Disregard WTC before the Housing Benefit childcare deduction by MaxGhenis · Pull Request #2776 · PolicyEngine/policyengine-uk

MaxGhenis · 2026-10-31T22:29:51Z

Also fixes #2777.

What and why

The model now disregards up to £17.10 a week of Working Tax Credit (WTC) in Housing Benefit before it deducts childcare. Under Schedule 5 paragraph 56 of the Housing Benefit Regulations 2006, the authority administering Housing Benefit disregards that amount for a family that meets a paragraph 17(2) work condition when its earnings can't fund the additional earnings disregard. The new housing_benefit_wtc_disregard takes the lesser of WTC and £889.20 a year (£1,929.20 in 2020-21). housing_benefit_childcare_earnings_limit lets childcare draw only on the WTC left after it, and housing_benefit_applicable_income deducts it alongside the earnings disregard. The existing gov.dwp.tax_credits.active parameter switches it off from model year 2025, when tax credits end.

In the issue's example, a lone parent earns £5,200 a year, has a £1,300 standard disregard, pays £4,160 for childcare and gets £100 of WTC. The model now disregards the £100 and deducts £3,900 for childcare, where main deducts £4,000.

The model also settles the WTC disregard in each accommodation segment of the calculation that #2234 added to housing_benefit_entitlement. Where a reform adds a defined_for condition to an amount the segments rebuild and keeps the model's formula, each segment now applies the condition as the annual output does (#2777).

The regulations

Under paragraph 56, the authority disregards up to £17.10 of WTC, "any amount of working tax credit up to £17.10", except "in a case which falls under sub-paragraph (1) of paragraph 17 of Schedule 4". Under paragraph 17(1), it raises the earnings disregard by £17.10 where "his net earnings equal or exceed the total" of the other earnings disregards, "the amount of child care charges calculated as deductible under regulation 27(1)(c)" and £17.10. Under regulation 40(2), it leaves out of other income "any sum, where applicable, specified in Schedule 5". Under regulation 27(1)(c), it then deducts childcare from earnings and, where earnings fall short and the family gets a tax credit, from "those earnings plus whichever credit" applies. In its BW2 guidance, DWP tells authorities to make that deduction "as the final step in the earnings calculation" (W2.131), to offset the rest against WTC or CTC "in addition to the existing disregards" (W2.181), and never to offset it against other income (W2.182).

The authority follows the same order under the pension-age and Northern Ireland regulations, and the model applies one calculation to all four:

Regulations	Additional earnings test	WTC disregard	Childcare deduction
SI 2006/213 (working age)	Sch 4 para 17	Sch 5 para 56	reg 27(1)(c)
SI 2006/214 (pension age)	Sch 4 para 9	Sch 5 para 21	reg 30(1)(c)
SR 2006/405 (NI working age)	Sch 5 para 17	Sch 6 para 58	reg 24(1)(c)
SR 2006/406 (NI pension age)	Sch 5 para 9	Sch 6 para 22	reg 28(1)(c)

Under SI 2020/371 reg 5 and SR 2020/53 reg 5, which the Secretary of State made, the authority read £37.10 for £17.10 in both disregards, in all four sets of regulations, from 6 April 2020 to the end of 4 April 2021. That period spans the 52 weeks from Monday 6 April 2020 to Sunday 4 April 2021. The model's fiscal year 2020 runs to 5 April 2021, and the model takes the £1,929.20 for those 52 weeks from the dated worker-disregard parameter it already has.

Where a claimant gets only the savings credit of Pension Credit, the authority may modify the Secretary of State's income figure under SI 2006/214 reg 27(4) "only in so far as necessary to take into account" the items in its list. The list includes "childcare charges taken into account under regulation 30(1)(c)" and the Schedule 4 paragraph 9(1) additional earnings disregard, but not the Schedule 5 paragraph 21 WTC disregard, and SR 2006/406 reg 25(4) has the same list. So on that route the model corrects the childcare charge for the disregarded WTC and deducts no WTC disregard from income.

Limit, deduction, WTC disregard

The paragraph 17 total includes the childcare deduction, and under regulation 27(2) that deduction draws on the WTC left after the paragraph 56 disregard, so each decision needs the other. The model settles them in one direction:

housing_benefit_childcare_earnings_limit keeps back the WTC that the authority would disregard under paragraph 56 if childcare took the whole capped charge. It reads neither the deduction nor the housing_benefit_wtc_disregard formula.
housing_benefit_applicable_income_childcare_element takes the lesser of the capped charge and that limit.
housing_benefit_wtc_disregard and the additional earnings disregard test paragraph 17 with the deduction in force in each segment.

For the deduction the model calculates, the WTC the limit keeps back equals the WTC disregard. Write E for the earnings the standard and accommodation disregards leave, C for the capped charge, a for the additional amount, R for the lesser of WTC and a, and T for CTC. If E ≥ C + a, earnings reach the paragraph 17(1) total with the whole charge and with any smaller deduction, and neither test disregards WTC. Otherwise the limit keeps back R, and the deduction comes to D = min(C, E + (WTC − R) + T). Either D = C, or D < C and D ≥ E; either way E < D + a, so earnings fall short of the paragraph 17(1) total and the model disregards R. Keeping back no WTC would only enlarge the deduction, and earnings would still fall short, so R gives the only consistent answer. test_limit_keeps_back_the_wtc_disregard_in_force checks the identity for every generated family, and the segment property checks it within each segment.

A reform can replace the deduction with a formula that reads the limit. With half the capped charge, capped at the limit, the issue's family gets a £2,080 deduction and the additional earnings disregard, and income of £1,030.80. Where the limit binds, the WTC it keeps back stays consistent: with £2,000 of earnings, a £260 standard disregard, a £9,100 charge and £1,000 of WTC, the deduction comes to £1,850.80 and the model disregards £889.20 of WTC.

Supplied and replaced values

WTC disregard. The limit reads the WTC disregard only where reading it reads no other variable. That covers an amount that a dataset, set_input or a scenario supplies, and a disregard that a reform neutralizes. It also covers one that a reform leaves without a formula, an adds or subtracts aggregate or a defined_for mask, which the engine (policyengine-core) uprates or carries over from its own earlier inputs. The model applies each of these to income, to the limit and in every segment, and neutralizing housing_benefit_wtc_disregard gives main's calculation back. The engine computes any other replacement through a formula, an aggregate (adds or subtracts, each a list or the path of a parameter that lists variables) or a mask it calculates first, and any of these could read the deduction. The model applies such a replacement to income and, as an annual amount, in every segment, but the limit keeps back what the whole-charge test gives, so a forwarding replacement leaves the limit's figure unchanged. When a reform only adds a defined_for mask and keeps the model's formula, the annual output takes the mask, and each 2026-27 segment applies the same condition to its own WTC disregard (see Reform conditions in each segment).
Childcare deduction and its limit. The model applies a deduction that a dataset, set_input or a reform supplies or replaces, or the lesser of the capped charge and a supplied or replaced limit, in every segment, and reads it in both paragraph 17 tests. Where the deduction itself comes from a dataset, set_input or a reform, only a replacement formula that calls the limit reads it. The limit still keeps back the WTC that the whole-charge test gives, so it can differ from main: with the issue's family and a supplied nil deduction, it comes to £3,900 where main gives £4,000, while the model disregards no WTC.
Earnings disregard. A supplied housing_benefit_applicable_income_disregard stays as supplied, and the model still makes the WTC decision's own paragraph 17 test from the earnings inputs, because a supplied total doesn't show whether it includes the additional amount.

Accommodation segments

From 5 October 2026, the paragraph 17(3)(a) total also includes the paragraph 18 specified or temporary accommodation disregard. The Secretary of State made that change by SI 2026/753 reg 2(4), and replaced that regulation by SI 2026/978 reg 2(2) before it came into force on the date in reg 1. The Department for Communities made the same change in Northern Ireland by SR 2026/157 reg 2(4)(a). Since #2234, the model settles paragraph 17 separately before and after that date and combines the awards by days. The authority's paragraph 56 decision depends on the paragraph 17 outcome, so the model now gives each segment its own WTC disregard and its own limit. hb_segment_deductions returns the segment's earnings disregard, WTC disregard and childcare deduction, housing_benefit_entitlement deducts all three in each segment, and the annual housing_benefit_wtc_disregard combines the segments' amounts by days.

Tax credits ended on 5 April 2025, before the accommodation disregard starts, so no current-law award changes. Take a reform that sets gov.dwp.tax_credits.active to true in 2026, and a single claimant aged 40 in specified accommodation. The claimant earns £500, has a £260 standard disregard, £2,000 of WTC and no childcare or other income, pays rent of £10,400 and has a nil applicable amount. Main gives £9,022.21, and the model now gives £9,600.19:

From 6 April to 4 October 2026 (182 days), earnings of £500 fall short of £260 plus £889.20, so the model disregards £889.20 of WTC. Income comes to £2,500 − £260 − £889.20 = £1,350.80, and the award to £10,400 − 0.65 × £1,350.80 = £9,521.98.
From 5 October 2026 to 5 April 2027 (183 days), the £4,041.96 accommodation amount takes the £240 of earnings left, and earnings again fall short. Income comes to £2,500 − £500 − £889.20 = £1,110.80, and the award to £9,677.98.
Combined, (182 × £9,521.98 + 183 × £9,677.98) / 365 = £9,600.19.

The model needs a WTC disregard in each segment. With £2,000 of earnings, the same claimant's earnings cover £260 plus £889.20 before 5 October, so the model gives the additional earnings disregard and no WTC disregard. After 5 October the £1,740 accommodation amount leaves earnings short, and the model disregards £889.20 of WTC. The award comes to £9,114.03. A WTC disregard settled once from the annual accommodation amount would give both disregards before 5 October, and an award of £9,402.23. With £4,000 of earnings it would give neither disregard after 5 October, and £8,176.03 where the model gives £8,465.81.

Deducting the day-weighted annual amount in every segment matches the per-segment award while both segments stay on the taper. Once income after 5 October falls below a £2,000 applicable amount, it gives £10,268.74 instead of £10,124.25. Reading the childcare limit with that annual amount gives £7,394.22 instead of £7,266.74 for a lone parent whose limit binds after 5 October.

Reform conditions in each segment

policyengine-core masks a variable that has a defined_for condition: it keeps the amount where the condition exceeds zero and gives the variable's default elsewhere. The segment helpers rebuild amounts without calling those variables, so hb_apply_defined_for applies the same condition to each rebuilt amount. It reads the condition from the variable the simulation holds, calculates it for the year and maps it to the variable's entity. In each 2026-27 segment, and in the single segment of every other year:

Amount	Where the model applies the condition
Accommodation disregard	`hb_accommodation_segments`, to each segment's full amount
Earnings disregard	`hb_segment_earnings_disregard`
WTC disregard	`hb_segment_wtc_disregard`
Childcare limit and deduction	`hb_segment_childcare_deduction`; `hb_fixed_childcare_deduction` applies the deduction's condition where a supplied or replaced limit sets it
Applicable income	`housing_benefit_entitlement`, to each segment's income

The model still uses a supplied, neutralized or replaced annual amount in every segment, whatever its condition gives. When the WTC or earnings disregard's condition fails, the limit still keeps back the WTC that the whole-charge test gives and still subtracts the standard and accommodation disregards. To let those amounts fund childcare, a reform must change the limit too.

Take a reform that adds defined_for = "is_couple" to housing_benefit_wtc_disregard and keeps its formula, in 2026 with tax credits kept. A lone parent aged 60 continues a Housing Benefit award with a child aged 8 in specified or temporary accommodation, paying £10,400 of London council rent. They work 16 hours a week for £11,000, draw a £10,000 private pension, pay £12,000 for qualifying childcare and report £2,000 of WTC and £4,000 of CTC. From those reports the model calculates WTC of £6,530.35 and CTC of £3,646.49. It attributes £883.14 of the £1,686 of income tax to the wages, so net earnings come to £10,116.86 and income before deductions to £11,000 + £10,000 − £1,686 + £10,176.84 = £29,490.84. The lone parent disregard gives £1,300 (£25 a week, means_test/income_disregard/lone_parent.yaml) and the childcare cap £9,100 (£175 a week, working_tax_credit/elements/childcare_1.yaml). From 5 October the accommodation disregard gives £4,041.96 (£77.73 a week, specified_or_temporary_accommodation/lone_parent/older.yaml). The worker amount comes to £889.20 (£17.10 a week, income_disregard/worker.yaml) and the applicable amount to £4,968.60 (£95.55 a week, allowances/lone_parent/older.yaml). Net earnings fall short of £11,289.20 before 5 October and £15,331.16 after, so the model disregards £889.20 of WTC in each segment, and the condition takes it away:

From 6 April to 4 October (182 days), income comes to £29,490.84 − £1,300 − £9,100 = £19,090.84, and the award to £10,400 − 0.65 × (£19,090.84 − £4,968.60) = £1,220.54.
From 5 October (183 days), income comes to £29,490.84 − £5,341.96 − £9,100 = £15,048.88, and the award to £3,847.82.
Combined, the award comes to £2,537.78, with no benefit cap reduction, and annual income to £17,064.32. Without the condition in each segment, the model paid £3,115.76, 0.65 × £889.20 = £577.98 more, and household net income rose by the same amount. A reform that neutralizes the WTC disregard gives the same award and household net income as the condition. In 2024, with one segment, the condition gives £1,113.09 where the model gives £1,691.07.

A reform that replaces the earnings disregard's formula and adds a condition that reads the WTC disregard, the childcare limit or, in 2026-27, the accommodation disregard makes policyengine-core raise CycleError. The engine calculates the condition first, and with the replacement in place the condition's amount reads the annual earnings disregard the replacement gives, so the condition reads its own result. For a claimant in specified accommodation we found five such cases: a condition that reads the accommodation disregard, the WTC disregard or the limit in 2026-27, and one that reads the WTC disregard or the limit in 2024. Without the replacement, the model calculates each of these conditions. A condition that reads its own variable or applicable income raises CycleError with or without a replacement.

