Funding Teams vs Funding Individuals: Which Model Produces Better Science?

Getting your Trinity Audio player ready...

Research funding is often presented as a choice between supporting individual investigators and financing research teams. Both models are necessary, but they solve different problems.

Individual funding protects intellectual independence, supports unconventional ideas, and makes responsibility relatively clear. Team funding allows researchers to combine specialized skills, equipment, data, and institutional capacity that no single person could provide.

The best funding system should therefore not choose one model exclusively. It should fund scientific work at two levels:

1. Reward individuals according to their attributable contributions.
2. Provide shared resources when a research objective genuinely requires coordinated teamwork.

This distinction matters because a research team produces results collectively, but the effort, originality, and responsibility inside the team are rarely distributed equally.

What Is Individual Research Funding?

Individual research funding allocates money primarily to a named scientist or investigator. The recipient may use the funding to conduct research, hire assistants, acquire equipment, or develop a research program.

The central object being evaluated is the person’s ability and scientific contribution, rather than only a predefined project.

Funding individuals can take several forms:

• investigator grants;
• research fellowships;
• salaries or stipends;
• unrestricted long-term research support;
• awards based on past scientific output;
• funding for independent researchers.

Some individual awards still support laboratories and collaborators. “Funding an individual” does not necessarily mean expecting one person to perform every task alone. It means giving a particular researcher sufficient autonomy to decide how the work should develop.

What Is Team-Based Research Funding?

Team funding allocates money to a formally organized group working toward a shared objective. The proposal normally describes the team’s leadership, work packages, budget, milestones, and division of responsibilities.

This model is appropriate when research requires:

• several distinct disciplines;
• expensive shared infrastructure;
• large datasets or distributed data collection;
• clinical, engineering, or field operations;
• coordinated software development;
• standardized experimental protocols;
• integration of multiple technical components.

For example, the US National Institute of General Medical Sciences describes its Collaborative Program Grant for Multidisciplinary Teams as support for ambitious projects that cannot be addressed through separate individual-investigator grants. Its program explicitly requires a highly integrated team rather than a loose collection of unrelated subprojects.

The US National Science Foundation similarly uses convergence-research programs for complex problems that require deep integration across disciplines.

The Case for Funding Teams

Teams can combine specialized knowledge

Many modern research problems exceed the competence of any one person. A biomedical project may require molecular biology, statistics, clinical expertise, software engineering, data governance, and specialized instrumentation.

A team allows each contributor to work in an area where they have comparative expertise.

This is particularly important when the output must be not merely discovered but also validated, engineered, deployed, or translated into practice.

Teams can undertake projects at greater scale

Some research objectives are structurally collective. A particle detector, longitudinal medical study, astronomical survey, genomic database, or large formal-software library cannot realistically be created by one investigator.

Funding the team as a coordinated unit can reduce duplicated infrastructure and ensure that its components are technically compatible.

Teams may sustain complex projects

Large scientific systems require maintenance, documentation, quality control, administration, and succession planning. A project dependent on one person may become fragile if that researcher becomes unavailable or changes direction.

A well-organized team can preserve institutional knowledge and maintain long-lived scientific infrastructure.

Collaboration now dominates much of science

The 2025 National Academies report The Science and Practice of Team Science notes that collaborative authorship has become a defining feature of modern science and engineering. It builds on an earlier finding that more than 90% of scientific and engineering publications were coauthored.

This does not prove that every collaboration is efficient. It does show that funding systems must be capable of supporting genuinely collective work.

The Weaknesses of Team Funding

Team grants can conceal unequal contributions

A grant may be awarded to a laboratory, consortium, university, or principal investigator, even though much of the substantive work is performed by postdoctoral researchers, programmers, technicians, data curators, or junior collaborators.

The team’s success may increase the reputation of its most visible leaders without proportionally rewarding less visible contributors.

This is partly why the CRediT Contributor Role Taxonomy distinguishes 14 types of contribution, including conceptualization, software, data curation, methodology, validation, supervision, and writing.

