A grant application can be judged on its scientific originality, methodological strength and potential impact. Yet a new study in Nature Human Behaviour draws attention to another force that may shape who receives research funding: the human tendency to favor people who resemble ourselves. This phenomenon, known as homophily, is widespread in social life. People often form relationships with others who share their background, identity, experiences or institutional environment. In grant evaluation, the same tendency could influence decisions that are expected to be based primarily on merit.
The study, “Homophily in grant evaluations,” by economists Nicola Fuchs-Schündeln and Fabian Holub, examines how similarity between evaluators and applicants may enter the funding process. Grant competitions are commonly designed to make scientific selection more objective. Applications are assessed by experts, scores are recorded, and panels compare proposals according to formal criteria. But evaluation is not performed by machines. It depends on human judgments about ambition, feasibility, credibility and importance—judgments that can be affected by prior knowledge and personal interpretation. The central issue raised by the research is whether evaluators respond differently when an applicant’s profile, background or career path feels familiar.
Homophily does not necessarily mean deliberate favoritism. An evaluator may consciously apply the same standards to every proposal while still finding a familiar research trajectory easier to understand or more convincing. Shared experiences can create a sense of intellectual affinity, making an applicant’s goals appear especially realistic or important. Similarity may also affect how evaluators interpret uncertainty. A proposal from someone whose career path resembles the evaluator’s may be viewed as evidence of promise, while an equally strong proposal from a less familiar background may require more explanation before it receives the same level of confidence.
This distinction is technically important because grant evaluation involves substantial discretion. Reviewers rarely possess complete information about an applicant’s future performance. They infer likely success from publications, institutional affiliations, recommendations, previous funding and the clarity of a proposed project. These signals are imperfect proxies for scientific quality. When evidence is incomplete, evaluators may rely more heavily on recognition and professional familiarity. In statistical terms, the decision-maker is estimating an applicant’s expected productivity under uncertainty. If similarity changes that estimate without reflecting actual differences in ability, it can introduce systematic bias into the selection process.
The research also speaks to a broader problem in social science: separating homophily from genuine differences in quality. Applicants who resemble evaluators may work in the same fields, use similar methods or submit proposals that are easier for a particular expert to assess. A positive association between similarity and funding could therefore arise because the match improves the accuracy of evaluation rather than because it creates favoritism. To identify homophily, researchers must account for observable characteristics that influence both the composition of review panels and the strength of applications. The challenge is to determine whether similarity continues to matter after differences in research area, experience and proposal quality are considered.
That challenge makes grant evaluations a valuable setting for studying bias. Funding decisions have clear outcomes: an application receives support or it does not, and reviewers often produce numerical scores or rankings. These records allow researchers to compare decisions across applicants and evaluators. A statistical analysis can examine whether the same applicant characteristics receive different assessments depending on who conducts the review. It can also test whether patterns persist across fields or become stronger when information about an applicant is more ambiguous. Such analysis cannot reveal every private thought behind a score, but it can show whether evaluator-applicant similarity is consistently associated with measurable differences in treatment.
The implications extend beyond individual researchers. Research grants determine which questions can be pursued, which laboratories can hire staff and which early-career scientists can establish independent careers. A small difference in the probability of receiving funding may compound over time. Successful applicants gain preliminary results, larger teams, stronger publication records and greater visibility, all of which can improve their chances in future competitions. If homophily affects the first step, it may contribute to cumulative advantage across an entire career. The result could be a scientific system in which familiar profiles become increasingly overrepresented, not necessarily because they produce better research, but because early judgments shape later opportunities.
The possibility of homophily also complicates the debate over diversity in science. Diversity is often defended on the grounds that different experiences generate new questions and approaches. If evaluators systematically favor applicants who resemble themselves, however, the process may narrow the range of ideas before research is even conducted. A panel composed of experts with similar backgrounds could unintentionally reward conventional topics, familiar methods or established professional networks. Conversely, a more diverse evaluator pool may expose proposals to a wider range of interpretations. The benefit would not simply be representational. It could improve the intellectual breadth of the research portfolio by changing which forms of evidence and innovation are recognized as promising.
The findings raise practical questions for funding agencies, although no single reform can eliminate judgment from peer review. Agencies could diversify panels, rotate evaluators, standardize scoring criteria and separate assessments of the proposal from assessments of the applicant’s track record. Structured review forms may reduce the influence of vague impressions, while anonymizing parts of an application could limit the role of identity and institutional prestige. Statistical monitoring could also identify whether evaluator characteristics are linked to systematic differences in scores. These measures have trade-offs: anonymity can hide information that is genuinely relevant, and rigid scoring systems may make it harder to recognize unconventional but valuable research. The goal is therefore not to remove expertise, but to make its exercise more transparent and accountable.
“Homophily in grant evaluations” places a familiar social instinct inside one of the most consequential decision-making systems in modern science. Its message is not that reviewers are necessarily acting in bad faith, nor that similarity always produces unfair decisions. Rather, it highlights how ordinary mechanisms of recognition and trust can influence judgments that appear purely technical. As research funding becomes increasingly competitive, understanding these mechanisms is essential. The credibility of science depends not only on the quality of the discoveries it produces, but also on whether the path to those discoveries gives diverse researchers and ideas a fair opportunity to be seen.
Subject of Research: Homophily and similarity-driven bias in scientific grant evaluations
Article Title: Homophily in grant evaluations
Article References:
Fuchs-Schündeln, N., Holub, F. Homophily in grant evaluations. Nat Hum Behav (2026). https://doi.org/10.1038/s41562-026-02554-w
Image Credits: AI Generated
DOI: https://doi.org/10.1038/s41562-026-02554-w
Keywords: homophily, grant evaluation, peer review, research funding, scientific bias, evaluator similarity, diversity in science, academic careers
