Conflict of interest is one of those topics that most grants programmes think they've handled — and most haven't. There's a form somewhere. Reviewers sign it at the start of the process. It says something about relationships with applicants. The file goes in a folder.
That's not a conflict of interest process. That's documentation of a process that doesn't exist.
The difference matters, because conflicts of interest in grant assessment — when they're not properly managed — undermine the integrity of the entire round. An unsuccessful applicant who discovers that a successful applicant was connected to a panel member has grounds to challenge the decision. A board member who learns that the assessment process didn't capture a material conflict has governance concerns. A media inquiry into how a particular organisation was funded can unravel quickly if the COI documentation doesn't hold up.
This guide is about building a conflict of interest process that actually works — not just one that ticks a box.
Conflicts of interest in grant assessment fall into three categories.
Financial conflicts arise when a reviewer has a direct or indirect financial stake in the outcome. This includes: being employed by an applicant organisation, being a director or trustee of an applicant, having a family member employed by an applicant, or having a financial interest in an applicant's work (investments, contracts, etc.). Financial conflicts are the easiest to define and the most serious.
Relational conflicts arise from personal or professional relationships that could — or could reasonably appear to — influence a reviewer's judgment. This includes close personal friendships with applicant staff, prior employment relationships, collaborative professional relationships, or family connections. Relational conflicts are the most common and the most under-declared, because people underestimate the influence of relationship on judgment.
Reputational conflicts arise when a reviewer has publicly taken a position that could compromise their neutrality on a particular application. A reviewer who has publicly criticised an applicant organisation, or who has a history of public dispute with its leadership, has a reputational conflict even if there's no financial or personal relationship. These are the rarest, but they're important in specialist fields where reviewers and applicants frequently interact.
A useful heuristic for reviewers who are uncertain whether to declare is the Three-Ring COI Test. Ask: would a reasonable person looking at this situation think it could influence my judgment, or think it could appear to influence my judgment?
The test has three rings of increasing subjectivity:
Ring 1 (automatic recusal): Financial conflicts. Employment, directorships, family financial interests. These don't require a judgment call — if they exist, the reviewer steps down from that application.
Ring 2 (declare and discuss): Close personal relationships, prior employment within the last three years, collaborative relationships that are ongoing. These should be declared. Whether recusal is required depends on the nature of the relationship and the significance of the grant. This is the ring where the programme manager or COI officer makes a determination.
Ring 3 (professional acquaintance, borderline): Professional acquaintance, occasional collaboration, having reviewed the applicant's work in another context. These should be noted if the reviewer has any doubt, but often don't require recusal — just documentation.
The key principle is: when in doubt, declare. Undeclared conflicts that surface later are far more damaging than declared conflicts that are managed transparently.
Most programmes collect COI declarations at the wrong time. Asking reviewers to declare conflicts when they sit down to assess an application is too late. By then, they may have already read the application, formed impressions, and (consciously or unconsciously) applied that judgment.
The correct sequence is:
Before the round opens: All reviewers should complete a general COI declaration covering their current and recent employment, directorships, family connections to organisations in your funding area, and any other relationships that could be relevant. This declaration should be on file before any application details are shared.
When the applicant list is released: Before applications are opened for review, reviewers should see the full list of applicants and formally confirm whether they have a conflict with any of them. This is the point where application-specific conflicts are identified. Reviewers should not begin assessment until this step is complete.
If a conflict emerges during assessment: If a reviewer encounters an application and recognises a conflict they didn't identify at the list-review stage, they should stop assessment of that application immediately, declare the conflict, and the programme manager should determine whether any assessment they've already completed needs to be set aside.
When a conflict is declared, you need a clear protocol for what happens next.
Full recusal: The reviewer has no involvement in that application. They don't review it, they don't see the assessment, they leave the room during panel discussion, and they don't see the outcome. This is the appropriate response for financial conflicts and significant relational conflicts.
Partial management: In some cases — particularly where a reviewer has subject matter expertise that's difficult to replace — you may determine that they can continue to assess applications where they have no conflict, but are fully recused from the conflicted application. The partial management approach requires careful documentation: what was the conflict, what was the determination, who made it, and what the reviewer's scope of involvement was.
Panel-level disclosure: For minor conflicts (Ring 3), the reviewer may be permitted to participate but the conflict is disclosed to the full panel, which then exercises collective judgment. This is appropriate for professional acquaintances and minor prior relationships, not for material conflicts.
Whatever the determination, it should be documented. The record should show: the declared conflict, who made the management determination, what the determination was, and any conditions attached.
A COI process that doesn't leave an audit trail isn't a process — it's a verbal agreement. The documentation you need is:
This documentation should be retained for at least five years after the round closes, and potentially longer for significant grants. If a decision is ever challenged — through an Official Information Act request, a complaint to a regulator, or litigation — this is what you'll produce.
The COI obligations apply differently depending on the reviewer's role.
Paid independent assessors (contractors engaged specifically for the round) should be treated as the highest-risk category. They often have extensive sector relationships, they're frequently engaged precisely because of their subject-matter expertise, and that expertise means they know everyone. Require full pre-round declarations and application-specific review. Consider whether the same assessor should review applications from organisations in their primary professional network.
Panel members are typically volunteers from the sector, which means they're likely to know applicants personally. This doesn't disqualify them — it's why they're on the panel. But it means COI management needs to be explicit and well-documented. A panel where three of five members recuse themselves from the same high-value application needs a protocol for how that application is still assessed fairly.
Board members often have the most complex conflict landscapes — they sit on multiple boards, have long sector careers, and are typically senior figures who know many applicant organisations. Board-level conflicts are often the hardest to manage because boards tend to think of themselves as decision-makers rather than assessors. If your board approves grant recommendations, board members need the same COI process as assessors.
Process and documentation are necessary but not sufficient. The deeper challenge in COI management is cultural.
In close-knit sectors — and the grants sector is almost always close-knit — people underestimate the influence of relationship on judgment. A reviewer who has known an applicant for ten years doesn't think of that as a conflict. They think of themselves as objective, and they are probably trying to be. But objective trying isn't the same as objective outcome, and the perception of conflict is as damaging to institutional integrity as actual bias.
Building a COI culture means normalising declaration rather than treating it as an accusation. Reviewers should understand that declaring a conflict is a professional obligation, not an admission of impropriety. Programmes that cultivate this culture do better on COI management than programmes that treat the form as an annual compliance ritual.
That starts with leadership. If programme managers and board chairs declare their own conflicts openly and matter-of-factly, that models the expected behaviour. If COI is treated as an embarrassing subject, it will be under-declared.
If you can only do one thing to improve your COI process, make it earlier.
Move your application-specific COI review to before assessors open any applications. Send the applicant list. Collect declarations. Close the loop with the appropriate management determination. Then — and only then — distribute applications to reviewers.
That single change, applied consistently, would eliminate most of the COI risk in most grants programmes. Everything else — the three-ring test, the audit trail, the cultural shift — is built on that foundation.