Scope of the Prohibition

Scope of the Prohibition

Model Rule 3.7(a) states a prohibition and then three exceptions: a lawyer shall not act as an advocate at a trial in which the lawyer is likely to be a necessary witness, except where the testimony relates to an uncontested issue, the testimony relates to the nature and value of legal services rendered in the case, or disqualification of the lawyer would work substantial hardship on the client.¹ The first exception, for testimony on an uncontested issue, shows how the rule is built: each exception lifts a bar that already applies.² Every fact pattern therefore begins with a threshold question, whether the prohibition reaches the lawyer at all.

The reach is narrower than the shorthand that a lawyer may not be a witness in her own case. Two conditions must both be met, and neither suffices alone: the lawyer must be acting as an advocate at a trial, and the lawyer must be likely to be a necessary witness at that trial.³ A lawyer can know the underlying events better than anyone, and can even be a named witness, without triggering the rule when either condition is missing. The pairing follows the harm. Confusion over which role is speaking arises only when one person holds both before the same factfinder, so a lawyer who testifies but never argues, or argues but never testifies, has no second role to blur.

Each condition carries limiting words that decide close cases. “Advocate at a trial” leaves out much ordinary legal work, and “likely” and “necessary” leave out lawyers whose evidence is merely relevant or available from another source. The trial setting comes first, because it decides which parts of a lawyer’s work the rule touches at all.

Advocate at Trial

The rule names its setting, and the setting is a trial: a lawyer shall not “act as an advocate at a trial” in which the lawyer is likely to be a necessary witness.⁴ The text says nothing of investigating facts, drafting pleadings, or preparing discovery. Students often read the rule as a ban on all advocacy once the lawyer’s testimony looks likely, and the text goes no further than the trial itself.

The anchor to trial follows from the harm the rule targets. When one person both testifies and argues, each role damages the other. A lawyer who is both counsel and witness becomes more easily impeachable for interest and so may be a less effective witness; opposing counsel may be handicapped in challenging the credibility of a lawyer who also argues the case; and the lawyer lands in the “unseemly and ineffective position of arguing his own credibility,” because an advocate’s function is to advance the cause of another while a witness’s is to state facts objectively.⁵ Interviews, drafting, and discovery all happen away from the factfinder, so none of them puts the lawyer’s partisan voice and factual voice before the same audience.

Silence about pretrial work in Rule 3.7 does not leave that work unregulated, because other rules reach it. Rule 4.2 bars a lawyer from communicating about the subject of the representation with anyone the lawyer knows to be represented in the matter, absent the other lawyer’s consent,⁶ and Rule 3.8(b) and (c) require a prosecutor to take reasonable steps to facilitate an accused’s access to counsel and forbid procuring a waiver of important pretrial rights from an unrepresented accused.⁷ Those rules, and not Rule 3.7, address how a lawyer may deal with represented persons and unrepresented accused persons before trial.

A lawyer who sat in on a negotiation can still help litigate the dispute that follows, as this example shows.

The trial the rule covers includes a trial before a judge. The Model Rule says “a trial” without limiting it to juries, and the reason holds up: most of the difficulties of taking on the roles of advocate and witness are present whether the testimony goes to a jury or a judge, because the roles are entirely irreconcilable and should not be undertaken by a single individual.⁹

A California appellate decision involving a lawyer who was counsel in a custody and visitation dispute and also headed the household whose conditions were contested shows the point under a narrower state rule. The court’s footnote quotes California’s rule 5-210, which barred a member from acting as an advocate “before a jury” that would hear the member’s testimony (the court supplied the italics) and listed the client’s informed written consent among its exceptions.¹⁰ The decision nonetheless set out exceptions worded as in Model Rule 3.7(a)(2) and (3), for testimony on the nature and value of legal services and for substantial hardship to the client, and it relied on the rationale quoted above from Donaldson.¹¹

A court deciding a disqualification motion before trial can look to what has already happened in the case. The California opinion quotes a passage asking whether, at one point, the lawyer was “arguing as an attorney, testifying as a witness, making an offer of proof for his client, or maybe for himself as a witness,” and observing that the confusion “already has been manifested in this case in a preliminary law and motion matter.”¹² State wording also matters. The California rule quoted in that decision was limited to jury trials and so was narrower on its face than the Model Rule; a question that names a state rule takes that rule’s own terms, while a question that cites only the Model Rule takes its trial-stage trigger. A later section on state variations returns to these differences.

Likely Necessary Witness

Two qualifiers, “likely” and “necessary,” decide whether a lawyer who would act as trial advocate falls under the prohibition, and they do so before any exception comes into play.¹⁴ “Likely” asks how probable it is that the lawyer will take the witness stand; “necessary” asks whether the trial needs what the lawyer would say. The rule requires both. Only a lawyer who already fits that description needs an exception, the first of which covers testimony on an uncontested issue.¹⁵ Treating every lawyer with personal knowledge as disqualified skips the question that matters, since a lawyer can know a great deal about a dispute and still be neither likely to testify nor needed as a witness.

The probability qualifier fits the harm the rule targets, which arises only if the lawyer actually testifies. A court can find likelihood in the lawyer’s position in the facts. In the California custody and visitation dispute, the lawyer headed a household that was the subject of controversy and conflicting averments, and the court called it a virtual certainty that his testimony would be necessary.¹⁶ That position in the contested events made his testimony close to inevitable before the trial began.

Necessity asks whether anyone else can supply the same facts. In Johnston, a court of appeals reviewed a refusal to let a government prosecutor testify at a pretrial suppression hearing and called allowing an advocate to testify “a situation to be avoided if possible,” permitted only “in extraordinary circumstances and for compelling reasons, usually where the evidence is not otherwise available.”¹⁷ The Fourth Circuit in Fogel likewise said that courts have generally frowned on counsel’s testifying, but that when “the testimony is important and no other witness would be able to supply it,” it may be allowed.¹⁸ Both decisions ask whether an advocate may testify at all, while Rule 3.7 asks whether the advocate must step aside, but the common question is whether the evidence is available elsewhere.

The availability test cuts both ways. Where other witnesses or sources can prove the facts, stepping aside gains the trial nothing, because the factfinder will hear them from someone who is not also arguing the case, and testimony that would merely help the client is not necessary if someone else can give it. Where the lawyer is the only source, the court cannot remove the witness role without losing the evidence, so the only way to keep one person from both arguing and testifying before the same factfinder is to change who argues. Necessity thus limits the prohibition to cases in which the role conflict cannot be avoided any other way.

A related question arises when a lawyer already in the case seeks to testify, and evidence law answers it, not the ethics rules. A lawyer for a party is not per se incompetent to testify, but the court has wide discretion to bar the testimony, especially when the lawyer testifies in her client’s favor.¹⁹ Judges routinely exercise that discretion where other sources of evidence on the fact are available or where the need for the lawyer’s testimony could easily have been foreseen and avoided.²⁰ A later section develops the line between the two bodies of law. For now, availability elsewhere and avoidability are the two themes that run through the cases that follow.

A lawyer can create his own necessity, and Givens shows how. The government described that case as raising a conflict-of-interest issue, and the court answered that the label was not quite apposite. The case was not one in which a lawyer for one defendant might have to cross-examine a former client who had turned state’s evidence, and not one in which the lawyer was implicated in his own client’s wrongdoing. “Instead of a conflict between different clients’ interests or between a client’s interests and his attorney’s self-interest, the problem here is a conflict of courtroom roles, of blurred distinctions between the roles of advocate and witness.” The lawyer created that problem by interviewing a witness without another person present.²¹ The characterization shows that the advocate-witness rule protects the structure of the trial and serves a purpose separate from the conflict-of-interest rules, a separation that matters again when this chapter reaches lawyers in the same firm.

Two interviews with identical content show how the outcome turns on who can prove what was said.

Both lawyers hold the same knowledge, and the difference lies in whether anyone else can prove it. Importance counts as well. Fogel speaks of testimony that is “important,” and in Givens the court, noting that the lawyer’s testimony would have been relevant only to a collateral issue, the credibility of one government witness, thought it unlikely that prejudice would have been large.²² A later section on enforcement returns to that observation.

Lineups show the same pattern from the defense side. In State v. Caldwell, a state court held that defense counsel may not testify about events at the lineup unless he withdraws from the case, because “[a]n advocate who becomes a witness is in the unseemly and ineffective position of arguing his own credibility.”²³ The discussion of Wade points toward a way to avoid that bind. The Court there left open whether substitute counsel might suffice where notification and presence of the suspect’s own counsel would result in prejudicial delay, and the commentary concludes that substitute counsel would suffice where sufficiently apprised of the circumstances to represent the defendant effectively.²⁴ Read together, the two suggest that who attends a lineup bears on who may later have to testify about it.

A footnote in Elliott v. Google reaches a similar result for survey evidence: the court agreed with the district court that, if the surveys were admitted, Elliott’s counsel would need to withdraw in order to offer testimony on the survey results.²⁵ The footnote states only that conclusion, and it shows that when counsel’s own testimony would be the vehicle for evidence, advocacy and testimony cannot share one lawyer at trial.

The interview, the lineup, and the survey each produce a necessary witness the same way: the lawyer is the only person positioned to supply some fact. A second person present at the interview, such as an investigator or a colleague, can testify to what happened, so the lawyer’s own testimony is unnecessary and the lawyer keeps the ability to try the case.²⁶ A lawyer who goes alone creates a need that could easily have been foreseen and avoided, which is a factor judges weigh against allowing the testimony.²⁷