Invariants

For every generated population, with assessed earnings, disregards, childcare, credits and work conditions as inputs:

The WTC disregard lies between zero and the lesser of WTC and the statutory amount. It falls to zero when the work condition fails, when earnings meet paragraph 17(1), and when tax credits stop.
No benefit unit gets both the additional earnings disregard and a positive WTC disregard within a segment. Annual amounts can both exceed zero: with £2,000 of earnings in specified accommodation, the year has £443.38 of additional earnings disregard and £445.82 of WTC disregard. Exclusivity holds when a dataset or set_input supplies the deduction or its limit, and example tests cover a neutralized deduction, one with no formula and replacement formulas, one of which reads the limit.
A deduction the model calculates never exceeds the capped charge, or the earnings and credits the disregards leave. A supplied deduction stays as supplied: £25,000 stays £25,000 against a £4,160 charge.
Where the model calculates the deduction, its limit and the WTC disregard, the WTC the limit keeps back equals the WTC disregard, for the year and within each segment.
Assessed income equals, and never falls below, an independent scalar calculation of the regulations. That calculation settles paragraph 17 by trying both outcomes.
In 2026-27 the award equals the day-weighted sum of the awards the same calculation gives in each segment, and the annual WTC disregard equals the day-weighted sum of the segments' amounts.
Where the model disregards no WTC in any segment, every output equals main's, except the limit where a dataset or set_input supplies the deduction. Under current law that covers every family.
In a year with one accommodation amount, supplying the deduction the model calculates changes no result. In 2026-27, where the accommodation disregard changes, the model applies the supplied annual amount in every segment: replaying the lone parent's calculated £4,160 moves the award from £7,266.74 to £7,394.22.
A reform that replaces the childcare deduction, its limit or the WTC disregard, each alone, with one that forwards the model's own amount meets no circular definition and changes no output. For the WTC disregard, that holds whether the replacement forwards through a formula or through adds or subtracts, each listing the forwarding variable or naming a parameter that lists it. It holds in 2024 for families with and without WTC, and in 2026-27 with tax credits kept for families outside specified or temporary accommodation. On the example populations, every 2024 output stays identical and every 2026-27 output agrees within £0.001, the float32 rounding of the day weights. The computed replacements of the WTC disregard leave all six outputs of the eight test families unchanged in both years, and the generated populations allow 2p. In specified accommodation the model applies the replacement's annual amount in every segment, as for any replacement: a forwarding replacement of the WTC disregard, through a formula or an aggregate, gives £10,268.74 where the model gives £10,124.25.
The override check that segments use (hb_uses_annual_override) gives main's answer for all 1,200 combinations of the states the engine reads: neutralization, a default or replacement formula, adds and subtracts (none, empty, a list or a parameter path), a defined_for mask, uprating and a supplied input. Where only a defined_for mask joins the model's own formula, the check keeps the segment route, and each segment helper applies the condition itself.

Under a reform that gives each of the accommodation, earnings and WTC disregards, the childcare deduction, its limit and applicable income its own condition, which reads none of these amounts, for every generated population with or without tax credits:

Each annual disregard equals the day-weighted sum of its segment amounts.
Where a condition fails, its amount takes the variable's default in the annual output and in every segment, unless the population supplies the amount. Where applicable income's condition fails, the award equals the maximum benefit less 65% of the excess of the variable's default over the applicable amount.
A benefit unit whose conditions all hold keeps the model's outputs, award included.
The award agrees with an independent scalar calculation within each segment, combined by days.
In each year but 2026-27, the award the single segment gives equals the award the model builds from annual income.

Test plan

For policy changes: the 13 cases in housing_benefit_wtc_disregard.yaml and the Python examples take their expected values from the regulations, and their comments work through the arithmetic by hand.
make format: Ruff 0.17.2 formats and checks every changed Python file, ruff check --select F811,F821 policyengine_uk passes, and git diff --check passes.
Changelog fragments changelog.d/2730.fixed.md and changelog.d/2777.fixed.md

The tests build their own households and never load survey data. Rows that name 2f087ae7c ran at this branch's head, which merges main at 52b30809f; the other rows name the commit they ran at, and CI runs the rest of the suite.

Check	Result
`test_housing_benefit_wtc_disregard.py`: examples, Hypothesis properties, differential tests and replacement tests	55 passed at `2f087ae7c`
`test_housing_benefit_wtc_disregard_segments.py`: segment examples and properties	18 passed at `2f087ae7c`
`test_housing_benefit_segment_masks.py`: a reform condition on each amount in each segment, #2777's reform, and the couples-only WTC disregard for a lone parent in 2026-27 and 2024	30 passed at `2f087ae7c`
`housing_benefit_wtc_disregard.yaml`	13 passed at `2f087ae7c`
`test_housing_benefit_accommodation_dates_properties.py`	6 passed at `2f087ae7c`
`test_housing_benefit_childcare_properties.py`, which reads the limit	5 passed at `2f087ae7c`
Housing Benefit YAML directory	597 passed at `2f087ae7c`
`policyengine_uk/tests/code_health`	3,092 passed at `2f087ae7c`
`test_housing_benefit_accommodation_dates.py`, with a 3,600 s cap	38 passed in 2,945 s at `6dc47664d`
The two original test modules and the WTC YAML file on main (`2da212abc`)	55, 18 and 13 failed; the reproducer deducts £4,000 for childcare against £3,900
The couples-only lone parent at `2f087ae7c` with the five model files back at `1d4a3e475`	2026-27 fails: the award comes to £3,115.76 against £2,537.78. 2024 passes on the award and fails only where it reads the segment's WTC disregard, £889.20 against nil

The differential ran main (2da212abc) and this branch at fdc13f128 over the same 15,552 family-years: 1,944 families in each of 2018, 2020, 2021, 2024, 2025, 2026 and 2027, and in 2026 with tax credits kept. Since fdc13f128 the model code changes which replacements of the WTC disregard the limit reads, and applies reform conditions in each segment; none of these populations replaces the WTC disregard or adds a condition. Of each 1,944, 1,296 families have a calculated deduction, 324 a supplied nil deduction and 324 a supplied £1,000 limit. Earnings, WTC, childcare, the standard disregard, the work condition, pension age and specified accommodation vary; other income comes to £1,000, rent to £10,400 and the applicable amount to £2,000, and regions alternate between London and Northern Ireland. It compares the childcare deduction, earnings disregard, childcare limit, applicable income and award.

Family-years	Where the model disregards WTC in no segment	Where it disregards WTC in a segment
Calculated deduction	8,962: all 44,810 values equal main's exactly	1,406: every one moves by more than 1p
Supplied £1,000 limit	2,328: all 11,640 values equal main's exactly	264: every one moves
Supplied nil deduction	2,340: 11,578 of 11,700 values equal main's; the other 122 come from the limit, which keeps back the WTC the whole-charge test gives	252: every one moves

None of the 5,832 family-years in 2025, 2026 and 2027 under current law changes.

Of 44 mutations across the changed model files for #2730, 42 make at least one test fail. The other two add a floor or a cap to the paragraph 17 test, which can't change its answer while the additional amount stays positive. Each mutation runs the main module, the segment module and the WTC YAML file in full, with Hypothesis shrinking off. The 15 that touch the override helpers or their call sites ran in full at 6dc47664d. The other 29 ran at fdc13f128, or for the main module and YAML file at 9c2adcde9, whose model code and tests match fdc13f128. Since then the model code changes which replacements of the WTC disregard the limit reads, and applies reform conditions in each segment; the test modules keep every earlier test. For #2777, all 10 mutations of the condition helper and its call sites make a test fail; each ran the main module, the segment module and the segment-mask module in full at 2f087ae7c, with Hypothesis seed 0.

Mutation results

File	Mutation	Result
`housing_benefit_childcare_earnings_limit.py`	The shared paragraph 17 test leaves out the accommodation disregard	Caught: main module 3 failed; segment module 9 failed. (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	`>=` becomes `>` in the paragraph 17 test	Caught: main module 4 failed; WTC YAML 1 failed. (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	The paragraph 17 test leaves out the standard disregard	Caught: main module 13 failed; segment module 1 failed; WTC YAML 4 failed. (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	The paragraph 17 test floors net earnings at zero	Survives, as intended: with a positive additional amount it can't change the answer (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	The paragraph 17 test caps the accommodation amount at the earnings left	Survives, as intended: with a positive additional amount it can't change the answer (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	The WTC decision leaves the accommodation disregard out of the paragraph 17 test	Caught: main module 3 failed; segment module 7 failed. (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	The disregard applies when earnings meet paragraph 17(1)	Caught: main module 18 failed; segment module 8 failed; WTC YAML 3 failed. (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	The disregard ignores the work condition	Caught: main module 2 failed; segment module 2 failed; WTC YAML 1 failed. (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	The disregard takes all WTC, uncapped	Caught: main module 9 failed; segment module 7 failed; WTC YAML 4 failed. (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	A negative WTC award gives a negative disregard	Caught: main module 1 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_childcare_earnings_limit.py`	The disregard ignores the end of tax credits	Caught: main module 4 failed; segment module 2 failed; WTC YAML 1 failed. (at `fdc13f128`)
`housing_benefit_childcare_earnings_limit.py`	The cap uses the weekly amount	Caught: main module 17 failed; segment module 8 failed; WTC YAML 7 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_childcare_earnings_limit.py`	The limit reads the WTC disregard variable again	Caught: main module 32 failed; segment module 10 failed; WTC YAML 10 failed. The forwarding-reform tests fail. The baseline itself meets a circular definition. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_childcare_earnings_limit.py`	The WTC kept back tests a supplied or replaced deduction in place of the capped charge	Caught: main module 5 failed; segment module 1 failed. (at `6dc47664d`)
`housing_benefit_childcare_earnings_limit.py`	The WTC kept back tests the deduction variable in place of the capped charge	Caught: main module 32 failed; segment module 10 failed; WTC YAML 10 failed. The forwarding-reform tests fail. The baseline itself meets a circular definition. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_childcare_earnings_limit.py`	The limit ignores a supplied, neutralized or input WTC disregard	Caught: main module 3 failed; segment module 1 failed; WTC YAML 1 failed. (at `6dc47664d`)
`housing_benefit_childcare_earnings_limit.py`	The limit reads a replacement WTC disregard formula	Caught: main module 4 failed; segment module 3 failed. (at `6dc47664d`)
`housing_benefit_childcare_earnings_limit.py`	A segment's limit keeps back the annual amount, not its own	Caught: segment module 2 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_childcare_earnings_limit.py`	The annual limit adds the segments' WTC without day weights	Caught: segment module 1 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_childcare_earnings_limit.py`	Childcare draws on the disregarded WTC	Caught: main module 10 failed; segment module 3 failed; WTC YAML 4 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_wtc_disregard.py`	The override check ignores a supplied or replaced limit	Caught: main module 3 failed; segment module 2 failed; WTC YAML 1 failed. (at `6dc47664d`)
`housing_benefit_wtc_disregard.py`	The override check ignores a supplied or replaced deduction	Caught: main module 9 failed; segment module 2 failed; WTC YAML 1 failed. (at `6dc47664d`)
`housing_benefit_wtc_disregard.py`	A supplied deduction reads the capped charge instead	Caught: main module 10 failed; segment module 2 failed; WTC YAML 1 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_wtc_disregard.py`	A supplied limit sets the deduction without the charge cap	Caught: main module 3 failed; segment module 2 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_wtc_disregard.py`	Each segment's deduction reads the annual limit	Caught: segment module 3 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_wtc_disregard.py`	Segments discard a supplied annual WTC disregard	Caught: main module 5 failed; segment module 9 failed. (at `6dc47664d`)
`housing_benefit_wtc_disregard.py`	Segments ignore a supplied, neutralized or replaced annual WTC disregard	Caught: segment module 9 failed. (at `6dc47664d`)
`housing_benefit_wtc_disregard.py`	The annual WTC disregard tests the capped charge, not the deduction in force	Caught: main module 12 failed; segment module 2 failed; WTC YAML 2 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_wtc_disregard.py`	The annual amount adds segments without day weights	Caught: segment module 7 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_wtc_disregard.py`	The annual amount comes from the annual accommodation amount	Caught: segment module 5 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_applicable_income_disregard.py`	A segment's WTC disregard tests the capped charge, not the deduction in force	Caught: segment module 2 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_applicable_income_disregard.py`	Segments settle the WTC disregard from the annual accommodation amount	Caught: segment module 6 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_applicable_income_disregard.py`	The earnings disregard tests the capped charge, not the deduction	Caught: main module 9 failed; segment module 1 failed; WTC YAML 2 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_applicable_income.py`	Annual income leaves out the WTC disregard	Caught: main module 9 failed; segment module 1 failed; WTC YAML 2 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_entitlement.py`	Segment income leaves out the WTC disregard	Caught: segment module 9 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_entitlement.py`	Each segment deducts the day-weighted annual WTC disregard	Caught: segment module 2 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	The override helper runs the formula of a neutralized variable	Caught: main module 2 failed. (at `6dc47664d`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	The helper lets the limit read an adds or subtracts aggregate	Caught: main module 8 failed; segment module 4 failed. (at `6dc47664d`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	The helper lets the limit read an adds aggregate	Caught: main module 4 failed; segment module 2 failed. (at `6dc47664d`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	The helper lets the limit read a subtracts aggregate	Caught: main module 4 failed; segment module 2 failed. (at `6dc47664d`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	The helper lets the limit read an aggregate named by a parameter path	Caught: main module 4 failed; segment module 2 failed. (at `6dc47664d`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	The helper lets the limit read a variable with a defined_for mask	Caught: main module 1 failed. (at `6dc47664d`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	The helper keeps an uprated input from the limit	Caught: main module 1 failed. (at `6dc47664d`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	Segments recalculate an aggregate replacement instead of applying its annual amount	Caught: segment module 4 failed. (at `6dc47664d`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	`hb_apply_defined_for` ignores every condition	Caught: segment-mask module 17 failed; main and segment modules pass. (#2777, at `2f087ae7c`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	A failed condition gives nil instead of the variable's default	Caught: segment-mask module 1 failed; main and segment modules pass. (#2777, at `2f087ae7c`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	The condition holds wherever it differs from zero, not where it exceeds zero	Caught: segment-mask module 1 failed; main and segment modules pass. (#2777, at `2f087ae7c`)
`housing_benefit_specified_or_temporary_accommodation_disregard.py`	Segments leave the accommodation disregard's condition out	Caught: segment-mask module 5 failed; main and segment modules pass. (#2777, at `2f087ae7c`)
`housing_benefit_applicable_income_disregard.py`	Segments leave the earnings disregard's condition out	Caught: segment-mask module 7 failed; main and segment modules pass. (#2777, at `2f087ae7c`)
`housing_benefit_wtc_disregard.py`	A supplied or replaced limit sets the deduction without the deduction's condition	Caught: segment-mask module 2 failed; main and segment modules pass. (#2777, at `2f087ae7c`)
`housing_benefit_wtc_disregard.py`	Segments leave the limit's condition out	Caught: segment-mask module 3 failed; main and segment modules pass. (#2777, at `2f087ae7c`)
`housing_benefit_wtc_disregard.py`	Segments leave the childcare deduction's condition out	Caught: segment-mask module 3 failed; main and segment modules pass. (#2777, at `2f087ae7c`)
`housing_benefit_wtc_disregard.py`	Segments leave the WTC disregard's condition out	Caught: segment-mask module 9 failed; main and segment modules pass. (#2777, at `2f087ae7c`)
`housing_benefit_entitlement.py`	Segment income leaves applicable income's condition out	Caught: segment-mask module 4 failed; main and segment modules pass. (#2777, at `2f087ae7c`)

No test on main changes its expected value. #2719, #2700, #2727, #2731, #2735 and #2234 have merged, and this branch includes them. At 2f087ae7c, git merge-tree --write-tree --name-only against each open PR that edits this branch's files gives:

Treat income derived from capital as capital in the legacy means tests #1995 (7381a46c7) edits housing_benefit_applicable_income.py, and Git reports a conflict there: both PRs extend the last sentence of the variable's documentation string. Git reports the same conflict against 1d4a3e475. Whoever merges second keeps both sentences. Treat income derived from capital as capital in the legacy means tests #1995 also changes the income tax that Housing Benefit deducts, so once it lands the lone parent's figures above will change.
Value means-tested capital less 10% sale expenses and secured debt #1969, Fix HB and national CTR non-dependant deductions #2007, Give benefit-unit members aged 16 or over a bedroom under LHA rules #2010, Carry unused Income Tax trading losses forward once #2661, Bar income-related ESA at pensionable age and on Pension Credit #2807 and Count UC carer awards for severe disability and Scottish CTR premiums #2827 edit policyengine_uk/utils/supplied_inputs.py, and Git merges this branch with each of them without a conflict.
Give UK simulations a working subsample #2265 and Count only supplied inputs that are still stored in their tier #2308 edit policyengine_uk/utils/supplied_inputs.py and conflict with main itself; against this branch Git reports the same files as against main, so this branch adds no conflict.

