Skip to content

Fix uk-ai-study post: scope inequality claim, correct decile labels, inline conclusion - #1141

Merged
MaxGhenis merged 1 commit into
mainfrom
fix/uk-ai-study-post-review
Aug 7, 2026
Merged

Fix uk-ai-study post: scope inequality claim, correct decile labels, inline conclusion#1141
MaxGhenis merged 1 commit into
mainfrom
fix/uk-ai-study-post-review

Conversation

@MaxGhenis

Copy link
Copy Markdown
Contributor

Review fixes to the live uk-ai-study post, verified against the attested canonical outputs in PolicyEngine/uk-ai-study (values_generated.tex / scenarios.csv at 5d75106):

  1. Internal consistency: the inequality section said "inequality moves in one direction" — contradicted by the post's own wage-margin family, where the Gini falls (−0.44pp). Now scoped to displacement scenarios with the exception stated.
  2. The wage-margin flip is now in prose (£26.6bn vs £18.2bn fiscal; +0.10 vs +1.81pp poverty; −0.44 vs +1.04pp Gini) — it previously lived only inside a table iframe.
  3. Decile denominator corrected: transition shares are of people in person-level deciles of size-adjusted household income; households don't transition to unemployment.
  4. Conclusion inlined as text (was an iframe — invisible to search/RSS/screen readers), with its exposure-vs-uniform comparison corrected from "roughly a quarter more" to "roughly a third more" (18.2 vs 13.9 = +31%) and the canonical figures added.
  5. Central headline numbers added as body text (£18.2bn, +1.85pp AHC, +1.04pp Gini) — previously embed-only.
  6. OBR link now points at the specific March 2026 EFO.

cc @vahid-ahmadi — all numbers from your attested build; happy to adjust wording.

🤖 Generated with Claude Code

…inline conclusion

Review fixes to the published uk-ai-study research post:

- Scope the inequality-section claim to displacement scenarios and state
  the wage-margin exception (the Gini falls, -0.44pp, under wage cuts) —
  as written it contradicted the post's own adjustment-margin family
- State the wage-margin flip in prose (GBP 26.6bn vs 18.2bn fiscal,
  +0.10 vs +1.81pp poverty, -0.44 vs +1.04pp Gini): the study's most
  policy-relevant nuance previously lived only inside a table iframe
- Correct the transition-gradient denominator: shares are of people in
  person-level deciles of size-adjusted household income, not
  "households transitioning to unemployment"
- Inline the conclusion as text instead of an iframe (invisible to
  search, scrapers, RSS and screen readers) and correct its
  exposure-vs-uniform comparison to "roughly a third more"
  (canonical 50-draw means: 18.2 vs 13.9 = +31%), adding the canonical
  figures
- Add the central headline numbers (GBP 18.2bn, +1.85pp AHC poverty,
  +1.04pp Gini) as body text — previously they existed only in embeds
- Link the OBR receipts claim to the specific March 2026 EFO

All figures verified against the attested outputs in
PolicyEngine/uk-ai-study (values_generated.tex / scenarios.csv at
5d75106).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
@vercel

vercel Bot commented Aug 7, 2026

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
policyengine-calculator-next Ready Ready Preview Aug 7, 2026 2:54pm
policyengine-website Ready Ready Preview Aug 7, 2026 2:54pm

Request Review

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant