Reviewing a Grant Proposal

Reviewing a proposal is not reviewing a paper, and applying paper-review habits produces reports that panels cannot use.

Schemes differ substantially in criteria, weighting and process. Read the assessment criteria you were sent and score against those, not against your own. For a concrete operational example of how a workplace tool approaches a related coordination problem, see further details.

The differences that matter

There is no revision. The applicant cannot respond, cannot clarify, cannot fix. A comment that would prompt a revision in a paper simply counts against them here.

You are ranking, not assessing. The panel funds a fixed number. Your score is a position in a queue, and a report saying "this is good" without differentiation is not usable.

It is prospective. You are judging whether work will be done, not whether it was done well. Feasibility and track record therefore carry weight they do not carry in paper review.

Money is being allocated, and a proposal that is over- or under-resourced is a real problem rather than a presentational one.

The consequences are large. For an early-career applicant a grant decision can determine whether they continue in research.

What to assess

Is the question worth answering? Would the answer matter, and to whom.

Can it be done? Is the design capable of producing the answer, in the time, with the resources. This is where most proposals are weakest and where experienced reviewers concentrate.

Can these people do it? Track record, team composition, environment.

Is the plan realistic? Timeline, milestones, dependencies. A three-year plan containing four years of work is a common failure.

Is the budget justified and proportionate? Both padding and under-asking are problems, and both are visible.

Does it fit the scheme? A strong proposal in the wrong call should be scored against the call's criteria, not against your view of its merit.

What to write

Be specific about the weaknesses, because the panel is comparing and "somewhat unclear" does not discriminate.

Distinguish fatal from fixable-in-principle. A design that cannot answer the question is different from a proposal that describes a good design badly. Both may be low scores; the panel needs to know which.

Say what is genuinely strong, and why. Panels are ranking, and a report that lists only concerns for a proposal you scored highly is contradictory and unhelpful.

Address each criterion you were asked about, in order. Panel members are comparing across reports and a report that ignores the structure is harder to weigh.

Say where your expertise ends. If you cannot assess the technical core, the panel needs to know. See reviewing outside your area.

Score consistently with the text. A glowing report with a middling score, or the reverse, is the most common problem panels report.

Things to be careful about

Do not penalise a proposal for not being the project you would do. The most common failure, exactly as in paper review.

Do not require preliminary data beyond what the scheme asks for. Some calls exist precisely to fund work that has none.

Watch the track record criterion for structural bias. Publication counts, prestige of previous institutions and network visibility all favour people who have already had resources. Many schemes now ask you to assess outputs relative to opportunity — read that instruction and apply it.

Career breaks and non-linear paths are common and are usually explicitly provided for. Do not treat a gap as a gap.

Do not review a proposal from someone you cannot assess impartially. Recent collaborator, competitor on the same question, your own former student. Declare it and decline. See declining academic requests.

And do not use the ideas. Proposals are confidential, they contain unpublished plans, and this is one of the clearer obligations in the whole system.

The asymmetry worth remembering

The applicant spent weeks on this. Frequently more, alongside everything else, with a low expected probability of success.

They will read your report, in most schemes, and they cannot reply.

A dismissive two-line review is a poor return on that, whatever the merit of the proposal. Specific criticism helps them resubmit; vague criticism does not.

And rejection is the normal outcome. Success rates in many schemes are low enough that most good proposals are rejected, which is worth holding in mind when writing about one that was not funded.

Timing

Review by the deadline. Panel schedules are fixed and there is no slack.

Decline promptly if you cannot. A late review either delays the panel or is excluded, and both are worse than an early refusal.

For panel members

Read the proposal, not only the reports. Reviewers vary in calibration and a panel that averages scores mechanically is averaging noise.

Notice the shape of the funded set — by career stage, by institution, by subfield. The pattern is visible if anyone looks and it usually is not looked at.

Discuss the borderline cases, which is where the decisions actually are.

And give feedback. For an unsuccessful applicant, a few sentences of panel commentary is the difference between a resubmission that improves and one that repeats.

The short version

No revision, ranking not assessment, prospective rather than retrospective.

Feasibility is where proposals are weakest and where your attention is most useful.

Be specific, because the panel is discriminating between proposals and vague comments do not.

Score consistently with what you wrote, and say where your expertise ends.

And remember they cannot reply — a report that is dismissive costs someone weeks of work and returns nothing.

For an independent authoritative reference, consult COPE guidance for publication-ethics guidance for editors and reviewers.