Model Basic Health Program as separate coverage (fixes #8112)

[MaxGhenis]

Rebased + review-fixed replacement for #8116 (daphnehanse11's original fork PR — fork PRs can't run CI here, so repushed to upstream).

Summary

• Adds gov/hhs/basic_health_program/ coverage modeling (statutory 133–200% FPL, with NY's 250% FPL expansion). Active states: MN (2015), NY (2016+), OR (2025+), DC (2026+).
• Excludes is_basic_health_program_eligible people from ACA PTC eligibility, so simulated aca_ptc zeros out for Essential Plan / MinnesotaCare / OHP Bridge / Healthy DC Plan enrollees. This is the upstream fix that closes the NY per-state ACA calibration gap seen in policyengine-us-data integration tests.
• Moves MN/NY/OR/DC adult and parent Medicaid income ceilings back to Medicaid-only levels so that above-Medicaid-line residents in those states land in BHP rather than silently losing coverage.

Review-driven fixes on top of #8116

Codex review caught two bugs on the original PR which are fixed in this version:

1. No income floor. BHP relied on ~medicaid_eligible alone, which only works in Medicaid-expansion states. Adding TX (non-expansion) to active_states would have qualified sub-133% FPL adults as BHP-eligible. Added explicit income_floor = 1.33 parameter per 42 USC §18051(b)(2)(B).
2. Sub-annual cutover dates on a YEAR-defined variable. NY (2015-04-01), OR (2024-07-01), and NY expansion (2024-04-01) cutovers never triggered in annual sims, so e.g. 2024 NY at 230% FPL and 2024 OR at 180% FPL both returned false when they should return true. Moved effective dates to Jan 1 of the first full BHP year for each state.

Added 5 YAML regression tests covering the floor, the NY 2016 first-full-year, the OR 2024 not-yet-active case, the NY 2025 250% FPL expansion boundary, and the NY 2024 at-230% FPL boundary.

Credits

Original implementation: @daphnehanse11 in #8116 (issue #8112). This PR preserves the original three commits from that branch, adds one review-fix commit, and pushes to upstream so CI can run.

Test plan

• uv run python policyengine_us/tests/test_batched.py policyengine_us/tests/policy/baseline/gov/hhs/basic_health_program --batches 1 → 11 passed
• uv run python policyengine_us/tests/test_batched.py policyengine_us/tests/policy/baseline/gov/hhs/medicaid --batches 1 → 167 passed
• uv run python policyengine_us/tests/test_batched.py policyengine_us/tests/policy/baseline/gov/aca/eligibility --batches 1 → 57 passed
• make format / ruff check clean
• Once merged + published: policyengine-us-data integration test_aca_calibration should drop NY from 511% to <100% error

Fixes #8112. Supersedes #8116.

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (20705b0) to head (ca325ca).
⚠️ Report is 30 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff              @@
##             main     #8127       +/-   ##
============================================
+ Coverage   85.36%   100.00%   +14.63%     
============================================
  Files           3         4        +1     
  Lines          41        57       +16     
  Branches        2         0        -2     
============================================
+ Hits           35        57       +22     
+ Misses          6         0        -6     

Flag	Coverage Δ
unittests	100.00% <100.00%> (+14.63%)	⬆️

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

☔ View full report in Codecov by Sentry.
📢 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.