Skip to content

Conflict-of-Interest Disclosure and Recusal

Disclosure-and-recusal protocol — instantiates Capture-Resistant Institutional Design

Requires decision-makers to declare financial, relational, and career ties to the regulated actor and steps the conflicted party out of any decision those ties touch, so a captured individual can inform but not decide.

Most anti-capture mechanisms work at the level of the institution; this one works at the level of the person. Conflict-of-Interest Disclosure and Recusal requires each decision-maker to declare their ties to the regulated actor — holdings, employment, family, prior and prospective jobs, interlocking board seats — screens those declarations against the matter in front of them, and removes anyone whose ties touch it from the decision. Its defining discipline is a clean split it draws through each conflicted individual: their input and their authority are separated, so a person with valuable but compromised expertise may still inform the body and may not decide. It is the scalpel of the set — precise, person-by-person, decision-by-decision — where the funding firewall and the ex parte rule are structural walls.

Example

A food-safety advisory committee is convened to judge a novel sweetener. Before it meets, every member files a disclosure. The screen flags three: one holds shares in an ingredient supplier that would benefit, one has consulted for the applicant within the look-back window, and one runs a lab funded by a rival manufacturer. The first two recuse from the vote — they may answer the committee's technical questions, then leave before deliberation — while the third's rival funding is disclosed and weighed in the open.

The additive is then judged by members whose interests it does not touch. Crucially, the compromised expertise is not thrown away: the recused consultant's knowledge of the manufacturing process is still available as testimony. What is withheld is his hand on the decision. That is the whole move — keep the expert's information, quarantine the expert's stake.

How it works

What sets this protocol apart is that it targets ties, not contacts or budgets:

  • Standing disclosure obligation. Members declare assets, employment, family interests, and — the part naïve regimes miss — interlocks: shared directorships, co-investments, and pending job talks with the regulated actor.
  • Match against the matter. A screen compares each declared tie to the specific decision at hand; a conflict is relational, not absolute, so the same member may be clear on one matter and recused on the next.
  • Tiered response. Depending on materiality: disclose-and-proceed, recuse from the vote, or recuse entirely with no access to the file.
  • Inform-but-don't-decide. The conflicted party is separated into an advisory role — testimony, not deliberation — so expertise survives while the bias is walled off from authority.

Tuning parameters

  • Disclosure scope — assets only, or relationships, family, past employment, and future job discussions too. Wider scope catches interlocks and the revolving-door tie but costs privacy and can deter qualified people from serving.
  • Materiality threshold — how large a stake triggers recusal. A low threshold is strict but can empty the room of everyone who understands the field.
  • Recusal depth — leave the vote, leave the room, or lose access to the file entirely. Deeper is cleaner but forfeits more expertise.
  • Look-back and look-forward window — how far past and prospective ties count; the pending-job-offer is the classic hard case.
  • Verification — self-declared honor system versus disclosures checked against holdings and registry data. Verified is far stronger and far heavier.

When it helps, and when it misleads

Its strength is that it reaches the person-level channel the structural mechanisms miss, and it does so without discarding scarce expertise — it quarantines the stake, not the knowledge. It is cheap, legible, and legally familiar.

Its failure modes are twin and opposite. Set too strict, recusal empties the panel of the only people who understand the industry, handing decisions to the less-qualified; set too loose, it becomes box-ticking. And disclosure is not neutralization — a conflict everyone acknowledges and then ignores can be worse than a hidden one, because naming it seems to license the biased advice that follows.[1] Interlocks and cultural capture also slip through, since they resist reduction to a declarable line-item. The discipline that keeps it honest is to verify disclosures, tier the response to genuine materiality, and treat a declared conflict as a trigger for action, never a substitute for it.

How it implements the components

Conflict-of-Interest Disclosure and Recusal fills the individual-conflict slice of the archetype:

  • conflict_and_interlock_screen — the disclosure-and-matching engine that flags direct conflicts and the interlocking ties naïve screens miss.
  • advisory_role_separation — recusal enforces the line between a conflicted party informing the body and holding authority over its decision.

It does not insulate the institution's money from the regulated actor — that is the Funding Firewall Rule — nor govern the timing of staff crossing to or from industry (Revolving-Door Cooling-Off Period), nor log the contacts themselves (the Privileged Access Log).

Notes

Recusal handles the ties a member holds now; it stops at the temporal seam where a person crosses between the institution and the industry — the pending offer, the recent departure. That gap belongs to the Revolving-Door Cooling-Off Period, and the two are designed to abut without overlapping.

References

[1] Research on conflict-of-interest disclosure has found a perverse effect: revealing a conflict can license the discloser to give more biased advice while leaving the recipient feeling forewarned rather than protected. It is why disclosure is treated here as a trigger for recusal, not as a remedy in itself. ↩