No other open PR edits the override helpers' file, the limit's file, the three test modules or the WTC YAML file. #2301 and #2302 no longer edit housing_benefit_applicable_income.py.

Reproducer

Reproducer: policyengine_uk/tests/test_housing_benefit_wtc_disregard.py::test_issue_2730_reserves_wtc_before_childcare, policyengine_uk/tests/test_housing_benefit_wtc_disregard_segments.py::test_wtc_disregard_applies_before_and_after_5_october, policyengine_uk/tests/test_housing_benefit_segment_masks.py::test_issue_2777_reform_masks_each_segment, policyengine_uk/tests/test_housing_benefit_segment_masks.py::test_a_couples_only_wtc_disregard_reaches_a_lone_parents_award

axiom: TheAxiomFoundation/rulespec-uk#510 queued

🤖 Generated with Claude Code

Apply the Schedule 5 WTC disregard separately from earnings disregards and reserve it before the childcare deduction. Add statutory examples, properties, differential checks and historical cutoff coverage. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The WTC disregard tested the full capped childcare charge, while the additional earnings disregard reads the childcare deduction. A supplied deduction, a supplied limit or a reform that replaces either could then give a benefit unit both disregards. The WTC disregard now reads that deduction whenever it differs from the model's own calculation, and keeps the cycle-free full-charge test otherwise. Tests add the accommodation disregard and supplied deductions to the oracle, which now settles paragraph 17 by trying both outcomes. They add a property that the two disregards never both apply and a differential test that supplying the calculated deduction changes nothing. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The WTC disregard compares the same net earnings and accommodation amount as the additional earnings disregard. A cap on the accommodation amount would not change the test, since earnings already fall short where it binds. New tests cover a deduction with no formula and a negative WTC award. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Brings in #2234's accommodation segments. The only conflict sat in the supplied-input registry, where both sides add the two childcare names; main's side lists them with #2234's three new names, so the merge keeps it. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

#2234 settles Schedule 4 paragraph 17 separately before and after the accommodation disregard starts on 5 October 2026, but its segment income left out the WTC disregard. Paragraph 56 follows paragraph 17's outcome, so each segment now has its own WTC disregard: the childcare limit takes it, hb_segment_deductions returns it, and entitlement deducts it. The annual variable combines the segments by days. The WTC and earnings decisions now share one paragraph 17 test and one check for a supplied or replaced childcare deduction, #2234's hb_uses_annual_override. Under a reform that keeps tax credits in 2026-27, a claimant in specified accommodation with £500 of earnings and £2,000 of WTC now gets £9,600.19 rather than £9,022.21. Current-law awards do not change: tax credits end before the accommodation disregard starts. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The WTC disregard is exactly nil where paragraph 56 does not apply, so the checks now count any positive amount rather than one above 2p. A pinned example gives earnings of £1,149.20, which cover the £260 standard disregard, a supplied nil deduction and the £889.20 additional amount exactly, with a 1p childcare charge and 1p of WTC. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Float32 stores £1,149.20 as £1,149.19995, so a plain float comparison would find that earnings fall short of the £260 standard disregard plus £889.20 and grant the WTC disregard instead of the additional earnings disregard. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The populations now supply the earnings and tax-credit limit as well as the deduction, and the statutory oracle caps the deduction at the charge under a supplied limit. A pinned case gives a £10,000 limit against a £4,160 charge: the deduction stays at £4,160, the additional earnings disregard applies, and paragraph 56 does not. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

After 5 October a lone parent with £8,000 of earnings, £3,000 of other income, £2,000 of WTC and a £4,160 charge has paragraph 56 disregard £889.20, so childcare can draw on only £3,768.84. The award comes to £7,266.74; a limit read with the day-weighted annual WTC disregard would give £7,394.21. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The earnings and tax-credit limit now keeps back the WTC that Sch 5 para 56 disregards by testing Sch 4 para 17 with the whole capped childcare charge, so it reads neither the childcare deduction nor the housing_benefit_wtc_disregard formula. The WTC disregard and the earnings disregard then test para 17 with the deduction in force in each segment. For the model's own deduction the two tests agree, so the limit, the deduction and the disregard keep their values, and the direction runs one way: limit, deduction, WTC disregard. A reform that replaces the deduction, its limit or the WTC disregard with a formula that forwards to the model's own no longer meets a circular definition, and a replacement deduction can read the limit. A WTC disregard that an input supplies, or that a reform neutralizes, still reaches the limit and every segment. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

A supplied childcare deduction keeps its value above the charge and the limit, the annual additional earnings and WTC disregards can both be positive though no segment has both, and supplying the calculated annual deduction in a segmented year applies it in every segment. The test modules now build their forwarding reforms on demand, so they import where housing_benefit_wtc_disregard does not exist. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

In a year with two accommodation segments, the annual earnings and tax-credit limit keeps back the day-weighted WTC disregard, as each segment's limit keeps back its own. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The limit read a WTC disregard with no formula as a supplied amount, but core still computes adds and subtracts there, and calculates a defined_for mask before anything else, so an aggregate identity replacement cycled. The limit now reads the WTC disregard only where reading it reads no other variable: a neutralized variable, one with no formula, aggregate or mask (which core uprates, carries over or defaults from its own inputs), or a supplied input. hb_uses_annual_override still treats any variable without a formula as an annual replacement in every segment. Tests cover formula, adds and subtracts replacements, each as a list or a parameter path, in 2024 and in 2026-27 with tax credits kept, and an uprated input replacement with and without a mask. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Each reform system holds about 38 MB, and caching all eleven kept the main test module above the 2 GB cap the mutation runs use. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Each computed replacement now runs one simulation against outputs the model computes once for 2024 and once for 2026-27. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

github-actions · 2026-10-31T23:38:46Z

Aggregate impact on certified data: pass

Base 52b30809ff against head b61ab7a3fc, both on enhanced_frs_2024_25.h5 (policyengine-uk-data 1.58.1, sha256 056e80fd8e91), fiscal years 2025-26 to 2031-32. Each year is a full microsimulation run on each side.

Checks: 0 broken by this change, 1 failing on both sides, 0 fixed. Moves flagged for review: 0 (thresholds in report_spec.yaml; a flag does not fail the job).

Workflow run and full report

Notes

In the years a figure from fewer than 10 records is suppressed (spending.disabled_students_allowance), every total containing it, and the decile incomes, are rebuilt from their other components, so no published total includes it. Those totals are marked †. A change is withheld where the two sides leave out different figures and can't be lined up.

Moves flagged for review

None.

Spending by benefit and service

Measure	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Disabled Students' Allowance	suppressed	suppressed	suppressed	suppressed	suppressed	suppressed	suppressed

Caseloads

Measure	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Disabled Students' Allowance	suppressed	suppressed	suppressed	suppressed	suppressed	suppressed	suppressed

Standard reform costings (change in gov_balance)

Measure	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Raise Class 1 main employee NICs rate to 10%	0	0	0	0	0	suppressed	suppressed
Raise additional rate by 3pp	0	0	0	0	0	suppressed	suppressed

The report suppresses a reform's costing and target change in these years, for the reasons below (privacy.py):

Raise Class 1 main employee NICs rate to 10%, 2030-31 to 2031-32 (base and head): the reform changes income_tax in fewer than 10 records.
Raise additional rate by 3pp, 2030-31 to 2031-32 (base and head): the reform changes universal_credit in fewer than 10 records.

Standard reforms: change in each reform's target

Measure	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Raise Class 1 main employee NICs rate to 10%: change in national_insurance	0	0	0	0	0	suppressed	suppressed
Raise additional rate by 3pp: change in income_tax	0	0	0	0	0	suppressed	suppressed

Household net income by decile

No household's net income changes in any year.

No change in any year: receipts by tax head; government balance, each component counted once; model totals; household income totals; population; poverty rates; people in poverty; inequality; mean household net income in each side's own deciles.

Checks failing on both base and head (1)

VAT excluding private school fees is never negative (sign:vat): 2025-26: 76 units below zero, a weighted £-53,888,684 in total; 2026-27: 76 units below zero, a weighted £-55,777,260 in total; 2027-28: 76 units below zero, a weighted £-57,727,917 in total

Counted-once check

Each of 92 components was raised by £1 per unit in 2026-27, and every head of its totals was read household by household: its own head must rise by exactly £1, no other tax head may move, no other benefit may rise, and the model's own total must move by what its heads moved. A head that breaks a rule is measured again twice: with every other head held at its baseline value, and with every other head held at its value after the first £1 while the component rises again. It fails where either still shows the response. 0 failed; 0 were skipped.

widowed_parents_allowance: pension_credit responds only through other heads in fewer than 10 households (other heads that moved there: carers_allowance; gov_spending; not a failure)
widowed_parents_allowance: pension_credit responds only through other heads in fewer than 10 households (other heads that moved there: carers_allowance; household_benefits; not a failure)

Provenance

Side	Commit	Runtime	Environment hash	Partition hash	Time
base	main `52b30809ff`	uk 2.172.0 (files `f8400aea019e`), core 3.34.1, py 3.13.18 aarch64	`18199314be15`	`7cb61324a46b`	2062.59s
head	#2776 merged into main `b61ab7a3fc`	uk 2.172.0 (files `27900fb6fb5f`), core 3.34.1, py 3.13.18 aarch64	`18199314be15`	`7cb61324a46b`	2083.48s

Dataset: hf://policyengine/policyengine-uk-data-private/enhanced_frs_2024_25.h5@2526edbbb687a3b54dbb11272e5389e4b62d2ac2, sha256 056e80fd8e9151d78f432a25cde863340baee81304c6216a37fb575ce8563784 (verified on both sides), policyengine-uk-data 1.58.1.

Receipts by tax head: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Final charge on cash interest in non-cash ISAs	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Lifetime ISA withdrawal charge	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Income tax	£294.46bn	£315.23bn	£335.82bn	£351.28bn	£366.96bn	£385.36bn	£400.00bn
National Insurance (employee and self-employed)	£51.70bn	£54.14bn	£56.13bn	£57.73bn	£59.76bn	£61.64bn	£63.51bn
National Insurance (employer)	£148.71bn	£155.37bn	£160.88bn	£165.09bn	£172.06bn	£177.36bn	£182.57bn
Employer NI change borne by capital	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Employer NI change borne by consumers	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Capital gains tax	£12.60bn	£13.15bn	£13.60bn	£14.11bn	£14.67bn	£15.24bn	£15.80bn
Student loan repayments	£7.82bn	£8.13bn	£8.39bn	£8.58bn	£8.79bn	£9.04bn	£9.29bn
VAT excluding private school fees	£249.37bn	£256.87bn	£266.33bn	£272.70bn	£278.97bn	£285.24bn	£291.59bn
Change in VAT from its baseline	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
VAT on private school fees	£1.95bn	£2.01bn	£2.09bn	£2.14bn	£2.20bn	£2.26bn	£2.31bn
Change in private school VAT from its baseline	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Fuel duties	£24.71bn	£24.13bn	£26.00bn	£26.83bn	£26.51bn	£25.78bn	£24.63bn
Alcohol duties	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Tobacco duties	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Vehicle excise duties	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Betting and gaming duties	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
TV licence fee	£4.26bn	£4.40bn	£4.41bn	£4.42bn	£4.43bn	£4.44bn	£4.45bn
Council tax	£49.21bn	£52.16bn	£54.89bn	£57.95bn	£61.16bn	£64.47bn	£67.91bn
High value council tax surcharge	£0.00bn	£0.00bn	£0.00bn	£0.48bn	£0.49bn	£0.50bn	£0.51bn
Northern Ireland domestic rates	£0.55bn	£0.58bn	£0.60bn	£0.63bn	£0.66bn	£0.69bn	£0.72bn
Business rates	£34.68bn	£36.20bn	£38.00bn	£39.46bn	£42.27bn	£43.25bn	£44.29bn
Stamp duty land tax on the household's own purchases	£6.88bn	£7.32bn	£7.66bn	£8.06bn	£8.47bn	£8.91bn	£9.35bn
Stamp duty land tax paid by companies	£4.72bn	£4.57bn	£5.29bn	£6.03bn	£6.37bn	£6.92bn	£7.45bn
Land transaction tax (Wales)	£0.07bn	£0.08bn	£0.08bn	£0.09bn	£0.10bn	£0.10bn	£0.11bn
Land and buildings transaction tax (Scotland)	£0.31bn	£0.32bn	£0.33bn	£0.35bn	£0.36bn	£0.38bn	£0.40bn
Unspecified tax change borne by capital	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Unspecified tax change borne by consumers	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Wealth tax	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Wealth tax excluding the main residence	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Land value tax	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Carbon tax	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn

Spending by benefit and service: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Lifetime ISA bonus	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
State Pension	£121.68bn	£125.92bn	£126.72bn	£129.49bn	£132.99bn	£136.66bn	£140.42bn
Pension Credit	£6.66bn	£7.03bn	£7.29bn	£7.48bn	£7.60bn	£7.84bn	£8.03bn
Winter Fuel Payment	£1.86bn	£1.85bn	£1.79bn	£1.77bn	£1.78bn	£1.78bn	£1.78bn
Pension Age Winter Heating Payment (Scotland)	£0.19bn	£0.20bn	£0.20bn	£0.21bn	£0.21bn	£0.22bn	£0.22bn
Universal Credit	£75.48bn	£80.52bn	£82.95bn	£84.96bn	£86.11bn	£87.61bn	£88.21bn
Housing Benefit	£8.56bn	£8.89bn	£9.23bn	£9.46bn	£9.69bn	£9.99bn	£10.27bn
Council Tax Reduction	£2.07bn	£2.20bn	£2.31bn	£2.45bn	£2.58bn	£2.72bn	£2.86bn
Employment and Support Allowance (income-related)	£0.79bn	£0.82bn	£0.84bn	£0.86bn	£0.88bn	£0.90bn	£0.92bn
Employment and Support Allowance (contribution-based)	£5.20bn	£5.37bn	£5.56bn	£5.69bn	£5.82bn	£5.95bn	£6.09bn
Jobseeker's Allowance (income-based)	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Jobseeker's Allowance (contribution-based)	£0.29bn	£0.30bn	£0.31bn	£0.31bn	£0.32bn	£0.33bn	£0.34bn
Income Support	£0.11bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Working Tax Credit	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Child Tax Credit	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Incapacity Benefit	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Personal Independence Payment	£26.95bn	£28.03bn	£29.05bn	£30.09bn	£30.75bn	£31.51bn	£32.21bn
Disability Living Allowance	£7.49bn	£7.79bn	£8.07bn	£8.36bn	£8.55bn	£8.76bn	£8.95bn
Attendance Allowance	£8.68bn	£9.03bn	£9.36bn	£9.69bn	£9.90bn	£10.15bn	£10.38bn
Carer's Allowance	£4.48bn	£4.66bn	£4.83bn	£5.00bn	£5.11bn	£5.27bn	£5.42bn
Carer Support Payment (Scotland)	£0.31bn	£0.32bn	£0.33bn	£0.34bn	£0.35bn	£0.36bn	£0.37bn
Carer's Allowance Supplement (Scotland)	£0.00bn	£0.04bn	£0.04bn	£0.05bn	£0.05bn	£0.05bn	£0.05bn
Severe Disablement Allowance	£0.03bn	£0.03bn	£0.03bn	£0.03bn	£0.03bn	£0.03bn	£0.03bn
Industrial Injuries Disablement Benefit	£0.50bn	£0.52bn	£0.54bn	£0.55bn	£0.56bn	£0.58bn	£0.59bn
Armed Forces Compensation Scheme	£0.61bn	£0.63bn	£0.65bn	£0.66bn	£0.68bn	£0.69bn	£0.71bn
War Widow's or Widower's Pension	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Child Benefit	£16.19bn	£16.82bn	£17.39bn	£18.00bn	£18.39bn	£18.84bn	£19.25bn
Scottish Child Payment	£0.49bn	£0.52bn	£0.53bn	£0.55bn	£0.56bn	£0.57bn	£0.56bn
Two Child Limit Payment (Scotland, counterfactual)	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Healthy Start vouchers	£0.05bn	£0.05bn	£0.05bn	£0.05bn	£0.05bn	£0.05bn	£0.05bn
Maternity Allowance	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Statutory Maternity Pay	£3.30bn	£3.41bn	£3.54bn	£3.62bn	£3.70bn	£3.78bn	£3.87bn
Statutory Paternity Pay	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Statutory Sick Pay	£1.12bn	£1.16bn	£1.20bn	£1.23bn	£1.26bn	£1.29bn	£1.32bn
Sure Start Maternity Grant	£0.01bn	£0.01bn	£0.01bn	£0.01bn	£0.01bn	£0.01bn	£0.01bn
Bereavement Support Payment	£0.67bn	£0.69bn	£0.71bn	£0.73bn	£0.75bn	£0.76bn	£0.78bn
Widowed Parent's Allowance	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Tax-Free Childcare	£0.56bn	£0.57bn	£0.57bn	£0.58bn	£0.60bn	£0.61bn	£0.59bn
Extended childcare entitlement	£2.99bn	£3.50bn	£3.59bn	£3.71bn	£3.78bn	£3.86bn	£3.93bn
Universal childcare entitlement	£2.06bn	£2.22bn	£2.29bn	£2.37bn	£2.43bn	£2.48bn	£2.53bn
Targeted childcare entitlement	£0.50bn	£0.51bn	£0.52bn	£0.54bn	£0.55bn	£0.57bn	£0.56bn
Childcare Offer for Wales	£0.26bn	£0.28bn	£0.28bn	£0.29bn	£0.30bn	£0.31bn	£0.32bn
Care to Learn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Childcare Grant	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Parents' Learning Allowance	£0.20bn	£0.20bn	£0.20bn	£0.20bn	£0.20bn	£0.20bn	£0.20bn
Adult Dependants' Grant	£0.06bn	£0.06bn	£0.04bn	£0.04bn	£0.04bn	£0.04bn	£0.04bn
Travel Grant	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Disabled Students' Allowance	suppressed	suppressed	suppressed	suppressed	suppressed	suppressed	suppressed
16 to 19 Bursary Fund	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
NHS spending	£203.89bn	£204.30bn	£204.56bn	£204.91bn	£205.34bn	£205.85bn	£206.35bn
State education spending	£0.02bn	£0.02bn	£0.02bn	£0.02bn	£0.02bn	£0.02bn	£0.02bn
Transport subsidy spending	£25.05bn	£25.10bn	£25.14bn	£25.18bn	£25.23bn	£25.30bn	£25.36bn
Cost-of-living support payment	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Energy Bills Rebate	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Household Energy Payment	£0.00bn	£2.11bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Energy Price Guarantee subsidy	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Basic income	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Disability basic income	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn
Other public spending budget change	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn

Caseloads: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Lifetime ISA bonus	0k	0k	0k	0k	0k	0k	0k
State Pension	11,264k	11,145k	10,819k	10,700k	10,723k	10,750k	10,776k
Pension Credit	1,486k	1,480k	1,472k	1,452k	1,450k	1,455k	1,464k
Winter Fuel Payment	7,844k	7,783k	7,473k	7,375k	7,391k	7,409k	7,427k
Pension Age Winter Heating Payment (Scotland)	803k	802k	800k	797k	799k	801k	802k
Universal Credit	6,342k	6,347k	6,336k	6,297k	6,292k	6,271k	6,223k
Housing Benefit	1,249k	1,245k	1,244k	1,227k	1,227k	1,229k	1,232k
Council Tax Reduction	3,176k	3,195k	3,200k	3,214k	3,224k	3,235k	3,242k
Employment and Support Allowance (income-related)	71k	71k	71k	71k	71k	71k	71k
Employment and Support Allowance (contribution-based)	719k	720k	721k	722k	724k	726k	728k
Jobseeker's Allowance (income-based)	0k	0k	0k	0k	0k	0k	0k
Jobseeker's Allowance (contribution-based)	67k	67k	67k	67k	67k	68k	68k
Income Support	25k	0k	0k	0k	0k	0k	0k
Working Tax Credit	0k	0k	0k	0k	0k	0k	0k
Child Tax Credit	0k	0k	0k	0k	0k	0k	0k
Incapacity Benefit	0k	0k	0k	0k	0k	0k	0k
Personal Independence Payment	3,567k	3,575k	3,579k	3,585k	3,593k	3,602k	3,611k
Disability Living Allowance	1,291k	1,294k	1,296k	1,298k	1,300k	1,304k	1,307k
Attendance Allowance	1,695k	1,698k	1,700k	1,703k	1,707k	1,711k	1,715k
Carer's Allowance	1,034k	1,037k	1,038k	1,040k	1,042k	1,060k	1,062k
Carer Support Payment (Scotland)	71k	71k	71k	72k	72k	72k	72k
Carer's Allowance Supplement (Scotland)	0k	71k	71k	72k	72k	72k	72k
Severe Disablement Allowance	4k	4k	4k	4k	4k	4k	4k
Industrial Injuries Disablement Benefit	137k	137k	137k	138k	138k	138k	139k
Armed Forces Compensation Scheme	93k	93k	94k	94k	94k	94k	94k
War Widow's or Widower's Pension	0k	0k	0k	0k	0k	0k	0k
Child Benefit	8,123k	8,120k	8,120k	8,130k	8,145k	8,161k	8,178k
Scottish Child Payment	351k	352k	352k	353k	350k	348k	337k
Two Child Limit Payment (Scotland, counterfactual)	0k	0k	0k	0k	0k	0k	0k
Healthy Start vouchers	200k	200k	200k	201k	201k	201k	202k
Maternity Allowance	0k	0k	0k	0k	0k	0k	0k
Statutory Maternity Pay	333k	333k	334k	334k	335k	336k	337k
Statutory Paternity Pay	0k	0k	0k	0k	0k	0k	0k
Statutory Sick Pay	181k	182k	182k	182k	182k	183k	183k
Sure Start Maternity Grant	26k	26k	26k	26k	26k	26k	26k
Bereavement Support Payment	98k	98k	98k	98k	98k	99k	99k
Widowed Parent's Allowance	0k	0k	0k	0k	0k	0k	0k
Tax-Free Childcare	1,043k	1,027k	1,014k	1,009k	1,011k	1,009k	993k
Extended childcare entitlement	795k	800k	796k	797k	798k	798k	792k
Universal childcare entitlement	1,071k	1,072k	1,072k	1,073k	1,076k	1,079k	1,074k
Targeted childcare entitlement	165k	156k	154k	154k	155k	155k	150k
Childcare Offer for Wales	39k	39k	39k	39k	39k	39k	39k
Care to Learn	0k	0k	0k	0k	0k	0k	0k
Childcare Grant	0k	0k	0k	0k	0k	0k	0k
Parents' Learning Allowance	98k	98k	98k	99k	99k	99k	99k
Adult Dependants' Grant	18k	18k	12k	12k	12k	12k	12k
Travel Grant	0k	0k	0k	0k	0k	0k	0k
Disabled Students' Allowance	suppressed	suppressed	suppressed	suppressed	suppressed	suppressed	suppressed
16 to 19 Bursary Fund	0k	0k	0k	0k	0k	0k	0k
Cost-of-living support payment	0k	0k	0k	0k	0k	0k	0k
Energy Bills Rebate	0k	0k	0k	0k	0k	0k	0k
Household Energy Payment	0k	21,065k	0k	0k	0k	0k	0k
Energy Price Guarantee subsidy	0k	0k	0k	0k	0k	0k	0k
Basic income	0k	0k	0k	0k	0k	0k	0k
Disability basic income	0k	0k	0k	0k	0k	0k	0k

Government balance, each component counted once: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Government receipts, each component once	£892.00bn	£934.66bn	£980.50bn	£1,015.93bn	£1,054.24bn	£1,091.56bn	£1,124.89bn
Government spending	£529.38bn †	£545.66bn †	£550.77bn †	£559.51bn †	£567.20bn †	£575.95bn †	£583.60bn †
Government balance, each receipt once	£362.62bn †	£389.00bn †	£429.74bn †	£456.42bn †	£487.04bn †	£515.61bn †	£541.29bn †
Benefits in household_benefits but not in gov_spending	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn	£0.00bn

Model totals: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
gov_tax	£892.00bn	£934.66bn	£980.50bn	£1,015.93bn	£1,054.24bn	£1,091.56bn	£1,124.89bn
household_tax	£491.97bn	£520.41bn	£551.21bn	£576.01bn	£601.00bn	£626.70bn	£648.42bn
pre_budget_change_household_tax	£484.15bn	£512.28bn	£542.82bn	£566.95bn	£591.72bn	£617.16bn	£638.61bn
gov_spending	£529.38bn †	£545.66bn †	£550.77bn †	£559.51bn †	£567.20bn †	£575.95bn †	£583.60bn †
household_benefits	£529.38bn †	£545.66bn †	£550.77bn †	£559.51bn †	£567.20bn †	£575.95bn †	£583.60bn †
gov_balance	£362.62bn †	£389.00bn †	£429.74bn †	£456.42bn †	£487.04bn †	£515.61bn †	£541.29bn †

Household income totals: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Household market income	£1,637.96bn	£1,702.77bn	£1,757.35bn	£1,806.41bn	£1,855.97bn	£1,912.43bn	£1,970.44bn
Household net income	£1,636.69bn †	£1,687.82bn †	£1,715.45bn †	£1,747.49bn †	£1,778.69bn †	£1,817.02bn †	£1,859.69bn †
Household net income (HBAI definition)	£1,449.57bn	£1,499.83bn	£1,530.86bn	£1,565.17bn	£1,597.89bn	£1,635.60bn	£1,676.61bn

Population: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
People	69,245k	69,383k	69,473k	69,592k	69,738k	69,912k	70,080k
Benefit units	35,110k	35,181k	35,226k	35,286k	35,360k	35,449k	35,534k
Households	31,046k	31,108k	31,149k	31,201k	31,267k	31,345k	31,420k

Poverty rates: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Absolute poverty, before housing costs: everyone	16.3%	15.5%	16.4%	16.3%	16.5%	16.5%	16.6%
Absolute poverty, before housing costs: children	24.7%	22.1%	23.3%	22.9%	23.1%	23.2%	23.5%
Absolute poverty, before housing costs: working-age adults	14.1%	13.8%	14.7%	14.7%	14.8%	14.8%	15.0%
Absolute poverty, before housing costs: pensioners	13.7%	13.4%	14.0%	13.8%	13.9%	13.9%	13.9%
Absolute poverty, after housing costs: everyone	21.6%	21.0%	22.0%	22.6%	22.9%	23.1%	23.4%
Absolute poverty, after housing costs: children	33.3%	31.3%	32.3%	32.5%	32.9%	33.4%	34.2%
Absolute poverty, after housing costs: working-age adults	19.9%	19.8%	20.8%	21.6%	21.9%	22.0%	22.3%
Absolute poverty, after housing costs: pensioners	12.7%	12.4%	13.2%	13.2%	13.3%	13.4%	13.6%
Relative poverty, before housing costs: everyone	17.0%	16.0%	16.4%	16.3%	16.4%	16.3%	16.6%
Relative poverty, before housing costs: children	25.9%	22.9%	23.3%	22.9%	23.0%	22.9%	23.5%
Relative poverty, before housing costs: working-age adults	14.5%	14.1%	14.6%	14.7%	14.7%	14.7%	15.0%
Relative poverty, before housing costs: pensioners	14.5%	14.0%	13.9%	13.8%	13.8%	13.9%	13.9%
Relative poverty, after housing costs: everyone	21.4%	21.0%	21.3%	21.3%	21.3%	21.5%	21.7%
Relative poverty, after housing costs: children	33.1%	31.1%	31.2%	31.0%	31.1%	31.2%	31.8%
Relative poverty, after housing costs: working-age adults	19.9%	19.8%	20.2%	20.4%	20.5%	20.6%	20.8%
Relative poverty, after housing costs: pensioners	12.5%	12.4%	12.2%	11.7%	11.7%	11.8%	12.0%

People in poverty: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Absolute poverty, before housing costs: everyone	11,278k	10,779k	11,402k	11,365k	11,476k	11,513k	11,651k
Absolute poverty, before housing costs: children	3,688k	3,314k	3,496k	3,445k	3,480k	3,502k	3,549k
Absolute poverty, before housing costs: working-age adults	5,965k	5,893k	6,308k	6,362k	6,420k	6,430k	6,521k
Absolute poverty, before housing costs: pensioners	1,624k	1,572k	1,598k	1,558k	1,576k	1,580k	1,581k
Absolute poverty, after housing costs: everyone	14,944k	14,600k	15,310k	15,721k	15,937k	16,141k	16,426k
Absolute poverty, after housing costs: children	4,975k	4,682k	4,837k	4,888k	4,945k	5,040k	5,179k
Absolute poverty, after housing costs: working-age adults	8,463k	8,461k	8,971k	9,344k	9,491k	9,584k	9,706k
Absolute poverty, after housing costs: pensioners	1,506k	1,457k	1,502k	1,490k	1,501k	1,517k	1,542k
Relative poverty, before housing costs: everyone	11,750k	11,084k	11,388k	11,364k	11,416k	11,415k	11,651k
Relative poverty, before housing costs: children	3,872k	3,423k	3,496k	3,445k	3,466k	3,454k	3,549k
Relative poverty, before housing costs: working-age adults	6,159k	6,013k	6,301k	6,362k	6,389k	6,389k	6,521k
Relative poverty, before housing costs: pensioners	1,720k	1,648k	1,590k	1,557k	1,562k	1,572k	1,581k
Relative poverty, after housing costs: everyone	14,849k	14,552k	14,776k	14,809k	14,874k	15,014k	15,234k
Relative poverty, after housing costs: children	4,942k	4,660k	4,673k	4,663k	4,679k	4,707k	4,806k
Relative poverty, after housing costs: working-age adults	8,426k	8,440k	8,711k	8,823k	8,873k	8,967k	9,062k
Relative poverty, after housing costs: pensioners	1,482k	1,452k	1,392k	1,323k	1,322k	1,339k	1,365k