A team is not a single moral or economic agent. It is a network of individuals whose contributions must still be identified.

Funding may follow hierarchy rather than merit

In many team grants, money flows through an institution or principal investigator. That structure gives leaders considerable influence over hiring, authorship, salaries, and access to resources.

A funder may therefore believe it is supporting an excellent team while actually reinforcing an internal hierarchy that it cannot observe.

Researchers who contributed the original idea may receive less authority than senior administrators or prestigious investigators added to make the proposal appear credible.

Large teams create coordination costs

Team science requires communication, conflict resolution, compatible incentives, clear leadership, and shared standards. The National Academies has identified factors such as disciplinary diversity, geographic dispersion, large membership, changing composition, and task interdependence as recurring challenges for scientific teams.

Adding more participants does not automatically increase productivity. At some point, additional specialization may be offset by meetings, reporting, negotiation, and integration costs.

Teams can encourage conservative research

A major study published in Nature analyzed more than 65 million papers, patents, and software products. It found that smaller teams tended to introduce more disruptive ideas, while larger teams more often developed and extended established lines of work.

This should not be interpreted as proof that small teams are always better. Development, replication, scaling, and consolidation are essential scientific activities. The result instead suggests that science needs an ecology containing both small exploratory efforts and large developmental teams.

A funding system that consistently favors large consortia may become good at extending accepted paradigms while neglecting ideas that could replace them.

The Case for Funding Individuals

Individuals can pursue unconventional ideas

Novel research often begins before a complete team can be assembled or before a conventional project plan can be written.

An individual researcher may notice an unexplored definition, theoretical connection, anomaly, algorithm, or methodological problem. Requiring that person to first construct a consortium can delay the work and dilute its intellectual direction.

Individual funding is especially important in mathematics, theoretical computer science, philosophy of science, and other fields where a major contribution may initially require concentrated reasoning rather than organizational scale.

Individual funding creates autonomy

When researchers receive stable support, they can change direction in response to evidence. They do not need to pretend that the exact sequence of future discoveries is already known.

This is one reason to distinguish funding demonstrated scientific contribution from funding research promises. Conventional project grants often ask committees to predict which proposal will succeed. Person-centered or output-based funding can instead support researchers whose public work already demonstrates value.

Responsibility is easier to attribute

When support is assigned to an individual, the funder can more clearly ask:

• What did this researcher contribute?
• Was the work public and attributable?
• Did the researcher create a useful result?
• Did later work depend on it?
• Is the researcher continuing to produce substantive work?

This does not eliminate evaluation errors, but it avoids treating membership in a successful institution or consortium as sufficient evidence of personal merit.

Individual funding can include outsiders

Team grants are often administered through eligible institutions. Independent researchers, unaffiliated programmers, citizen scientists, and researchers outside prestigious networks may be unable to apply or may need an institutional intermediary.

A system that evaluates people through their public research and software can be more open to contributors without conventional credentials. This is one of the proposed advantages of AI Internet-Meritocracy for independent researchers and other underrepresented contributors.

The Weaknesses of Individual Funding

Individual funding also presents serious risks.

A funder may overestimate a charismatic or highly visible researcher. Individual rewards can encourage personal branding, excessive competition, strategic self-citation, or reluctance to share credit.

It may also underfund essential collective goods. An individual salary does not automatically finance laboratory equipment, research participants, cloud infrastructure, data acquisition, regulatory work, or technical staff.

Finally, discoveries often depend on earlier contributions by many people. Excessive emphasis on one “genius” can replace an inaccurate institutional story with an equally inaccurate heroic-individual story.

The solution is not to deny individual merit. It is to model merit as divisible, attributable, and dependency-aware.

Teams and Individuals Are Not Opposites

The apparent choice between teams and individuals is partly misleading.

Every team consists of individuals. Most individually funded researchers also rely on collaborators, previous publications, software maintainers, reviewers, technicians, and shared infrastructure.

