Implement Utah Child Care Assistance Program (CCAP)

[hua7450]

Summary

Implements Utah Child Care Assistance (CC) — the state's CCDF child care subsidy (Utah DWS Office of Child Care, Utah Admin. Code R986-700) — in PolicyEngine. The program is modeled as the Employment Support (ES) Child Care subsidy: a per-child market-rate reimbursement cap (Table 3) reduced by a sliding-scale family copayment (Table 4), gated by an 85%-of-state-median-income test and the federal CCDF asset limit.

Closes #9061

Regulatory Authority

• Utah Admin. Code R986-700 — Child Care Assistance (full rule; HTML): https://utrules.elaws.us/uac/r986-700
  • Sections used: -702 (child age, immigration), -707 (copayment), -709 (ES CC activity hours), -710 (income/asset limits, countable income, deductions), -713 (amount of payment, lower-of rule), -742 (Enhanced Subsidy Grant — not modeled).
  • Cornell LII per-section mirror: https://www.law.cornell.edu/regulations/utah/Utah-Admin-Code-R986-700-710
• Utah DWS Table 3 — Maximum Monthly Child Care Payments (local market-rate caps; PDF, single page — no page anchor), effective October 1, 2024: https://jobs.utah.gov/occ/provider/table30824.pdf
• Utah DWS Table 4 — Child Care Income Eligibility and Co-Payment (HTML), effective April 1, 2026: https://jobs.utah.gov/Infosource/eligibilitymanual/Tables,_Appendicies,_and_Charts/Tables,_Appendicies,_and_Charts/Table_4_-_Child_Care_Income_Eligibility_and_Co-Payment.htm
• Utah DWS Office of Child Care subsidy fact sheet (PDF, 2 pages; footer OCC_2_09/24/25), income limits effective October 1, 2025: https://jobs.utah.gov/customereducation/services/childcare/occsubsidyfact.pdf#page=1
• 45 CFR 98.20 (federal CCDF asset/eligibility limit): https://www.law.cornell.edu/cfr/text/45/98.20

Income Eligibility Tests

Test	How modeled	Source
Countable income ≤ 85% of state median income for household size	ut_ccap_income_eligible compares ut_ccap_countable_income against smi_rate (0.85) × hhs_smi. Design decision: the eligibility ceiling is computed dynamically from the existing hhs_smi variable rather than reading the encoded Table 4 Group-16 dollar amounts, so the ceiling tracks annual SMI updates. The copay income-group classification (REQ-024) still uses the encoded Table 4 income-group upper bounds — only the eligibility ceiling is dynamic.	R986-700-710(3); fact sheet; Table 4
Family assets ≤ CCDF limit ($1,000,000)	Reuses federal is_ccdf_asset_eligible (no state parameter)	R986-700-710; 45 CFR 98.20
Child under age 13 (under 18 if special needs)	ut_ccap_eligible_child with child_age_limit (13) and disabled_child_age_limit (18), using is_disabled as the special-needs proxy	R986-700-702; fact sheet
Child is a U.S. citizen or authorized non-citizen	ut_ccap_eligible_child reuses federal is_ccdf_immigration_eligible_child; parent SSN/immigration status is not gated	R986-700-702; fact sheet
Work/activity hours: single parent ≥15 hr/wk; two-parent one ≥15 and other ≥30	ut_ccap_activity_eligible with single_parent_min_hours (15), two_parent_first_min_hours (15), two_parent_second_min_hours (30)	R986-700-709; fact sheet

Income Deductions & Exemptions

Countable income (ut_ccap_countable_income) starts from ut_ccap_gross_income — an adds-list of earned + unearned parent income defined in income/sources.yaml (R986-700-710) — then applies the following, all read at the monthly period:

Item	How modeled	Source
Count earned + unearned income of parent(s); both parents count	income/sources.yaml adds-list summed at SPMUnit	R986-700-710
Child support paid directly to a parent = child's unearned income	child_support_received included in sources.yaml	R986-700-710
Exclude SSI benefits	ssi omitted from sources.yaml (never added)	R986-700-710
Exclude earned income of a child who is not a parent	Subtract minor (non-parent) earned income below child_earned_income_exclusion_age (18)	R986-700-710
Deduct first $50 of child support received/month	deductions/child_support_received_deduction (50)	R986-700-710
Deduct court-ordered child support and alimony paid out	child_support_expense + alimony_expense (full deduction, no parameter)	R986-700-710
Deduct $100 per person with countable earned income	deductions/earned_income_deduction (100) × number of earners	R986-700-710
Deduct $100 automatic medical deduction	deductions/medical_deduction (100) per SPM unit	R986-700-710

