Decisions
,
Grant rate
,
≥1 issue granted
Remand rate
,
sent back to AOJ
Deny rate
,
fully denied
AMA share
,
post-2019 reform
VLJs · offices
,
judges · regional
Disposition by year (volume + mix)
Outcome rates by year
Disposition mix
Top conditions · grant rate
Representation · grant rate
Top regional offices · grant rate
ⓘ How these rates are defined (denominators)
Every rate's denominator is all decisions in scope (after any filters). A single decision can adjudicate multiple issues, so we report two grant lenses:
- Any-grant rate, % of decisions with ≥1 issue granted (
n_granted>0). This is the headline Grant rate card and the green line in “Outcome rates by year.” It’s the veteran-relevant “did I win anything?” measure. - Full-grant rate, % whose overall disposition is “granted” (every adjudicated issue granted).
- Remand / Deny rate cards, % whose overall disposition is purely remanded / denied.
- “Outcome rates by year” uses the consistent any-issue basis for all three lines (≥1 issue granted / remanded / denied), so a “mixed” decision counts in each category it touches and lines may sum to >100%.
- mixed = a decision with more than one distinct outcome across its issues (e.g. one granted, one remanded).
Breakdown by
full-corpus aggregates · year filtering in V2These are historical outcome frequencies across decided appeals, descriptive, not predictions about any individual claim. Each any-grant rate shows a 95% confidence interval and its Δ from the national baseline; groups with fewer than 30 decisions are suppressed as unreliable. Veterans Law Judge names are preliminary (heuristic, not yet roster-matched).
Ask the corpus
Answers are grounded in the corpus KPIs + Title 38, with legal-information-not-advice guardrails; every statute/regulation citation is verified against the dated Title 38 store (unverified cites are struck through).