Every grant that a funder awards does more than pay for an experiment. According to a perspective article published on 23 October 2026 in Frontiers in Systems Biology, funding mechanisms act as incentive systems that quietly shape what scientists study, how they work, and which contributions their careers reward. The article, written by Lillian Parr, Nadia Sarfraz, Sylvina M. Raver and Nadia M. Penrod of the Milken Institute Science Philanthropy Accelerator for Research and Collaboration in Washington, DC, argues that the design of funding programs is a powerful and underused lever for steering scientific behavior, and that philanthropic funders are uniquely positioned to experiment with it.
The authors’ central claim is that funding mechanisms communicate two things at once. The first is explicit: the priorities, eligibility criteria and review criteria that state what science a funder wants to advance. The second is implicit, emerging from the structure of the mechanism and the pattern of funding decisions over time. These implicit signals tell researchers which kinds of science are valued and which researchers and institutions are considered most likely to produce important advances. Because researchers respond to those signals, the authors argue, every funding decision influences the questions scientists ask, the risks they take, how they collaborate, what they share, and which contributions are ultimately recognized.
The perspective draws on existing research to describe how current incentive structures skew scientific work. When hiring, promotion and grant renewal prioritize publications over replication, novelty over rigor and speed over collaboration, investigators respond rationally to those incentives. Publications, citations and grant funding are commonly used as indicators of achievement, while awards, journal reputation and institutional affiliation shape perceptions of status. Yet many researchers already recognize that activities such as replication, data sharing, mentorship, collaboration and building shared resources are essential for a strong scientific ecosystem. The challenge, the authors write, is not persuading scientists that these activities matter; it is ensuring that the systems governing scientific careers consistently recognize and reward them.
The article also describes how these dynamics can become self-reinforcing. Visibility attracts funding, talent and recognition, which further concentrates opportunities among already prominent researchers, institutions and fields. Research on scientific prizes illustrates the point: prizes do not merely recognize achievement but reinforce networks of influence, shaping which researchers and ideas gain visibility. As a result, the authors suggest, researchers may gravitate toward dominant paradigms and well-connected networks, where recognition and collaboration are more accessible, rather than pursuing ideas that challenge prevailing views or create friction with influential figures.
A key section of the article examines how every type of funding mechanism embodies a different set of assumptions about what drives scientific progress, each with trade-offs. Investigator-based awards assume that talented researchers are best positioned to identify important questions, and that flexibility lets them pursue those questions effectively, but selection based on standard achievement indicators may reinforce existing advantages. Project-based grants assume that promising ideas can be evaluated through defined research plans, enabling rigorous assessment, but may favor incremental advances over bold breakthroughs. Milestone-driven funding can accelerate targeted efforts but may constrain adaptation when discoveries lead research in unexpected directions. Prize mechanisms direct attention and confer prestige, but their influence depends on which achievements and values they signal as worthy of recognition.
The mechanisms used to evaluate proposals carry their own incentives. Peer review assumes that expert judgment can identify promising science and allocate resources well. But because reviewers evaluate work through the lens of their own expertise, experiences and disciplinary norms, the authors note that peer review can also introduce or reinforce biases in favor of established approaches, institutions or research paradigms. None of these mechanisms is inherently wrong, the article argues; the point is that each creates different conditions for how scientific communities generate, share and build knowledge over time.
Why single out philanthropy? The authors argue that scientific progress depends on a diverse funding ecosystem in which different funders bring different priorities, risk tolerances and time horizons. Public agencies provide foundational support for basic and translational research under requirements for accountability and public oversight. Industry funding advances discoveries toward applications such as therapeutics and diagnostics, with decisions shaped by scientific opportunity, unmet needs, development feasibility and potential financial return. Philanthropic funding often supports areas outside traditional priorities, including emerging fields, underfunded diseases and high-risk approaches with uncertain timelines, and it typically operates with more flexible accountability structures. That flexibility, the authors contend, enables philanthropy to test how different ways of selecting, supporting and evaluating researchers shape scientific behavior and research communities over time.