Inequality: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Gini coefficient, equivalised HBAI income before housing costs	0.3080	0.3052	0.3058	0.3058	0.3063	0.3069	0.3085

Mean household net income in each side's own deciles: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Decile 1: mean household_net_income per household	£17,681 †	£17,997 †	£17,915 †	£18,152 †	£18,307 †	£18,531 †	£18,717 †
Decile 2: mean household_net_income per household	£27,939 †	£29,023 †	£28,711 †	£29,013 †	£29,276 †	£29,691 †	£30,150 †
Decile 3: mean household_net_income per household	£34,320 †	£35,485 †	£36,073 †	£36,611 †	£37,197 †	£37,737 †	£38,187 †
Decile 4: mean household_net_income per household	£38,933 †	£40,687 †	£41,723 †	£42,727 †	£43,334 †	£44,132 †	£44,432 †
Decile 5: mean household_net_income per household	£43,239 †	£45,567 †	£46,014 †	£46,959 †	£47,712 †	£48,584 †	£50,282 †
Decile 6: mean household_net_income per household	£47,927 †	£49,210 †	£50,892 †	£51,939 †	£52,945 †	£53,942 †	£54,680 †
Decile 7: mean household_net_income per household	£55,662 †	£56,571 †	£57,442 †	£57,670 †	£58,729 †	£59,657 †	£61,082 †
Decile 8: mean household_net_income per household	£68,789 †	£70,803 †	£71,858 †	£73,466 †	£73,880 †	£75,320 †	£77,178 †
Decile 9: mean household_net_income per household	£78,890 †	£81,013 †	£80,861 †	£82,646 †	£85,464 †	£86,946 †	£89,509 †
Decile 10: mean household_net_income per household	£117,005 †	£120,339 †	£124,742 †	£126,632 †	£125,971 †	£128,600 †	£130,938 †

Standard reform costings (change in gov_balance): head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Raise basic rate by 1pp	£6.41bn	£6.66bn	£6.63bn	£6.85bn	£7.03bn	£7.25bn	£7.49bn
Raise higher rate by 1pp	£3.61bn	£3.87bn	£4.00bn	£4.22bn	£4.44bn	£4.71bn	£4.83bn
Raise personal allowance by ~800GBP/year	£-3.80bn	£-3.92bn	£-3.98bn	£-4.04bn	£-4.10bn	£-4.23bn	£-1.95bn
Raise child benefit by 25GBP/week per additional child	£-1.99bn	£-1.75bn	£-1.55bn	£-1.38bn	£-1.30bn	£-1.18bn	£-1.07bn
Reduce Universal Credit taper rate to 20%	£-20.77bn	£-21.32bn	£-21.75bn	£-21.99bn	£-22.34bn	£-22.75bn	£-22.94bn
Raise Class 1 main employee NICs rate to 10%	£10.89bn	£11.30bn	£11.71bn	£12.01bn	£12.31bn	suppressed	suppressed
Raise VAT standard rate by 2pp	£24.89bn	£25.71bn	£26.63bn	£27.24bn	£27.87bn	£28.50bn	£29.14bn
Raise additional rate by 3pp	£2.90bn	£3.15bn	£3.22bn	£3.41bn	£3.61bn	suppressed	suppressed
Raise the Universal Credit standard allowance by 10%	£-3.06bn	£-3.18bn	£-3.26bn	£-3.35bn	£-3.44bn	£-3.49bn	£-3.50bn
Raise the State Pension by 10%	£-7.78bn	£-8.01bn	£-8.24bn	£-8.47bn	£-8.82bn	£-8.55bn	£-8.89bn

Standard reforms: change in each reform's target: head levels

Measure (head)	2025-26	2026-27	2027-28	2028-29	2029-30	2030-31	2031-32
Raise basic rate by 1pp: change in income_tax	£6.55bn	£6.86bn	£6.79bn	£7.00bn	£7.20bn	£7.44bn	£7.66bn
Raise higher rate by 1pp: change in income_tax	£3.61bn	£3.92bn	£4.01bn	£4.23bn	£4.45bn	£4.72bn	£4.84bn
Raise personal allowance by ~800GBP/year: change in income_tax	£-3.93bn	£-4.07bn	£-4.13bn	£-4.18bn	£-4.22bn	£-4.40bn	£-1.75bn
Raise child benefit by 25GBP/week per additional child: change in child_benefit	£2.33bn	£2.14bn	£1.94bn	£1.76bn	£1.65bn	£1.52bn	£1.40bn
Reduce Universal Credit taper rate to 20%: change in universal_credit	£20.74bn	£21.31bn	£21.78bn	£22.04bn	£22.37bn	£22.78bn	£22.99bn
Raise Class 1 main employee NICs rate to 10%: change in national_insurance	£11.15bn	£11.63bn	£12.00bn	£12.29bn	£12.61bn	suppressed	suppressed
Raise VAT standard rate by 2pp: change in vat	£24.69bn	£25.51bn	£26.42bn	£27.03bn	£27.65bn	£28.28bn	£28.91bn
Raise additional rate by 3pp: change in income_tax	£2.90bn	£3.15bn	£3.22bn	£3.42bn	£3.62bn	suppressed	suppressed
Raise the Universal Credit standard allowance by 10%: change in universal_credit	£3.06bn	£3.19bn	£3.28bn	£3.37bn	£3.43bn	£3.49bn	£3.51bn
Raise the State Pension by 10%: change in state_pension	£10.68bn	£11.17bn	£11.28bn	£11.60bn	£12.03bn	£12.47bn	£12.97bn

A reform that adds defined_for to a Housing Benefit deduction and keeps the model's formula masked only the annual output. In 2026-27 the segment helpers rebuilt each period's amount without the condition, and the award tapered income from those amounts. hb_apply_defined_for reads the condition from the variable the simulation holds and applies it as core does, so each segment's accommodation, earnings and WTC disregards, childcare deduction, childcare limit and applicable income are nil where it fails. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

One test each for the accommodation, earnings and WTC disregards, the childcare deduction, its limit and applicable income, with the condition failing and holding; the issue's reform; a supplied deduction; and a condition, on benefit units or people, that holds for one of two benefit units in a household. Hypothesis properties check that annual disregards equal their day-weighted segments, failed conditions give nil, units whose conditions hold keep the model's outputs, the award matches an independent calculation within each segment, and one-segment years give the annual route's award. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The annual earnings disregard summed hb_segment_deductions(...)[0], and that helper calculated the WTC disregard the caller discarded. With the segment mask in place, a reform giving the WTC disregard a defined_for condition that reads the earnings disregard raised CycleError: the earnings formula reached the WTC condition. hb_segment_earnings_disregard now returns the earnings disregard and the childcare deduction, and hb_segment_deductions adds the WTC disregard for entitlement. The amounts don't change. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Conditions of -1, 1 and 0 with a default of 7 give 7, 99 and 7 from Core and from hb_apply_defined_for, and applicable income with a £17 default reaches each 2026-27 segment. The tests fail if the helper uses nil instead of the variable's default or tests the condition with != 0. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The limit's WTC reserve ignores a condition on the WTC disregard, and the limit subtracts the standard disregard whatever the earnings disregard's condition gives. The docstrings and M4 now say so, and a test pins both cases with a limit a reform author can substitute. M2 states that a failed condition gives the variable's default, and that the award is the maximum benefit after non-dependant deductions. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Docstrings now say what each helper does and that a failed condition gives the variable's default. The changelog names the amounts the helpers rebuild. No code path changes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The registration sat where #2765 adds self_employment_income, so the two conflicted. The set's order means nothing; this place merges cleanly with #1969, #2661, #2765 and #2776. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Name the annual overrides each helper keeps, state the award floor in M2 and limit M4 to the generated cases. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

juaristi22

Reviewed 1d4a3e4757f20b0b6cf8d36ad038784f1a9d2bc9. One confirmed blocker remains.

C1 — Respect the new WTC disregard's mask inside accommodation segments. At housing_benefit_wtc_disregard.py:53–57, adding only defined_for = "is_couple" retains the original formula code, so the annual-override check misses the mask and the segment path recalculates an unmasked disregard.

Executed 2026 reform case, with tax credits kept active: a single continuing-HB parent aged 60 with child 8, £11,000 wages/16 hours, £10,000 private pension, £12,000 qualifying childcare and £10,400 London council rent in specified/temporary accommodation; known WTC £2,000 and CTC £4,000. The false mask correctly gives annual WTC disregard £0 and assessed income £17,064.32, matching a neutralized-disregard control. Yet final HB remains £3,115.76 instead of £2,537.78: £577.98 too much, also entering household net income. Childcare is £9,100 and the benefit-cap reduction is zero in both controls. The 2024 one-segment control passes.

The older helper limitation recorded in #2777 is now exposed by this new variable's segment call. Please preserve the mask in segment deductions and add a final-HB regression while retaining cycle protection for replacement formulas.

Validation: 97 repository tests passed; the focused two-segment reproduction fails on the final-award contract. The corrected native control and original harness setup error are retained separately. Independent policy/source review is complete; exact-head hosted CI is green and does not cover this counterexample. Local runtime: Core 3.33.2.

MaxGhenis · 2026-11-03T05:49:11Z

Picking this up for the UK hub. Reply here if you're mid-change and I'll stop.

Generated by Claude Code

MaxGhenis · 2026-11-03T06:48:41Z

Picking this up for the UK hub as brief 826, which replaces brief 814. Reply here if you're mid-change and I'll stop.

Generated by Claude Code

…ng Benefit A reform that adds defined_for = "is_couple" to housing_benefit_wtc_disregard and keeps its formula must take the disregard away in both 2026-27 accommodation segments, so the final award and household net income equal a neutralized-disregard control. The 2024 case checks the one-segment year. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LwQq6JE7NGFKUqVxTbNfqw

MaxGhenis · 2026-11-03T07:29:47Z

PR_TITLE.txt

Disregard WTC before the Housing Benefit childcare deduction

Generated by Claude Code

MaxGhenis · 2026-11-03T07:29:48Z

PR_BODY.md

Fixes #2730

Also fixes #2777.

## What and why

The model now disregards up to £17.10 a week of Working Tax Credit (WTC) in Housing Benefit before it deducts childcare. Under Schedule 5 paragraph 56 of the Housing Benefit Regulations 2006, the authority administering Housing Benefit disregards that amount for a family that meets a paragraph 17(2) work condition when its earnings can't fund the additional earnings disregard. The new `housing_benefit_wtc_disregard` takes the lesser of WTC and £889.20 a year (£1,929.20 in 2020-21). `housing_benefit_childcare_earnings_limit` lets childcare draw only on the WTC left after it, and `housing_benefit_applicable_income` deducts it alongside the earnings disregard. The existing `gov.dwp.tax_credits.active` parameter switches it off from model year 2025, when tax credits end.

In the issue's example, a lone parent earns £5,200 a year, has a £1,300 standard disregard, pays £4,160 for childcare and gets £100 of WTC. The model now disregards the £100 and deducts £3,900 for childcare, where main deducts £4,000.

The model also settles the WTC disregard in each accommodation segment of the calculation that #2234 added to `housing_benefit_entitlement`. Where a reform adds a `defined_for` condition to an amount the segments rebuild and keeps the model's formula, each segment now applies the condition as the annual output does (#2777).

### The regulations

Under [paragraph 56](https://www.legislation.gov.uk/uksi/2006/213/schedule/5/paragraph/56/2024-04-06), the authority disregards up to £17.10 of WTC, "any amount of working tax credit up to £17.10", except "in a case which falls under sub-paragraph (1) of paragraph 17 of Schedule 4". Under [paragraph 17(1)](https://www.legislation.gov.uk/uksi/2006/213/schedule/4/paragraph/17/2024-04-06), it raises the earnings disregard by £17.10 where "his net earnings equal or exceed the total" of the other earnings disregards, "the amount of child care charges calculated as deductible under regulation 27(1)(c)" and £17.10. Under [regulation 40(2)](https://www.legislation.gov.uk/uksi/2006/213/regulation/40/2024-04-06), it leaves out of other income "any sum, where applicable, specified in Schedule 5". Under [regulation 27(1)(c)](https://www.legislation.gov.uk/uksi/2006/213/regulation/27/2024-04-06), it then deducts childcare from earnings and, where earnings fall short and the family gets a tax credit, from "those earnings plus whichever credit" applies. In its [BW2 guidance](https://assets.publishing.service.gov.uk/media/5a7c7cf7ed915d6969f4538d/hbgm-bw2-assessment-of-income.pdf), DWP tells authorities to make that deduction "as the final step in the earnings calculation" (W2.131), to offset the rest against WTC or CTC "in addition to the existing disregards" (W2.181), and never to offset it against other income (W2.182).

The authority follows the same order under the pension-age and Northern Ireland regulations, and the model applies one calculation to all four:

| Regulations | Additional earnings test | WTC disregard | Childcare deduction |
|---|---|---|---|
| SI 2006/213 (working age) | [Sch 4 para 17](https://www.legislation.gov.uk/uksi/2006/213/schedule/4/paragraph/17) | [Sch 5 para 56](https://www.legislation.gov.uk/uksi/2006/213/schedule/5/paragraph/56/2024-04-06) | [reg 27(1)(c)](https://www.legislation.gov.uk/uksi/2006/213/regulation/27/2024-04-06) |
| SI 2006/214 (pension age) | [Sch 4 para 9](https://www.legislation.gov.uk/uksi/2006/214/schedule/4/paragraph/9/2024-04-06) | [Sch 5 para 21](https://www.legislation.gov.uk/uksi/2006/214/schedule/5/paragraph/21/2024-04-06) | [reg 30(1)(c)](https://www.legislation.gov.uk/uksi/2006/214/regulation/30/2024-04-06) |
| SR 2006/405 (NI working age) | [Sch 5 para 17](https://www.legislation.gov.uk/nisr/2006/405/schedule/5/paragraph/17/2024-04-06) | [Sch 6 para 58](https://www.legislation.gov.uk/nisr/2006/405/schedule/6/2024-04-06) | [reg 24(1)(c)](https://www.legislation.gov.uk/nisr/2006/405/regulation/24/2024-04-06) |
| SR 2006/406 (NI pension age) | [Sch 5 para 9](https://www.legislation.gov.uk/nisr/2006/406/schedule/5/paragraph/9/2024-04-06) | [Sch 6 para 22](https://www.legislation.gov.uk/nisr/2006/406/schedule/6/2024-04-06) | [reg 28(1)(c)](https://www.legislation.gov.uk/nisr/2006/406/regulation/28/2024-04-06) |

Under [SI 2020/371 reg 5](https://www.legislation.gov.uk/uksi/2020/371/regulation/5/2021-01-01) and [SR 2020/53 reg 5](https://www.legislation.gov.uk/nisr/2020/53/regulation/5/2024-04-06), which the Secretary of State made, the authority read £37.10 for £17.10 in both disregards, in all four sets of regulations, from 6 April 2020 to the end of 4 April 2021. That period spans the 52 weeks from Monday 6 April 2020 to Sunday 4 April 2021. The model's fiscal year 2020 runs to 5 April 2021, and the model takes the £1,929.20 for those 52 weeks from the dated worker-disregard parameter it already has.