The real policy questions are:

• At what level should money be administered?
• At what level should scientific merit be evaluated?
• Who controls shared resources?
• How should credit and compensation be divided?
• Can contributors move between projects without losing support?
• How should previously invisible enabling work be recognized?

A sound system may fund a team operationally while evaluating its members individually.

For example, a grant could pay for a shared telescope, database, laboratory, or software platform. Separate contribution-based payments could then reward the people who design experiments, develop software, prove theorems, maintain datasets, perform replications, or create later improvements.

How AIIM Could Combine Both Models

AI Internet-Meritocracy proposes funding researchers and free-software developers according to measurable contribution rather than institutional status alone.

Under such a model, the basic recipient of merit-based compensation should usually be the individual contributor. However, this does not require abandoning team science.

AIIM could support teams through several complementary mechanisms.

Reward each contributor separately

The system could evaluate who contributed ideas, proofs, experiments, code, data, documentation, replication, reviewing, coordination, or dissemination.

The total value of a team’s output would not need to be divided equally. Nor would every coauthor automatically receive the full value of the publication.

Recognize dependency structures

Research outputs form dependency graphs. A later result may depend on an earlier theorem, dataset, library, experimental method, or conceptual framework.

AIIM could allocate recognition not only to the final visible product but also to the contributors whose earlier work made it possible.

This would allow collaboration without requiring every contributor to belong to one formal project or institution.

Fund shared costs separately from personal rewards

Some expenses are genuinely collective:

• equipment;
• laboratory space;
• computing;
• data collection;
• participant recruitment;
• project administration;
• security audits;
• publication infrastructure.

These should be evaluated as project or infrastructure costs rather than disguised as rewards for personal scientific merit.

In other words:

Fund people for their contributions, and fund teams for the shared resources that their work requires.

Allow teams to form dynamically

Traditional grants often define the team before the research begins. However, important collaborators may appear later, and some initially listed participants may contribute little.

A continuous funding system could allow contributors to join, leave, or interact across project boundaries while their rewards follow demonstrated work.

This would replace the rigid grant consortium with a more flexible scientific network.

A Better Hybrid Funding Architecture

A balanced research-funding system should contain at least three layers.

Funding layer	Primary purpose	Typical recipient
Personal support	Autonomy, livelihood, continued research	Individual contributor
Project funding	Specific coordinated objective	Small or large team
Infrastructure funding	Shared tools, facilities, datasets, and platforms	Institution, DAO, consortium, or service provider

These layers should not be collapsed into one grant.

A team may deserve substantial project resources without every member deserving the same personal reward. Conversely, a highly valuable individual may deserve continuing support even when not currently leading a formal project.

Conclusion: Fund Collaboration Without Erasing Contributors

Funding teams is necessary for science that depends on scale, specialization, and shared infrastructure. Funding individuals is necessary for autonomy, disruptive ideas, accountability, and intellectual diversity.

The mistake is to assume that choosing a team as the administrative recipient means that the team should also be treated as an indivisible unit of merit.

A better principle is:

Scientific cooperation should be collective, but scientific credit and compensation should remain divisible.

AIIM could implement this principle by evaluating individual contributions across team boundaries while separately financing the shared resources that teams genuinely need.

Such a system would not force researchers to choose between independence and collaboration. It would make collaboration economically possible without allowing hierarchy, authorship order, or institutional prestige to determine who receives recognition.

References

• National Academies of Sciences, Engineering, and Medicine, The Science and Practice of Team Science, 2025.
• National Research Council, Enhancing the Effectiveness of Team Science, 2015.
• Wu, L., Wang, D., and Evans, J. A., “Large Teams Develop and Small Teams Disrupt Science and Technology”, Nature, 2019.
• National Institute of General Medical Sciences, Collaborative Program Grant for Multidisciplinary Teams.
• National Information Standards Organization, CRediT Contributor Role Taxonomy.