Benefit Calculation

ut_ccap (SPMUnit, MONTH) implements the R986-700-713 lower-of rule reduced by the Table 4 copayment:

ut_ccap = max_(min_(billed childcare expenses, Σ per-child market-rate caps) − copay, 0)

• Per-child market-rate cap (Table 3): rates/market_rate.yaml is a breakdown keyed by ut_ccap_provider_type (5 provider types) × ut_ccap_age_group (6 age bands, derived from age-in-months via rates/age_group.yaml). ut_ccap_market_rate looks up the monthly cap per child.
• Lower-of rule (simplified): family subsidy = min_(spm_unit_pre_subsidy_childcare_expenses, Σ per-child caps). Billed expenses proxy the provider's actual established rate; the separate hourly-unit-cost × approved-hours path is not modeled (documented simplification, per scope Key Decision 5). Per-child caps are pooled to the SPM unit.
• Sliding-scale copayment (Table 4): ut_ccap_copay looks up copay/amount.yaml, keyed by household size (2 through 12+, bounded by copay/min_household_size and copay/max_household_size) × income group (ut_ccap_income_group, 16 brackets classified against copay/income_limits.yaml) × child-count group (ut_ccap_child_count_group: 1 / 2 / >2 children in care). The full published dollar table is encoded directly (interior bracket boundaries are irregular and not derived).
• Group 1 = $0 copay: the lowest income group (upper bound ≈ 100% FPG monthly) is encoded as $0 in copay/amount.yaml.
• Copay exceeds cost of care ⇒ ineligible: handled naturally by the max_(… − copay, 0) floor and reflected in ut_ccap_eligible via a copay-affordability check.

The state YEAR aggregator ut_child_care_subsidies (adds = ["ut_ccap"], defined_for = StateCode.UT) feeds the federal child_care_subsidies, and is registered in child_care_subsidy_programs.yaml.

Requirements Coverage

All 21 in-scope requirements are implemented with parameter, variable, and test coverage.

