Monthly Archives: January 1984

1984-1985 – Impeaching Defendants with Their Prior Convictions: Reconsidering the Dangerous Propensities of Character Evidence after People v. Castro, 18 UC Davis Law Review 681-719 (OCR)

FORMATTED PDF: 1984-1985 – Leon Letwin 18 U.C. Davis L. Rev. 681 1984-1985 Impeaching Defendants With Their Prior Convictions – OCR

18 U.C. Davis L. Rev. 681 1984-1985

Impeaching Defendants With Their

Prior Convictions: Reconsidering the

Dangerous Propensities of Character

Evidence After People v. Castro

Leon Letwin*

Proposition 8, adopted in 1982, called for the admission of “any”

prior felony conviction “without limitation” as impeachment evidence. In

People v. Castro, the California Supreme Court held that only convictions

involving “moral turpitude” were admissible under the terms of the

Proposition, despite its seemingly unqualified language. Even as to this

class of convictions, admissibility further depended on whether the offer

survived the trial court’s discretionary power to exclude evidence when

its prejudicial potential outweighed its probative value. The decision gave

little guidance as to the meaning of “moral turpitude”; as to how the

trial court was to exercise its discretion; or as to the scope of appellate

review of the trial court’s discretion. This Article identifies four models

for assessing the probative value of such convictions, including the majority’s

“moral turpitude” standard, and criticizes each. It takes the view

that such convictions are all but irrelevant in the special context of a

defendant testifying on her own behalf Ideally, this would argue for the

flat exclusion of such convictions; absent such a rule, however, Evidence

Code section 352 provides a sufficient protective mechanism if trial

courts are encouraged to employ it to guard against abuses in the use of

prior convictions and are reinforced in this effort by adequate appellate



Among Alexis de Tocqueville’s memorable observations was his

proposition that political issues ultimately reach American courts enshrouded

as legal issues.1 He might have added that they often erupt in

* Professor of Law, University of California, Los Angeles. Ph.B 1948, University of

Chicago; LL.B. 1952, Wisconsin; LL.M. 1968, Harvard University.

1 2 A. DE TOCQUEVILLE, DEMOCRACY IN AMERICA 290 (H. Reeve ed. 1954).

— 18 U.C. Davis L. Rev. 681 1984-1985

University of California, Davis

the less obviously political terrain of procedural and evidentiary rules.

A case in point is the current controversy over the admissibility of a

criminal defendant’s prior convictions for impeachment purposes.2

From one perspective, this is a matter of technical, if not abstruse, evidentiary

doctrine located in the complex domain of character evidence

and of concern primarily to legal practitioners. Prior convictions are

one branch of the much debated topic of using character evidence as a

basis for inferring conduct on a particular occasion.’

2 In its most recent form, the issue was triggered by the adoption of Proposition 8, a

California constitutional initiative, on June 8, 1982. That Proposition, among numerous

other things, amended article I of the California Constitution by adding § 28(f),

providing in part: “Any prior felony conviction of any person in any criminal proceeding,

whether adult or juvenile, shall subsequently be used without limitation for purposes

of impeachment or enhancement of sentence in any criminal proceeding.” CAL.

CONST. art. I, § 28(0.

For a general treatment of impeachment by prior conviction, see 3A J. WIGMORE,

EVIDENCE IN TRIALS AT COMMON LAW §§ 980, 980a, 985-987 U. Chadbourn rev.

  1. 1970) [hereafter J. WIGMORE]. For a review of the debate surrounding impeachment

by prior conviction and its resolution in Rule 609 of the Federal Rules of Evidence,


at 609-1 to -45 (1982). An interesting recent consideration of character evidence’s uses

and abuses is found in Uviller, Evidence of Character to Prove Conduct: Illusion,

Illogic and Injustice in the Courtroom, 130 U. PA. L. REV. 845 (1982); see also J.

WEINSTEIN & M. BERGER, supra, at 609-44 to -45 (extensive bibliography); Ladd,

Credibility Tests – Current Trends, 89 U. PA. L. REV. 166 (1940); Mendez,

California’s New Law on Character Evidence: Evidence Code Section 352 and the

Impact of Recent Psychological Studies, 31 UCLA L. REV. 1003, 1003-41 (1984).