Where a claimant gets only the savings credit of Pension Credit, the authority may modify the Secretary of State's income figure under [SI 2006/214 reg 27(4)](https://www.legislation.gov.uk/uksi/2006/214/regulation/27/2024-04-06) "only in so far as necessary to take into account" the items in its list. The list includes "childcare charges taken into account under regulation 30(1)(c)" and the Schedule 4 paragraph 9(1) additional earnings disregard, but not the Schedule 5 paragraph 21 WTC disregard, and [SR 2006/406 reg 25(4)](https://www.legislation.gov.uk/nisr/2006/406/regulation/25/2024-04-06) has the same list. So on that route the model corrects the childcare charge for the disregarded WTC and deducts no WTC disregard from income.

### Limit, deduction, WTC disregard

The paragraph 17 total includes the childcare deduction, and under regulation 27(2) that deduction draws on the WTC left after the paragraph 56 disregard, so each decision needs the other. The model settles them in one direction:

1. `housing_benefit_childcare_earnings_limit` keeps back the WTC that the authority would disregard under paragraph 56 if childcare took the whole capped charge. It reads neither the deduction nor the `housing_benefit_wtc_disregard` formula.
2. `housing_benefit_applicable_income_childcare_element` takes the lesser of the capped charge and that limit.
3. `housing_benefit_wtc_disregard` and the additional earnings disregard test paragraph 17 with the deduction in force in each segment.

For the deduction the model calculates, the WTC the limit keeps back equals the WTC disregard. Write E for the earnings the standard and accommodation disregards leave, C for the capped charge, a for the additional amount, R for the lesser of WTC and a, and T for CTC. If E ≥ C + a, earnings reach the paragraph 17(1) total with the whole charge and with any smaller deduction, and neither test disregards WTC. Otherwise the limit keeps back R, and the deduction comes to D = min(C, E + (WTC − R) + T). Either D = C, or D < C and D ≥ E; either way E < D + a, so earnings fall short of the paragraph 17(1) total and the model disregards R. Keeping back no WTC would only enlarge the deduction, and earnings would still fall short, so R gives the only consistent answer. `test_limit_keeps_back_the_wtc_disregard_in_force` checks the identity for every generated family, and the segment property checks it within each segment.

A reform can replace the deduction with a formula that reads the limit. With half the capped charge, capped at the limit, the issue's family gets a £2,080 deduction and the additional earnings disregard, and income of £1,030.80. Where the limit binds, the WTC it keeps back stays consistent: with £2,000 of earnings, a £260 standard disregard, a £9,100 charge and £1,000 of WTC, the deduction comes to £1,850.80 and the model disregards £889.20 of WTC.

### Supplied and replaced values

- **WTC disregard.** The limit reads the WTC disregard only where reading it reads no other variable. That covers an amount that a dataset, `set_input` or a scenario supplies, and a disregard that a reform neutralizes. It also covers one that a reform leaves without a formula, an `adds` or `subtracts` aggregate or a `defined_for` mask, which the engine (policyengine-core) uprates or carries over from its own earlier inputs. The model applies each of these to income, to the limit and in every segment, and neutralizing `housing_benefit_wtc_disregard` gives main's calculation back. The engine computes any other replacement through a formula, an aggregate (`adds` or `subtracts`, each a list or the path of a parameter that lists variables) or a mask it calculates first, and any of these could read the deduction. The model applies such a replacement to income and, as an annual amount, in every segment, but the limit keeps back what the whole-charge test gives, so a forwarding replacement leaves the limit's figure unchanged. When a reform only adds a `defined_for` mask and keeps the model's formula, the annual output takes the mask, and each 2026-27 segment applies the same condition to its own WTC disregard (see [Reform conditions in each segment](#reform-conditions-in-each-segment)).
- **Childcare deduction and its limit.** The model applies a deduction that a dataset, `set_input` or a reform supplies or replaces, or the lesser of the capped charge and a supplied or replaced limit, in every segment, and reads it in both paragraph 17 tests. Where the deduction itself comes from a dataset, `set_input` or a reform, only a replacement formula that calls the limit reads it. The limit still keeps back the WTC that the whole-charge test gives, so it can differ from main: with the issue's family and a supplied nil deduction, it comes to £3,900 where main gives £4,000, while the model disregards no WTC.
- **Earnings disregard.** A supplied `housing_benefit_applicable_income_disregard` stays as supplied, and the model still makes the WTC decision's own paragraph 17 test from the earnings inputs, because a supplied total doesn't show whether it includes the additional amount.

### Accommodation segments

From 5 October 2026, the paragraph 17(3)(a) total also includes the paragraph 18 specified or temporary accommodation disregard. The Secretary of State made that change by [SI 2026/753 reg 2(4)](https://www.legislation.gov.uk/uksi/2026/753/regulation/2), and replaced that regulation by [SI 2026/978 reg 2(2)](https://www.legislation.gov.uk/uksi/2026/978/regulation/2) before it came into force on the date in [reg 1](https://www.legislation.gov.uk/uksi/2026/753/regulation/1). The Department for Communities made the same change in Northern Ireland by [SR 2026/157 reg 2(4)(a)](https://www.legislation.gov.uk/nisr/2026/157/regulation/2). Since #2234, the model settles paragraph 17 separately before and after that date and combines the awards by days. The authority's paragraph 56 decision depends on the paragraph 17 outcome, so the model now gives each segment its own WTC disregard and its own limit. `hb_segment_deductions` returns the segment's earnings disregard, WTC disregard and childcare deduction, `housing_benefit_entitlement` deducts all three in each segment, and the annual `housing_benefit_wtc_disregard` combines the segments' amounts by days.

Tax credits ended on 5 April 2025, before the accommodation disregard starts, so no current-law award changes. Take a reform that sets `gov.dwp.tax_credits.active` to true in 2026, and a single claimant aged 40 in specified accommodation. The claimant earns £500, has a £260 standard disregard, £2,000 of WTC and no childcare or other income, pays rent of £10,400 and has a nil applicable amount. Main gives £9,022.21, and the model now gives £9,600.19:

- From 6 April to 4 October 2026 (182 days), earnings of £500 fall short of £260 plus £889.20, so the model disregards £889.20 of WTC. Income comes to £2,500 − £260 − £889.20 = £1,350.80, and the award to £10,400 − 0.65 × £1,350.80 = £9,521.98.
- From 5 October 2026 to 5 April 2027 (183 days), the £4,041.96 accommodation amount takes the £240 of earnings left, and earnings again fall short. Income comes to £2,500 − £500 − £889.20 = £1,110.80, and the award to £9,677.98.
- Combined, (182 × £9,521.98 + 183 × £9,677.98) / 365 = £9,600.19.

The model needs a WTC disregard in each segment. With £2,000 of earnings, the same claimant's earnings cover £260 plus £889.20 before 5 October, so the model gives the additional earnings disregard and no WTC disregard. After 5 October the £1,740 accommodation amount leaves earnings short, and the model disregards £889.20 of WTC. The award comes to £9,114.03. A WTC disregard settled once from the annual accommodation amount would give both disregards before 5 October, and an award of £9,402.23. With £4,000 of earnings it would give neither disregard after 5 October, and £8,176.03 where the model gives £8,465.81.

Deducting the day-weighted annual amount in every segment matches the per-segment award while both segments stay on the taper. Once income after 5 October falls below a £2,000 applicable amount, it gives £10,268.74 instead of £10,124.25. Reading the childcare limit with that annual amount gives £7,394.22 instead of £7,266.74 for a lone parent whose limit binds after 5 October.

### Reform conditions in each segment

policyengine-core masks a variable that has a `defined_for` condition: it keeps the amount where the condition exceeds zero and gives the variable's default elsewhere. The segment helpers rebuild amounts without calling those variables, so `hb_apply_defined_for` applies the same condition to each rebuilt amount. It reads the condition from the variable the simulation holds, calculates it for the year and maps it to the variable's entity. In each 2026-27 segment, and in the single segment of every other year:

| Amount | Where the model applies the condition |
|---|---|
| Accommodation disregard | `hb_accommodation_segments`, to each segment's full amount |
| Earnings disregard | `hb_segment_earnings_disregard` |
| WTC disregard | `hb_segment_wtc_disregard` |
| Childcare limit and deduction | `hb_segment_childcare_deduction`; `hb_fixed_childcare_deduction` applies the deduction's condition where a supplied or replaced limit sets it |
| Applicable income | `housing_benefit_entitlement`, to each segment's income |

A supplied, neutralized or replaced annual amount still applies as supplied in every segment, whatever its condition gives. The limit still keeps back the WTC that the whole-charge test gives, and still subtracts the standard and accommodation disregards, when the WTC or earnings disregard's condition fails; a reform that wants those amounts to fund childcare changes the limit too.

Take a reform that adds `defined_for = "is_couple"` to `housing_benefit_wtc_disregard` and keeps its formula, in 2026 with tax credits kept. A lone parent aged 60, continuing a Housing Benefit award with a child aged 8, works 16 hours a week for £11,000, draws a £10,000 private pension, pays £12,000 for qualifying childcare and £10,400 of London council rent in specified or temporary accommodation, and reports £2,000 of WTC and £4,000 of CTC. From those reports the model calculates WTC of £6,530.35 and CTC of £3,646.49. It attributes £883.14 of the £1,686 of income tax to the wages, so net earnings come to £10,116.86 and income before deductions to £11,000 + £10,000 − £1,686 + £10,176.84 = £29,490.84. The lone parent disregard gives £1,300 (£25 a week, `means_test/income_disregard/lone_parent.yaml`), the childcare cap £9,100 (£175 a week, `working_tax_credit/elements/childcare_1.yaml`), the accommodation disregard £4,041.96 from 5 October (£77.73 a week, `specified_or_temporary_accommodation/lone_parent/older.yaml`), the worker amount £889.20 (£17.10 a week, `income_disregard/worker.yaml`) and the applicable amount £4,968.60 (£95.55 a week, `allowances/lone_parent/older.yaml`). Net earnings fall short of £11,289.20 before 5 October and £15,331.16 after, so the model disregards £889.20 of WTC in each segment, and the condition takes it away:

- From 6 April to 4 October (182 days), income comes to £29,490.84 − £1,300 − £9,100 = £19,090.84, and the award to £10,400 − 0.65 × (£19,090.84 − £4,968.60) = £1,220.54.
- From 5 October (183 days), income comes to £29,490.84 − £5,341.96 − £9,100 = £15,048.88, and the award to £3,847.82.
- Combined, the award comes to £2,537.78, with no benefit cap reduction, and annual income to £17,064.32. Without the condition in each segment, the model paid £3,115.76, 0.65 × £889.20 = £577.98 more, and household net income rose by the same amount. A reform that neutralizes the WTC disregard gives the same award and household net income as the condition. In 2024, with one segment, the condition gives £1,113.09 where the model gives £1,691.07.

A reform that replaces the earnings disregard's formula and adds a condition that reads the WTC disregard, the childcare limit or, in 2026-27, the accommodation disregard meets a circular definition, and policyengine-core raises `CycleError`. The engine calculates the condition first, and with the replacement in place the condition's amount reads the annual earnings disregard the replacement gives, so the condition reads its own result. For a claimant in specified accommodation that covers five cells, three in 2026-27 and two in 2024. The same conditions on the model's own formula run, and a condition that reads its own variable or applicable income raises `CycleError` with or without a replacement.

### Invariants

For every generated population, with assessed earnings, disregards, childcare, credits and work conditions as inputs:

- The WTC disregard lies between zero and the lesser of WTC and the statutory amount. It falls to zero when the work condition fails, when earnings meet paragraph 17(1), and when tax credits stop.
- No benefit unit gets both the additional earnings disregard and a positive WTC disregard within a segment. Annual amounts can both exceed zero: with £2,000 of earnings in specified accommodation, the year has £443.38 of additional earnings disregard and £445.82 of WTC disregard. Exclusivity holds when a dataset or `set_input` supplies the deduction or its limit, and example tests cover a neutralized deduction, one with no formula and replacement formulas, one of which reads the limit.
- A deduction the model calculates never exceeds the capped charge, or the earnings and credits the disregards leave. A supplied deduction stays as supplied: £25,000 stays £25,000 against a £4,160 charge.
- Where the model calculates the deduction, its limit and the WTC disregard, the WTC the limit keeps back equals the WTC disregard, for the year and within each segment.
- Assessed income equals, and never falls below, an independent scalar calculation of the regulations. That calculation settles paragraph 17 by trying both outcomes.
- In 2026-27 the award equals the day-weighted sum of the awards the same calculation gives in each segment, and the annual WTC disregard equals the day-weighted sum of the segments' amounts.
- Where the model disregards no WTC in any segment, every output equals main's, except the limit where a dataset or `set_input` supplies the deduction. Under current law that covers every family.
- In a year with one accommodation amount, supplying the deduction the model calculates changes no result. In 2026-27, where the accommodation disregard changes, the model applies the supplied annual amount in every segment: replaying the lone parent's calculated £4,160 moves the award from £7,266.74 to £7,394.22.
- A reform that replaces the childcare deduction, its limit or the WTC disregard, each alone, with one that forwards the model's own amount meets no circular definition and changes no output. For the WTC disregard, that holds whether the replacement forwards through a formula or through `adds` or `subtracts`, each listing the forwarding variable or naming a parameter that lists it. It holds in 2024 for families with and without WTC, and in 2026-27 with tax credits kept for families outside specified or temporary accommodation. On the example populations, every 2024 output stays identical and every 2026-27 output agrees within £0.001, the float32 rounding of the day weights. The computed replacements of the WTC disregard leave all six outputs of the eight test families unchanged in both years, and the generated populations allow 2p. In specified accommodation the model applies the replacement's annual amount in every segment, as for any replacement: a forwarding replacement of the WTC disregard, through a formula or an aggregate, gives £10,268.74 where the model gives £10,124.25.
- The override check that segments use (`hb_uses_annual_override`) gives main's answer for all 1,200 combinations of the states the engine reads: neutralization, a default or replacement formula, `adds` and `subtracts` (none, empty, a list or a parameter path), a `defined_for` mask, uprating and a supplied input. Where only a `defined_for` mask joins the model's own formula, the check keeps the segment route, and each segment helper applies the condition itself.

