What
policyengine_uk_data/targets/sources/housing.py builds the national calibration target housing/total_mortgage as £1,100/month × 12 × 7.5m owner-occupiers with a mortgage (~£99bn), with reference_url pointing at the ONS bulletin "Private rent and house prices, UK: March 2026":
https://www.ons.gov.uk/economy/inflationandpriceindices/bulletins/privaterentandhousepricesuk/march2026
That bulletin publishes PIPR average private rents (UK £1,374/month at February 2026, etc.) and house-price statistics. It carries no average mortgage payment figure at all — the citation does not support the value. The module docstring lists sources for rent and tenure counts but names no source for the £1,100, and the 7.5m mortgagor-household count is likewise an unpinned ~ approximation with no published table behind it.
How this was found
During the uk-data → Chronicle migration (chronicle#133, chronicle#159), porting housing/total_mortgage was refused because no publisher source exists for the value; it is recorded as a first-class blocked row in chronicle's docs/pe-uk-source-checklist.md. Review on that PR flagged that the mis-citation is live in this repo and deserves its own issue rather than living only in a PR body.
Nearest sourced anchor (not a drop-in replacement)
English Housing Survey 2023-24 rented sectors, Annex Table 1.7: mean weekly housing costs for all mortgagors = £222.42/week ≈ £964/month. Caveats: England only (not UK), and per the table's note 5, EHS mortgagor costs include households set to a nominal £0.02/week (shared owners with no outstanding loan, equity-release mortgages at 55+), which depresses the mean. So it anchors the order of magnitude but cannot simply be substituted.
Ask
Rebuild housing/total_mortgage on a real published source — both the per-household payment and the mortgagor-household count need pinned publisher tables — or retire/replace the target. Until then the target inherits a citation that does not support it.
What
policyengine_uk_data/targets/sources/housing.pybuilds the national calibration targethousing/total_mortgageas £1,100/month × 12 × 7.5m owner-occupiers with a mortgage (~£99bn), withreference_urlpointing at the ONS bulletin "Private rent and house prices, UK: March 2026":https://www.ons.gov.uk/economy/inflationandpriceindices/bulletins/privaterentandhousepricesuk/march2026
That bulletin publishes PIPR average private rents (UK £1,374/month at February 2026, etc.) and house-price statistics. It carries no average mortgage payment figure at all — the citation does not support the value. The module docstring lists sources for rent and tenure counts but names no source for the £1,100, and the 7.5m mortgagor-household count is likewise an unpinned
~approximation with no published table behind it.How this was found
During the uk-data → Chronicle migration (chronicle#133, chronicle#159), porting
housing/total_mortgagewas refused because no publisher source exists for the value; it is recorded as a first-class blocked row in chronicle'sdocs/pe-uk-source-checklist.md. Review on that PR flagged that the mis-citation is live in this repo and deserves its own issue rather than living only in a PR body.Nearest sourced anchor (not a drop-in replacement)
English Housing Survey 2023-24 rented sectors, Annex Table 1.7: mean weekly housing costs for all mortgagors = £222.42/week ≈ £964/month. Caveats: England only (not UK), and per the table's note 5, EHS mortgagor costs include households set to a nominal £0.02/week (shared owners with no outstanding loan, equity-release mortgages at 55+), which depresses the mean. So it anchors the order of magnitude but cannot simply be substituted.
Ask
Rebuild
housing/total_mortgageon a real published source — both the per-household payment and the mortgagor-household count need pinned publisher tables — or retire/replace the target. Until then the target inherits a citation that does not support it.