‘ The particular use of character evidence of concern in this Article is to assist in

evaluating the truthfulness of a witness’s testimony at trial. CAL. EVID. CODE §§ 780,

785-788 (West 1966). Character evidence can also be used to draw inferences about a

person’s primary conduct – whether the defendant committed the charged criminal

act. The fact, for example, that a person had committed several robberies might be

considered useful evidence to show that she possessed a general disposition toward lawlessness,

violence, or dishonesty. This disposition, in turn, might indicate that she committed

a similar crime on the charged occasion. This use of character evidence to prove

conforming conduct is, however, generally barred by California law. CAL. EVID. CODE

  • 1101(a) (West 1966); see also People v. Spearman, 25 Cal. 3d 107, 116, 599 P.2d 74,

78, 157 Cal. Rptr. 883, 887 (1979); People v. Antick, 15 Cal. 3d 79, 96-97, 538 P.2d

43, 54-55, 123 Cal. Rptr. 475, 486-87 (1975); People v. Beagle, 6 Cal. 3d 441, 453,

492 P.2d 1, 8, 99 Cal. Rptr. 313, 320 (1972). This general doctrine is subject to the

exceptions set forth in CAL. EVID. CODE §§ 1102, 1103 (West 1966 & Supp. 1985).

Although I have distinguished between two potential uses of character evidence, they

are, analytically, two aspects of a single technique: the offer of character evidence as

circumstantial proof of conforming behavior on a given occasion. In one use, the evidence

is thought to elucidate a defendant’s primary behavior; in the other, it is offered

to elucidate the defendant’s behavior in court as a witness on her own behalf, i.e., to

help resolve the veracity of her testimony. Though the two uses are analytically the

[Vol. 18:681

— 18 U.C. Davis L. Rev. 682 1984-1985

Impeaching Defendants

Of course, this issue is not merely one of concern to practitioners;

many people do take their character evidence seriously. The subject has

excited continuing, indeed irrepressible, interest in the courts, among

scholars,’ and in the overtly political arena. The admissibility of prior

convictions for impeachment purposes was “extremely controversial” 5

and the most fully considered issue as the Federal Rules of Evidence

worked their way through Congress. In 1982, the subject erupted as a

question for direct voter resolution in the politically charged context of

a California constitutional initiative on Proposition 8, the so-called

“Victims’ Bill of Rights.”6 The initiative was aimed at changing major

portions of the criminal process, including some of the rules limiting

the use of prior felony convictions that had developed in California case

law.7 With virtually no information about those limits provided to the

voters,’ they were asked, among other things, to decide whether prior

same, practical and historical reasons have combined to produce separate rules, doctrines,

and analyses to cover each.

Still unresolved is the extent to which the enactment of another provision of Proposition

8, CAL. CONST. art. I, § 28(d) (the “truth in evidence” provision), effectively abolishes

present statutory rules excluding character evidence, such as CAL. EvID. CODE §§

787, 1101(a) (West 1966). Article I, § 28(d) suggests this abolition by providing that

relevant evidence may not be excluded from a criminal trial. However, § 28(d) explicitly

preserves CAL. EviD. CODE § 352 (West 1966). Under § 352, courts are empowered

to exclude evidence on a case-by-case basis when the prejudicial tendency of the

evidence is adjudged to outweigh its probative value. But the very reasons underlying

the categorical, statutory exclusion of such evidence in the first place, under provisions

such as CAL. EvID. CODE §§ 787, 1101(a) (West 1966), are largely those that are

recognized as compelling grounds for exclusion on an ad hoc basis under § 352. A

likely scenario, then, is that in the course of successive applications of § 352, rules will

emerge that will more or less restore the exclusionary rules now crystallized in such

provisions as CAL. EvID. CODE §§ 787, 1101(a) (West 1966).

4 See, e.g., supra note 2.

‘ J. WEINSTEIN & M. BERGER, supra note 2, 1 609[04], at 609-4; id. at 609-70.

During congressional consideration of the issue, Representative Hogan spoke of a “raging

debate” on the subject. Id. at 609-13 (quoting Representative Hogan). Indeed, the

intensity of the dispute is suggested by the fact that several senators, dissatisfied with

some of the Supreme Court’s proposed provisions for dealing with prior convictions, in

response sought to modify the Court’s rulemaking power. Id. I 609[01], at 609-50 to


6 See supra note 2. Proposition 8 also involved restitution to victims, safe schools,

criminal evidence, bail, and a variety of changes in the Penal and Welfare and Institutions


‘ People v. Castro, 38 Cal. 3d 301, 308-13, 696 P.2d 111, 115-18, 211 Cal. Rptr.

719, 723-26 (1985).