Under a reform that gives each of the accommodation, earnings and WTC disregards, the childcare deduction, its limit and applicable income its own condition, which reads none of these amounts, for every generated population with or without tax credits:

- Each annual disregard equals the day-weighted sum of its segment amounts.
- Where a condition fails, its amount takes the variable's default in the annual output and in every segment, unless the population supplies the amount. Where applicable income's condition fails, the award equals the maximum benefit less 65% of the excess of the variable's default over the applicable amount.
- A benefit unit whose conditions all hold keeps the model's outputs, award included.
- The award agrees with an independent scalar calculation within each segment, combined by days.
- In each year but 2026-27, the award the single segment gives equals the award the model builds from annual income.

## Test plan

- [x] For policy changes: the 13 cases in `housing_benefit_wtc_disregard.yaml` and the Python examples take their expected values from the regulations, and their comments work through the arithmetic by hand.
- [x] `make format`: Ruff 0.17.2 formats and checks every changed Python file, `ruff check --select F811,F821 policyengine_uk` passes, and `git diff --check` passes.
- [x] Changelog fragments `changelog.d/2730.fixed.md` and `changelog.d/2777.fixed.md`

The tests build their own households and never load survey data. Rows that name `2f087ae7c` ran at this branch's head, which merges main at `52b30809f`; the other rows name the commit they ran at, and CI runs the rest of the suite.

| Check | Result |
|---|---|
| `test_housing_benefit_wtc_disregard.py`: examples, Hypothesis properties, differential tests and replacement tests | 55 passed at `2f087ae7c` |
| `test_housing_benefit_wtc_disregard_segments.py`: segment examples and properties | 18 passed at `2f087ae7c` |
| `test_housing_benefit_segment_masks.py`: a reform condition on each amount in each segment, #2777's reform, and the couples-only WTC disregard for a lone parent in 2026-27 and 2024 | 30 passed at `2f087ae7c` |
| `housing_benefit_wtc_disregard.yaml` | 13 passed at `2f087ae7c` |
| `test_housing_benefit_accommodation_dates_properties.py` | 6 passed at `2f087ae7c` |
| `test_housing_benefit_childcare_properties.py`, which reads the limit | 5 passed at `2f087ae7c` |
| Housing Benefit YAML directory | 597 passed at `2f087ae7c` |
| `policyengine_uk/tests/code_health` | 3,092 passed at `2f087ae7c` |
| `test_housing_benefit_accommodation_dates.py`, with a 3,600 s cap | 38 passed in 2,945 s at `6dc47664d` |
| The two original test modules and the WTC YAML file on main (`2da212abc`) | 55, 18 and 13 failed; the reproducer deducts £4,000 for childcare against £3,900 |
| The couples-only lone parent at `2f087ae7c` with the five model files back at `1d4a3e475` | 2026-27 fails: the award comes to £3,115.76 against £2,537.78. 2024 passes on the award and fails only where it reads the segment's WTC disregard, £889.20 against nil |

The differential ran main (`2da212abc`) and this branch at `fdc13f128` over the same 15,552 family-years: 1,944 families in each of 2018, 2020, 2021, 2024, 2025, 2026 and 2027, and in 2026 with tax credits kept. Since `fdc13f128` the model code changes which replacements of the WTC disregard the limit reads, and applies reform conditions in each segment; none of these populations replaces the WTC disregard or adds a condition. Of each 1,944, 1,296 families have a calculated deduction, 324 a supplied nil deduction and 324 a supplied £1,000 limit. Earnings, WTC, childcare, the standard disregard, the work condition, pension age and specified accommodation vary; other income comes to £1,000, rent to £10,400 and the applicable amount to £2,000, and regions alternate between London and Northern Ireland. It compares the childcare deduction, earnings disregard, childcare limit, applicable income and award.

| Family-years | Where the model disregards WTC in no segment | Where it disregards WTC in a segment |
|---|---|---|
| Calculated deduction | 8,962: all 44,810 values equal main's exactly | 1,406: every one moves by more than 1p |
| Supplied £1,000 limit | 2,328: all 11,640 values equal main's exactly | 264: every one moves |
| Supplied nil deduction | 2,340: 11,578 of 11,700 values equal main's; the other 122 come from the limit, which keeps back the WTC the whole-charge test gives | 252: every one moves |

None of the 5,832 family-years in 2025, 2026 and 2027 under current law changes.

Of 44 mutations across the changed model files for #2730, 42 make at least one test fail. The other two add a floor or a cap to the paragraph 17 test, which can't change its answer while the additional amount stays positive. Each mutation runs the main module, the segment module and the WTC YAML file in full, with Hypothesis shrinking off. The 15 that touch the override helpers or their call sites ran in full at `6dc47664d`. The other 29 ran at `fdc13f128`, or for the main module and YAML file at `9c2adcde9`, whose model code and tests match `fdc13f128`. Since then the model code changes which replacements of the WTC disregard the limit reads, and applies reform conditions in each segment; the test modules keep every earlier test. For #2777, all 10 mutations of the condition helper and its call sites make a test fail; each ran the main module, the segment module and the segment-mask module in full at `2f087ae7c`, with Hypothesis seed 0.

<details>
<summary>Mutation results</summary>

| File | Mutation | Result |
|---|---|---|
| `housing_benefit_childcare_earnings_limit.py` | The shared paragraph 17 test leaves out the accommodation disregard | Caught: main module 3 failed; segment module 9 failed. (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | `>=` becomes `>` in the paragraph 17 test | Caught: main module 4 failed; WTC YAML 1 failed. (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | The paragraph 17 test leaves out the standard disregard | Caught: main module 13 failed; segment module 1 failed; WTC YAML 4 failed. (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | The paragraph 17 test floors net earnings at zero | Survives, as intended: with a positive additional amount it can't change the answer (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | The paragraph 17 test caps the accommodation amount at the earnings left | Survives, as intended: with a positive additional amount it can't change the answer (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | The WTC decision leaves the accommodation disregard out of the paragraph 17 test | Caught: main module 3 failed; segment module 7 failed. (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | The disregard applies when earnings meet paragraph 17(1) | Caught: main module 18 failed; segment module 8 failed; WTC YAML 3 failed. (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | The disregard ignores the work condition | Caught: main module 2 failed; segment module 2 failed; WTC YAML 1 failed. (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | The disregard takes all WTC, uncapped | Caught: main module 9 failed; segment module 7 failed; WTC YAML 4 failed. (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | A negative WTC award gives a negative disregard | Caught: main module 1 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_childcare_earnings_limit.py` | The disregard ignores the end of tax credits | Caught: main module 4 failed; segment module 2 failed; WTC YAML 1 failed. (at `fdc13f128`) |
| `housing_benefit_childcare_earnings_limit.py` | The cap uses the weekly amount | Caught: main module 17 failed; segment module 8 failed; WTC YAML 7 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_childcare_earnings_limit.py` | The limit reads the WTC disregard variable again | Caught: main module 32 failed; segment module 10 failed; WTC YAML 10 failed. The forwarding-reform tests fail. The baseline itself meets a circular definition. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_childcare_earnings_limit.py` | The WTC kept back tests a supplied or replaced deduction in place of the capped charge | Caught: main module 5 failed; segment module 1 failed. (at `6dc47664d`) |
| `housing_benefit_childcare_earnings_limit.py` | The WTC kept back tests the deduction variable in place of the capped charge | Caught: main module 32 failed; segment module 10 failed; WTC YAML 10 failed. The forwarding-reform tests fail. The baseline itself meets a circular definition. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_childcare_earnings_limit.py` | The limit ignores a supplied, neutralized or input WTC disregard | Caught: main module 3 failed; segment module 1 failed; WTC YAML 1 failed. (at `6dc47664d`) |
| `housing_benefit_childcare_earnings_limit.py` | The limit reads a replacement WTC disregard formula | Caught: main module 4 failed; segment module 3 failed. (at `6dc47664d`) |
| `housing_benefit_childcare_earnings_limit.py` | A segment's limit keeps back the annual amount, not its own | Caught: segment module 2 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_childcare_earnings_limit.py` | The annual limit adds the segments' WTC without day weights | Caught: segment module 1 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_childcare_earnings_limit.py` | Childcare draws on the disregarded WTC | Caught: main module 10 failed; segment module 3 failed; WTC YAML 4 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_wtc_disregard.py` | The override check ignores a supplied or replaced limit | Caught: main module 3 failed; segment module 2 failed; WTC YAML 1 failed. (at `6dc47664d`) |
| `housing_benefit_wtc_disregard.py` | The override check ignores a supplied or replaced deduction | Caught: main module 9 failed; segment module 2 failed; WTC YAML 1 failed. (at `6dc47664d`) |
| `housing_benefit_wtc_disregard.py` | A supplied deduction reads the capped charge instead | Caught: main module 10 failed; segment module 2 failed; WTC YAML 1 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_wtc_disregard.py` | A supplied limit sets the deduction without the charge cap | Caught: main module 3 failed; segment module 2 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_wtc_disregard.py` | Each segment's deduction reads the annual limit | Caught: segment module 3 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_wtc_disregard.py` | Segments discard a supplied annual WTC disregard | Caught: main module 5 failed; segment module 9 failed. (at `6dc47664d`) |
| `housing_benefit_wtc_disregard.py` | Segments ignore a supplied, neutralized or replaced annual WTC disregard | Caught: segment module 9 failed. (at `6dc47664d`) |
| `housing_benefit_wtc_disregard.py` | The annual WTC disregard tests the capped charge, not the deduction in force | Caught: main module 12 failed; segment module 2 failed; WTC YAML 2 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_wtc_disregard.py` | The annual amount adds segments without day weights | Caught: segment module 7 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_wtc_disregard.py` | The annual amount comes from the annual accommodation amount | Caught: segment module 5 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_applicable_income_disregard.py` | A segment's WTC disregard tests the capped charge, not the deduction in force | Caught: segment module 2 failed. The forwarding-reform tests fail. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_applicable_income_disregard.py` | Segments settle the WTC disregard from the annual accommodation amount | Caught: segment module 6 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_applicable_income_disregard.py` | The earnings disregard tests the capped charge, not the deduction | Caught: main module 9 failed; segment module 1 failed; WTC YAML 2 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_applicable_income.py` | Annual income leaves out the WTC disregard | Caught: main module 9 failed; segment module 1 failed; WTC YAML 2 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_entitlement.py` | Segment income leaves out the WTC disregard | Caught: segment module 9 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_entitlement.py` | Each segment deducts the day-weighted annual WTC disregard | Caught: segment module 2 failed. (segment module at `fdc13f128`; main module and YAML at `9c2adcde9`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | The override helper runs the formula of a neutralized variable | Caught: main module 2 failed. (at `6dc47664d`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | The helper lets the limit read an adds or subtracts aggregate | Caught: main module 8 failed; segment module 4 failed. (at `6dc47664d`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | The helper lets the limit read an adds aggregate | Caught: main module 4 failed; segment module 2 failed. (at `6dc47664d`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | The helper lets the limit read a subtracts aggregate | Caught: main module 4 failed; segment module 2 failed. (at `6dc47664d`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | The helper lets the limit read an aggregate named by a parameter path | Caught: main module 4 failed; segment module 2 failed. (at `6dc47664d`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | The helper lets the limit read a variable with a defined_for mask | Caught: main module 1 failed. (at `6dc47664d`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | The helper keeps an uprated input from the limit | Caught: main module 1 failed. (at `6dc47664d`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | Segments recalculate an aggregate replacement instead of applying its annual amount | Caught: segment module 4 failed. (at `6dc47664d`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | `hb_apply_defined_for` ignores every condition | Caught: segment-mask module 17 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | A failed condition gives nil instead of the variable's default | Caught: segment-mask module 1 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | The condition holds wherever it differs from zero, not where it exceeds zero | Caught: segment-mask module 1 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
| `housing_benefit_specified_or_temporary_accommodation_disregard.py` | Segments leave the accommodation disregard's condition out | Caught: segment-mask module 5 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
| `housing_benefit_applicable_income_disregard.py` | Segments leave the earnings disregard's condition out | Caught: segment-mask module 7 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
| `housing_benefit_wtc_disregard.py` | A supplied or replaced limit sets the deduction without the deduction's condition | Caught: segment-mask module 2 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
| `housing_benefit_wtc_disregard.py` | Segments leave the limit's condition out | Caught: segment-mask module 3 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
| `housing_benefit_wtc_disregard.py` | Segments leave the childcare deduction's condition out | Caught: segment-mask module 3 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
| `housing_benefit_wtc_disregard.py` | Segments leave the WTC disregard's condition out | Caught: segment-mask module 9 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
| `housing_benefit_entitlement.py` | Segment income leaves applicable income's condition out | Caught: segment-mask module 4 failed; main and segment modules pass. (#2777, at `2f087ae7c`) |
</details>

No test on main changes its expected value. #2719, #2700, #2727, #2731, #2735 and #2234 have merged, and this branch includes them. At `2f087ae7c`, `git merge-tree --write-tree --name-only` against each open PR that edits this branch's files gives:

- #1995 (`7381a46c7`) edits `housing_benefit_applicable_income.py`, and Git reports a conflict there: both PRs extend the last sentence of the variable's documentation string. Git reports the same conflict against `1d4a3e475`. Whoever merges second keeps both sentences. #1995 also changes the income tax that Housing Benefit deducts, so once it lands the lone parent's figures above will change.
- #1969, #2007, #2010, #2661, #2807 and #2827 edit `policyengine_uk/utils/supplied_inputs.py`, and Git merges this branch with each of them without a conflict.
- #2265 and #2308 edit `policyengine_uk/utils/supplied_inputs.py` and conflict with main itself; against this branch Git reports the same files as against main, so this branch adds no conflict.

No other open PR edits the override helpers' file, the limit's file, the three test modules or the WTC YAML file. #2301 and #2302 no longer edit `housing_benefit_applicable_income.py`.

## Reproducer

Reproducer: policyengine_uk/tests/test_housing_benefit_wtc_disregard.py::test_issue_2730_reserves_wtc_before_childcare, policyengine_uk/tests/test_housing_benefit_wtc_disregard_segments.py::test_wtc_disregard_applies_before_and_after_5_october, policyengine_uk/tests/test_housing_benefit_segment_masks.py::test_issue_2777_reform_masks_each_segment, policyengine_uk/tests/test_housing_benefit_segment_masks.py::test_a_couples_only_wtc_disregard_reaches_a_lone_parents_award

axiom: TheAxiomFoundation/rulespec-uk#510 queued

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Generated by Claude Code

MaxGhenis · 2026-11-03T07:30:28Z

Ready for hub review

Brief 826 (replaces 814). Head 2f087ae7c349abcb7651e1b48b099230dd57f1d7 on branch claude/reform-masks-accommodation-a3bc5v. The harness pinned pushes to that claude/ branch. I opened no PR and pushed nothing to fix-2730-hb-wtc-disregard or fix-2777-hb-segment-masks.

Branch history: af4819b14 (the nine #2777 commits on 1d4a3e475) → b8372df8d merges origin/main at 52b30809f → 2f087ae7c adds the review's case.

