UK local-surface facts for the 2025 calibration: household counts by nation and area, communal-establishment residents by age, mid-2025 population estimates, UC households by area for calendar 2025, and BRMA geography and rent inputs (PolicyEngine/microcosm#1123, PolicyEngine/microcosm#1090) #313

Follow-up scope: The remaining Scottish Census 2022 ward tenure-by-bedroom export and public UV404 output-area tenure ingestion are tracked in #326. PR #314 implements the other requested inputs and closes this issue when merged.

Summary

PolicyEngine/microcosm#1123 lists the defects a K=25 local-area build found on Microcosm's constituency and local-authority target surface. Four of them are facts-side.

• No household count below the UK total. The feed's only household control is ons.households_total at K02000001 (ONS Families and households 2025, LFS). Every area's census household count is uprated by one UK factor. That factor ignores that Scotland's census is a year later and that the nations grow at different rates. Nothing binds a nation's household total, so a solve can lose Northern Ireland's households without any row noticing.
• Ages in student areas cannot be reached. The local age rows come from the mid-year estimates, which include residents of communal establishments (student halls, care homes, prisons, armed forces bases). Microcosm's frame is private households only. The feed has no communal-establishment count by age. First calibrated rowwise UK local candidate (#495 increment 6) microcosm#762 (A1) already recorded that an adjustment "would be a chronicle fact request, never a runtime correction".
• Two local families sit on an older or a narrower period than the publishers now offer.
  • The ages are mid-2024. Mid-2025 is now out for local authorities in all four nations.
  • UC households by area is a May 2025 snapshot. Stat-Xplore now carries every month of 2025, which would match the national UC rows' calendar-2025 window.
• Private rents follow the region, not the area. Every household placed in a local authority keeps its source household's rent; half of local authorities have a 2025 mean private rent more than 15 % away from their region's. Section 5 asks for the inputs of a Microcosm stage that draws each private renter's rent from its assigned area's BRMA distribution. The same inputs serve the Scottish BRMA rent targets that Scottish private-rent targets: PIPR is BRMA-grain, so council cells cannot bind by translation microcosm#1090 left open.

Chronicle owns the preserved artifacts, source cells and emitted facts. Microcosm keeps the decisions on which rows become targets, any uprating or roll-forward to 2025, how the communal adjustment is applied, and which household control wins when sources disagree.

1. Household counts by nation and local area, 2024–2025

What the feed has

• The UK total from ONS Families and households in the UK 2025 (LFS).
• Census household counts per area: Census 2021 for England, Wales and Northern Ireland; Scotland's Census 2022.
• ons-households-by-type-country-2025 is archived, but only its Scotland worksheet is ported (one S92000003 fact).

Requested

• Scotland. NRS "Households and dwellings in Scotland, 2025" (published 30 July 2026, accredited official statistics): households and dwellings at mid-2025 by council area, with the small-area release by data zone.
  • Scotland total: 2.57m households, 2.76m dwellings.
  • Release: https://nrscotland.gov.uk/publications/households-and-dwellings-in-scotland-2025/ (data file household-est-25-data.xlsx).
  • Small areas: https://nrscotland.gov.uk/publications/small-area-statistics-on-households-and-dwellings-2025/
• Wales, estimate. Welsh Government "Household estimates: mid-2024" (published 28 January 2026, accredited): households at 30 June 2024 by local authority (1.39m in total).
  • Method: Census 2021 household membership rates applied to the ONS mid-year estimates. The release also revises mid-2022 and mid-2023, so port those years as well.
  • https://www.gov.wales/household-estimates-mid-2024-html
• Wales, projection. The 2022-based local-authority household projections, which give mid-2025 directly. Mark them as a projection.
  • https://www.gov.wales/local-authority-household-projections-wales-2022-based-html
• England. ONS "Household projections for England: 2022-based" (published 28 October 2025), Table 406: households by local authority and higher areas, mid-2022 to mid-2047. England is 24,440,280 at mid-2025.
  • England has no annual local-authority household estimate. The ONS user guide treats the LFS series as England's official estimate at national and regional level only. That is why the projection is requested.
  • https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationprojections/datasets/householdprojectionsforengland
• Northern Ireland. Land & Property Services housing stock by district (April 2025 and April 2026 counts, published by DoF NI).
  • NISRA's official household figures are the 2016-based projections, which Census 2021 has already overtaken. A dwelling-stock series is the best available way to roll the 2021 district counts forward.
  • https://www.finance-ni.gov.uk/topics/housing-stock-statistics
• GB countries and English regions, survey-based. The remaining worksheets of the archived householdsbyregionscountries2025.xlsx (sha 42073679…): England, Wales and the nine English regions, all-households rows, 2015 to 2025, with the publisher's coefficient-of-variation code and 95 % interval as fact metadata.
  • 2025 values: England 24.329m ±0.159m; Wales 1.404m ±0.051m; Scotland 2.520m ±0.071m.

