The structure of public research funding, 2015–2025: concentration, cross-funder co-funding, and the funding→output relationship in a reconciled NIH/NSF/EC/UKRI grant graph

We assemble a reconciled graph of the global public-research economy — 887,016 grants from the U.S. National Institutes of Health (NIH), the U.S. National Science Foundation (NSF), the European Commission (EC, via CORDIS), and UK Research and Innovation (UKRI), 2015–2025 — in which recipient organizations are merged on ROR identifiers, investigators on ORCID, and 156,877 research outputs are linked to the grants that funded them through 285,604 OpenAlex acknowledgement edges.

Working only from statistics that are robust to known entity-resolution noise (grant and work counts, country- and funder-level aggregates, the ROR-resolved subset), we characterize three structural features of the funding system. (1) Concentration. Across 4,840 ROR-resolved recipient institutions the distribution of grants is extreme: Gini = 0.929 (95% bootstrap CI [0.919, 0.936]); the top 1% of institutions hold 53.6% of all grants and the top 10% hold 92.0%. (2) Cross-funder structure. Of 156,877 funded works, 26.3% acknowledge two or more distinct funders; the dominant co-funding pair is intra-European (ERC↔EC, 17,502 shared works), with EC↔NSF (3,430) and NIGMS↔NSF (2,944) the leading trans-Atlantic and cross-agency ties. (3) The funding→output relationship. Restricting to the three funders with output linkage (NIH, NSF, EC), the count-based productivity rate ranges from 0.94 linked works per $1M (NIBIB) to 0.05 (NCATS), with NSF at 0.72 and the ERC at 0.67; these differences track each funder's mission (basic-science vs. translational/infrastructure) rather than efficiency.

We state the corpus and entity-resolution limitations plainly — most importantly that the country distribution reflects the NIH-heavy composition of the corpus, not a measurement of global funding — and we release all code, data references, and a Zenodo-ready metadata record.