Writer check

fix-2777-hb-segment-masks has no commits beyond af4819b14, and fix-2730-hb-wtc-disregard has none beyond 1d4a3e475.
After 05:48Z on 3 Nov, PR Disregard WTC before the Housing Benefit childcare deduction #2776 has only the old pick-up comment (6596493880) and no new reviews.
One claude/ branch has a newer commit: claude/pr-2807-review-2813-carry-knc8d1 (06:12Z), which is Bar income-related ESA at pensionable age and on Pension Credit #2807/Model the 2026 income-related ESA abolition and its UC claims #2813 Pension Credit and ESA work. It edits supplied_inputs.py and some Housing Benefit eligibility and take-up files, but it doesn't touch this job's model files or tests and doesn't name Disregard WTC before the Housing Benefit childcare deduction #2776 or Apply eligibility-only Housing Benefit reforms to each 2026-27 period #2777. I treated it as another job and went ahead.

Findings

1. #2777's fix and whether its tests cover C1. git log 1d4a3e475..af4819b14 adds hb_apply_defined_for (housing_benefit_specified_or_temporary_accommodation_disregard.py:18-32). Each segment helper calls it for its own amount:

accommodation disregard, :120
earnings disregard, housing_benefit_applicable_income_disregard.py:58
fixed deduction, limit, deduction and WTC disregard, housing_benefit_wtc_disregard.py:33,46,52,70
segment income, housing_benefit_entitlement.py:120

test_housing_benefit_segment_masks.py already covered a mask on the WTC disregard reaching each segment and the award (SINGLE_MASKS "WTC disregard", test_issue_2777_reform_masks_each_segment), so the mechanism behind C1 was covered. It did not cover the review's case: a condition on a core variable (is_couple) for a lone parent with childcare, a pension and model-calculated tax credits, checked through final housing_benefit and household_net_income. I added that case.

CycleError cells at 2f087ae7c. I probed each of the six masked variables as the target, with a condition reading each of the six as the source, comparing a mask on the model's formula against a forwarding replacement formula. The claimant was the specified-accommodation single claimant, in 2026 with tax credits kept and in 2024. Reform-only cells (the mask runs, the replacement raises CycleError):

2026: the earnings disregard replaced, with a condition reading the accommodation disregard, the WTC disregard or the childcare limit. That's 3 cells.
2024: the earnings disregard replaced, with a condition reading the WTC disregard or the limit. That's 2 cells.

That makes five, not the six the brief names. I couldn't find a list of the six in the repo, the commits or the PR/issue comments. Any cell whose condition reads its own variable or applicable income cycles with or without the replacement. The probe is a scratch script, not a test. I added no test for these cells, following the hub's 1 Nov call.

2. Merging main. b8372df8d merges origin/main (52b30809f, 115 commits after 1d4a3e475) with no conflicts. Of main's changes since 1d4a3e475, only policyengine_uk/utils/supplied_inputs.py touches a file this PR changes: main added self_employment_income, miscellaneous_income_trading_allowance_deduction and trading_allowance_deduction to SUPPLIED_INPUT_VARIABLES, and this PR adds one entry. No other main change touches Housing Benefit files. git merge-tree --write-tree --name-only HEAD origin/main → fd676ff73… with no conflicted files.

3. The review's case. test_a_couples_only_wtc_disregard_reaches_a_lone_parents_award (test_housing_benefit_segment_masks.py, end of file) covers 2026 with tax credits kept, plus the 2024 one-segment control. Masked outputs equal a neutralized-disregard control for the WTC disregard, earnings disregard, childcare deduction, applicable income, benefit cap reduction, housing_benefit and household_net_income. The model's own award exceeds the masked one by £577.98, and household net income moves by the same amount.

Where my figures differ from the review's. The figures match only when the £2,000 WTC and £4,000 CTC are reported awards (working_tax_credit_reported and child_tax_credit_reported). From those reports the model calculates WTC £6,530.35 and CTC £3,646.49 (2024: £6,079.55 and £4,000).
- Supplying the awards themselves (working_tax_credit, child_tax_credit and tax_credits) gives annual income of £12,887.48, not £17,064.32.
- Without tax_credits the model pays no tax credits at all, because is_WTC_eligible needs a reported award.
- With the reported awards, every review figure reproduces exactly: WTC disregard nil, income £17,064.32, childcare £9,100, cap reduction nil, HB £2,537.78 against £3,115.76.
Hand working. Net earnings come to £10,116.86: £11,000 less 11,000/21,000 of the £1,686 of tax. Income before deductions is £29,490.84. The parameter files:
- lone_parent.yaml: £25/wk → £1,300
- childcare_1.yaml: £175/wk → £9,100
- specified_or_temporary_accommodation/lone_parent/older.yaml: £77.73/wk → £4,041.96 from 5 October
- worker.yaml: £17.10/wk → £889.20
- allowances/lone_parent/older.yaml: £95.55/wk → £4,968.60
- withdrawal_rate.yaml: 0.65
Before 5 October (182 days), income is £19,090.84 and the award £1,220.54. After 5 October (183 days), income is £15,048.88 and the award £3,847.82. Combined, the award comes to £2,537.78. The test comment shows the working, and the PR body has it too.
Revert check. In a scratch worktree of 2f087ae7c, I checked out the five Housing Benefit model files at 1d4a3e475. I also added an identity hb_apply_defined_for stub, because the test module imports it.
- 2026 fails: housing_benefit comes to 3,115.7595 against 2,537.78.
- 2024 fails only where it reads the segment's WTC disregard (£889.20 against nil). With that one assertion dropped, the 2024 award passes, as the review found.

4. The body. It adds "Also fixes #2777." and replaces the old defined_for limitation sentence. The override-check invariant now says the check keeps the segment route for a mask, and each helper applies the condition. A new section, "Reform conditions in each segment", has the helper table, the worked lone parent and the CycleError cells. It also adds #2777's invariants (M1–M5), the masks-module rows and the revert row in the test plan, and 10 mutation rows labelled (#2777, at 2f087ae7c). I redid the overlap paragraph and added the new test to Reproducer:.

Tests at `2f087ae7c`

Check	Result
`test_housing_benefit_segment_masks.py -n 4`	30 passed
`test_housing_benefit_wtc_disregard.py -n 4`	55 passed
`test_housing_benefit_wtc_disregard_segments.py -n 4`	18 passed
`test_housing_benefit_accommodation_dates_properties.py`	6 passed
`test_housing_benefit_childcare_properties.py`	5 passed
`housing_benefit_wtc_disregard.yaml`	13 passed
Housing Benefit YAML directory	597 passed
`policyengine_uk/tests/code_health -n 4`	3,092 passed
`check_fix_pr.py` on the posted title and body, with this branch's changes and empty labels	PASS, 4 changed executable tests
`uvx ruff@0.17.2 format --check .`, `ruff check .`, `ruff check --select F811,F821 policyengine_uk`	all pass
`git diff --check 1d4a3e475..HEAD`	clean
`git merge-tree --write-tree --name-only HEAD origin/main`	no conflicts

policyengine_uk/tests/policy/baseline/gov/dwp/housing_benefit doesn't exist. The Housing Benefit YAML directory is policyengine_uk/tests/policy/baseline/finance/benefit/family/housing_benefit, and I ran that instead.

Mutants (#2777, at `2f087ae7c`)

Each mutant ran the segment-mask, segment and main modules with -n 4 and Hypothesis seed 0. All 10 were caught, each by the segment-mask module; the other two modules passed every time.

Mutation	Segment-mask module
The helper ignores every condition	17 failed
A failed condition gives nil, not the default	1 failed
The condition holds where it is `!= 0`, not `> 0`	1 failed
No accommodation condition in segments	5 failed
No earnings condition in segments	7 failed
No deduction condition on the fixed-limit path	2 failed
No limit condition	3 failed
No deduction condition	3 failed
No WTC condition	9 failed
No condition on segment income	4 failed

Open-PR overlap at `2f087ae7c`

Treat income derived from capital as capital in the legacy means tests #1995 (7381a46c7) conflicts in housing_benefit_applicable_income.py: both PRs extend the documentation string's last sentence. It conflicts the same way against 1d4a3e475. This is new against the old body, which said Git merges cleanly. Treat income derived from capital as capital in the legacy means tests #1995 also changes the tax that Housing Benefit deducts, so the lone parent's figures will move once it lands.
Value means-tested capital less 10% sale expenses and secured debt #1969, Fix HB and national CTR non-dependant deductions #2007, Give benefit-unit members aged 16 or over a bedroom under LHA rules #2010, Carry unused Income Tax trading losses forward once #2661, Bar income-related ESA at pensionable age and on Pension Credit #2807 and Count UC carer awards for severe disability and Scottish CTR premiums #2827 merge cleanly.
Give UK simulations a working subsample #2265 and Count only supplied inputs that are still stored in their tier #2308 conflict with main itself, in the same files.
Add the CTC qualifying young person period after leaving education #2301 and Exclude young Income Support claimants from PC, CB and CTC QYP status #2302 no longer edit this branch's files.

Skipped

test_housing_benefit_accommodation_dates.py (3,600 s cap): the body keeps its 6dc47664d row.
The 15,552-family differential: the body keeps its fdc13f128 figures. None of its populations adds a condition.
The Housing Benefit: reserve WTC for the Schedule 5 paragraph 56 disregard before childcare #2730 mutation table: kept at its earlier commits.
I loaded no survey data.
Title unchanged.
Aggregate-report figures are left for the hub.

Judgment calls

I built the review's "supplied" WTC and CTC as reported awards, because that reproduces every review figure and supplied awards don't. I didn't assert absolute household net income, since it depends on unrelated parts of the model. I asserted it against the neutralized control and the £577.98 gap instead.
I reported five CycleError cells as measured, not the brief's six.

Generated by Claude Code

MaxGhenis · 2026-11-03T07:42:51Z

Thanks for the lone parent case. At 2f087ae7c, each segment helper applies a reform's defined_for condition to the amount it rebuilds, the way policyengine-core masks the annual output. With your couples-only condition, Housing Benefit comes to £2,537.78 in 2026-27, as in your neutralized control and £577.98 below the model's own award, and household net income falls by the same amount. The 2024 one-segment case gives £1,113.09. test_a_couples_only_wtc_disregard_reaches_a_lone_parents_award checks both years through final Housing Benefit and household net income.

We reproduced your figures when we entered the £2,000 of WTC and £4,000 of CTC as reported awards (working_tax_credit_reported and child_tax_credit_reported), from which the model calculates WTC of £6,530.35 and CTC of £3,646.49. A replacement formula without a condition still takes the annual route in every segment. A reform that replaces the earnings disregard's formula and adds a condition that reads the WTC disregard, the limit or the accommodation disregard raises CycleError, because that condition then reads its own result. We list the five cases we found in the description.

juaristi22

Re-reviewed exact head 2f087ae7c349abcb7651e1b48b099230dd57f1d7. Review COMPLETE, with C1 resolved, no new findings and no material evidence gaps.

The segment helpers now apply the reform's defined_for mask. The coherent reported-award lone-parent regression reaches final Housing Benefit and household net income: the couples-only mask matches neutralisation (£2,537.78 HB in the 2026 tax-credits-active counterfactual), £577.98 below the unrestricted award. The 2024 control also passes. 127 focused checks passed in the matched locked environment; unchanged legal evidence and the current Axiom queue record were checked.

One qualification to my earlier wording: the original diagnostic mixed reported credits with supplied component awards, so it should not be read as certifying a £6,000 aggregate tax-credit award. The coherent reported-award regression independently supports the fix.

The full exact-head Python suite and Aggregate impact are still running, so I am withholding approval pending the applicable required checks. I will check CI again without repeating the completed review if the head remains unchanged.

MaxGhenis and others added 20 commits October 30, 2026 18:44

Name the segment tests for what they check

a07ed43

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Name every amount the full-charge comparison covers

44825de

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Cite DWP's guidance on which credits meet childcare charges

d1d8b37

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Merge origin/main into fix-2730-hb-wtc-disregard

97ada90

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Build the segment tests' previous-path systems on demand

9c2adcd

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Build the replacement test systems per test, not once per module

b25e79a

Each reform system holds about 38 MB, and caching all eleven kept the main test module above the 2 GB cap the mutation runs use. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Compute the model's outputs once per year in the replacement tests

6dc4766

Each computed replacement now runs one simulation against outputs the model computes once for 2024 and once for 2026-27. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

MaxGhenis mentioned this pull request Oct 31, 2026

Apply eligibility-only Housing Benefit reforms to each 2026-27 period #2777

Open

Release the Housing Benefit test modules' caches

1d4a3e4

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

MaxGhenis marked this pull request as ready for review November 1, 2026 01:03

MaxGhenis and others added 6 commits October 31, 2026 22:49

Tighten the hb_apply_defined_for docstring

90b4622

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

MaxGhenis and others added 2 commits November 1, 2026 05:25

Rewrap the segment-mask module docstring

2cafac7

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Finish the Housing Benefit helper docstrings

af4819b

Name the annual overrides each helper keeps, state the award floor in M2 and limit M4 to the generated cases. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

This was referenced Nov 1, 2026

Bar income-related ESA at pensionable age and on Pension Credit #2807

Draft

Count Maternity Allowance in the benefit cap #2811

Open

juaristi22 requested changes Nov 2, 2026

View reviewed changes

claude added 2 commits November 3, 2026 06:49

Merge main into the Housing Benefit WTC disregard

b8372df

MaxGhenis requested a review from juaristi22 November 3, 2026 07:42

juaristi22 reviewed Nov 3, 2026

View reviewed changes

Navigation Menu

Repository navigation

Disregard WTC before the Housing Benefit childcare deduction - #2776

Disregard WTC before the Housing Benefit childcare deduction#2776
MaxGhenis wants to merge 32 commits into
mainfrom
fix-2730-hb-wtc-disregard

MaxGhenis commented Oct 31, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Oct 31, 2026 •

edited

Loading

Uh oh!

juaristi22 left a comment

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

juaristi22 left a comment

Uh oh!

This branch has not been deployed

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

MaxGhenis commented Oct 31, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What and why

The regulations

Limit, deduction, WTC disregard

Supplied and replaced values

Accommodation segments

Reform conditions in each segment

Invariants

Test plan

Reproducer

Uh oh!

github-actions Bot commented Oct 31, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Aggregate impact on certified data: pass

Notes

Moves flagged for review

Spending by benefit and service

Caseloads

Standard reform costings (change in gov_balance)

Standard reforms: change in each reform's target

Household net income by decile

Uh oh!

juaristi22 left a comment

Choose a reason for hiding this comment

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

MaxGhenis commented Nov 3, 2026

Writer check

Findings

Tests at 2f087ae7c

Mutants (#2777, at 2f087ae7c)

Open-PR overlap at 2f087ae7c

Skipped

Judgment calls

Uh oh!

MaxGhenis commented Nov 3, 2026

Uh oh!

juaristi22 left a comment

Choose a reason for hiding this comment

Uh oh!

This branch has not been deployed

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

MaxGhenis commented Oct 31, 2026 •

edited

Loading

github-actions Bot commented Oct 31, 2026 •

edited

Loading

Tests at `2f087ae7c`

Mutants (#2777, at `2f087ae7c`)

Open-PR overlap at `2f087ae7c`