The Political Economy of Science Funding

• In Event: Innovation Policy, R&D Systems, and Science Funding

Abstract

Science funding systems are undergoing a structural transformation from a postwar model centered on federal support toward a pluralistic political economy in which governments, firms, philanthropic organizations, and military actors jointly allocate resources and shape the direction of research. Despite this shift, empirical research remains fragmented across sectors, while prevailing theories and policy narratives continue to assign canonical roles, such as government supporting basic science, industry driving applied innovation, and philanthropy addressing social needs, without systematic validation. From a political economy perspective, this is a critical gap: science funding is not merely an input, but a mechanism of allocation and agenda setting that structures which problems receive attention and how scientific effort is distributed. This paper addresses this gap by providing a system-level empirical mapping of how heterogeneous funders position themselves across the scientific landscape, while explicitly separating military actors from broader government funding to examine their distinct roles, and assessing the extent to which these positions align with canonical expectations.

We construct a large-scale dataset linking disambiguated funding acknowledgments to the global scientific literature, covering over 13.3 million government-funded publications, 3.27 million philanthropic-funded publications, and 0.84 million corporate-funded publications, alongside tens of thousands of distinct funders (32,767 philanthropic, 11,687 corporate, and 8,913 government entities, within which military actors are separately identified). To locate funding within the structure of knowledge production, we embed publications in a high-dimensional semantic space and characterize them along three dimensions central to the political economy of science: (1) innovation profile (novelty, conventionality, and interdisciplinarity), (2) disciplinary domain, and (3) societal orientation via the United Nations Sustainable Development Goals (SDGs). We then estimate machine-learning–augmented models to recover sector-specific allocation patterns and positional signatures.

The results show that science funding constitutes a structured allocation system rather than a uniform distribution of support. Across disciplines, funders exhibit differentiated patterns consistent with distinct institutional logics: government funding is broadly distributed but leans toward engineering and physical sciences; corporate funding concentrates in applied, health, and profit-oriented fields; philanthropy is selectively clustered in health and social domains; and military funding, analyzed separately from other government actors, is narrowly concentrated in engineering and physics. A similar division of labor emerges across societal domains: government aligns with infrastructure and system-level goals, corporate funding with economic and technological development, and philanthropy with social and institutional priorities, while military funding remains narrowly targeted.

However, these empirically observed roles do not fully conform to canonical expectations. Across sectors, funding is more strongly associated with conventional research than with frontier novelty, and corporate funding, rather than government, is the only sector positively associated with novelty, though modestly. Moreover, sectors are not cleanly separable but occupy overlapping regions of the knowledge space, indicating a system characterized by partial differentiation rather than clear functional specialization.

Taken together, the findings advance a political economy account of science funding in which heterogeneous funders act as interdependent allocators of scientific effort, shaping the topology of knowledge production through their institutional priorities and constraints.

Author

• Jin Ai, Rutgers University - Newark
  Presenting Author