Replication data and code for "What people say, what pools need: willingness to pay surveys against thirteen requirements for launching a voluntary health insurance pool"

Replication materials for the paper "What people say, what pools need: willingness to pay surveys against thirteen requirements for launching a voluntary health insurance pool" by D. M. Dror. Four published sources, written for different readers, between them name thirteen arguments a voluntary health insurance pool must know before it opens: a practitioner indicator handbook, the international supervisory standard for inclusive insurance, the guidance of the actuarial profession, and a published capital-requirements model built on the operating data of two schemes in India and Nepal. The paper sets out the thirteen, records which sources name each, imply it, or are silent, classifies each by when it can be known, and locates the willingness-to-pay survey within the requirement. This deposit lets a reader reproduce every number the paper's own analyses rest on, offline and from deposited files. The literature search occupies one folder and is deposited in full, in two dated arms. The first arm, 12 June 2026, searched PubMed and Europe PMC with four query families and returned 11,110 records, 4,855 unique after de-duplication, 190 priority records reducing to 156 after de-duplication on title, of which 55 were read in full and 101 classified off-target. The second arm, 5 August 2026, searched OpenAlex, Crossref and the Directory of Open Access Journals and returned 1,817 records, 1,385 unique, 719 in the universe, 188 priority and 57 carrying a stated-versus-actual signal; all 57 were read and each disposition is recorded with its reason. No study in either arm sets a declared willingness to pay against the same households' subsequent payment and renewal. The verbatim queries, the run logs, the records retrieved at every stage and the completed screens are all included. The income-direction classification of section 4.1 rests on a corpus of 4,851 coded abstracts, deposited with its coding scheme and every trigger phrase, so that any classification is auditable. Income is a positive correlate of the declared sum in 22 of the 26 primary-empirical abstracts stating an explicit sign, and in 57 of 61 once abstracts naming income as a determinant without a sign are added. Two scripts reproduce the paper's numbers. scripts/replicate_all.R locates the data beside itself, needs base R only, writes nothing to disk, and reports 22 checks passed against the deposited files. scripts/verify_published_arithmetic.R recomputes every arithmetic operation the paper performs on figures printed in the cited sources, from inputs typed out of those tables, and requires no deposited file at all. The six-district correlations of section 4.1 derive from a household survey conducted between 2008 and 2010 under Netherlands Organisation for Scientific Research WOTRO Integrated Programme grant W 01.65.309.00. The raw microdata is not deposited, since it carries identifying detail about households in named villages; the derived correlation and regression statistics are deposited and reproduce every coefficient the paper reports, together with the two co-author source workbooks. Raw microdata is available from the author on reasonable request under the original grant terms. Figures drawn from other authors' published papers are documented with their exact table or page for verification against the cited article and are not redistributed. The repository is organised into literature_search, india_survey, scripts and outputs, with README.md, PROVENANCE_MAP.md mapping every claim in the paper to its source, CODEBOOK.md documenting every variable, MANIFEST.txt carrying an MD5, a size and a row count for every file, CHANGELOG.md, and LICENSE.txt.