Repository navigation
UK spine: JSA claimants 2.8× the DWP caseload — legacy-benefit reporting after the UC migration #869
Description
Found while measuring the first calibrated rowwise UK local candidate (#762, receipts R8–R15 on uk-rowwise-candidate-762-b). The row is a reviewed measure exclusion under #762 A16, expiring 2026-10-03, tracked on #736.
Measured on spine-m (FRS 2024-25, policyengine-uk 2.92.1, calibration year 2025), single-block engine resolution: dwp.jsa_claimants — spine 199k against DWP 71k (2.8× at initialization). Block-insensitive, so the miss is real. The related OBR JSA spend row (obr.jobseekers_allowance, 3.4× at initialization) is reachable by reweighting and is not excluded, which says the spine has too many JSA rows rather than too much JSA per row.
Likely cause: legacy-benefit reporting after the UC migration. FRS respondents still report JSA where the live claim is UC with searching-for-work conditionality (or new-style contributory JSA is conflated with the retired income-based JSA); the spine's legacy-proxies / take-up stages then keep them as JSA claimants.
Ask: reconcile the spine's JSA claimant count against the DWP JSA caseload (contributory vs income-based, the latter structurally near zero in 2025) and reclassify FRS-reported income-based JSA to UC where the respondent's circumstances say UC; keep contributory JSA as reported.
Acceptance: the row sits within 25 % at initialization on the next spine and the A16 exclusion retires.
Activity
Vahid's PR B review round 1: bind the signed report to the attempt, p…
Vahid's PR B review round 1: bind the signed report to the attempt, p…
juaristi22 commentedon Sep 16, 2026
Re-adjudicated to 2026-12-08 in #937. DWP's 71,000 at August 2025 is New Style (contributory) JSA after the August-2025 methodology change; FRS 2024-25 reports 63k JSA benefit units at design weights (37k contributory over 17 records, 23k income-based over 9), so the source reproduces the count. The spine carries 178k contributory and 92k income-based reporters (GB, design weights) because 161k of the 181k contributory reporters are the SPI support channel's synthetic households; the stage-6
legacy_jobseeker_proxy(967k) is consumed by nothing. On Q50f the row measures 34k after the solve pares the reporters to fit the OBR JSA spend line. ESA: FRS 2024-25 reports 1.06m ESA units against DWP's 1.0m, the spine 791k (the SPI channel under-carries them); DWP's contributory (620k) and income-related (180k) figures are single-type counts and the total includes both-type and credits-only claimants, now declared on the three rows' notes. The repair is the SPI channel's fill, not a JSA rule.juaristi22 commentedon Oct 14, 2026
Status 2026-10-14, for the next triage round.
The direction reversed after #1100:
spi_benefit_coherencezeroes JSA on SPI rows, so the row is now under DWP's 71k New Style count, not over. On the published cut it starts at 1.65× and ends at −33%; on #1100's arm E it starts at −73% and ends at −36% (#796 audit, 2026-10-06). Two defects to fix together: the zeroing also drops about half of the FRS-reported JSA that SPI copies carried, and the exclusion's reason compares contributory plus income-based JSA with a contributory-only measure. The exclusion expires 2026-12-08, and its "next spine with the SPI repair" trigger has passed.Next: re-measure on the post-#1256 spine; decide whether SPI copies keep their twin's contributory JSA report (as #1256 keeps a no-record twin's values); then restate the reason or retire the entry. The two ESA count rows belong in the same pass (contributory kept, income-related zeroed since #1100).