[juaristi22]

Part of #435 (item 1) and #420 (items 12 and 13), the published facts Microcosm's next data PR (PolicyEngine/microcosm#1095) reads that are not yet in Chronicle. Still to come:

• Not published yet:
  • NEET's November 2026 edition (UK facts for Microcosm's next data PR: Scotland's second homes and empty dwellings, and NEET's November edition #435 item 2), which ONS releases on 26 November;
  • Ofgem's January–March 2027 cap (UK facts for four November events: young people and work, the PIP review, the January 2027 price cap and the Scottish Budget 2027-28 #420 item 10), due by 25 November.
• Blocked: Registers of Scotland's sales over £1 million and £2 million (UK facts for four November events: young people and work, the PIP review, the January 2027 price cap and the Scottish Budget 2027-28 #420 item 13's other part). The RoS property market report page refused the download (HTTP 403).

Five packages add 2,219 facts. Each comes from an official publication under the Open Government Licence v3.0, written by a generator script from pinned bytes; no microdata is read or stored.

Scotland's second homes and empty dwellings (#435 item 1)

scotgov-second-homes-empty-properties-2025 (396 facts), from the Scottish Government's Second homes and empty properties in September 2025 (published 16 December 2025). It ports all six tables of the web table, for Scotland and its 32 council areas, at September 2024 and September 2025:

• second homes;
• empty properties (empty for more than 6 months and liable for council tax);
• long-term empty properties (12 months or more);
• long-term empty properties with a discount below 10% or an increase;
• unoccupied exemptions;
• second homes plus empty properties.

Microcosm levels Scotland's council tax bills to the liability of every chargeable dwelling. These counts let it take out the second homes and empty dwellings, which no survey household lives in.

Semantics:

• Each class is its own concept, scotgov.council_tax.<class>, quoting the cover sheet's definition and the workbook notes on it.
• Each year is month YYYY-09 ("Data presented for September each year"), the month scotgov-council-tax-bands-2024 and -2025 give CTAXBASE's chargeable dwellings.
• Second homes and empty properties are chargeable_dwelling, unoccupied exemptions exempt_dwelling. The provenance class is administrative (the issue's admin_aggregate).
• gov.scot's caveats travel with the facts. Before the guidance was strengthened in September 2024, empty homes "will therefore be overstated in such instances" (note 10). South Ayrshire still includes "a very small number" of short-term empties (note 11), and Na h-Eileanan Siar's 2024 figures "are not as up to date as they would normally be" (note 13).

Parity:

• Councils add to Scotland in every table and year but one misprinted cell. For long-term empty properties with a discount below 10% or an increase at September 2025, the Scotland row prints 25,073; its councils add to 25,703, the figure the edition's own text gives. The cell is ported as printed, and the generator fails if gov.scot corrects it.
• Classes add up. Second homes plus empty properties is the total in every area and year, and the subsets nest.
• No revisions. The September 2024 edition prints the same cells for every year to 2024.
• Year-on-year changes reproduce. The key points' changes from 2024 match the packaged cells: second homes −679, empty properties +915, long-term empty +741, unoccupied exemptions −2,790.

At September 2024, Scotland had 21,606 second homes and 43,538 empty properties, of which 31,596 were long-term empty.

Not in the package:

• Years before 2024, which UK facts for Microcosm's next data PR: Scotland's second homes and empty dwellings, and NEET's November edition #435 does not ask for.
• Splits by band or rate, and Band D equivalents by class. gov.scot publishes no split by council tax band or by discount or premium rate beyond the long-term empty table. CTAXBASE's tables give chargeable dwellings by band only, so Microcosm's share has to be taken in dwellings.

Payment methods by region (#420 item 12)

Three packages from DESNZ's quarterly domestic energy price statistics (updated 29 September 2026), for a Direct Debit, standard credit and prepayment draw in Microcosm (PolicyEngine/uk-energy-price-cap-jan-2027#4):

• desnz-qep-electricity-payment-methods-2026 (627 facts): table 2.4.2, standard electricity;
• desnz-qep-time-of-use-electricity-payment-methods-2026 (627): table 2.4.3, Economy 7 and other time of use electricity;
• desnz-qep-gas-payment-methods-2026 (561): table 2.5.2, gas.

They give the percentage of domestic customers paying by standard credit, Direct Debit and prepayment. Coverage is each quarter from April–June 2024 to April–June 2026, plus the annual sheets for 2024 and 2025.

• Filters: payment method (standard_credit, direct_debit, prepayment), fuel, and the metering arrangement (standard, economy_7_and_time_of_use), which keeps tables 2.4.2 and 2.4.3 apart.
• Regions: the 14 former PES areas, under the desnz:qep:pes ids and value ids the QEP unit cost packages already use. Gas uses them too, "for consistency with the electricity tables". The electricity tables add England and Wales, Scotland, Great Britain, Northern Ireland and the United Kingdom; the gas table stops at Great Britain.
• Register: the label register gains the four nation rows these packages share.
• Shares, not counts: each region's three shares are rounded percentages that add to 100 within 2 points. The survey of suppliers is "uplifted to Ofgem's total customer numbers".
• Annual figures: table 2.4.3's annual figures "are averages of the rounded quarterly figures".
• Not ported: earlier periods, the fixed-tariff sheets, and the supplier-type tables 2.4.1 and 2.5.1.

Bands I and J (#420 item 13)

scotgov-council-tax-high-value-bands-consultation-2026 (8 facts), from the Scottish Government's Council Tax High-Value Property Bands (Mansion Tax): consultation (6 July 2026), Figures 2 and 3. It holds the published low and high estimates for Scotland:

• Band I (more than £1 million and up to £2 million): 12,680 to 16,740 properties at 1 April 2026 values, and £9.13m to £12.05m a year of additional revenue at the paper's illustrative rates on 2026-27 Band D charges.
• Band J (more than £2 million): 760 to 1,070 properties, and £2.74m to £3.85m a year.

How the facts are typed:

• Filters: band is council_tax_band (I, J) and the bound is statistic (low_estimate, high_estimate).
• Properties are an observation at month 2026-04.
• Revenue is a source_projection for fiscal year 2026-27, as the OBR's council tax surcharge costing is.
• Provenance: both are model_output. The paper calls its figures "illustrative".

The chart labels print the low estimates for Band I and J, then the high ones. The tests check that order against the paper's own totals: "approximately 15,000 properties" and "between £12 million and £16 million per year".

Checks

• Generators: --check reports each generator's committed YAML up to date.
• Tests:
  • the three package test files (31);
  • test_dimension_value_register.py (7);
  • test_issue_376_one_label_per_value.py (48);
  • test_chronicle_bundle.py with the whole-bundle test deselected (34);
  • test_chronicle_boundaries.py (48);
  • test_issue_270_qep_energy_prices.py (17).
• Register: the static check finds no problems.
• Bundle pins move by the packages' own counts:
  • facts 2,245,294 (+2,219), packages 506, tables 448;
  • desnz 354,399, scotgov 7,058;
  • households 680,028, dwellings 708,936, government 14,075.
• Unchanged: periods (610) and semantic duplicate keys (3,930). The period, entity and source pins each sum to the fact count. The whole-bundle test runs in CI.
• R2: all five raw files are uploaded with chronicle publish-raw and read back with chronicle verify-raw.

🤖 Generated with Claude Code