Add disabled adult dependent care expenses to the CDCC base

[hua7450]

Closes #8786

Summary

A small, self-contained fix for #8786: the Child and Dependent Care Credit drew care expenses only from tax_unit_childcare_expenses, which is child-scoped (children under 18). A disabled adult qualifying individual under IRC § 21 — a dependent or spouse age 18 or older — was therefore attributed $0 of care expenses and received a $0 credit. Because ks_cdcc is a share of the federal credit, Kansas inherited the same gap.

Approach

Rather than broadening the childcare pipeline (which would ripple into ~40 CCDF/state CCAP programs, SNAP/TANF dependent-care deductions, and the DC KCCATC), this routes adult dependent care through a separate, dedicated input and leaves childcare untouched:

• care_expenses / pre_subsidy_care_expenses (pre-existing, previously-unused person variables) now represent care for a disabled adult dependent or spouse, redefined as YEAR to match the rest of the CDCC chain.
• cdcc_relevant_expenses adds care_expenses to tax_unit_childcare_expenses before the § 21(c) dollar cap and the earned-income cap: min(childcare + adult care, cdcc_limit, min_head_spouse_earned).
• The childcare and childcare-subsidy pipeline is unchanged, so child-only programs are unaffected.

Result (TY 2024)

A disabled adult dependent with an earning spouse (so the earned-income test is met), $6,000 of care via care_expenses:

	value
tax_unit_childcare_expenses	0 (pipeline untouched)
care_expenses	6,000
cdcc_relevant_expenses	3,000
cdcc	600
ks_cdcc	300

Scope / follow-up

• Adult-care dollars must be supplied via care_expenses; the childcare input stays children-only by design.
• The disabled-spouse case (where the disabled person is a non-earning spouse) still returns $0 until the § 21(d)(2) deemed-income work in the follow-up PR (the remainder of Improve federal CDCC conformance with IRC § 21 (deemed income, qualifying-individual count, filing status) #8704) — the expense now reaches the base, but the earned-income test needs deemed income to survive.

Files

• variables/household/expense/childcare/care_expenses.py, pre_subsidy_care_expenses.py — YEAR, relabeled for disabled-adult care
• variables/gov/irs/credits/cdcc/cdcc_relevant_expenses.py — add care_expenses to the base
• Tests: care_expenses.yaml, cdcc_relevant_expenses.yaml, CDCC integration.yaml, ks_cdcc.yaml

Tests

New cases for care_expenses passthrough, the combined CDCC base, and federal + Kansas end-to-end (12 pass locally). CI runs the full suite.

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (20652c5) to head (b5dd136).
⚠️ Report is 60 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##              main     #8746   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            5         3    -2     
  Lines           63        34   -29     
=========================================
- Hits            63        34   -29     

Flag	Coverage Δ
unittests	100.00% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[PavelMakarchuk]

are you sure this is safe?

[PavelMakarchuk]

@hua7450 I traced every reference to care_expenses and pre_subsidy_care_expenses on main — the only usage anywhere was care_expenses's own formula reading pre_subsidy_care_expenses; no other variable consumes either (the CO/MA/MT hits are local Python vars coincidentally named care_expenses, assigned from tax_unit_childcare_expenses/pre_subsidy_childcare_expenses). So the MONTH→YEAR change looks safe to me — nothing relied on the monthly period. Can you confirm you're comfortable it's safe? If so I'll approve.

(Minor: the comment in cdcc_relevant_expenses.py still says "person-level monthly input summed to the year" — now stale since it's a YEAR input.)

[hua7450]

Make it YEAR so it is consistent with the childcare_expenses

[PavelMakarchuk]

A bit worried about month to year change