Semantics to carry

• provenance_class: estimate (NRS, Welsh Government), projection (ONS England, Welsh projections), survey aggregate (LFS) or administrative dwelling stock (LPS).
• Universe: private households. NRS removes communal establishments and multi-household dwellings from council-tax records. LPS counts dwellings, not households.
• Reference date: 30 June for NRS, Welsh Government and ONS; April–June 2025 for the LFS; April for LPS.
• Boundary vintage of the area codes. The ONS projection also has a 2023-boundary variant.

Closes: PolicyEngine/microcosm#1123 item 6 (nation household rows and the agreement check between grains), and a per-nation replacement for the single UK uprating factor on the area household and tenure cells.

2. Communal-establishment residents by age, Census 2021/2022

What the feed has: census2021_ts001_oa (household versus communal residents, by output area, with no age split). Nothing by age.

Requested

• England and Wales. Census 2021 RM120, "Residence type by sex by age".
  • Household versus communal-establishment residents × sex × 24 age bands.
  • Nomis NM_2220_1 at 2024 Westminster constituencies (geography TYPE172; 575 constituencies, checked) and at local authorities.
  • https://www.nomisweb.co.uk/datasets/c2021rm120
• England and Wales, optional. TS048 (communal residents by establishment type, including "Other establishment: Education"), at the same geographies.
  • Examples: Leeds Central and Headingley 13,481; Sheffield Central 11,680; Bristol Central 6,863.
  • https://www.ons.gov.uk/datasets/TS048
• Scotland. Scotland's Census 2022 UV101a (residence type by sex by age, 20 bands), at UK Parliamentary Constituencies 2024 and council areas. UV414 (establishment type) is optional.
  • https://statistics.ukdataservice.ac.uk/dataset/scotland-s-census-2022-uv101a-usual-resident-population-by-sex-by-age-20
• Northern Ireland, district level. NISRA Census 2021 CT0105 (communal residents by sex, age and position in the establishment), by local government district.
  • https://www.nisra.gov.uk/publications/ct0105-communal-establishment-residents-sex-age-and-position-establishment
• Northern Ireland, ward level. NISRA MS-F04, communal tables by ward in broad age bands. This is the only route to Assembly-area or constituency counts (for example Belfast South and Mid Down), by aggregating wards.
  • https://www.nisra.gov.uk/publications/census-2021-main-statistics-communal-establishment-tables

Semantics to carry

• Census day (21 March 2021; 20 March 2022 for Scotland).
• Residence type as a dimension (household / communal establishment).
• Usual residents.
• Students are counted at their term-time address.
• Age bands as published.

Closes: PolicyEngine/microcosm#1123 item 5. Microcosm would take the communal residents out of the mid-year-estimate age rows, holding the census communal count constant as the ONS household projections do.

3. Mid-2025 population estimates (vintage refresh of the local age rows)

What the feed has: mid-2024 everywhere (ons/lad_population_by_age_2024, ons/pcon24_population_by_age_2024, nrs/pcon24_population_by_age_2024, nisra/pcon24_population_by_age_2024, ons/mye_2024_uk).

