dissenting:
Appellant Jackson raises three issues in this appeal. The majority, in abbreviated fashion, dismisses each in turn. I find its perfunctory treatment of Jackson’s claims especially troubling because of the importance of some of the questions presented.
The rules surrounding testimony by the defendant permit the introduction of character and prior conduct evidence, evidence unrelated to the enumerated charges. This practice carries with it the risk that jurors may convict a defendant not for the offense he is charged with but because of a general belief that he is a “bad actor.” For example, nothing is so damaging to a defendant’s chances for acquittal as the introduction of evidence that he has previously been convicted of a felony. Given these realities, we have an affirmative obligation to scrutinize carefully the introduction of evidence not directly relevant to the crime charged. Because I believe that at least two of appellant’s objections to the use of such evidence are correct, I dissent.
1. At trial, the prosecution introduced evidence from a fourteen year old bar disciplinary hearing. I agree with the majority that there is no specific time limitation on prior bad act evidence and that the district court has broad discretion in weighing the admissibility of evidence. That being said, I have grave doubts about the introduction of such attenuated evidence against the defendant. As Judge, now Justice, Kennedy said for the court, “Since ... [the defendant’s] involvement was more than ten years old, it detracted only minimally from his credibility.... Disclosing it in court was likely to unduly prejudice the jury.” United States v. Kennedy, 714 F.2d 968, 973 (9th Cir.1983). Fourteen years is substantially more remote than any other bad act evidence ever upheld in this circuit and approximately two years more remote than any bad act ever admitted in any other circuit, if we are to judge by the majority opinion. It seems clear to me that the majority is pushing the limits of discretion, if it has not already broken them. While it is not necessary for me to say which, because in my opinion the way in which the cross-examination on this point was conducted constitutes error, I would suspect the latter to be the case.
The district court not only allowed the defendant to be questioned by the prosecution regarding the old disciplinary action, it permitted the prosecution to impeach him with extrinsic evidence of his prior bad act. The use of such extrinsic evidence is prohibited. The language of Federal Rule of Evidence 608(b) is clear. “Specific instances of the conduct of a witness, for the purpose of attacking or supporting his credibility ... may not be proved by extrinsic evidence.” By its explicit terms, the Rule forbids the precise practice that the district court permitted in this case. See United States v. Whitehead, 618 F.2d 523, 529 (4th Cir.1980) (documents of bar disciplinary proceedings).
*1451The majority attempts to elude the restraints imposed by Rule 608(b) by means of two arguments. First, it argues that the defendant’s signed statement was never introduced into evidence and, thus, falls beyond the reach of the Rule. The majority’s description of events is somewhat disingenuous, because the signed statement was marked into evidence and was a focal point of the cross-examination.
At the beginning of the cross-examination, Jackson was questioned about misappropriating funds from a former client. When he attempted to answer the questions in a manner inconsistent with the signed statement, the prosecution confronted him with the statement and impeached him in front of the jury.
Q: “Sir, you asked him to loan you $2,500 for a period of two weeks. Is that correct?”
A: “No, that is not correct. I know what you are reading from, but I am under oath, also. You asked me a question, and I am answering it.”
After the prosecution placed the statement in front of the defendant and forced him to acknowledge his signature, the following exchange completed this portion of cross-examination.
Q: “Do you admit or deny that you wilfully misappropriated the sum of $2,500 for your own use from Mr. Randolph?”
A: “Well, I signed that statement. Yes.”
Q: “Do you admit or do you deny it?”
A: “Do you want to know what happened, or do you want to know whether I signed the statement admitting it? Yes, I signed it admitting it.”
These exchanges make clear that the prosecution effectively used the written statement at trial. In doing so, it proved the witness’ prior bad conduct through the use of extrinsic evidence. Although the written statement was ultimately not introduced at trial (nor did it need to be since the prosecution had accomplished its purpose without formal introduction), this is irrelevant to the question of whether facts were proved through extrinsic evidence within the meaning of the Rule. It is perfectly apparent to me that marking the statement and forcing Jackson to acknowledge its accuracy to the jury constituted extrinsic proof of the bad act in contravention of the language of the Rule.
Moreover, the prosecution’s use of the evidence at trial not only contradicts the plain language of the rule, it also contravenes the policy underlying the Rule. The purpose of the Rule is three-fold: to limit the degree to which trials become mired in collateral issues, to reduce the surprise of a witness suddenly faced with evidence which would require rebuttal witnesses or documentation, and to curtail the risk of prejudice which “unavoidably attends the introduction of evidence of specific bad acts, since juries are likely to misuse such evidence, particularly (though not only) when the witness in question is a party.” 3 D. Louisell & C. Mueller, Federal Evidence § 306 (1979). See also 3 J. Weinstein & M. Berger, Weinstein’s Evidence § 608[05] (1988).
The use of the evidence in this case implicates all three concerns. First, after the prosecution used the statements to impeach the defendant, the district court would, in fairness to Jackson, have had to let him introduce rebuttal evidence about the bar disciplinary hearing and the purported misappropriation. If, in fact, Jackson had an excuse or contradictory evidence, he had the right to present that evidence. This could have turned the case into a mini-trial over only slightly relevant evidence and mired the participants in a lengthy, collateral dispute. Second, the fact that, in this case, there was no such collateral litigation could conceivably reflect the defense’s surprise about the evidence. Appellant may simply not have had the time or resources to accumulate the evidence necessary to rebut the collateral charge. This is, of course, one of the evils the Rule was specifically designed to avoid. Third, the fact that the prior bad act was the subject of a bar disciplinary proceeding and that the defendant confessed to judgment raises a strong possibility that the jury would misuse and overemphasize the evidence against the defendant.
