Followups: VA CCSP FDH rate age dimension; VA/NJ/CT CCAP countable-income child exclusions

[MaxGhenis]

Two lower-priority findings from yesterday's structural bug hunt of newly-merged state benefit code (#7998, #7881, #7811, #7935, #7933, #7778, #7729). Fixed the HIGH / MEDIUM items in #8067, #8068, #8071, #8072, #8073. These two remain:

1. VA CCSP FDH rate missing age-group dimension

File: policyengine_us/variables/gov/states/va/dss/ccsp/rates/va_ccsp_daily_mrr.py:25-27

fdh_rate = p.family_day_home[region]                # flat per region
center_rate = p.center[region][age_group]           # varies by age

Parameter tree confirms FDH has no va_ccsp_care_age_group breakdown. If actual VDOE FDH rates do vary by age (plausible for infants vs school-age), all FDH age groups collapse to the same rate and subsidies are misestimated.

Needs statutory research — VDOE's published FDH reimbursement tables.

2. VA/NJ/CT CCAP/CCSP/C4K countable income: missing child-earnings exclusion

Files:

• policyengine_us/variables/gov/states/va/dss/ccsp/income/va_ccsp_countable_income.py
• policyengine_us/variables/gov/states/nj/njdhs/ccap/income/nj_ccap_countable_income.py
• policyengine_us/variables/gov/states/ct/oec/c4k/income/ct_c4k_countable_income.py (or equivalent)

All three sum employment / self-employment across the whole SPM unit via adds=[...] without filtering out children. YAML comments note the statutory exclusion ("earnings of child under 18", "earnings under $25/month") but the implementation treats every member's earnings as countable. Small effect for most households but material for families with working teens.

Scope: filter earnings by is_tax_unit_head_or_spouse or by age.