Context

PolicyEngine/policyengine-us-data#1162 added direct CHIP enrollment calibration targets and CHIP take-up plumbing in the old data repo. That PR is now closed as superseded by the Populace migration, but Populace has not yet reproduced the direct CHIP target.

Current Populace state:

• cms_medicaid total_medicaid_enrollment maps to medicaid_enrolled.
• cms_medicaid total_medicaid_chip_enrollment maps to medicaid_enrolled,chip_enrolled.
• There is no first-class direct CHIP enrollment target mapping to chip_enrolled.

That means the combined Medicaid+CHIP target can fit while the Medicaid/CHIP split is still wrong.

Latest published release checked:

• Release: populace-us-2024-c86a631-6e1bcd0271a5-20260619T002242Z
• Runtime: PolicyEngine-US 1.729.0
• Dataset sha256: f0af25192d6c8a7efc2638da2bd8ec4278b066a1092cc89ef2275811efaff11d

Derived from the release diagnostics:

• National CHIP implied by Medicaid+CHIP - Medicaid target: 7.31M
• National CHIP implied by final estimates: 4.95M
• Relative miss: -32.3%

Some state split examples:

• CA: target-derived CHIP 1.23M, final-derived CHIP 9k
• IL: target-derived CHIP 291k, final-derived CHIP 0
• NC: target-derived CHIP 348k, final-derived CHIP 0
• TX: target-derived CHIP 356k, final-derived CHIP 809k
• NY: target-derived CHIP 676k, final-derived CHIP 1.07M

Proposed Work

Add direct CHIP enrollment targets to the Populace US fiscal target surface.

Expected shape:

• Add a Ledger/CMS target mapping for total_chip_enrollment or equivalent CMS measure.
• Map it to chip_enrolled.
• Include state and national target rows where available.
• Add a target coverage requirement so direct CHIP enrollment cannot silently disappear.
• Add source-coverage/active-alias coverage if this requires a separate Ledger package alias.
• Add tests parallel to the current Medicaid and Medicaid+CHIP target tests.

Acceptance Criteria

• The compiled US fiscal target registry includes direct CHIP enrollment targets.
• The release target profile coverage gate requires direct CHIP enrollment.
• Calibration diagnostics report direct CHIP enrollment fit separately from Medicaid and Medicaid+CHIP combined.
• The Medicaid/CHIP split no longer has to be inferred only by subtracting two separate target rows.

Related

• Superseded old-repo PR: Add CHIP enrollment calibration targets policyengine-us-data#1162
• Existing Populace Medicaid+CHIP mapping: cms_medicaid total_medicaid_chip_enrollment -> medicaid_enrolled,chip_enrolled