UK facts for the microcosm#1095 ports: HMRC Stamp Tax Statistics and the corporate share of SDLT, DWP's Move to UC claim rates, the geography of OBR table 4.9's DWP lines, and HBAI with vintages #357

Summary

PolicyEngine/microcosm#1095 tracks the fixes Microcosm ports from the policyengine-uk-data wave. Five of
them wait for facts the feed does not carry, because Microcosm compiles its targets and its validation
references from Chronicle facts only:

1. HMRC Annual Stamp Tax Statistics 2024 to 2025, Tables 6b, 8a and 8b.
2. The corporate share of SDLT in OBR's forecast years.
3. DWP's Move to Universal Credit claim rates, by combination of legacy benefits.
4. The geography of OBR table 4.9's DWP lines: Great Britain, or England and Wales.
5. Households Below Average Income for FYE 2025, with confidence intervals, in a package that keeps
   each release.

Chronicle owns the preserved files, the source cells and the emitted facts. Microcosm keeps the
decisions about how each fact is bound.

1. HMRC Annual Stamp Tax Statistics 2024 to 2025

Why. Two Microcosm fixes wait for these tables.

• Additional-dwelling and non-residential purchases are not drawn. Microcosm draws only whether a
  household bought its main home. policyengine-uk then treats the household's whole other-property
  stock as an additional purchase at the higher rates (additional_residential_property_purchased treats a household's whole other-property stock as an additional purchase policyengine-uk#2238).
  Impute additional-dwelling and non-residential purchases, calibrated to HMRC's HRAD statistics policyengine-uk-data#573 draws those purchases and sets their level from these tables.
  Microcosm would bind the same publisher figures as targets.
• obr.sdlt covers every buyer. The feed's only SDLT fact is OBR's receipts line (EFO March 2026,
  sheet 3.8), which includes companies, and Microcosm fits household SDLT to it.
  SDLT calibration target includes corporate SDLT, which policyengine-uk adds separately policyengine-uk-data#564 subtracts non-natural persons' share; Microcosm cannot until
  that share is a fact.

Source

• Publication page: https://www.gov.uk/government/statistics/uk-stamp-tax-statistics
• The corrected tables (ODS): https://assets.publishing.service.gov.uk/media/69aef72fc78869bf8eb8a589/Annual_Stamps_2024_25_correction.ods

Requested

• Table 6b, higher rates on additional dwellings: transactions, value and receipts, in total and by
  price band, with the surcharge's own receipts where the table separates them.
• Table 8a, residential transactions by type of buyer (individuals and non-natural persons):
  transactions, value and receipts.
• Table 8b, non-residential transactions by type of buyer: the same three measures, in total and by
  price band.
• Every financial year each table carries, so a later edition's revisions can be seen.

PolicyEngine/policyengine-uk-data#573 and PolicyEngine/policyengine-uk-data#564 quote figures from these tables. Please read them from
the workbook rather than from those issues.

Semantics to carry

• Coverage as the tables state it (the publication's notes give England and Northern Ireland,
  transactions of £40,000 or more), on each fact.
• Period: financial years ending in March.
• Rounding as published (the notes describe transactions to the nearest 100 and values and receipts
  to the nearest £5m), so band rows need not sum exactly to the headline rows.
• Whether Table 6b's receipts are gross of refunds on replacement main residences, and whether they
  cover every buyer.
• Edition: this is the corrected release. Keep the edition on each fact.

2. The corporate share of SDLT in OBR's forecast years

obr.sdlt is an all-buyer line. For years after HMRC's latest outturn, a household target needs a
corporate figure for the same year.

Requested: whichever split of SDLT receipts by type of buyer the Economic and fiscal outlook
publishes beside sheet 3.8. PolicyEngine/policyengine-uk#2322 cites a table for projecting the
corporate statistic. If the EFO has no buyer-type split, please say so on this issue, and Microcosm will
carry HMRC's outturn share forward under its own declared rule.

3. DWP's Move to Universal Credit claim rates

Why. From policyengine-uk 2.128.0, a family moves onto Universal Credit when DWP closes a legacy
benefit it receives, tax credits from 2025-26, and would_claim_uc_at_legacy_closure says whether it
claims (PolicyEngine/policyengine-uk#2119). Microcosm leaves the input at the engine's default, true, so
its 2025 Universal Credit targets are fitted on the assumption that every family moved off a legacy
benefit claims. PolicyEngine/policyengine-uk-data#534 draws the input at DWP's claim rates for each
combination of legacy benefits. Microcosm would draw it at the same rates, from Chronicle facts.

