Give benefit-unit members aged 16 or over a bedroom in the LHA size criteria

[MaxGhenis]

Fixes #2008

Stacked on #2006 (base benefit-unit-rent-liability), which is stacked on #1896. Retarget to main once both merge.

Problem

The LHA size criteria gave no bedroom to a member of the benefit unit aged 16 or over who is not the claimant or partner. The commonest case is a young person at school or college living with a parent.

LHA_allowed_bedrooms (Universal Credit) and housing_benefit_LHA_allowed_bedrooms (Housing Benefit) counted three things:

• the claimant or couple;
• non-dependants or other occupiers aged 16 or over outside the benefit unit;
• children's rooms for members under 16.

So a lone parent with a 17-year-old at college was assessed for one bedroom (category B) instead of two (C). main subtracts every benefit-unit member aged 16 or over, which has the same effect.

The law

Universal Credit. UC Regs 2013 Sch 4 para 10(1): "A renter is entitled to one bedroom for each of the following categories of persons in their extended benefit unit":

• (a) the renter (or joint renters);
• (b) "a qualifying young person for whom the renter or either joint renter is responsible";
• (c) "a non-dependant who is not a child";
• (d)-(f) children.

Para 9(2) makes anyone else aged 16 or over who normally lives with the renter a non-dependant. The exception relevant here is (g): "a child or qualifying young person for whom no-one in the renter's extended benefit unit is responsible". An example is a young person looked after by a local authority (reg 4(6)). A child is a person under 16 (WRA 2012 s.40); a qualifying young person is defined in reg 5.

Housing Benefit. HB Regs 2006 and HB (SPC) Regs 2006 reg 13D(3) (SI 2006/213, SI 2006/214): "one bedroom for each of the following categories of occupier ... (b) a person who is not a child". A child is "a person under the age of 16" (reg 2(1) of each).

Change

	Bedroom for a member aged 16 or over who is not the claimant or partner
UC (LHA_allowed_bedrooms)	a qualifying young person the renter is responsible for (is_child_or_qualifying_young_person_for_universal_credit, para 10(1)(b)); any other member as a non-dependant (para 10(1)(c)), except a qualifying young person no one is responsible for (para 9(2)(g))
HB (housing_benefit_LHA_allowed_bedrooms)	any such member (reg 13D(3)(b))

The counts of non-dependants and occupiers from other families are unchanged, as is bedrooms_for_children.

Why the UC rule also covers non-dependants. The issue's UC example is the qualifying young person. A benefit-unit member aged 16 or over who is not a qualifying young person is, in law, a non-dependant with their own bedroom. lha_renter_has_non_dependant already treats them as one. On the enhanced FRS that group is 19k people in 20 survey households, against 1.95m qualifying young people.

Invariants

Hypothesis properties are in test_lha_bedroom_young_people_properties.py: 25 examples each, derandomized, over populations of up to 4 households with sharers, boarders, lodgers and non-dependant families.

1. Monotonicity. Adding a person aged 16 to 19 to any family never lowers any family's UC or HB bedrooms. This holds whether the claimant and partner are supplied or inferred.
2. Exact effect. When the added person is not the claimant or partner:
   • their family gains one HB bedroom;
   • their family gains one UC bedroom, unless they are a qualifying young person no one is responsible for;
   • the household head's family gains one bedroom under both schemes if the person joins a non-dependant's family, and one HB bedroom if they join a boarder's or lodger's family;
   • nothing else changes.
3. Reference agreement (differential). Every family's bedrooms equal an independent count of the size criteria. Children's rooms in that count are the fewest rooms holding them two to a room (same sex, or both under 10), found by brute force. This also checks bedrooms_for_children's greedy pairing against the optimum on every generated family.

Tests

• New LHA_young_person_bedroom.yaml, 7 cases at 2026:

  • qualifying young people aged 17 and 19;
  • an 18-year-old non-dependant;
  • a looked-after qualifying young person: no UC bedroom;
  • a couple with a young person, a boy of 12 and a girl of 8: 4 bedrooms, E;
  • a partner aged 19: no extra bedroom;
  • a pension-age couple's grandchild at college under HB (SPC): HB meets a £200-a-week rent in full.

  The £ expectations use a rent of £200 a week. That is between the published Maidstone one- and two-bedroom rates (VOA LHA April 2024, held for 2025-26: B £172.60, C £208.27), so they hold whatever the model's rate table says. On the pre-change formulas, 5 of the 7 cases fail. The 2 that pass are the regression guards for the looked-after and partner cases.

• LHA_claimant_composition.yaml: two cases asserted "existing bedroom formula = 1: B", for a 17-year-old apprentice non-dependant and an 18-year-old qualifying young person. They are now C, citing para 10(1)(c) and (b).

