Concept
Credibility Cross-Examination
Scope of examination
Credibility Cross-Examination
The first sentence of Rule 611(b) opens two areas to cross-examination: the subject matter of the direct examination and matters affecting the witness's credibility.1 The earlier sections of this chapter dealt with the first area and with how generously courts measure the subject of a direct. The second area runs on a different footing. A credibility question needs no connection to anything the direct covered, and it is allowed even when it opens an entirely new subject, such as whether the witness holds a grudge against a party, has a financial stake in the case, or has been convicted of a crime like perjury or fraud.2
A party that calls a witness asks the jury to believe two things: that the witness is credible and that the facts the witness related are accurate.3 Cross-examination exists to weaken the direct testimony, and it can do so in two ways. It can show that the facts are incomplete or inaccurate, or it can suggest that the witness cannot be trusted to be truthful and accurate. Because the second route is always open, credibility is always implicitly at issue at trial.4 If credibility questions had to be tied to the substance of the direct, a cross-examiner could never expose a grudge or a financial interest, since a party rarely opens its own witness's direct by announcing the witness's bias.
Two limits remain. The cross-examiner must challenge credibility solely by interrogating the witness and cannot call outside witnesses during the direct examiner's case-in-chief, though evidence relating to the credibility of opposing witnesses may be offered later, in the cross-examiner's own case.5 And the trial judge keeps control. Under Rule 611(a), the court exercises reasonable control over the mode and order of examining witnesses so as to make procedures effective for determining the truth, avoid wasting time, and protect witnesses from harassment or undue embarrassment.6 A question that falls within the credibility category can therefore still be restricted if it only harasses the witness or burns time without helping the jury judge the testimony.
Credibility attacks take several forms. The first subsection covers facts that give a witness a reason to shade the truth, and the second covers ways of testing the direct testimony itself.
Bias and Interest
Bias cross-examination lives in the credibility clause: a lawyer who wants the jury to discount a witness for partiality may ask about it whether or not the direct touched the subject.7 Maryland's rules name the targets. A witness may be attacked by proof that the witness is biased, prejudiced, interested in the outcome of the proceeding, or has a motive to testify falsely.8 The Maryland decisions below apply Maryland Rule 5-611(b)(1), whose scope language tracks the federal rule: cross is limited to the subject matter of the direct and to credibility, and the court may in its discretion permit inquiry into additional matters as if on direct.9
Partiality can be shown in two ways. Evidence about the circumstances of the witness's situation can make it probable on its face that the witness leans toward one side, and the witness's own conduct and statements can support an inference that the feeling exists.10 Neither route depends on the direct. A witness's stake belongs to the witness's situation, not to the accident, contract, or crime on trial, so a direct examiner who stays on those events has no reason to mention it. The cross-examiner needs a good-faith basis for the matter a partiality question asserts, and may ask about the witness's statements as well as the witness's situation and conduct.11
Experts supply the clearest examples because they are paid for their time. In Goldberg v. Boone, the Maryland court recalled an earlier case in which defense counsel showed that the plaintiff's expert had testified fifty to sixty times for malpractice plaintiffs, about twenty-five of those times for clients of the plaintiff's lawyer, that 80 percent of his appearances were for plaintiffs, and that he had earned between $30,000 and $50,000 from testifying in the preceding year, most of it from that lawyer. The court upheld the ruling that let the jury hear the total compensation because the question sought to expose potential bias.12 Maryland has long treated the cutting off of such questions as error: in Wise v. Ackerman, decided in 1892, the court held that a trial judge abused his discretion by sustaining objections to questions about remarks the plaintiff's expert had made in an unrelated case about his ability to obtain large verdicts.13
The Goldberg court collected the lines of questioning that expose an expert's bias, and they make a working checklist for a defense lawyer cross-examining a plaintiff's causation expert:
- how much the expert is being paid for the services in the present case;
- how often the expert testifies in similar kinds of cases;
- whether the expert customarily testifies for plaintiffs or for defendants;
- whether a particular party or attorney frequently employs the expert, and how much income that work produces; and
- what amount or percentage of the expert's total income comes from lawyer referrals or testimony.14
A direct examination about causation never mentions money, and none of these questions needs it to, because each tests whether the expert has a financial reason to favor the side that retained him. Not every jurisdiction goes this far. Authorities disagree about questions on the amount of earlier compensation from the same party, the share of the witness's income that comes from testifying, and the mere fact of prior testimony for similar litigants. The better view leaves those inquiries to the trial judge's discretion and permits them, because financial interest can exert a powerful biasing effect and many professional experts earn large fees by spending more time in court than in the laboratory or hospital.15
Stouffer v. State marks the other edge. The appellant there argued that evidence about a man named Baltimore would place the police investigation in context and show the police's bias in building their case. The Maryland appellate court found that no proffer ever explained why the evidence was relevant, and it observed that Baltimore never testified, so his bias was not in issue.16
Stouffer yields two conditions for a bias question. The first is a proffer, an explanation to the judge of the theory on which the answer shows bias. Maryland courts give wide latitude for bias or prejudice but do not let questioning stray into collateral matters that obscure the trial issues and confuse the fact finder, so the judge balances, and a judge who is never told the theory has nothing to weigh.17 The second condition is a link to a witness who actually testified. Bias is a trait of a witness, and its only use is to help the jury decide how far to credit what that witness said, so a payment to someone who never took the stand colors no testimony. In the expert cases the theory and the testifying witness were plain from the question itself, while in Stouffer neither was present.
Appellate courts review the extent of this cross for abuse of discretion resulting in substantial harm, and in practice they find abuse more often when the trial judge cut the examination short than when the judge let it run long.18 Matters of partiality are also not collateral, so the witness's answers do not bind the cross-examiner and other evidence may contradict them, though Rule 403 and the protection against harassment or undue embarrassment in Rule 611(a)(3) still apply.19
In a criminal case the latitude is widest when the defense probes the partiality of a government witness. The accused has a constitutional right to cross-examine on any matter that explains, modifies, or discredits the direct testimony, and foreclosing inquiry into dropped charges, coindictee status, pending charges, paid-informant status, immunity, promised or expected leniency, a pending sentencing, protective custody, or a cooperation plea agreement may be reversible error.20 A witness whose own liberty depends on the prosecutor's satisfaction has an evident reason to give the account the prosecutor wants, and the jury can discount that account only if it learns of the arrangement.
The question on an exam is therefore not whether the direct covered the fee or the deal. It is whether the inquiry exposes the partiality of a witness who testified, and whether counsel can explain why it matters.
Testing the Direct Testimony
Bias exposes a reason for a witness to shade the truth, but the credibility clause also lets a cross-examiner test the testimony itself.22 The examiner can ask whether the witness's memory and perception match the confidence of the direct, whether the witness has told the story differently before, and whether the witness's record raises doubt about honesty. Each of these attacks may open a topic the direct never mentioned, and each passes the scope limit because its aim is to help the jury weigh the testimony already given.
The measure of relevance changes with the aim. When cross targets the historical merits, the question is whether the answer sheds light on an issue in the case. When the cross-examiner attacks credibility instead, the question is whether the answer aids the trier of fact in appraising the witness's credibility and assessing the probative value of the direct testimony, and an answer can meet that test while bearing on no merits issue.23 Credibility inquiry therefore extends to whatever tends to explain, modify, qualify, discredit, or otherwise shed light on the direct, whether or not the same matter also supports the adversary's case in chief.24
Exploratory questions are the clearest example. The cross-examiner, who often has not had the chance to interview the witness, asks about details somewhat remote from the main inquiry in order to try out the witness's ability to remember detailed facts like those recited on direct, the capacity to perceive accurately, or the willingness to tell the truth without distortion or exaggeration. This kind of inquiry is part of the tradition and art of cross-examination, and some of the most famous successful cross-examinations are of this variety.25 A rule that demanded a merits connection for every question would cripple the technique, because the examiner cannot know before asking what the answer will reveal about the witness.
Prior inconsistent statements and convictions share a feature with the exploratory question: neither has any necessary relevance to the historical merits, and courts recognize that a rule strictly limiting cross to merits questions would cripple both.26 A witness who has told the story differently before gives the jury a reason to doubt the version on the stand, and a witness convicted of a crime that raises doubts about honesty, such as perjury or fraud, gives the jury a reason to question the witness's word generally.27 Because the credibility clause admits either subject even though the direct never touched it, a scope objection fails. The rules on how these two kinds of impeachment may be carried out are a separate subject, and this unit is concerned only with the scope gate.
The cross in United States v. Lara tested the direct from two directions. Perry took the stand in his own defense and denied that the Vandergroen incident had anything to do with the Latin Kings, although two government witnesses had tied it to the gang's code. Because the direct included that denial, the government was entitled to test its veracity, and one way to do so was to ask about other crimes Perry had committed under the gang's auspices.28 Perry's direct also tried to exonerate his codefendant Lara by portraying him as unaware, until it was too late, that the carjacking or the killing would happen. On cross, the government sought to show that this version conflicted with Perry's earlier statement to the police, and it suggested that he was covering for Lara out of a perceived duty as a Latin King not to “rat” on fellow gang members. The question about correspondence in which Perry solicited punishment for his loose-lipped girlfriend was asked in that setting, and the court found no abuse of discretion in the judge's ruling that these lines of questioning were not beyond the scope of the direct.29 Both lines attacked the reliability of what Perry said on direct. The court added that, whether or not the questions fell within the scope of the direct, it could not call the overruling of the objections an abuse of discretion, because the questions took up only a fraction of the cross and bore a close relationship to major trial issues.30 A later section takes up that discretion.
Testing an expert has one more tool. When a publication is used to impeach an expert, it is admitted only to attack the quality of the expert's reasoning, not as substantive evidence, and on request the judge gives a limiting instruction under Rule 105.31
Because credibility questions are often exploratory, the cross-examiner frequently cannot say in advance how an answer will bear on the witness, and too tight a rein would rob the examination of its force. The same looseness risks unfair prejudice and wasted time, so the trial judge keeps discretionary control over how far such questioning goes. Appellate courts overturn that control only for abuse resulting in substantial harm to the complaining party, and they find abuse more often when the judge cut the examination short than when the judge let it run long. In a criminal case they are especially likely to find it when the judge restricted cross probing a prosecution witness's bias.32 The judge's general authority under Rule 611(a) points the same way: the court controls the examination to make it effective for determining the truth, to avoid wasting time, and to protect witnesses from harassment or undue embarrassment.33
The second step does the real work, because the question that fails is almost always one for which the cross-examiner cannot name a theory. The usual error is to demand a merits connection for a credibility question, when the test is whether the answer helps the jury weigh this witness. The opposite error is to treat the word “credibility” as automatic permission, when a question about a third party's drinking habits tests nothing the witness said.