REQ	Description	Param	Variable	Test
REQ-001	Child under 13 (under 18 if special needs)	eligibility/child_age_limit (13), disabled_child_age_limit (18)	ut_ccap_eligible_child	ut_ccap_eligible_child 1-6
REQ-004	Child citizen / authorized non-citizen (reuse)	— (reuse)	ut_ccap_eligible_child → is_ccdf_immigration_eligible_child	ut_ccap_eligible_child 7-9
REQ-006	Activity hours: single ≥15; two-parent 15 & 30	single_parent_min_hours (15), two_parent_first_min_hours (15), two_parent_second_min_hours (30)	ut_ccap_activity_eligible	ut_ccap_activity_eligible 1-7
REQ-009	Countable income ≤ 85% SMI (dynamic)	income/smi_rate (0.85)	ut_ccap_income_eligible	ut_ccap_income_eligible 1-4 + integration
REQ-010	Assets ≤ CCDF $1,000,000 (reuse)	— (reuse)	ut_ccap_eligible → is_ccdf_asset_eligible	ut_ccap_eligible 4
REQ-012	Count earned + unearned of both parents	income/sources.yaml	ut_ccap_gross_income, ut_ccap_countable_income	ut_ccap_countable_income 1, 2, 10
REQ-013	Child support to parent = child's unearned income	income/sources.yaml (child_support_received)	ut_ccap_gross_income	ut_ccap_countable_income 5, 6, 14
REQ-014	Exclude SSI benefits	income/sources.yaml (omits ssi)	ut_ccap_gross_income	ut_ccap_countable_income 8 + integration 5
REQ-015	Exclude non-parent minor's earned income	income/child_earned_income_exclusion_age (18)	ut_ccap_countable_income	ut_ccap_countable_income 3, 4, 11
REQ-017	Deduct first $50 child support received	deductions/child_support_received_deduction (50)	ut_ccap_countable_income	ut_ccap_countable_income 5, 6, 14
REQ-018	Deduct child support & alimony paid out	— (child_support_expense + alimony_expense)	ut_ccap_countable_income	ut_ccap_countable_income 7, 12
REQ-019	Deduct $100 per earner	deductions/earned_income_deduction (100)	ut_ccap_countable_income	ut_ccap_countable_income 1, 2, 13
REQ-020	Deduct $100 automatic medical	deductions/medical_deduction (100)	ut_ccap_countable_income	all ut_ccap_countable_income cases
REQ-021	Per-child subsidy = lower of (cap, billed expenses)	—	ut_ccap	ut_ccap 1, 2
REQ-022	Market-rate cap by provider type (5) × age band (6)	rates/market_rate.yaml, rates/age_group.yaml	ut_ccap_market_rate, ut_ccap_provider_type, ut_ccap_age_group	ut_ccap_market_rate (all 5×6), ut_ccap_age_group
REQ-023	Copay subtracted from standard subsidy	—	ut_ccap	ut_ccap 1, 2, 6, 8
REQ-024	Copay sliding scale: size (2-12+) × income group (16) × #children (1/2/>2)	copay/amount.yaml, copay/income_limits.yaml, copay/min_household_size, copay/max_household_size	ut_ccap_copay, ut_ccap_income_group, ut_ccap_child_count_group	ut_ccap_copay 1-10, ut_ccap_income_group 1-12, ut_ccap_child_count_group 1-6
REQ-025	Group 1 = $0 copay	copay/amount.yaml (group 1 = 0)	ut_ccap_copay	ut_ccap_copay 1 + integration 3, 9, 10
REQ-026	Copay > cost of care ⇒ ineligible	—	ut_ccap_eligible (copay-affordable check)	ut_ccap_eligible 7, 8; ut_ccap 3, 4, 7; integration 4
REQ-028	Tiered/formula benefit, monthly units	—	ut_ccap (MONTH, formula)	ut_ccap (structural)
REQ-029	YEAR aggregator + registry registration	—	ut_child_care_subsidies (adds ut_ccap); registered in child_care_subsidy_programs.yaml	ut_child_care_subsidies 1, 2

Not Modeled

REQ	What	Why excluded
REQ-002	Child needs ≥8 hours of care/month	Care-need hours are not in the model; assume any family with billed care meets it
REQ-003	Applicant is parent / foster / relative / guardian	No PE flag distinguishes relative/guardian caretaker status; assumed satisfied for the SPM unit
REQ-005	Parent SSN / work-authorization not required	Informational — the formula intentionally never gates on parent immigration or SSN; no code needed
REQ-007	Net income ≥ minimum wage for hours worked	Verification rule not simulatable without per-parent wage/hours
REQ-008	Education/training pathway (24-month limit)	Not tracked in the model; ES-CC working pathway is the modeled scope
REQ-011	12-month certification period	Informational — PE computes a snapshot month; no monthly re-test logic needed
REQ-016	Exclude DHHS independent-living grants to minor parents	No corresponding PE income variable to exclude
REQ-027	Enhanced Subsidy Grant (CCQS quality supplement, R986-700-742)	Deferred by design; no quality-tier input in the model

Historical Notes

• Parameter vintages and backdating. Table 3 (rates/market_rate.yaml) is effective October 1, 2024; Table 4 (copay/amount.yaml, copay/income_limits.yaml) is effective April 1, 2026; the 85% SMI fact-sheet income limits are effective October 1, 2025. Every parameter entry is placed at a single 2020-01-01 date — PolicyEngine's backdating floor — as a documented approximation. Earlier editions of Tables 3 and 4 are not encoded, so periods before their true effective dates use the current tables. Each file's header comment documents this.
• Documented source typo. In Table 4, the published lower bound for household size 5, group 3 reads 3,644.01, an apparent typo for 3,664.01 (the size 5, group 2 upper bound is 3,664). Only upper bounds are encoded in income_limits.yaml, so the typo has no effect on the model, but it is documented in that file's header comment for future maintainers.
• Identical provider rows. In Table 3, the License-Exempt home (family, friend, and neighbor) and Licensed Commercial Preschool rows publish identical rates (495/443/437/416/413/385). They are kept as distinct ut_ccap_provider_type enum values because the source table lists them as separate provider categories.