*1452Moreover, the procedure envisioned by Rule 608(b) was that the cross-examiner could question the witness on the prior acts but ultimately would have to “take the answer of the witness.” If the witness does not come forth with an answer that the attorney finds satisfactory, he make seek an admission through further cross-examination. See 3 Louisell & Mueller, supra. Irrespective of whether the answer from the stand is truthful or not, the Rule plainly bars further inquiry into the substance of the bad act, either by use of evidence derived from the witness or other sources.1
The majority also claims that Rule 608(b) is not properly at issue here because the signed statements constitute a prior inconsistent statement under Rule 801(d)(2) and Rule 613(b). Rule 801 simply defines the written statements as nonhearsay. The Rule is one of competency, not of admissibility. Rule 613(b) lays out prerequisites for impeachment evidence of witnesses, prerequisites that may or may not have been met in this case. However, Rule 613(b), like all the other rules of evidence, is still subject to the specific exclusion principles set forth elsewhere in the rules. Since Rule 608(b) specifically requires exclusion of the statements here, the fact that the statements would have been otherwise admissible under 613(b) or 801(d)(2) is irrelevant.
2. At trial, John Speidel testified that appellant was “short on funds” and was suffering from “financial difficulty.” He further stated that Jackson “couldn’t pay for things he needed to have done.” As the majority opinion accurately recounts, the prosecutor argued to the jury that the defendant’s poverty was the motive for Jackson to violate federal law.
The majority opinion contains an accurate recitation of the various opinions on the question of introducing evidence of poverty to support theories of criminal motivations. But my colleagues, like the evidence against appellant, are “somewhat vague” as to the policy behind the rule and the source of our reluctance to admit proof of mere impecuniosity.
The prosecution introduced the evidence of Jackson’s general financial shortcomings to demonstrate that he had a motive to commit the charged crimes. On a purely theoretical level, I have little quarrel with the government’s argument that there is some statistical correlation between poverty and crime. Many of the problems of our society, both in terms of crime or otherwise, can be traced to systemic and pervasive poverty, especially in the inner cities. Nevertheless, even if there is a link between poverty and crime,2 we have consistently excluded generalized evidence of poverty on the ground that this evidence wrongly casts suspicion on the disadvantaged and affords them unequal treatment with members of more privileged classes; “the practical result ... [of admitting evidence of poverty] ... would be to put a poor person under so much unfair suspicion and at such a relative disadvantage that for reasons of fairness this argument has seldom been countenanced as evidence of the graver crimes ...” 2 Wigmore, Evidence, § 392 (Chadbourn rev. 1979).3 See also United States ex rel. Mertz v. New Jersey, 423 F.2d 537, 542 (3d Cir.1970). I suspect *1453that poor people are already sufficiently disadvantaged by our criminal justice system that we do not need to invite images of stereotype and prejudice into the jury box.
There are times, however, when federal courts have admitted proof of financial need in support of motivation. These cases involve more than mere impecuniosity; rather, they rely on the defendant’s specific and immediate need for funds. See United States v. Feldman, 788 F.2d 544, 556-57 (9th Cir.1986) (overdrawn bank account); United States v. Saniti, 604 F.2d 603, 604 (9th Cir.1979) (per curiam) (drug habit). Thus, the evidence involves substantially more than the mere suggestion that those who are poor are more likely to turn to crime. When the prosecution has attempted to introduce generalized evidence of monetary need that is shared by a great number of people, the federal appellate courts have been virtually unanimous in their condemnation of that evidence. See, e.g., United States v. Zipkin, 729 F.2d 384 (6th Cir.1984); United States v. Reed, 700 F.2d 638, 642-43 (11th Cir.1983).
The majority recognizes the distinction between generalized evidence of poverty shared by many people and the specific and immediate need that marks admissibility. However, without explanation, it baldly states that “short on funds” comes closer to the latter than to the former. This claim borders on the fanciful. The evidence in no way suggested that Jackson was operating under an imminent possibility of financial collapse but instead reflected a status that millions of poor and, indeed, middle class people share these days. It seems clear to me that appellant’s predicament falls within the cases barring admission of evidence. While the error standing alone may not require reversal, when combined with the other errors described above, it warrants a new trial. I dissent.
. Some courts have rejected poverty evidence on strict relevancy grounds. “It is doubtless true that in a large class of cases the poverty or pecuniary embarrassments of a party accused of crime cannot be shown as substantive evidence of guilt. The reasons for the exclusion of such evidence is that in those cases there is no certain or known connection between the facts offered to be proved and the conclusion which is sought to be established by it. It does not follow, because a man is destitute, that he will steal, or that when embarrassed with debt and incapable of meeting his engagements he will commit forgery ...” Commonwealth v. Jefferies, 7 Allen 548, 565 (Mass.1863). Although Chief Justice Bigelow’s theory in Jefferies has a certain attraction, I would side with the more modern approach which argues for exclusion not on relevancy grounds but on the theory that the prejudicial effect outweighs the probative value under Fed.R.Evid. 403.
.Professor Wigmore argues that the rule barring admission of evidence of general poverty should apply principally to crimes of violence and may not apply to merely “peculative crimes *1453(such as larceny or embezzlement).” I, however, believe that the principles supporting the general rule adhere with equal force in the latter context. In fact, because the relationship between poverty and economic crimes may be even stronger in the minds of a jury, the prejudicial effect in such cases is undoubtedly greater.