Source

• DWP Stat-Xplore, the "Households invited to Move to Universal Credit" dataset: the table by legacy
  benefit combination (Move legacy-benefit reporters onto Universal Credit at Move to UC claim rates policyengine-uk-data#534 cites "MtUC Households 3"). https://stat-xplore.dwp.gov.uk/
• The publication: Move to Universal Credit statistics, official statistics under development on the
  people sent managed migration notices.
  https://www.gov.uk/government/collections/move-to-universal-credit-statistics

Requested

• Households sent a migration notice, by combination of legacy benefits, with each outcome: claimed
  Universal Credit, did not claim, and still in progress.
• The all-households row.
• By the month the notice was sent, or cumulatively to the latest month, whichever the table carries.
• Every release the series has, since it is under development and revised.

Semantics to carry

• Coverage: Great Britain.
• DWP's definitions of each outcome, including whether a claim made after the legacy award closed
  counts as claimed, and how in-progress cases are counted.
• The notice period covered and the date of the extract.
• The statistics' status, official statistics under development, on each fact.

4. The geography of OBR table 4.9's DWP lines

Why. The feed stamps every fact from table 4.9 (EFO March 2026 detailed forecast tables,
expenditure) with the United Kingdom (K02000001). But DWP's lines in it cover Great Britain and
residents overseas (DWP's benefit expenditure and caseload tables, Spring 2026, Notes 3), and Northern
Ireland's social security has rows of its own. Since executive competence moved to the Scottish
Government, DWP reports Attendance Allowance, Carer's Allowance, DLA and PIP, and Winter Fuel Payment for
England and Wales only (Carer's Allowance from 3 September 2018; DLA, PIP and Attendance Allowance from
1 April 2020; Winter Fuel Payment from 1 April 2024). Microcosm now applies those scopes itself when it
binds the rows (PolicyEngine/microcosm#1100). The scope belongs on the facts, so every consumer reads the
lines as DWP counts them.

Source

• OBR, Economic and fiscal outlook, March 2026, detailed forecast tables: expenditure, table 4.9.
  https://obr.uk/efo/economic-and-fiscal-outlook-march-2026/
• DWP, Benefit expenditure and caseload tables 2026, the notes on coverage.
  https://www.gov.uk/government/publications/benefit-expenditure-and-caseload-tables-2026

Requested

• Each DWP line of table 4.9 stamped with the geography its figures cover: Great Britain with residents
  overseas, or England and Wales for the four benefits above from the dates DWP gives (with residents
  overseas wherever DWP's notes include them), and the years before those dates at Great Britain.
• The Northern Ireland rows stamped Northern Ireland.
• Child Benefit stays United Kingdom: it is an HMRC line that covers Northern Ireland.

5. Households Below Average Income, FYE 2025, with vintages

Why. Microcosm checks its income distribution against HBAI as validation only, never as a target
(the hbai_basis of PolicyEngine/microcosm#968). The check is to move to FYE 2025, the year the FRS
2024-25 spine describes, and to report each statistic as inside or outside DWP's confidence interval.
DWP revises earlier years in later releases, so the reference has to keep each release's values.

Source

• Households below average income: for financial years ending 1995 to 2025.
  https://www.gov.uk/government/statistics/households-below-average-income-for-financial-years-ending-1995-to-2025

Requested

• Median equivalised household income, before and after housing costs.
• The share of individuals in relative low income (below 60% of the median), before and after housing
  costs.
• The 95% confidence interval DWP publishes for each.
• FYE 2025 and the earlier years the release carries, with the release on every fact, so a later
  release's revisions sit beside it.

Semantics to carry

• Coverage: the United Kingdom, as HBAI states it.
• Period: financial years ending in March.
• The equivalisation scale, the housing-cost definitions and the confidence-interval method, as the
  release's methodology states them.

What this closes

• UK: track the defects uk-data is fixing that microcosm still has microcosm#1095: "Additional-dwelling and non-residential purchases are not drawn" (with
  Stop charging property stocks as purchases when a household buys its home policyengine-uk#2241), and "obr.sdlt fits households to the SDLT that all buyers pay".
• UK: track the defects uk-data is fixing that microcosm still has microcosm#1095: "Legacy-benefit reporters never move to UC", the rates for the draw.
• UK: track the defects uk-data is fixing that microcosm still has microcosm#1095: the open request under "OBR table 4.9 benefit rows are measured over the
  UK".
• UK: track the defects uk-data is fixing that microcosm still has microcosm#1095: "HBAI validation", the source for the re-based check (UK scorecard: populace_uk_2023 vs enhanced_frs_2024_25 vs admin benchmarks (2026) microcosm#731).

🤖 Filed with Claude Code