Scottish, Welsh and NI lists of rents are copies of English BRMAs' lists #2036

policyengine_uk/parameters/gov/dwp/LHA/lha_list_of_rents.csv.gz holds the 2019 and 2020 lists of rents. Its Scottish, Welsh and Northern Ireland lists are not those nations' rents.

• They are copies. Every one of the 480 non-English (BRMA, category, year) lists is byte-identical to an English list from the same year. Each non-English (BRMA, category) took the English list, of any category, whose 2019 30th percentile is closest to the nation's April 2019 LHA rate (240 of 240 pairs).
• Provenance. policyengine-uk#787's data README, since deleted, said: "Scottish, Welsh and Northern Ireland BRMAs are matched from the closest English BRMA". The English lists match the VOA originals exactly.
• Full evidence: BRMA imputation draws Scottish, Welsh and NI households from copies of English rent lists policyengine-uk-data#515.

Where it matters

• On main, Scottish, Welsh and NI LHA rates are percentiles of these copies. In 2020 they are a median 8.6% (Scotland), 4.6% (Wales) and 6.1% (NI) below the published rates. Use published LHA determinations and apply Sch 3B paras 3 and 3A #2022 replaces them with published rates.
• On Use published LHA determinations and apply Sch 3B paras 3 and 3A #2022, utils/lha.py _percentile_ratios scales the published 30th percentile by each list's percentile ratio for percentile reforms. Commit aef9c24 there gives every non-English BRMA the median English ratio.
• Genuine Scottish lists exist. The Scottish Government released Rent Service Scotland's records under FOI 202200303624. Sheet Y-1 reproduces the published 30th percentile for April Y in 522 of 540 cells to the penny (April 2017-2022). No genuine 2019-20 lists were found for Wales or Northern Ireland.

Fix: replace the Scottish lists with the FOI records, so Scottish BRMAs use their own ratios. Keep Wales and NI on the English median until genuine lists turn up.