Repository navigation
From 2030-31 every UC LCWRA claimant gets the new-claimant amount; the protected amount stops with the 2026-29 modifier #2282
Description
What's wrong
From 2030-31 every Universal Credit LCWRA claimant gets the new-claimant amount (elements.disabled.amount), including pre-2026, severe conditions criteria and terminally ill claimants, who are paid the protected amount in law.
universal_credit_july_2025_reform (scenarios/uc_reform.py) sets the health element only for 2026 to 2029 (for year in range(2026, 2030)). After that, uc_LCWRA_element's formula pays elements.disabled.amount to everyone.
Reproducer (main 9475c95a)
A single claimant aged 30 with uc_limited_capability_for_WRA: true, in a situation (no claim-status input, so the protected amount applies):
| Tax year | Monthly LCWRA element |
|---|---|
| 2029-30 | £457.15 (protected amount, s. 4 floor) |
| 2030-31 | £235.86 (new-claimant amount, CPI-uprated on main) |
With #2284 (UC Act 2025 s. 3 freeze) the 2030-31 figure becomes £221.60, still the new-claimant amount.
#2248 noted this ("After 2029-30, the modifier stops and every claimant falls to elements.disabled.amount") and left it out of scope.
The law
- The UC Regs 2013 reg. 36 row for pre-2026, severe conditions criteria and terminally ill claimants has no end date. UC Act 2025 s. 2(1)(b) only relabels it.
- s. 3(1) disapplies SSAA 1992 s. 150(1) and (2) for the LCWRA element (both rows, reg. 27) only in the tax years ending 5 April 2026 to 5 April 2029. The review in 2029-30 governs the April 2030 rates again (s. 150(10)).
- s. 4 requires increases to the protected amount only for 2026-27 to 2029-30.
Expected
- From 2030-31 a protected claimant keeps the protected amount, uprated from its 2029-30 level under s. 150. That uprating is discretionary for UC amounts (s. 150(2)(b)); the model's convention is the benefit uprating index.
- A new claimant keeps
elements.disabled.amount. - The new-claimant share of claimants also needs a value for 2030 on (
post_2025_claimant_sharestops at 2029).
gov.dwp.universal_credit.elements.disabled.uprating (added by that fix) is the benefit index held through the s. 3 years. Uprating the 2029-30 protected amount by it from 2030-31 gives the s. 150 path.
Impact
This affects every 2030-31+ projection: UC spending, poverty and household incomes for LCWRA claimants.
From the main run behind #2284's impact table (enhanced_frs_2024_25.h5@1.58.0, main 9475c95a, a full microsimulation): uc_LCWRA_element, summed over every benefit unit with LCWRA, is £32.3bn in 2029-30 and £20.1bn in 2030-31, and universal_credit falls from £87.6bn to £84.5bn. That run does not isolate this bug, since other things change between the two years too; the size of the fix needs its own run.
Activity
MaxGhenis commentedon Oct 15, 2026
Spec for the fix (UK hub, 2026-10-15). The UC Act 2025 sections, read today at https://www.legislation.gov.uk/ukpga/2025/22, keep both reg. 36 rows past 2029-30:
The model's 2030 drop is therefore a bug against the law. A fix that corrects the model to the law merges on the usual gates, with its real-data impact stated in the PR.
MOCK reproduction on 2.130.1 (one LCWRA claimant aged 35 with no income, monthly element):
What the PR should do
policyengine_uk.data.economic_assumptions.PROJECTION_END_YEAR(2032 today). Give the years one named source, a constant or a parameter, that other code can read. The Scenario.from_reform(dict).simulation_modifier silently drops statutory changes on a built simulation #2426 guard reads it for its warning.universal_credit,uc_LCWRA_element, and the counts of protected and new claimants.Name the scenario's years; protected amount never below new-claimant …
Read the LCWRA uprating index at its new path
Document reforms to the index; correct the scenario's stated years
Limit the single-day reform caveat to non-calendar-year parameters