Carve-out: harness formalization moves to the popdgp methods paper; this paper becomes the fill dossier + first application

Decision

The portfolio review (2026-07-03) promotes the population-view harness from a section of this paper to its own methods paper + package: popdgp (Survey Methodology / JOS target). This paper stays the fill operator's dossier and the harness's first application.

Changes here

1. Trim the harness formalization (methodology §"The population-view harness") to a working summary + citation of the popdgp paper (in prep); keep everything needed to read the results tables.
2. Keep all experiments and results as-is — both harness profiles, the tail block, floors. The application is this paper's evidence; only the framework's general formalization moves.
3. Code extraction (coordinate with popdgp's extraction issue): experiments/views.py and the joint/tail metrics migrate to the popdgp package; this repo then depends on popdgp (pinned) and keeps only task-specific wiring. Behavior-preserving — regenerated tables must be bit-identical.
4. Background §"Evaluating imputed files": the formal-ancestry paragraph (Lohr–Raghunathan, Rässler, proper scoring, PRDC) largely moves to the popdgp paper; keep two sentences + citation.
5. Update the contributions list: contribution 2 becomes "first application of the population-view harness (popdgp)" rather than the framework itself.

Sequencing: this paper submits first (IJM); the popdgp paper cites it as the first application — the standard application/methods split, no cannibalization.

🤖 Generated with Claude Code

Start by reading the methodology section “The population-view harness,” the “Evaluating imputed files” background, and experiments/views.py. Coordinate with popdgp’s extraction issue, then trim the formalization, update the contributions list, preserve task-specific wiring, and verify regenerated tables are bit-identical.