Requested

• ONS UK local authorities. ONS mid-2025 UK estimates by local authority, age and sex (published 1 October 2026, includes Northern Ireland).
  • Or extend the Nomis NM_31_1 pull behind lad_population_by_age_2024 to date=2025, if Nomis now carries the Northern Ireland rows.
  • https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/datasets/populationestimatesforukenglandandwalesscotlandandnorthernireland
• Scotland council areas. NRS mid-2025 estimates (published 14 July 2026).
  • https://www.nrscotland.gov.uk/publications/mid-2025-population-estimates/
• Northern Ireland districts. NISRA 2025 mid-year estimates by district, single year of age (published 24 September 2026).
  • https://www.nisra.gov.uk/publications/2025-mid-year-population-estimates-northern-ireland-and-estimates-population-aged-85
• Scotland constituencies. NRS "Other geographies, mid-2011 to mid-2024 (2022 data zones)" (published 1 September 2026), UK Parliamentary Constituencies 2024.
  • It replaces the 2011-data-zone edition (24 February 2026) behind nrs/pcon24_population_by_age_2024, which predates the revised migration estimates.
  • Please confirm whether it carries single year of age.
  • https://nrscotland.gov.uk/publications/other-geographies-mid-2011-to-mid-2024-2022-data-zones/
• Not yet published, for the record:
  • England and Wales constituency mid-2025 (ONS gives December 2026 to January 2027).
  • Northern Ireland Assembly-area 2025.
  • Scotland small-area mid-2025 (winter 2026).

Semantics to carry: reference date 30 June 2025; revised back-years kept as their own vintage; boundary vintage.

Closes: the vintage half of PolicyEngine/microcosm#1123 item 2 for the age family. Microcosm moves the local-authority age rows to mid-2025 and keeps an explicit hold for constituencies until their mid-2025 lands.

4. UC households by constituency and local authority, every month of 2025

What the feed has: uc_households_by_constituency_may_2025, uc_households_by_constituency_children_may_2025, uc_households_by_local_authority_may_2025. These are single-month snapshots.

Requested

• The same three Stat-Xplore cubes for January 2025 to the latest loaded month.
• The UC release is quarterly since May 2026. The 18 August 2026 release carries households to May 2026, and the next, on 17 November 2026, adds to August 2026.
• https://www.gov.uk/government/collections/universal-credit-statistics

Semantics to carry:

• Month as the period.
• DWP revises recent months, so record the extraction date as the vintage.

Closes: the UC part of PolicyEngine/microcosm#1123 item 2. The national UC rows already bind calendar-2025 means; the area rows would follow the same window.

5. Private-rent inputs: area-placed rents (PolicyEngine/microcosm#1123 item 4) and Scottish BRMA rent targets (PolicyEngine/microcosm#1090)

Why: two Microcosm uses share these inputs.

• Item 4, the re-draw. Each private renter's rent would be drawn from the rent distribution of the Broad Rental Market Area (BRMA) that contains the household's assigned area, instead of carrying the source household's region-level rent.
  • The household's BRMA becomes a layer derived from its assigned output area (like its local authority and constituency), so BRMA, rent and the LHA rate all describe one place.
  • Today BRMA is drawn within the household's region, independently of its area.
• Scottish private-rent targets: PIPR is BRMA-grain, so council cells cannot bind by translation microcosm#1090, Scottish rent targets. PIPR publishes Scottish rents for 18 BRMAs, not for the 32 councils, so the 32 Scottish council rent cells are deferred.
  • Measure Scotland's BRMA × council overlap; keep BRMA rents out of council targets microcosm#1092 measured the BRMA × council overlap and kept the deferral.
  • The routes left open are BRMA-level target rows (12 × BRMA mean rent × BRMA private-renter households, with the model side summed over each household's BRMA) or a signed exception for the four councils that coincide with one BRMA.
  • Both need the same BRMA geography as item 4. The BRMA-level route also needs private-renter counts by BRMA and current BRMA rents.

Where the pieces end up is not decided. The OA-to-BRMA layer, the rent distributions and the BRMA renter counts are derived build artifacts that would stay in Microcosm with sidecar provenance, like the atomic-area supports. Several of their publisher originals are pinned today inside Microcosm (PolicyEngine/microcosm#1092's build tool, from a local cache) or in policyengine-uk-data (the PolicyEngine/policyengine-uk-data#516 BRMA table). Which copy each consumer reads will be agreed once Chronicle holds them. This section asks Chronicle to hold the originals, raw-only on the #269 precedent, plus the publisher aggregates as facts.

No published local-authority-to-BRMA lookup exists. BRMAs do not nest in local authorities (152 English BRMAs against 296 districts; 18 of Scotland's 32 councils fall in more than one BRMA). The link has to be built from boundaries and centroids.