• policyengine-core test policyengine_uk/tests/policy -c policyengine_uk: 1,507 passed.

• pytest policyengine_uk/tests --ignore=policyengine_uk/tests/microsimulation on 7c6f5f3: 2,044 passed, 19 skipped, 1 failed. The one failure, test_deterministic_variables.py::TestHeadOrdering::test_oldest_person_is_head_and_gets_rent, also failed on the base. Give sharers, boarders and lodgers their own rent, tenure and LHA category #2006 fixed it in b065a48, now merged here, and test_deterministic_variables.py passes 25/25 on the merged head. The 3 new Hypothesis properties pass at 25 examples each. A full re-run on the merged head (49866c2): 2,045 passed, 19 skipped, 0 failed.

• ruff format and ruff check on the changed files: clean.

• Microsimulation tests (make test-microsimulation) were not run. The impact below is a direct base-vs-branch run instead.

Impact (real microsimulation)

Enhanced FRS 2024-25 as released (policyengine-uk-data 1.57.4, enhanced_frs_2024_25.h5, sha256 03fe15e4…), year 2026. Base is #2006's head b065a48; the branch is 49866c2. The first runs, on 190482e and a9e1b4f, gave identical output. Same dataset in both runs.

Every figure below comes from at least 10 distinct survey households. A survey household is counted by its rent and members' ages, which every enhanced-FRS record shares with its FRS 2024-25 source household. Collisions only merge households, so the counts are lower bounds. Cells under 10 are suppressed.

Who is affected

• Benefit-unit members aged 16 or over who are not the claimant or partner: 1.97m people.
  • 1.95m are UC qualifying young persons their parent is responsible for.
  • 19k are not qualifying young persons, so they are non-dependants.
  • None are qualifying young persons no one is responsible for; the data has no looked-after flag.
• Benefit units whose UC and HB bedroom counts rise: 1.80m (770 survey households), +1: 1.63m; +2: 158k; +3: suppressed. None falls.
• Of those, 363k rent privately (130 survey households). Changes in their LHA category (identical under UC and HB): B→C 43k, C→D 222k, D→E 58k. Other transitions are suppressed.

Money (£ a year, 2026)

	Change
UC housing costs element, as computed for every renting benefit unit	+£558m
of which within UC awards (161k benefit units, 37 survey households)	+£308m
of which in benefit units with no UC award (no effect on income)	+£250m
Universal Credit paid (152k benefit units gain, 33 survey households; mean gain £1,800)	+£273m
Housing Benefit	changes in fewer than 10 survey households: suppressed
Household net income	+£273m
People in poverty (BHC and AHC)	changes in fewer than 10 survey households: suppressed

The roughly £35m between the element's rise within awards and UC paid is withheld by the benefit cap and the UC deductions rules. Each applies in fewer than 10 survey households, so they are not shown separately.

In fewer than 10 survey households the LHA cap falls when the bedroom count rises. In 5 of 200 BRMAs, the model's rate table, which is its 2020 list of rents uprated by the private rent index, prices a larger category below a smaller one. That is a separate, existing issue.

Known gaps (not changed here)

• Members of the benefit unit who are non-dependants (aged 16 or over, not a qualifying young person) now have a bedroom. They still have no UC, HB or CTR non-dependant deduction: the deduction variables count only people outside the claimant's benefit unit.
• Under-16 children of a non-dependant's family are not counted in the household head's size criteria: the model counts only non-dependants and occupiers aged 16 or over. Read with UC Sch 4 para 9(2)(g) and HB reg 13D(12), the law appears to count them as well. Worth checking.
• Foster and looked-after children under 16 are still counted in bedrooms_for_children. The foster carer's additional bedroom is not modelled (UC Sch 4 para 12; HB reg 13D(3A)(b)).
• The model's LHA rates for 2024 onward come from its 2020 list of rents uprated by the private rent index, not from the published rates. For example, Maidstone B is £182.07 a week in the model against £172.60 published (VOA, April 2024), and C is £228.70 against £208.27.

Documentation

The variables' documentation fields describe the new rule. The Housing Benefit docs notebook does not cover the size criteria, so no docs build is needed.

axiom: uk/regulations/uksi/2013/376/schedule/4/paragraph/10 (UC Sch 4 para 10(1)(b), (c)) encoded-correct, companion cases in TheAxiomFoundation/rulespec-uk#408 | HB Regs 2006 / HB (SPC) Regs 2006 reg 13D(3)(b): TheAxiomFoundation/rulespec-uk#406 queued (addendum: TheAxiomFoundation/rulespec-uk#406 (comment))

🤖 Generated with Claude Code