Files Added

policyengine_us/parameters/gov/states/ut/dwf/ccap/
  copay/amount.yaml
  copay/income_limits.yaml
  copay/max_household_size.yaml
  copay/min_household_size.yaml
  eligibility/child_age_limit.yaml
  eligibility/disabled_child_age_limit.yaml
  eligibility/single_parent_min_hours.yaml
  eligibility/two_parent_first_min_hours.yaml
  eligibility/two_parent_second_min_hours.yaml
  income/child_earned_income_exclusion_age.yaml
  income/deductions/child_support_received_deduction.yaml
  income/deductions/earned_income_deduction.yaml
  income/deductions/medical_deduction.yaml
  income/smi_rate.yaml
  income/sources.yaml
  rates/age_group.yaml
  rates/market_rate.yaml

policyengine_us/variables/gov/states/ut/dwf/ccap/
  ut_ccap.py
  ut_child_care_subsidies.py
  copay/ut_ccap_child_count_group.py
  copay/ut_ccap_copay.py
  copay/ut_ccap_income_group.py
  eligibility/ut_ccap_activity_eligible.py
  eligibility/ut_ccap_eligible.py
  eligibility/ut_ccap_eligible_child.py
  eligibility/ut_ccap_income_eligible.py
  income/ut_ccap_countable_income.py
  income/ut_ccap_gross_income.py
  rates/ut_ccap_age_group.py
  rates/ut_ccap_market_rate.py
  rates/ut_ccap_provider_type.py

policyengine_us/tests/policy/baseline/gov/states/ut/dwf/ccap/
  integration.yaml
  ut_ccap.yaml
  ut_child_care_subsidies.yaml
  copay/ut_ccap_child_count_group.yaml
  copay/ut_ccap_copay.yaml
  copay/ut_ccap_income_group.yaml
  eligibility/ut_ccap_activity_eligible.yaml
  eligibility/ut_ccap_eligible.yaml
  eligibility/ut_ccap_eligible_child.yaml
  eligibility/ut_ccap_income_eligible.yaml
  income/ut_ccap_countable_income.yaml
  rates/ut_ccap_age_group.yaml
  rates/ut_ccap_market_rate.yaml

Also modified: parameters/gov/hhs/ccdf/child_care_subsidy_programs.yaml (registers ut_child_care_subsidies), programs.yaml (adds UT to CCDF coverage and a Utah CCAP state implementation entry), and a changelog.d/ut-ccap.added.md fragment.

Test plan

• 113/113 tests pass locally (13 YAML files)
• CI passes

[DTrim99]

Program Review — Utah Child Care Assistance (new program)

PR #9063 — Utah Child Care Assistance (CCAP / Employment Support Child Care, CCDF). NEW state program. Consolidated from five independent read-only review passes (regulatory, Table 3/4 value audit, code pattern audit, test coverage, PR context).

Scope

Implements Utah's Child Care Assistance Program (CCAP) as a new state benefit under the federal CCDF child-care-subsidy aggregator:

• Eligibility — eligible child (under 13; up to 18 special-needs), 85% SMI net-income test, CCDF asset limit (reused), activity/work-hours test, and a copay-affordability gate (family ineligible if copay exceeds cost of care).
• Market-rate cap (Table 3) — 5 provider types × 6 age bands.
• Copay (Table 4) — copay amount + income-limit schedule across 11 household sizes × 16 income groups × 3 child-count columns.
• Subsidy + aggregation — ut_ccap = Σ_child min(expense, market_rate) − copay, floored at 0 (MONTH/SPMUnit) → ut_child_care_subsidies (YEAR wrapper, adds = ["ut_ccap"]) → federal child_care_subsidies.
• 15 new variables (13 with formulas + 2 input Enums), full parameter tree under gov.states.ut.dwf.ccap.*, plus programs.yaml + child_care_subsidy_programs.yaml wiring.

CI: 31/31 passing.

Branch status: 7 ahead / 425 behind main — VERY STALE. This is the one operational blocker: the branch must be rebased on main and re-tested before merge; merge conflicts are likely at 425 commits of drift.

Sources: Utah Admin. Code R986-700 (sections -702, -707, -709, -710, -713; law.cornell.edu), Table 3 market-rate PDF (table30824.pdf, eff. Oct 1 2024), Table 4 copay/income-limit HTM (eff. Apr 1 2026), and the DWS OCC subsidy fact sheet PDF (85% SMI). All three data sources confirmed reachable.