Already archived, no action:

• NRS "Census 2022 index" (nrs/census2022_index, sha 0e4096a3…): postcode to output area, and output area to ward and council. This is the same file Measure Scotland's BRMA × council overlap; keep BRMA rents out of council targets microcosm#1092 pins.
• PIPR July 2026 edition (ons/pipr_rents_by_area_june_2026), including the 18 Scottish BRMA rows (all-dwellings rental price only).

Requested: raw-only registrations (artifacts and checksums, no facts)

BRMA boundaries

• England. VOA "Broad rental market area boundary layer for GIS, applicable May 2020" (published 4 June 2020; GML in a ZIP despite the page calling it a shapefile). This is the latest English layer.
  • Page: https://www.gov.uk/government/publications/broad-rental-market-area-boundary-layer-for-geographical-information-system-gis-applicable-may-2020
  • File: https://assets.publishing.service.gov.uk/media/5ee0ef64d3bf7f1ec05388df/GML.zip
• Wales. The Rent Officers Wales BRMA layer (May 2012 boundaries, release of 23 September 2014).
  • The only copy found online is the archived deposit at Edinburgh DataShare, which encloses Rent Officers Wales' correspondence granting OGL: https://datashare.ed.ac.uk/bitstreams/c2fe4c38-fd3e-4498-a4f9-62ac32163311/download
  • If Chronicle's governance requires the publisher's own URL, please say so; a request to Rent Officers Wales would be the alternative.
• Scotland. Scottish Government BRMA polygons (2009 boundaries, revised 28 July 2015): https://maps.gov.scot/ATOM/shapefiles/SG_BroadRentalMarketAreas_2009.zip
  • Measure Scotland's BRMA × council overlap; keep BRMA rents out of council targets microcosm#1092 pins sha 4d121236…
• Scotland, postcode lookup. Rent Service Scotland's postcode-to-BRMA lookup (FOI 202300368850, published 15 November 2023). It corrects the polygon placement of border output areas: https://www.gov.scot/publications/foi-202300368850/
  • Measure Scotland's BRMA × council overlap; keep BRMA rents out of council targets microcosm#1092 pins sha 50366fea…

Centroids, for placing small areas in BRMAs

• England and Wales. ONS "Lower layer Super Output Areas (December 2021) EW Population Weighted Centroids" (V4). Each output area takes its LSOA's BRMA, as Weight FRS BRMA draws by census private-rented households policyengine-uk-data#516 does.
• Scotland. NRS "2022 Census Output Area Population Weighted Centroids": https://www.nrscotland.gov.uk/media/gm4dvdsv/output-area-2022-pwc.zip
  • Measure Scotland's BRMA × council overlap; keep BRMA rents out of council targets microcosm#1092 pins sha 509d7d93…

Lists of rents

• England. VOA (HMRC) "Shadow list of rents April 2026", 1 October 2024 to 30 September 2025 (published 1 April 2026; CSV, 15.6 MB).
  • 502,239 individual weekly net rents with columns NET_RENT, PERIOD, BRMA, LHA_TYPE: 152 BRMAs, LHA categories A–E.
  • It has no local-authority or postcode field.
  • https://www.gov.uk/government/publications/shadow-list-of-rents-april-2026-collated-1st-october-2024-30th-september-2025
• Wales. Welsh Government FOI release ATISN 25142, "Shadow list of rents", 1 October 2022 to 30 September 2023 (published 6 October 2025; Excel, 3 MB). It is two years older than England's list: https://www.gov.wales/node/71335
• Scotland. Rent Service Scotland market evidence, FOI 202200303624, "BRMA market evidence and calculations" (published 28 August 2023). Requested: the 2016–2021 workbook.
  • One row per letting, with BRMA, postcode, net rent, payment frequency, property type, dwelling type and LHA category. The year to September 2021 alone has 31,833 lettings.
  • The release also has a 2010–2015 workbook, which is not needed.
  • It is the only record-level Scottish rent list, and the only rent source finer than a BRMA, since its postcodes place lettings in councils.
  • Measure Scotland's BRMA × council overlap; keep BRMA rents out of council targets microcosm#1092's dispersion experiment pins it (scot_foi_brma_2016_2021.xlsx, sha dba86fa1…).
  • https://www.gov.scot/publications/foi-202200303624/

