DOJ's August 27 Radiation Exposure Compensation Act dashboard does not show that 99.6% of all new-era claims have been approved. It shows 9,653 approvals and 42 denials among disposed claims, while 41,702 of the 51,397 claims received under Public Law 119-21 remain pending. The published formula excludes that 81.1% pending share from its denominator.
The page also does not describe one 108,478-claim queue. It places a 51,397-claim post-2025-law cohort beside a separate 57,081-claim legacy cohort received before July 3, 2025. The two cohorts use different category labels and reflect different statutory periods, so combining them erases the state change the dashboard is designed to preserve.
The question this map answers
How should a reader reconstruct DOJ's RECA dashboard into cohorts, claim states, category definitions, and dated changes without turning a cumulative snapshot into an approval rate, a processing log, or an eligibility decision?
MyMap observed the DOJ awards page and program page at 4:04 AM Pacific Time on August 28, 2026. Both identify August 27 as their update date but publish no exact extraction time. The comparison baseline is DOJ's separately dated July 1 snapshot. The event dates are therefore July 1 and August 27; the later observation occurred August 28.
In this map, a cohort is the dashboard's own source grouping. A pending, approved, or denied value is a cumulative state count at one snapshot. Disposed means approved plus denied for purposes of DOJ's displayed percentage. A snapshot difference is the later cumulative value minus the earlier one; it is not a record of each submission or adjudication event between those dates.
Two cohorts sit on one page but use different schemas
The post-law table is labeled “Claims Received Under PL 119-21.” Public Law 119-21 was enacted July 4, 2025, reauthorized RECA, expanded affected populations, extended covered uranium work, and created the Manhattan Project Waste category. DOJ's program page describes four amended claimant categories: Downwinder, Onsite Participant, Uranium Worker, and Manhattan Project Waste.
The legacy table is labeled “Claims Received before 7/3/2025.” It preserves five category rows: Downwinder, Onsite Participant, Ore Transporter, Uranium Miller, and Uranium Miner. Its uranium occupations remain separate rather than appearing under the new table's consolidated Uranium Worker label.
That is a schema boundary, not a formatting detail. A reader can sum each table's states, but should not compare an individual legacy uranium row directly with the post-law consolidated row or assume a person counted in one category maps cleanly into another.
| Dashboard cohort | Total claims | Pending | Approved | Denied | Approved amount |
|---|---|---|---|---|---|
| under Public Law 119-21 | 51,397 | 41,702 | 9,653 | 42 | $674,890,015 |
| received before July 3, 2025 | 57,081 | 2 | 42,599 | 14,480 | $2,739,732,080 |
The source calls these claims, not unique people. It does not publish a person-level deduplication key, survivor relationship, resubmission flag, or cross-cohort link. The map therefore never relabels 108,478 claims as 108,478 claimants.
The 99.6% denominator excludes pending claims
DOJ labels the post-law percentage % Approved/of Disposed. Reproducing it requires dividing 9,653 approvals by 9,695 disposed claims: 9,653 / (9,653 + 42) = 99.6% after rounding. Pending claims are outside that denominator.
If the question is instead “what share of all post-law claims in this snapshot is approved?”, the calculation is 9,653 / 51,397 = 18.8%. Neither percentage predicts the final outcome of pending claims. They answer different questions:
- 99.6% describes the current mix among claims already labeled approved or denied.
- 18.8% describes approved claims as a share of every post-law claim in the snapshot.
- 81.1% describes the pending share and is the main reason those two approval percentages diverge.
The legacy cohort has almost no pending inventory, so its displayed 74.6% approved-of-disposed value is numerically close to approved-of-total. That does not make it a valid benchmark for the much newer cohort. Cohort age, eligibility rules, evidence requirements, and category mix differ.
July to August is a net state diff, not a throughput log
DOJ's July 1 post-law snapshot contained 40,542 total claims: 33,913 pending, 6,624 approved, and five denied, with $461,490,015 approved. By August 27, the published net changes were:
| Post-law state | July 1 → August 27 net change |
|---|---|
| total claims | +10,855 |
| pending | +7,789 |
| approved | +3,029 |
| denied | +37 |
| approved dollars | +$213,400,000 |
Those state differences balance: 7,789 + 3,029 + 37 = 10,855. They still do not reveal gross filings, individual processing time, withdrawals, corrections, appeals, or the path each claim took. A pending claim can become approved or denied while newly received claims enter pending; the public tables show endpoints, not event-level transitions.
The category changes reinforce that point. Downwinder total claims increased by 7,248; Manhattan Project Waste by 3,259; Uranium Worker by 300; and Onsite Participant by 48. Yet pending and disposed changes differ within every row. The dashboard supports a net category diff, not a claim-arrival rate or a forecast of when the current backlog will clear.
The legacy table shows why even a flat total is not a frozen state. Its total remained 57,081 between July 1 and August 27, while pending fell by one, approved rose by three, denied fell by two, and approved dollars rose by $300,000. The source does not explain the individual reclassifications. A model that assumes states only move from pending to one irreversible outcome would misread that revision.
Award amounts do not form one flat multiplier
The amended program page states a one-time $100,000 amount for qualifying Downwinders and Uranium Workers. Onsite Participant awards are also $100,000 but can be offset by amounts received from the Department of Veterans Affairs for the same illness.
Manhattan Project Waste uses a different rule. For a qualifying person living at filing, the amount is the greater of $50,000 or documented eligible out-of-pocket medical expenses. For a qualifying person who died before filing, the program page describes a $25,000 survivor payment under the stated priority. That variable structure is why multiplying every approval by $100,000 would not reproduce the dashboard.
RECA is also not a general causal finding. DOJ describes it as a non-adversarial alternative to litigation in which a claimant establishes the statutory diagnosis, location or work, and period criteria; the program does not require proof of causation. An approval confirms the program's eligibility determination, not a scientific conclusion about an individual's disease mechanism.
Confirmed, derived, and still unknown
The controlling sources confirm the two cohort labels; every August 27 state, category, and dollar total; the July 1 baseline; the approved-of-disposed label; the four amended categories; award rules; the electronic or mailed filing paths; and the December 31, 2027 claim deadline published by DOJ.
MyMap derives the 81.1% pending share, 18.8% approved-of-all share, denominator reconstruction, July-to-August net differences, category deltas, and warning that the snapshots are not an event log. Each derived value can be reproduced from the cited tables.
The sources leave unknown the exact extraction time, unique claimant count, individual state histories, gross filings during the interval, processing duration, reasons for denials or revisions, evidence completeness, portal-versus-mail mix, future awards, and the final form of revised regulations DOJ says it is working to issue during 2026.
Reproducible extraction method and next step
Save each dated table with cohort, category, pending, approved, denied, approved dollars, source URL, source update date, and observation time. First verify total = pending + approved + denied. Then compute approved-of-disposed and pending share with explicit formulas. Join snapshots only on the exact cohort and category label; flag taxonomy changes instead of forcing a match.
For the next update, compare a new DOJ snapshot to August 27 and report net state changes without calling them submissions or processing completions. Keep person-level eligibility work outside the aggregate map: applicants should use DOJ's current category instructions, official forms or claim portal, and claim-number evidence process. This article explains the feed; it does not determine eligibility, replace the controlling law, or provide legal or medical advice.