Philanthropic Funding in Scientific Innovation

• In Event: AI, Innovation Policy, and Global Sustainability

Abstract

The U.S. has long relied on federal and corporate investments to drive scientific progress. However, recent funding instability and shifting federal priorities have created uncertainty, especially in applied fields like clean energy, climate science, biomedical research, and inclusive studies. Basic science, with its delayed payoffs, also struggles to secure support. While corporate investments are substantial, they are primarily motivated by commercial returns. These changes reveal the vulnerability of scientific progress to political and economic forces. Yet, these research areas remain essential for long-term societal benefit. As traditional funding falters, one key question arises: who will support the next wave of scientific breakthroughs?

Philanthropy has historically stepped in where government and markets fall short. Amid unstable federal priorities and market-driven incentives, philanthropic funding supports high-risk, long-term research. However, its priorities and patterns remain poorly understood. As concern grows over stalled innovation, understanding the role of philanthropy is more pressing than ever. This study investigates philanthropic funding in science, focusing on (1) how resources are distributed across disciplines, (2) how philanthropic, corporate, and government funding operate and align across knowledge domains, and (3) what sets philanthropy apart. By examining these areas, we provide strategic insights to strengthen science funding ecosystems.

To meet these goals, we constructed a new dataset. Existing bibliometric platforms remain fragmented and inconsistent, often missing key contributors. While tools like IRIS UMETRICS, OpenAlex (OA), and Dimensions have improved funder disambiguation, they still overlook many philanthropic and corporate sources. We address this by integrating multiple sources: bibliometric databases and curated lists of U.S. philanthropic organizations, corporations, and government agencies. Our method uses heuristic-driven string matching to link acknowledgments in Web of Science (WoS) to organizations in OpenAlex and curated lists. In OpenAlex, funder names are normalized, acronyms are extracted, and documents are grouped based on lexical similarity through MinHash LSH, with a Jaccard similarity threshold of 0.95 to form connected clusters. The general matching process prioritizes exact name matches, followed by structured comparisons, including longest prefix or suffix matches (≥70% similarity), substring matches, and acronym-based associations. We verify matching results through manual checks and large language model-based reviews of high-frequency strings from over 1,000 publications.

Preliminary findings: Of 17 million WoS publications, 21% include funding information. About 60% of these are linked to 43,316 distinct funders across sectors. Government remains the dominant funder; however, nearly 30% of publications report philanthropic support, highlighting its salient and resilient role. Philanthropy also supports nearly five times more publications than corporate funders and aligns most with government in biomedical and life sciences. Topic-wise, government funding spans many fields, especially physical and multidisciplinary sciences, while corporate funding focuses on applied medical areas. Philanthropy tends to complement rather than disrupt existing funding patterns.

Next steps will examine how funding patterns evolve across the knowledge landscape. By clarifying these dynamics, this study seeks to inform funders, policymakers, and researchers, helping to design alternative strategies and support a more resilient, diversified ecosystem for scientific discovery.

Author

• Jin Ai, Rutgers University - Newark
  Presenting Author