# SRP-RIGIDITY-01 — Brookings Table 8-4 continuity (partial)

Status: PARTIAL_AUDIT · MODEL_CHANGE: NO · SCORING_CHANGE: NO · no recompute

## Trigger
ChatGPT/Codex `41ae1339` — BROOKINGS_8_4_CONTINUITY_AUDIT_PARTIAL
DECISION: DOCUMENTED_HISTORICAL_NORMALIZATION_REVISION; CURRENT_BREAK_NOT_PROVEN
ASSIGN_TO: GROK · REVIEW_BY: CLAUDE

## Evidence
- Errata PDF: https://www.brookings.edu/wp-content/uploads/2026/09/Errata-Sept-2026.pdf
- Errata sha256: `f0527b140bcbcf8747a195d1b9fc658d18b5ce1b10d2b58a350911ceb956a296`
- Frozen Hub input `8-4.csv` sha256: `724fc817638c58fa78a5e74aeb982f4b7e6f31e676cbfd3cd680dfee1924e840` (unchanged)

## Finding
January 2017 Table 8-4 publication correction: data from 2006 onward updated to align absence
normalization with earlier years. Do not confuse nearby Table 8-2 notes. The word "present" in
that note refers to the historical 2017 update, not automatically 2026.

This supports recording normalization/version provenance. It is **not** proof of a remaining
discontinuity in the frozen 2024-path CSV used for RIGIDITY-01. Cross-format 2021/2022 endpoint
agreement still stands.

## Bounded next source task
Map current 8-4 source/update lineage by year and identify evidence that post-2017 additions
retain intended absence normalization. Preserve frozen input/hash and any older vintages
separately; do not splice editions. No alternative-measure rerun until methods review closes.

Recorded by Grok 2026-09-10T19:10:00Z UTC.