Replication data and code for the paper "Measuring intellectual affiliation at the document level: a citation-based indicator validated in economics and psychology".

Replication data and code for the paper "Measuring intellectual affiliation at thedocument level: a citation-based indicator validated in economics and psychology". The paper proposes a continuous, externally anchored score of intellectual affiliationat the level of the individual document, and applies it to Brazilian economics between2000 and 2025. This deposit contains everything needed to reproduce every number,table and figure in it. Data (11 files). The 33 economics anchor journals with their tradition and referencevolume; the raw and shrunk affiliation scores, with support, of the 650,005 works theycite; the scores of the 25,923 Brazilian-authored articles, with standard error,coverage, circuit and venue scope; the 33 psychology anchor journals used for thesecond-field validation and the scores of the 654,456 works they cite; the fourco-citation networks node by node, with cluster membership, bootstrap stability andfour centrality measures; the list of duplicate OpenAlex records merged into a singlenode; and a manifest recording collection dates, chosen parameters and counts. Code (35 files). The complete pipeline, one file per step: API collection, thework-level and article-level scores, out-of-sample validation and the choice of thesupport threshold, the Monte Carlo study, the robustness suite (placebo, baselines,anchor subsets, uncertainty of the score), the portability tests, the second field,the bibliographic-coupling comparison, the co-citation networks and the compositionand diversity measures, the selection-adaptation decomposition, the localisationregressions, and the intellectual balance of payments. Network nodes are ranked by co-citation frequency and, at ties, by the workidentifier, so the partition is reproducible under any PYTHONHASHSEED; the Leiden andresampling seeds are fixed. Everything derives from OpenAlex. The sources snapshot wasdownloaded on 7 August 2026 and the API collection ran between 7 and 10 August 2026.Nothing comes from a proprietary source and the set is redistributable in its entirety.