‘ See id. at 310-12, 696 P.2d at 115-18, 211 Cal. Rptr. at 723-26. The Legislative

Analyst’s discussion of this provision, distributed to all voters in the CALIFORNIA BAL-


— 18 U.C. Davis L. Rev. 683 1984-1985

University of California, Davis

felony convictions should be admissible “without limitation. ‘

The issue was political in the most conventional sense: it surfaced in

a political context and was accompanied by strident law and order rhetoric.”°

The explicit aim of those favoring liberalized admissibility was

to “toughen” a system of criminal administration viewed as “soft” on

criminals and indifferent to the plight of victims. 1

The issue can also be seen, however, as political in another, more

subtle sense. The beliefs that people hold about the importance of character

in understanding criminal conduct are often related to a systematic

way of thinking about crime and personal responsibility. A strong

affinity for character as a behavioral explanation dovetails neatly with

a law and order orientation, emphasizing the wrongdoer’s free will, her

personal moral default, and the justness of her punishment.’2 Explanations

for crime as rooted in structural, social, and economic conditions

– discrimination, joblessness, lack of education, poverty, destructive

family situations, or drug culture – are rejected in favor of a perspective

focusing on a socially contextless individual suffering from a moral


PAMPHLET], was limited to the following paragraph: “Prior Convictions. The measure

would amend the State Constitution to require that information about prior felony convictions

be used without limitation to discredit the testimony of a witness, including that

of a defendant. Under current law, such information may be used only under limited

circumstances.” Id. at 54. The issue was not explored further in either the arguments

for or against Proposition 8. Id. at 34-35. A major obstacle to detailed public understanding

was the technical complexity of many of the issues and their sweeping range.

See supra note 6. That raised the issue of whether Proposition 8 violated the singlesubject

rule of CAL. CONST. art. II, § 8(d). A challenge to Proposition 8 along these

lines was tendered to, and rejected by, the California Supreme Court in Brosnahan v.

Brown, 32 Cal. 3d 236, 651 P.2d 274, 186 Cal. Rptr. 30 (1982). See Lowenstein,

California Initiatives and the Single-Subject Rule, 30 UCLA L. REV. 936 (1983).

See supra note 2 for the relevant portion of CAL. CONST. art. I, § 28(f.

See, e.g., CALIFORNIA BALLOT PAMPHLET, supra note 8, at 34 (arguments advanced

by Lieutenant Governor Mike Curb and then Attorney General George

Deukmejian in support of Proposition 8).

” See CAL. CONST. art. I, § 28(a) (preamble) (enacted as part of Proposition 8); see

also People v. Castro, 38 Cal. 3d 301, 311, 696 P.2d 111, 116, 211 Cal. Rptr. 719, 724


12 An illustration of this point is found in the testimony of Dr. Thomas Szasz, who,

testifying as a psychiatrist in opposition to a murder defendant’s insanity defense, said

that the defendant was not psychotic but instead was a “bad person” who had “lived a

life badly, stupidly and evilly from the time of her teens.” L.A. Times, Jan. 26, 1981,

Pt. I, at 18, col. 3. The tenor of this testimony reflects his general philosophic view,

which stresses an individual’s moral responsibility for behavioral choices and rejects

attempts to transform what he sees as ethical choices into false “scientific” issues of

health or illness. See T. SZASZ, LAW, LIBERTY, AND PSYCHIATRY (1963).

[Vol. 18:681

— 18 U.C. Davis L. Rev. 684 1984-1985

Impeaching Defendants

deficit.’ 3 Thus, the appeal of character evidence lies not only in its

value as probative evidence in determining behavior, but also in its validation

of the same concept of personal responsibility that underlies the

“wicked person” approach to criminal responsibility.

This is not to say that one must necessarily hold a law and order

perspective to value character evidence; one may see character as a

powerful explanatory device without rejecting liberal or even radical

social explanations for crime.” Indeed, one may believe that character

explains much errant behavior even if that character is socially or externally

determined. Thus, Marx believed that one way in which social

oppression victimized the oppressed was precisely through the mechanism

of personality malformation.’ 5 As a matter of contemporary social

fact, however, it is hard to deny a connection between an affinity for

character explanations of human behavior and law and order ways of

thinking about crime, criminals, society, and punishment. To recognize

this does not answer the question of the appropriate uses of prior convictions

or of character evidence in particular cases. Understanding this

appeal of the use of prior convictions does, however, suggest a context

for Proposition 8 and a perspective for understanding the arguments for

” This view likewise does not take kindly toward efforts to avoid personal responsibility

by attributing antisocial conduct to mental illness. See T. SZASZ, supra note 12.

Not surprisingly, Proposition 8 also abolished the diminished capacity defense, by adding

  • 25(a) to the California Penal Code; sharply restricted the insanity defense’s

scope, by adding § 25(b) to the Penal Code; and rendered inadmissible evidence “concerning

an accused person’s intoxication, trauma, mental illness, disease or defect” to

disprove the requisite mens rea of a criminal offense, by adding § 25(a) to the Penal


, As, for example, in the following statement:

Is it not a delusion to substitute for the individual with his real motives,

with multifarious social circumstances pressing upon him, the abstraction

of “free will” – one among the many qualities of man for man himself?

  • . . Is there not a necessity for deeply reflecting upon an alteration of the

system that breeds these crimes, instead of glorifying the hangman who

executes a lot of criminals to make room only for the supply of new ones?