Requested: facts

• PIPR, latest edition (16 September 2026, data to August 2026; next 21 October 2026). Rental price for all dwellings, by bedrooms (one, two, three, four or more) and by property type (detached, semi-detached, terraced, flat or maisonette).
  • Two separate breakdowns, not a cross.
  • Local authorities for England and Wales; the 18 BRMAs for Scotland; Northern Ireland as a whole.
  • The Scottish BRMA rows are the target side of Scottish private-rent targets: PIPR is BRMA-grain, so council cells cannot bind by translation microcosm#1090's BRMA-level route. The district rows validate the item 4 re-draw by area and bedrooms.
  • https://www.ons.gov.uk/economy/inflationandpriceindices/datasets/priceindexofprivaterentsukmonthlypricestatistics
• Scotland's Census 2022, households by detailed tenure and bedrooms, by 2022 electoral ward. The finest level at which this table is not suppressed.
  • It is a table-builder export with no stable URL. Database household2022ftb, all 355 wards; Measure Scotland's BRMA × council overlap; keep BRMA rents out of council targets microcosm#1092 pins its export as sha 03d6157d…
  • Please register the export and emit the ward counts as facts.
  • It gives the private-renter households that Scottish private-rent targets: PIPR is BRMA-grain, so council cells cannot bind by translation microcosm#1090's BRMA-level targets multiply by the BRMA mean rent, once wards are split across BRMAs by their output areas.
  • If NRS publishes tenure alone at output-area level, please say so: that would remove the ward split for the counts.
  • https://www.scotlandscensus.gov.uk/webapi/home
• Scottish Government "Private sector rent statistics, Scotland, 2010 to 2025" (25 November 2025): mean, median and quartiles by BRMA and bedrooms, year to September 2025.
  • It is Scotland's only current published rent distribution.
  • https://www.gov.scot/publications/private-sector-rent-statistics-scotland-2010-to-2025/

The 2025 data-zone household estimates in section 1 carry the census renter counts forward to the calibration year.

Not requested

• Northern Ireland. NIHE defines its BRMAs as sets of postcode districts on a web page with no open reuse licence, and publishes no list of rents. Northern Ireland has no local private-rent target rows.
• Census 2021 tenure × bedrooms by LSOA for England and Wales (the Weight FRS BRMA draws by census private-rented households policyengine-uk-data#516 inputs). Neither use needs private-renter counts by BRMA outside Scotland.

Semantics to carry

• The lists of rents are rent-officer evidence collected for LHA: lets in payment, by BRMA and LHA category. VOA describes the England list as new and renewal rents negotiated between landlord and tenant. They are not the LHA rate (the 30th percentile, frozen at the April 2024 rates).
• The payment period (weekly, monthly) is a field and should be kept, not normalised.
• Collection windows run October to September.
• PIPR, the lists of rents and the Scottish rent statistics share the rent-officer evidence base. They are different statistics, not independent sources.
• Boundary vintages as published.

Closes: the inputs for PolicyEngine/microcosm#1123 item 4 and for both routes left open in PolicyEngine/microcosm#1090.

Not requested (answered by the checks)

• HMRC income by area. Tables 3.14 (local authorities) and 3.15 (constituencies) for 2023-24, published 29 April 2026, are still the latest. 2024-25 is expected around February to March 2027. At source they agree with Table 3.6 within 0.5 % on all four measures:
  • employment income £1,079.2bn against £1,084.2bn, and 27.14m against 27.26m people;
  • self-employment income £106.8bn against £107.5bn, and 3.67m people in both.
  • The gap UK local build (K=25, 2026-10-06): data-quality defects on the target surface and the spine microcosm#1123 reports comes from Microcosm holding the area rows at 2023-24 while the national rows are projected to 2025. No Chronicle change is needed.
• CGT total. The £119.3bn is HMRC CGT statistics Table 1, 2024-25: individuals' gains before the annual exempt amount, excluding trusts. The revised 2023-24 value is £66.6bn. The feed carries the publisher's figures; which tax year stands for 2025 is a Microcosm period-rule choice. 2024-25 includes disposals brought forward ahead of the October 2024 rate change.