On Tuesday the VA
announced it had processed more than three million disability claims this fiscal year, a record, with a week still to go. The average decision now takes 75.6 days, down from 141.5 in January 2025. The backlog is down more than 70 percent. Accuracy is at 94 percent, the highest in two years. More than 60 percent of claims are granted.
Those are the numbers the press release leads with, and they describe a system that is faster than it has been in a generation. Then, near the end, one more figure: only about 3 percent of the Veterans Benefits Administration's decisions are reviewed by the Board of Veterans' Appeals.
Read those together and a different story appears.
The math the release doesn't do
If more than 60 percent of claims are granted, then something under 40 percent are denied or, in practice, granted at a lower rating or later date than the veteran sought. On three million decisions, that's north of a million outcomes the veteran didn't want.
Only about 3 percent of decisions reach the Board. The Board isn't the only review lane, and some veterans use higher-level review or supplemental claims without ever going to a judge. But the gap between "roughly 40 percent unfavorable" and "roughly 3 percent reviewed at the Board" is enormous, and no amount of lane-counting closes it. The overwhelming majority of veterans who get a decision they disagree with do nothing about it.
That's the fact underneath the release, and the VA published it.
Why 94% accuracy and 40% denials both fit
Accuracy, as the VA measures it, means the decision correctly applied the rules to the record in front of the rater. A denial can be perfectly accurate on a thin file and still wrong about the veteran, because the file never contained the
medical opinion or the
records that would have changed it. The accuracy figure measures the rater. It says nothing about whether the evidence was complete.
And 94 percent still leaves 6 percent. On three million decisions, that's roughly 180,000 decisions the VA's own quality review would call wrong. The inspector general, as we
wrote in August, tends to find more error than internal reviews do.
What speed does to the odds
Faster is good. A veteran waiting 75 days instead of 141 is a real improvement, and the department deserves credit for it. But the same release credits automated tools for streamlining routine work, and automation processes what a file contains rather than what it implies. We've
written about what that shift means for how a claim should be built. The short version is that a thin file gets a fast, accurate, unfavorable decision, and the system considers that a success.
The case for appealing, from the VA's own release
Put the pieces together. Four in ten decisions go against the veteran in some way. Independent review, when it happens, finds errors at high rates. And almost nobody asks for it. Every one of those facts argues the same direction: if your decision came back wrong, the odds that it's fixable are better than the 3 percent figure would suggest, because the 3 percent measures how many veterans asked for review, and says nothing about how many decisions deserved it.
A denial is a decision on a file, made in 75 days, by a system optimized for throughput. It says less about your condition than it feels like it does.
Read it closely, and if it doesn't hold up against your record,
the review lanes exist because decisions like that are ordinary.
Where we come in
We're a veteran-led firm, and our whole practice lives in the gap between the 40 percent and the 3 percent: the decisions that went wrong and the veterans who assumed nothing could be done. If you're holding one,
reach out and we'll tell you whether it deserves a second look.