Marx, Capital Punishment, N.Y. Daily Tribune, Feb. 18, 1853, quoted in Murphy,

Marxism and Retribution, 2 PHILOSOPHY & PUB. AFFAIRS 217 (1973), reprinted in

  1. WASSERSTROM, TODAY’S MORAL PROBLEMS 491 (1979). See generally Murphy,


” Marx and Engels spoke of a degraded proletariat as “the ‘dangerous class,’ the

social scum, that passively rotting mass thrown off by the lowest layers of old society.”

  1. MARX & F. ENGELS, THE COMMUNIST MANIFESTO 92 (S. Moore trans. 1967).

Thus, they conceived of a dehumanized mass, broken and brutalized by capitalism,

which internalized the worst “character” traits fostered by that system, precisely as

social product. Id.


— 18 U.C. Davis L. Rev. 685 1984-1985

University of California, Davis

admission of such evidence apart fr6rn technical doctrine.

In People v. Castro,” the court confronted the task of divining the

meaning of article I, section 28(f), 7 added to the California Constitution

with the adoption of Proposition 8. Section 28(f) purported to

make admissible “any” prior felony conviction “without limitation” for

impeachment purposes. The defendant had been tried for receiving stolen

property. After she testified in her own defense, the prosecution

introduced two prior convictions for impeachment purposes, one for

possession of heroin and the other for possession of heroin for sale. For

reasons to be examined later, these convictions would have been inadmissible

under the pre-Proposition 8 case law. 8 The issue in Castro

was whether the “without limitation” language of article I, section

28(f) abolished all judicial discretion to exclude such evidence. The

court rejected such an exuberantly literal reading of section 28(f).’ 9 It

reasoned, first, that not every prior felony conviction was relevant to


‘ As to convictions that were not, section 28(f) did not mandate

admission, despite its apparently unconditional language.2 Indeed,

had section 28(f) so intended, it would have violated federal due process

protections. 2 2 What convictions, then, were relevant? The court rea-

” 38 Cal. 3d 301, 696 P.2d 111, 211 Cal. Rptr. 719 (1985).

” See supra note 2.

” See infra notes 72-78 and accompanying text.

, Justice Kaus wrote a plurality opinion speaking for three of the justices; the remaining

four members of the court divided between three opinions, each concurring

and dissenting. As I read it, a majority of the justices ended up supporting the major

propositions in the plurality opinion. For convenience, therefore, I treat Kaus’s opinion

as though it speaks for the court. I recognize full well that this plethora of opinions

provides fertile soil for dispute as to proper nose-counting on one or another issue.

20 38 Cal. 3d at 313-14, 696 P.2d at 117-18, 211 Cal. Rptr. at 725-26; id. at 327-

30, 696 P.2d at 128-30, 211 Cal. Rptr. at 736-38 (Bird, C.J., concurring and


21 Id. at 313-14, 696 P.2d at 117-19, 211 Cal. Rptr. at 725-27; id. at 322, 696 P.2d

at 124, 211 Cal. Rptr. at 732 (Grodin, J., concurring and dissenting); id. at 330-32,

696 P.2d at 130-32, 211 Cal. Rptr. at 738-40 (Bird, C.J., concurring and dissenting).

22 Id. at 313-14, 696 P.2d at 117-19, 211 Cal. Rptr. at 725-27; id. at 322, 696 P.2d

at 124, 211 Cal. Rptr. at 732 (Grodin, J., concurring and dissenting); id. at 330-32,

696 P.2d at 130-32, 211 Cal. Rptr. at 738-40 (Bird, C.J., concurring and dissenting).

As to the constitutional issue, see Comment, Impeaching the Accused with Prior Convictions:

Does Proposition 8 Put Beagle in the Dog House?, 15 PAc. L.J. 301, 314-19

(1984) [hereafter Comment, Impeaching the Accused]; Note, Constitutional Problems

Inherent in the Admissibility of Prior Record Conviction Evidence for the Purpose of

Impeaching the Credibility of the Defendant Witness, 37 U. CIN. L. REV. 168 (1968).

But cf. People v. Beagle, 6 Cal. 3d 441, 454, 492 P.2d 1, 9, 99 Cal. Rptr. 313, 321

(1972) (specifically rejecting the constitutional route for barring admissibility). Spencer

[Vol. 18:681

— 18 U.C. Davis L. Rev. 686 1984-1985

Impeaching Defendants

soned that they consisted of those “necessarily involv[ing] moral turpitude,”

23 or put another way, those reflecting a “general readiness to do

evil.”2 Moreover, the moral turpitude determination was to be made

by consulting the indispensable elements of the crime in general, rather

than the specific and adventitious facts underlying the conviction.2″

Even as to convictions involving moral turpitude, the determination

of relevancy was merely the starting point for analysis. For such evidence,

as was true for evidence “across the board,”26 had to pass muster

under Evidence Code section 352.27 “[Ils there hard evidence that …

[the voters] intended . . to abrogate entirely the discretion of the trial

court under section 352, a traditional, inherent and, in truth, indispensable

tool of the law of evidence? ’28 There was not, the plurality opinion

responded. The voters had expressed “continued trust in the discretion

of the trial courts’ 29 despite the mandatory tone of section 28(f).