Verified correct

Rate/copay value audit — SAMPLE, 38 cells checked, 0 mismatch.

• Table 3: 17 market-rate cells across the 5 provider types × 6 age bands — all exact (full extracted grid legible; 15+ spot-checked to the cent).
• Table 4: 11 income-limit upper-bound cells + 10 copay rows — all exact.
• License-Exempt Home == Licensed Commercial Preschool equivalence is genuinely in the source — the PDF prints both rows with identical values (495/443/437/416/413/385). The PR encodes them as two separate provider blocks with identical values and documents it. Faithful, not an invented equivalence.
• The documented 3,644.01→3,664.01 is a REAL source typo, correctly handled — the transposition lives in a group lower bound; income_limits.yaml encodes only upper bounds, so it never enters the parameters. The PR's comment documents it accurately and the encoding is unaffected. No action needed.
• n/a-padding is sound — structurally-impossible child-count columns (size-2 can't have 2+ children; size-3 can't have 3+) are padded by repeating the largest published column; the child_count_group/spm_unit_size clamps mean the padded cells can never bind.
• 85% SMI confirmed verbatim on the DWS fact sheet — "household size is set at 85% of the state median income effective Oct. 1, 2025," corroborating income/smi_rate.yaml = 0.85. The reg (R986-700-710(6)(b)) delegates the percentage to the Department; citing both the fact sheet and the reg is the correct sourcing.

Regulatory confirmations (researched independently, then compared). Every material rule matches R986-700: net-countable ≤ %-of-SMI income test (710(6)(b), correctly on post-deduction income); the four deductions ($50 child support received, support/alimony paid, $100/earner, $100 automatic medical) (710(5)); copay-exceeds-cost ineligibility verbatim (707(3)); Group-1 (≤100% FPL) $0 copay (707(5)(a)); 15 / 30-15 activity hours (709); lesser-of market-rate payment (713); under-13 / under-18-special-needs child ages (702). Asset limit reuses is_ccdf_asset_eligible; immigration reuses is_ccdf_immigration_eligible_child; activity reuses meets_ccdf_activity_test — none reinvented.

Code cleanliness. No hard-coded policy values (every literal is 0/1/2/np.inf; all magnitudes are parameters). Breakdown metadata complete and correct (30 market-rate combos; 11×16×3 copay combos; integer keys via range(), string keys match enums). Entity/period/aggregation chain correct (MONTH ut_ccap → YEAR wrapper → federal roll-up; Person/YEAR expense read at MONTH auto-÷12 to match the monthly cap). All CCDF helpers reused not reinvented. Copay used once as an affordability gate and subtracted once from the subsidy — not double-counted. income_group floored to 1 outside UT; household size clamped. defined_for = StateCode.UT throughout. Changelog + programs.yaml present. Backdated cleanly to a single 2020-01-01 floor (no mid-series toggles).

Test thoroughness. All 13 formula-bearing variables have dedicated would-fail-if-broken tests (plus the 2 input Enums exercised as inputs); no zero-coverage formula. Every core mechanic pinned: 85% SMI boundary (below/at/above/zero + dynamic per-size ceiling in integration), market-rate provider×age lookup, copay schedule (group-1 zero, mid, top, size clamp, group-16 overflow, single-adult floor), all five eligibility gates individually, the copay≤cost-of-care boundary (Cases 7/8), per-child lower-of with per-child capping (strong regression pin against pooled-cap bugs), copay floor at 0 while still eligible, and end-to-end per-child→unit→YEAR→federal roll-up asserting the federal total. No invalid test periods; no non-functional boolean margins.

Critical (must fix)

None.

Should address

1. Factor the duplicated client-identification block into a shared helper. income/ut_ccap_gross_income.py:47-52, income/ut_ccap_countable_income.py:21-26, and eligibility/ut_ccap_activity_eligible.py:33-36 repeat verbatim:

   is_parent = person("is_parent", period.this_year)
   is_head_or_spouse = person("is_tax_unit_head_or_spouse", period.this_year)
   no_parent_present = spm_unit.sum(is_parent) == 0
   is_client = is_parent | (spm_unit.project(no_parent_present) & is_head_or_spouse)

   The scoping is correctly narrow (person-level is_parent; the head/spouse fallback fires only when no parent is present) — not an over-broad any(), and it is regulatorily defensible per R986-700-710(4). The issue is drift risk from verbatim triplication: extract a person-level ut_ccap_is_client helper. (Combines code-audit S + regulatory G4 — same client-scoping item.)

