Consulting to a Court on Eyewitness Identification Reform
A state court's criminal rules committee has invited Dr. P, an expert in eyewitness memory, to advise on a proposed rule requiring double-blind, sequential photo lineups. During the consultation Dr. P learns the committee chair is a sitting judge who has publicly criticized defense experts as 'junk science.' Dr. P is also retained as a paid expert in two pending cases in the same county, one of which will be heard by that judge. The committee asks Dr. P to draft the rule's language and to estimate how many convictions the rule would overturn annually.
1. Identify the role and conflict issues Dr. P faces and how they should be managed.
Read the worked answer
Dr. P occupies two roles: neutral consultant to a rulemaking body and retained expert in adversarial litigation. These create a conflict of interest and an appearance problem. The committee chair's public hostility to defense experts compounds it: Dr. P's recommendations could be characterized - fairly or not - as advocacy shaped by his litigation work, and his litigation work could be characterized as benefiting from a rule he helped write.
Management: (1) Disclose the retentions to the committee in writing, including the identity of the judge and the pending cases; (2) ask the committee to determine whether recusal from the affected cases or from the consultation is required; (3) if he continues consulting, keep the two roles firewalled - no use of committee deliberations in litigation, no use of case materials in the consultation; and (4) document the disclosures. If the committee cannot accommodate the conflict, declining the consultation is the cleaner course.
2. How should Dr. P respond to the request to estimate how many convictions the rule would overturn?
Read the worked answer
He should decline to produce the requested number as framed. The question asks for a policy prediction that the research literature cannot supply: studies establish that double-blind and sequential procedures reduce mistaken identifications under experimental conditions, but translating that into a count of overturned convictions in this jurisdiction requires data Dr. P does not have - the number of lineups conducted, the rate of contested identifications, the courts' application of harmless-error doctrine, and the retrofit of past cases.
Offering a number would be overclaiming and would lend false precision to a contested policy debate. The appropriate response is to state what the evidence supports (direction and approximate magnitude of error reduction in controlled studies), what it does not (jurisdiction-specific conviction counts), and what data the committee would need to model it. He can also flag that the rule's effect depends on implementation fidelity - a rule on paper without training and audit may produce little change.
If the committee wants a quantitative estimate, Dr. P should recommend that it be produced by someone with access to the jurisdiction's case data, with his role limited to specifying assumptions.