  1. Texas, 385 U.S. 554 (1967), frequently cited in constitutional discussions, is of

equivocal effect. See Comment, Impeaching the Accused, supra, at 314-19. Cases upholding

a constitutional challenge to admissibility are in short supply. See Hawaii v.

Santiago, 53 Hawaii 254, 492 P.2d 657 (1971).

2 38- Cal. 3d at 306, 696 P.2d at 113, 211 Cal. Rptr. at 721. Justice Grodin concurred

in this conclusion. Id. at 322, 696 P.2d 2d at 124, 211 Cal. Rptr. at 732

(Grodin, J., concurring and dissenting).

24 Id. at 314, 696 P.2d at 119, 211 Cal. Rptr. at 727.

25 Id. at 316-17, 696 P.2d at 120, 211 Cal. Rptr. at 728. None of the other opinions

specifically address this point, but Chief Justice Bird had previously committed herself

to this view as the author of People v. Spearman, 25 Cal. 3d 107, 599 P.2d 74, 157

Cal. Rptr. 883 (1979). See infra note 74 and accompanying text.

26 38 Cal. 3d at 306, 696 P.2d at 113, 211 Cal. Rptr. at 721.

2 CAL. EVID. CODE § 352 (West 1966) (court has discretion to exclude evidence “if

its probative value is substantially outweighed by the probability that its admission will

(a) necessitate undue consumption of time or (b) create substantial danger of undue

prejudice, of confusing the issue, or of misleading the jury”). Historically, “section 352

was intended to apply across the board, excluding no relevant and otherwise admissible

evidence from judicial weighing of prejudice against probative value.” Castro, 38 Cal.

3d at 306-07, 696 P.2d at 113, 211 Cal. Rptr. at 721. The court emphasized and

reemphasized this discretionary power: admissibility is “always subject to the trial

court’s discretion under section 352,” id. at 306, 696 P.2d at 113, 211 Cal. Rptr. at

721; if a conviction involves moral turpitude, “it is prima facie admissible, subject to the

exercise of trial court discretion,” id. at 316, 696 P.2d at 120, 211 Cal. Rptr. at 728;

and “[wie reemphasize that . . .admissibility is subject to trial court discretion under

section 352,” id. at 317, 696 P.2d at 120, 211 Cal. Rptr. at 728.

” 38 Cal. 3d at 309, 696 P.2d at 115, 211 Cal. Rptr. at 723.

29 Id. at 312, 696 P.2d at 117, 211 Cal. Rptr. at 725. In part, its decision that § 352

still conditioned the admissibility of prior convictions rested on its reconciliation of two

subdivisions of § 28. Subdivision (d) provides that “relevant evidence shall not be excluded

in any criminal proceeding” but that “[n]othing in this section shall affect . ..


— 18 U.C. Davis L. Rev. 687 1984-1985

University of California, Davis

Finally, the court rejected the pre-Proposition 8 “Antick line”30 of

cases, implying that these cases had introduced excessively rigid appellate

control over the trial courts’ exercise of discretion; or even if in fact

the Antick line had not suffered from this defect, so the matter had

been perceived by the drafters of article I, section 28(f).” In adopting

that provision, the voters, then, had asserted their power to reject the

Antick line.32

Applying its analysis to the case before it, the court concluded that

the defendant had been improperly impeached with a conviction for

simple possession of heroin; such offense did not exhibit moral turpitude.

Possession for sale, however, was an admissible conviction, because

it did. Reversal was not warranted, however, since the error had

not been prejudicial. 3

Castro expanded to an unknown degree the potential admissibility of

prior convictions. It also introduced marked uncertainty into the process.

It offered little guidance as to the scope of the offenses within the

moral turpitude standard; as to the relevance of such convictions in the

special context of the impeachment of a witness-defendant; or as to the

correct, “nonrigid” way for appellate courts to supervise the admissibility

process. The aftermath of Castro, therefore, provides a good occasion

for rethinking the genuine relevance of prior conviction evidence,

especially in the context of its use against a witness-defendant. Such

analysis is indispensable to the intelligent exercise of the discretion

mandated alike by sound evidentiary policy, by Evidence Code section

352, by Castro itself, and perhaps by the due process clause as well.