2. Independently corroborate 85% SMI via the Group-16 cross-check. Sourced value is confirmed verbatim on the fact sheet (see Verified correct), but the YAML comment's claim that "Table 4 Group-16 upper bounds equal 85% of SMI by household size" is worth a one-line confirmation: Group-16 upper bound ÷ 0.85 should reproduce HHS/Utah SMI by household size. income/smi_rate.yaml:10, ut_ccap_income_eligible. (Regulatory S1.)

3. Confirm the copay-affordability gate is intentionally on uncapped unit cost. ut_ccap_eligible gates on whole-unit spm_unit_pre_subsidy_childcare_expenses (uncapped) while the benefit builds from per-child min(expense, market_rate). R986-700-707(3) speaks to "the actual cost of child care" (unit-level), so gating on uncapped cost is defensible; the mild inconsistency can let a unit be "eligible" yet receive a benefit floored to 0. Add a one-line docstring note so future editors don't "fix" the apparent divergence. ut_ccap.py:22, ut_ccap_eligible.py:20-22. (Regulatory S3 + code-audit suggestion, deduped.)

4. Note the bare-input-expense population effect. The copay-affordability gate depends on spm_unit_pre_subsidy_childcare_expenses, a bare input defaulting to 0 in microsim unless imputed. Where expense = 0 the program is inert (benefit = min(expense, cap) − copay, floored). This is the established convention across every state CCAP, NOT a UT defect and not a code change here — record it as a population-level limitation (remedy is dataset imputation). (Regulatory S2.)

5. Add a paired test asserting License-Exempt-Home == Commercial-Preschool at the same age band. Both rows are priced identically, but tests pin them only at different bands (LICENSE_EXEMPT_HOME under_two=495/school_age=385; COMMERCIAL_PRESCHOOL four_year=416). A one-line divergence in one row at an untested band would pass CI. Add a paired case at one shared band. rates/ut_ccap_market_rate.yaml. (Test MODERATE 1.)

6. Force a size-2 n/a-padded copay cell to a TWO/THREE_PLUS child-count group. ut_ccap_copay.yaml reaches size-2 only via the one-child column; no case forces a padded TWO/THREE_PLUS cell to confirm it returns the same value. Low risk (padding never binds), but the would-fail-if-broken test is absent. copay/ut_ccap_copay.yaml. (Test MODERATE 2.)

Suggestions

1. Masked-multiply style — ut_ccap_gross_income.py:57-58 and ut_ccap.py:29 use value * bool_mask inside spm_unit.sum(...). Correct and vectorized; house style prefers spm_unit.sum(where(mask, value, 0)).
2. copay/ut_ccap_income_group.py:15-16 reaches into private parameter internals (next(iter(p.income_limits._children.values())), len(size_node._children)) to derive the group count; a group_count parameter or reused range(1, 17) is more robust to core changes.
3. copay/ut_ccap_income_group.py:21 — np.ones_like(size) on a possibly-float spm_unit_size; np.ones(len(size), dtype=int) is clearer intent.
4. ut_ccap_activity_eligible.py — reference is a bare string in parentheses (a plain str, not a 1-tuple); visually resembles a tuple missing its trailing comma. Cosmetic.
5. Pre-2026 copay values are not period-accurate — Table 4 (eff. Apr 1 2026) is backdated to 2020, so 2020–2025 copay/income-limit computations use forward-dated values. Intentional per the scope note; flagged for awareness. (copay/*; also rates/* from Table 3 eff. Oct 1 2024.)
6. Activity two-parent "second" (15h) boundary tested only from below (Case 3 = 30 & 15 pass; Case 5 = 29 & 29 fail). No case at highest=30, lowest=14 isolating the second bar binding independently when the first is exactly met. Minor; individual thresholds pinned elsewhere. (Test MODERATE 3 / polish.)
7. Documented modeling limitations (all disclosed in code comments, no action needed): activity min-wage-for-hours and education/training pathways only partially modeled; special-needs proxied by is_disabled, court-supervision and 8-hours/month-need conditions not tracked. (Regulatory G1/G2.)
8. Optional test hardening: no case varies the monthly benefit across months within the year for the annual roll-up. (Test LOW.)

