Young people and work review: an age test on the UC health element, a care-leaver input, and a never-worked flag #2888

Context

The independent review of young people and work publishes its final report in November 2026; its interim report came out on 28 May 2026. PolicyEngine/uk-young-people-work-review tracks what the model needs to score the likely recommendations. This issue collects the policyengine-uk side.

Data items belong to Microcosm, not here: limited capability for work-related activity from the FRS's linked UC status, full-time education, and youth calibration targets.

These are already in releases:

• the employer NICs zero rates for under-21s and apprentices under 25 (Zero-rate employer NI for under-21s and apprentices under 25 #2505, from 2.160.0);
• the April 2027 minimum wage rates (2.168.0);
• soc2020_major_group (Add soc2020_major_group, an input for the SOC 2020 major group of a person's main or last job #2856, from 2.173.0).

1. An age test on the UC health element

uc_LCWRA_element (variables/gov/dwp/universal_credit/disability_element/limited_work_ability/uc_LCWRA_element.py) has no age condition. The 2025 Pathways to Work green paper proposed delaying the health element until age 22, and the review may return to it (PolicyEngine/uk-young-people-work-review#2).

Proposal:

• A minimum age, for example gov.dwp.universal_credit.elements.disabled.minimum_age beside amount, at 0 in current law. Or a contributed block such as gov.contrib.dwp.uc_health_element_age_limit, with in_effect, age and applies_to_protected_claimants.
• uc_LCWRA_element counts a claimant's limited capability only at or above that age. A joint claim keeps the element if a member with limited capability is old enough.
• An option to exempt claimants who already receive the element, read from uc_receives_new_claimant_health_element.

The 2026–2032 rebalancing step (add_universal_credit_reform in scenarios/uc_reform.py) reads the formula before it sets the amounts: it takes has_health_element from the formula's result. So a gate inside the formula carries through the rebalancing years, and only changes to the amounts are shadowed (#2798). A test should pin that order.

Tests:

• a new claimant aged 21;
• a protected claimant aged 21;
• a new claimant aged 22;
• a couple with one member under the limit;
• a case in a rebalancing year.

2. Care leavers and the other shared accommodation rate exceptions

is_lha_shared_accommodation_rate_specified_renter (variables/gov/dwp/is_lha_shared_accommodation_rate_specified_renter.py) records that the other exceptions are not observed. Those are care leavers, hostel residents, MAPPA cases, people who have experienced domestic abuse or modern slavery, and foster parents. The only care-leaver input is the 16 to 19 Bursary Fund's (bursary_fund_16_to_19_in_care_or_care_leaver).

Proposal:

• A person input is_care_leaver, default false.
• Read it in the UC specified-renter test, and in Housing Benefit's shared accommodation test, which reads gov.dwp.LHA.shared_accommodation_age_threshold.
• Give each scheme's own age limit for care leavers, cited to its regulations.

No survey records care history, so datasets won't fill the input. It serves household calculations, and reforms that set it, such as an exemption the review may propose (PolicyEngine/uk-young-people-work-review#5).

3. A never-worked flag, derived (optional)

The review may target young people who have never worked (PolicyEngine/uk-young-people-work-review#4). The FRS asks the job questions of everyone who has ever had paid work, so Microcosm writes soc2020_major_group as NONE for adults with no main or last job.

A formula variable has_ever_worked, true when soc2020_major_group is not NONE, would name that selector. As a formula it needs no dataset input. Its documentation should say that the FRS question leaves out casual and holiday work.

4. A UC work allowance by age (only if the report proposes one)

The readiness checklist (PolicyEngine/uk-young-people-work-review#9) lists a youth work allowance. uc_work_allowance doesn't vary by age (its parameters were consolidated in #1651). A contributed amount by claimant age would let a reform set one.

Relationship to data work

Microcosm's data work for this event is still being scoped, and none of it is decided. If any goes ahead:

• Item 2. A care-leaver imputation would need item 2 first, because a dataset can write is_care_leaver only once the input exists. No survey asks about care history, so it would rest on DfE's care leaver statistics.

• Item 1. It acts on whoever the data flags with limited capability for work-related activity. That holds for today's disability-benefit proxy, and for the FRS's admin-linked UC status if Microcosm switches to it. Neither change waits for the other.

• Items 3 and 4. Neither needs data work: soc2020_major_group is already in the data.

• The rest of the event's data work needs nothing from this issue. That is:

  • youth targets (NEET, and UC claimants by age and conditionality regime);
  • a full-time education flag, since is_in_full_time_education already exists;
  • the base of employees under 21.

  The facts it would use are requested in UK facts for four November events: young people and work, the PIP review, the January 2027 price cap and the Scottish Budget 2027-28 chronicle#420.

Related

• Values the model stores at build don't follow inputs a caller sets later #2798: values the model stores at build don't follow inputs set later.
• Apply SI 2014/1230 reg 28 to the UC qualifying young person test #2845 and Apply SI 2014/1230 reg 28 in is_qualifying_young_person_for_universal_credit #2804: the UC qualifying young person test (SI 2014/1230 reg. 28).
• Narrow the reg. 60(1) education protection to claimants who failed the condition on the migration day #2878: the reg. 60(1) education protection.
• Apply each scheme's own non-dependant test to the LHA shared accommodation rate #2030: the shared accommodation rate's non-dependant test.