Part I of this Article outlines the doctrines concerning the use of

prior convictions both before and after Proposition 8. Part II analyzes

the relevance of prior convictions for impeachment purposes and concludes

that such evidence has remarkably little evidentiary value in the

special context of the witness-defendant. Part II also considers, and

criticizes, four theoretical models as to the relevance of prior conviction

evidence, including several that have played a prominent role in the

evolution of the California doctrine. It concludes that ideally the whole-

Evidence Code, Sections 352, 782 or 1103.” (emphasis added). The court held that

subdivision (f) remains subject to § 352, by virtue of the just quoted language of (d),

despite the apparently uncompromising language of admissibility found in subdivision

(f). Id. at 309-12, 696 P.2d at 115-17, 211 Cal. Rptr. at 723-25.

30 Id. at 308, 696 P.2d at 115, 211 Cal. Rptr. at 723. For Castro’s review of those

cases, see id. at 307-08, 696 P.2d at 114, 211 Cal. Rptr. at 722.

11 Id. at 308, 312, 696 P.2d at 114, 117, 211 Cal. Rptr. at 722, 725.

32 Id.

11 Id. at 318-19, 696 P.2d at 121-22, 211 Cal. Rptr. at 729-30.

[Vol. 18:681

— 18 U.C. Davis L. Rev. 688 1984-1985

Impeaching Defendants

sale exclusion of such convictions when used against witness-defendants

would be preferable to any of the four models.” Absent such a rule of

flat exclusion, however, it argues for the alert use of Evidence Code

section 352 to guard against the potential abuses of prior conviction

evidence offered for impeachment purposes.



  1. The Utility of Prior Convictions

The backdrop of the issue is quite familiar. The factfinder is often

presented with squarely conflicting testimony on crucial issues of fact.

To decide the case, the factfinder, whether judge or jury, must decide

whom to believe. Because the believability, or credibility, of a witness is

itself an issue of fact, litigants may resort to various forms of evidence

to attack an opposing witness’s credibility.” One entrenched form is

evidence of a witness’s character for truthfulness. 6 The common sense

underpinning of such evidence can be stated as follows: We conduct our

lives on the belief that people have identifiable character traits and that

knowing these traits will help us determine how a person will behave

on specific occasions. If evidence of one’s character is helpful in assessing

conduct in the world at large, why not in the courtroom as well?

And why not in respect to the particular activity of testifying as a witness

on the stand? Thus, if a person possesses a lax general commitment

to tell the truth, that fact is relevant to assessing the person’s

credibility as a witness.

Accepting this premise, as does California law,37 the issue arises as to

the degree of hospitality such evidence ought to be accorded. In many

respects, the adoption of article I, section 28(f) did not affect the answer

to this question. Proof of character can still be made through witnesses

I would recognize an exception when the defendant opens the door by falsely

testifying to the absence of a prior criminal record or of prior criminality. I exclude

from this discussion the subject of impeachment of nonparty witnesses because many of

the considerations in that context are different. For a pre-Proposition 8 case considering

that issue, see People v. Woodard, 23 Cal. 3d 329, 590 P.2d 391, 152 Cal. Rptr. 536


” CAL. EvID. CODE §§ 780, 785-788 (West 1966).

36 CAL. EVID. CODE § 780(e) (West 1966) identifies evidence of a witness’s “character

for honesty or veracity or their opposites” as relevant to credibility and to be admitted

“[e]xcept as otherwise provided by statute.”

Id.; see also People v. Castro, 38 Cal. 3d 301, 314, 696 P.2d 111, 119, 211 Cal.

Rptr. 719, 727 (1985) (quoting Justice Holmes in support of this approach).

1985] 689

— 18 U.C. Davis L. Rev. 689 1984-1985

University of California, Davis

prepared to render their opinion of the defendant’s character for veracity,

or by witnesses prepared to testify to the defendant’s reputation in

that regard. 8 The issue of concern here arises in the context of the

seemingly narrow question of whether the witness’s character may also

be proved by evidence of specific prior misdeeds reflecting on her disposition

toward truthfulness. The general answer, both before and after

the adoption of Proposition 8, has been to reject this form of proof. 9

However, to this general rule of preclusion, an exception was historically

annexed in favor of prior convictions. “° Thus, evidence of specific

acts to demonstrate the defendant’s mendacity was permitted under a

limited circumstance and in a special manner. The specific conduct

must have resulted in a felony conviction, “1 and the impeaching proof

was restricted to evidence of the conviction itself, rather than the detailed

facts of the prior crime.”2 For our purposes, however, the crucial

fact is that not all convictions were admissible.