Caveats

• The rate/copay audit is a SAMPLE (38 cells / 0 mismatch), not a full cell-by-cell verification. The full extracted Table 3 grid was legible and structure confirmed; Table 4 was sampled repeatedly via WebFetch. Confidence is high but the audit did not exhaustively check every one of the ~500 copay cells.
• The branch is 425 commits behind main — VERY STALE. It MUST be rebased on main and the full test suite re-run before merge. At 425 commits of drift, merge conflicts are likely (especially in programs.yaml and child_care_subsidy_programs.yaml, both high-churn shared files). CI's current 31/31 is against a stale base and does not guarantee a clean merge. This is the single operational item standing between the PR and merge.

Validation summary

Dimension	Result
Regulatory (R986-700)	0 critical; faithful to -702/-707/-709/-710/-713; 3 should (all verification/confirmation asks), 5 suggestion
Rate/copay value audit (SAMPLE)	38 cells / 0 mismatch; typo handled correctly; n/a-padding sound; 85% SMI verbatim; License-Exempt==Commercial equivalence genuine
Code pattern audit	0 critical; no hard-codes; breakdown/entity/period/reuse all correct; 3 should, 5 suggestion
Test coverage	0 critical gaps; all 13 formula vars + core mechanics pinned; 3 moderate (2 equivalence/padding + 1 activity boundary)
CI	31/31 passing (against a stale base — see Caveats)

Review severity: APPROVE

0 criticals across all five passes, the rate/copay sample is 0-mismatch, and the three things a reviewer would most suspect (the 3,644.01 typo, the n/a-padding, the 85% SMI figure) are each independently confirmed correct — the only operational item to close is the 425-behind staleness, resolved by a rebase + re-test, not a code fix.

Next steps

• /fix-pr 9063 — extract the shared ut_ccap_is_client helper (should-address 1), add the two moderate tests (should-address 5, 6), add the docstring/comment notes (should-address 3, 4), and record the 85% SMI cross-check (should-address 2).
• Rebase on main (425 behind) and re-run the full YAML/test suite before marking ready — resolve conflicts in programs.yaml and child_care_subsidy_programs.yaml.

🤖 Generated by /review-program (Claude Code)

[DTrim99]

Follow-up — review items verified at 2c116a2965

Re-checked the two fix commits (987666fb14, 2c116a2965) against the review above. All 6 should-address items are resolved, and the 425-behind staleness — the one operational blocker — is fully cleared (new head is 0 behind / 10 ahead of main).

• Client-identification helper (S1) — the triplicated is_parent | head-or-spouse block is now a single ut_ccap_is_client variable (Person/YEAR) consumed by ut_ccap_gross_income, ut_ccap_countable_income, and ut_ccap_activity_eligible, with the verbatim block deleted from each. The entity-projection fix in 2c116a2965 is correct — person.spm_unit.sum(is_parent) already broadcasts the unit total back to members, so the earlier spm_unit.project() inside a Person formula was redundant; the head-or-spouse cross-tax-unit fallback is preserved.
• 85% SMI cross-check (S2) — smi_rate.yaml comment now shows each Table 4 Group-16 upper bound ÷ 0.85 reproducing FFY2026 Utah SMI across sizes 2–12.
• Uncapped-cost gate (S3) and bare-input-expense population note (S4) — both documented in the ut_ccap_eligible docstring.
• License-Exempt == Commercial-Preschool at same age band (S5) — new ut_ccap_market_rate.yaml Case 14 asserts [437, 437] for two age-3 children across the two provider blocks — a genuine one-sided-edit-fails pin.
• Size-2 n/a-padded copay cell (S6) — new Case 11 forces ut_ccap_child_count_group: TWO and confirms the padded column repeats the one-child value (18).

The 8 suggestions were all optional (2 accept-by-design, the rest style/hardening) and left as-is — none blocking. You also pinned explicit work hours across the suite. Clean from this review's perspective. 👍

🤖 Generated by /review-program re-verification (Claude Code)