The article offers a concrete example of such experimentation: the Volkswagen Foundation’s Experiment! Initiative. Between 2013 and 2021, the Foundation supplemented expert review with partial randomization, selecting some scientifically vetted, eligible proposals through an interdisciplinary jury and others by lottery. The approach was intended to increase diversity and reduce potential bias in consensus-oriented conventional review, and the Foundation commissioned research to evaluate its effects. The results showed greater diversity among grantees, including higher shares of early-career researchers and women, while jury-selected and lottery-selected projects showed no difference in standard outcome metrics such as the number of publications and patent applications. Among participating researchers, 88 percent of respondents agreed that randomized selection clearly avoids conflicts of interest and unconscious bias, offering opportunity to topics and thematic and methodological approaches that may be underrepresented on review panels.
The authors then map four complementary targets for philanthropic investment. Investing in people, they argue, shapes how scientists approach research; long-term, flexible investigator awards can reduce pressure to optimize research agendas for short-term outputs and give scientists freedom to follow emerging evidence and pursue uncertain but potentially transformative directions. They cite comparative studies suggesting that investigators supported by the Howard Hughes Medical Institute’s Investigator program, which provides long-term flexible support rather than requiring detailed project plans, pursue more novel research directions and produce a wider distribution of scientific outcomes than similarly accomplished NIH-funded investigators. Investing in fields, meanwhile, means supporting research networks, evidence synthesis and collective efforts that help communities define shared priorities and build the visibility needed to sustain investment; the Chan Zuckerberg Rare As One initiative, which supports patient-led organizations building collaborative research networks in rare disease, is offered as an example.
Investing in norms and in infrastructure rounds out the framework. Norm investments, such as support for open data practices, reproducibility efforts and responsible research assessment, can help translate aspirational values into routine behavior; the article points to initiatives including the Declaration on Research Assessment, the Leiden Manifesto and the Hong Kong Principles, as well as narrative CV formats that give researchers space to describe mentorship, collaboration, data sharing and methods development. Recognition mechanisms such as the NIH Replication Prize, the AAAS Mentor Award and the Parasite and Symbiont Awards reward behaviors that traditional evaluation systems often overlook. Infrastructure investments, exemplified by the Gordon and Betty Moore Foundation’s Data-Driven Discovery Initiative, create shared public goods; an independent evaluation found that modest investments in widely shared software tools generated substantial scientific value. The article also notes that funding design shapes how work is organized, citing Wellcome Leap’s multidisciplinary, time-limited challenge teams as a model that incentivizes collaboration, accountability and urgency.
Finally, the authors turn to evaluation, arguing that it should be built into funding design rather than added afterward. They propose a three-question framework: what behaviors and norms does the mechanism encourage, who participates and benefits, and what persists beyond the award? Because no single measure can show that a funding mechanism caused a change in behavior, they recommend combining quantitative and qualitative evidence, including surveys, interviews and administrative data, and comparing outcomes across mechanisms or before and after design changes. They highlight the Howard Hughes Medical Institute’s revision of eligibility for its 2026 Investigator Program after identifying overrepresentation from a small number of institutions among previous awardees, as an example of evaluation revealing unintended consequences. The authors also assign scientists a role: understanding what a funding mechanism is designed to support, participating in evaluations, and making less visible contributions such as mentorship, data sharing and replication visible to funders. By treating funding mechanisms as design choices that can be tested and refined, the article concludes, philanthropy can build an evidence base for funding approaches that strengthen a scientific ecosystem that is more reliable, creative and capable of generating knowledge as a public good.
Subject of Research: How funding mechanism design shapes scientific incentives and the role of philanthropy in testing alternative funding approaches
Article Title: Funding mechanisms as incentive systems: a role for science philanthropy
Article References: Parr, L., Sarfraz, N., Raver, S. M., & Penrod, N. M. (2026). Funding mechanisms as incentive systems: a role for science philanthropy. Frontiers in Systems Biology, 6, Article 1939099. https://doi.org/10.3389/fsysb.2026.1939099
Image Credits: AI Generated
DOI: 10.3389/fsysb.2026.1939099
Keywords: science funding, philanthropy, research incentives, funding design, peer review, evaluation, replication, data sharing, mentorship, research infrastructure, scientific careers, Volkswagen Foundation
News Source: Drew Townsend. (October 28, 2026). How Science Philanthropy Could Rewire the Incentives Behind Research Funding. Scienmag.
