own words, “removes the prohibition of a forced, or standardized, distribution of performance rating levels.”
That prohibition had been on the books since the Carter administration. Not that is meant to impress you by any means. The old rule barred agencies from using “preestablished distributions of expected levels of performance (such as a requirement to rate on a bell curve).” You were supposed to be rated against the standard in your performance plan, not against the person at the next desk.
That is over. The new § 430.208 lets OPM “establish and maintain a standardized distribution of some or all rating levels,” and the preamble says the change “permits OPM to require and enforce an agency-wide and government-wide distribution of performance ratings.” The rule also says out loud that “comparing, categorizing, and ranking employees on the basis of their performance” are now acceptable methods for assigning a summary rating.
Two other pieces went by quietly and matter more than they look.
First, OPM deleted the provision at 5 C.F.R. § 430.207(c) — the one that required an agency to provide assistance whenever an employee’s performance fell below “Fully Successful” but above "Unacceptable." Help on the way down is no longer required. It is now optional.
Second, a rating of record is no longer bitchable. OPM’s position is that ratings “are not subject to negotiated grievance procedures or arbitration under 5 U.S.C. 7121, regardless of collective bargaining coverage.” Contracts already in force are honored until they expire. After that, nothing. OPM went further and prohibited unions from even negotiating for the right.
So: your rating can now be set partly by how you compare to your coworkers, nobody has to help you if you start slipping, and you cannot grieve the result.
The Numbers Are VA’s, Not OPM’s
Here is the part I want to be precise about, because precision is the whole point of this blog.
OPM’s rule does not contain a number. I read it. There is no percentage in it. OPM said it “intends only to require a limit on the upper rating levels,” that any distribution “will not impose limitations on any of the levels at or below Level 3,” and that it would be applied at the department level rather than the work unit. No figure.
The figures came from VA. From the all-hands.
Note the hedge in that last sentence, and note what it concedes. The 20% sub-cap on Level 5 is VA’s, by its own admission. Nobody made them do that.
The word “calibration” does not appear anywhere in OPM’s rule either. The calibration panels are VA’s construction too — “a group of senior agency employees” that takes a supervisor’s preliminary ratings and relative rankings and produces “final ratings of record.” VBA published three calibration factors to be applied: Performance Standard Achievement, Mission Impact & Contribution, and Accountability & Organizational Effectiveness.
Read that as plain English. Your supervisor, who watched you work all year, writes down what he thinks. Then a panel of people who did not watch you work adjusts it until the percentages come out right.
Who It Hits
Covered: non-bargaining unit employees classified as general schedule, wage leaders, and wage supervisors.
Excluded: title 38 employees rated under the proficiency system, the Veterans Health Administration’s Executive Career Field, wage grade employees, and bargaining unit employees covered under a collective bargaining agreement.
Sit with that list for a second, because it is the most interesting thing in the document.
The forced curve does not land on the unionized workforce. It lands on everybody outside the unit, which, at a regional office, means a great deal of the supervisory and management layer. The people being ranked into a fixed distribution this year are in substantial part the same people who set production expectations for everybody else.
I am not going to tell you that is poetic justice, because it is not justice and nothing good comes of it. A supervisor who knows that only two in ten of his subordinates can be rated outstanding, and only four in ten can be rated above fully successful, is a supervisor under pressure. Pressure rolls downhill. It always has. Always will.
And remember what OPM added while it was in there, a new mandatory critical element for every supervisor, titled “Holding Employees Accountable.”
"We Do Not Have Guidance Yet"
The local message is the one that got me. A regional office director wrote to his own employees the same afternoon, and this is what he had:
I want to give that director his due. That is a decent, honest paragraph, and it is more candor than the Board has shown me in three FOIA requests. He also wrote that none of it “changes how much we value this team or the work you do,” and I believe he meant it. That is real leadership, a far cry from what DC has been offering. (Doug Collins or District of Columbia. Dealer’s choice.)
But look at the sequence. Central announced a forced distribution to the entire workforce at 12:27 in the afternoon. Two hours later the people who have to run it were telling their employees they had not been told how. That is not a rollout. That is a press release with a field office attached. Maximum chaos.
What VA Chooses to Count
Now put the two halves together.
The Board cannot produce, in reportable form, the date your appeal was assigned to a judge. One in five legacy hearing cases from 2024 is still sitting more than two years after the Veteran testified. The Board’s answer to a FOIA request asking how its own queue works was that the data does not exist in a form anybody can report.
Meanwhile, that same department can tell you that no more than 40% of covered employees may be rated above fully successful, that no more than 20% may be outstanding, and that three named calibration factors will be applied by panel to make the numbers come out.
VA is not incapable of measurement. VA measures beautifully when it wants to. It has simply decided which things get measured, and a Veteran’s appeal aging on a Judge’s shelf is not one of them.
There is a word for an organization that counts its own people to two decimal places and cannot count how long the public waits. The word is not “understaffed.”
What Happens Next
Here is my prediction, and I will come back in a year and tell you whether I was wrong.
The curve will not make anybody faster. It will make the ratings quieter. Supervisors will spread the top marks around to keep the peace, the people genuinely carrying the load will get the same number as the people coasting, and the ones who can leave will leave — which is the thing VA cannot afford, because the person who knows how to work a forty-year-old Legacy file is not replaceable by a hiring announcement.
The birdie said production goes through the floor. I think he is right, and I think Veterans pay for it in exactly the place the Board’s own data points to, in the drawer, waiting on one judge, with nobody tracking the date.
Read It Yourself
Both sets of documents are below. Read them yourself and tell me I have it wrong.
The Board’s 202-page production is first. The response letter is Exhibit 1, the hearing reports are Exhibits 2 through 4, and the two handbook pages are Exhibit 5 — if you only read two pages of this, read those. Redacted only for e-mail addresses.
THE BOARD’S FULL RESPONSE TO FOIA 26-24867-F
202 pages. Exhibit 1 is the response letter, Exhibits 2–4 are the hearing reports, Exhibit 5 is the handbook. Redacted only for e-mail addresses.
The VA messages are second: the VBA all-hands and the regional office follow-up that came down the same afternoon. Redacted for the sender’s identity and every e-mail address.
THE VA MESSAGES, AS RECEIVED
The VBA all-hands message and the regional office follow-up. Redacted for the sender’s identity and every e-mail address.
Godspeed.