The following sections consider the contrasting approaches developed

before and since the adoption of Proposition 8 toward the breadth of

admissibility of prior convictions. Before proceeding to that subject,

however, it is worth noting the striking practical importance of how

broadly the scope of admissibility is defined. This is partly true for the

well-recognized reason that evidence about the defendant’s prior conviction

may trigger jury prejudice. 3 More significantly, this adverse news

” CAL. EVID. CODE § 1100 (West 1966) makes all forms of character evidence –

whether opinion, reputation, or specific act evidence – admissible to prove a relevant

character trait unless otherwise excluded by statute. Neither opinion nor reputation

evidence concerning credibility is generally excluded by statute. Therefore, such evidence

is admissible in principle, though subject to discretionary exclusion under id.

  • 352. See generally 3A J. WIGMORE, supra note 2, § 920.

” CAL. EVID. CODE § 787 (West 1966). See generally 3A J. WIGMORE, supra note

2, § 979. As to whether this result is changed by article I, § 28(d) of the California

Constitution, adopted with the passage of Proposition 8, see supra note 3.

40 CAL. EVID. CODE § 788 (West 1966) (originally codified at CAL. EVID. CODE

  • 2051 (West 1955)).

,’ Id.

42 People v. Spearman, 25 Cal. 3d 107, 116, 599 P.2d 74, 78, 157 Cal. Rptr. 883,

887 (1979) (deriving conclusion from CAL. EVID. CODE § 787 (West 1966)). This view

prevails also under Castro. See supra note 25 and accompanying text. See generally

MCCORMICK ON EVIDENCE § 43, at 98-99 (E. Cleary 3d ed. 1984) [hereafter

MCCORMICK ON EVIDENCE]; 3A J. WIGMORE, supra note 2, § 979.

4′ MCCORMICK ON EVIDENCE, supra note 42, § 43, at 99. The jury, though instructed

in the limited use of evidence, may be unwilling or unable to conform its

conduct to the instruction. Similar limitations, in other contexts, have been regarded as

“unmitigated fiction.” Krulewitch v. United States, 336 U.S. 440, 453 (1949); see also

People v. Fries, 24 Cal. 3d 222, 227-28, 594 P.2d 19, 23, 155 Cal. Rptr. 194, 198

[Vol. 18:681

— 18 U.C. Davis L. Rev. 690 1984-1985

Impeaching Defendants

about the defendant comes packaged in awesome form. The prior conviction

is not mere “opinion” or “reputation” testimony concerning the

defendant, dependent for its persuasive force on the vagaries of how an

opinion or reputation witness is perceived by the factfinder. Prior felony

convictions represent the judicial system’s official certification that

the defendant was guilty beyond a reasonable doubt of serious criminal

behavior. While the jury must still weigh the bearing of the evidence on

witness credibility, its factual validity is not easily controverted. In

short, prior convictions are surrounded by an air of authority that does

not normally accompany other forms of character evidence.”

The significance of prior convictions is not exhausted by their

weighty aura. They also possess a high degree of practical availability.

First, a substantial number of defendants have previously suffered one

or more convictions. 5 Thus, a fund of available impeaching evidence

exists. Furthermore, this evidence is readily available – generally no

(1979); People v. Antick, 15 Cal. 3d 79, 98-99, 539 P.2d 43, 55-56, 123 Cal. Rptr.

475, 487-88 (1975); State v. Santiago, 53 Hawaii 254, 258, 492 P.2d 657, 660 (1971).

In one survey, 98% of the criminal defense attorneys and 43% of the trial judges believed

jurors incapable of drawing the distinction called for in this type of instruction.

Note, To Take the Stand or Not to Take the Stand: The Dilemma of the Defendant

with a Criminal Record, 4 COLUM. J.L. & Soc. PROBS. 215, 218 (1968).

” Ironically, this fact is a principal justification for recognizing an exception in favor

of admitting prior convictions. Restricting proof of specific conduct to prior convictions

is said to benefit the defendant by sparing her the hazard of having to defend against

unanticipated and speculative collateral charges of misconduct that might be sprung at

trial. See People v. Castro, 38 Cal. 3d 301, 316-17, 696 P.2d 111, 120, 211 Cal. Rptr.

719, 728 (1985); MCCORMICK ON EvIDENCE, supra note 42, § 42.

41 The figures are not readily available. BUREAU OF CRIMINAL STATISTICS AND


CRIME AND DELINQUENCY IN CALIFORNIA 52 (1980) reports that of the

37,876 adult felons convicted in 1980, 14.9% had prior prison records which, by definition,

means they had suffered prior “felony” convictions, CAL. PENAL CODE § 17

(West 1970); 21.8% had no prior criminal records; and 63.2% had a “miscellaneous

prior record.” To an extent not indicated, this last category includes some, perhaps

many, with prior felony convictions – for example, those cases in which felony sentence

had been imposed and then suspended, or in which the defendant had been convicted

of a felony but with the imposition of sentence suspended. Of those defendants

convicted of felonies during 1980, then, the number who had had prior felony convictions

may have substantially exceeded 14.9%. Furthermore, these figures do not reveal

the prior convictions of defendants who pleaded not guilty and were subsequently acquitted.

For the years after 1980, the annual volumes of Crime and Delinquency in

California do not include analogous statistics. H. KALVEN, JR. & H. ZEISEL, THE

AMERICAN JURY 145 (1966) report that 47% of defendants in all criminal cases have

prior criminal records (22% for similar crimes, 25% for different crimes). However,

their data does not distinguish between prior felony and prior misdemeanor convictions.


— 18 U.C. Davis L. Rev. 691 1984-1985

University of California, Davis

further from the prosecution’s elbow than the nearest computer terminal

linked to the state’s repository of conviction data. In contrast, the

production of opinion or reputation testimony requires considerable labor.

The would-be impeacher must identify the potential witnesses, locate

them, evaluate them, induce them to testify, ensure their presence

at trial, and then hope they survive cross-examination. In short, opinion

and reputation evidence lacks both the automatic credibility and routine

availability generally associated with prior convictions. Indeed, in the

routine assembly-line criminal case “character evidence concerning

credibility” has probably come to mean no more, no less, than prior

conviction evidence.

  1. Admissibility of Prior Convictions: Pre-Proposition 8 Doctrines
  2. People v. Beagle Limitations on Admissibility of Prior Convictions

for Impeachment

Commencing with the 1965 adoption of the California Evidence

Code, California case law concerning prior convictions has evolved in a

series of stages. Until 1972, Evidence Code section 788 was widely understood

to authorize the admission of all felony convictions. Any felony

conviction would do, regardless of the crime, and courts had no discretionary

power of exclusion.”‘ A major change occurred in 1972 with the

California Supreme Court’s decision in People v. Beagle.47 The court

unanimously held that the admissibility of prior convictions was not

automatic; the evidence had to survive the balancing called for by Evidence

Code section 352.4″ Although Beagle declined to fashion hard and

fast rules concerning the exercise of this discretion, it delineated four

factors entitled to weight: 1) whether the conviction was for a crime

relating to credibility; 2) whether the conviction was too remote in time;

3) whether, in the case of a defendant-witness, the conviction was for

the same crime as the one for which the defendant stood charged; and

4) whether the admission of the conviction would unfairly discourage

the defendant from taking the stand to testify.49 The first two factors

46 See People v. Beagle, 6 Cal. 3d 441, 451-52, 492 P.2d 1, 7, 99 Cal. Rptr. 313, 319

(1972), and the cases cited therein, disapproved by Beagle.

4′ 6 Cal. 3d 441, 492 P.2d 1, 99 Cal. Rptr. 313 (1972).

41 Id. at 452-53, 492. P.2d at 8, 99 Cal. Rptr. at 320; see also People v. Castro, 38

Cal. 3d 301, 307, 696 P.2d 111, 113, 211 Cal. Rptr. 719, 721 (1985).

4′ 6 Cal. 3d at 453, 492 P.2d at 8, 99 Cal. Rptr. at 320. In identifying these factors,

the court adopted then Circuit Judge Burger’s approach delineated in Gordon v.

United States, 383 F.2d 936, 940-41 (D.C. Cir. 1967), cert. denied, 390 U.S. 1029

(1968); see also J. WEINSTEIN & M. BERGER, supra note 2, 609[03], at 609-68 to

[Vol. 18:681

— 18 U.C. Davis L. Rev. 692 1984-1985

Impeaching Defendants

identified questions about the relevance of the evidence. The last two

reflected a recognition of its prejudicial potential.

Disagreements over the proper application of Beagle spawned much

litigation in the following decade, often before lower courts that were

unsympathetic to restrictions on the use of such evidence.” The

California Supreme Court ultimately reaffirmed and refined the Beagle

criteria in its pre-Proposition 8 Antick line of cases, exemplified by

People v. Spearman.’ That case provides a useful foil for the evaluation

of People v. Castro,52 because, presumably, it illustrates the “rigidified”

review of trial court discretion that became such an important

part of the Castro critique.

  1. People v. Spearman: Evolution of the Beagle Doctrine

In People v. Spearman,” the defendant was charged with several

drug offenses, including possession of heroin for sale. Apparently he

was no stranger to drug trafficking, having been convicted of possession

for sale some two years earlier.’ Bringing the prior conviction to the

jury’s attention would, of course, be an unwelcome development from

the defendant’s viewpoint. The law at that time afforded the defendant

both good and bad news. The defendant’s character was inadmissible to

prove his criminal propensity: his prior conviction could be offered

neither to show a general criminal tendency nor a particular tendency

to commit drug offenses. 5 However, the conviction was admissible as

credibility evidence if it met the criteria set forth in Beagle. If admitted,

-69 (discussing Gordon and its forerunner, Luck v. United States, 348 F.2d 763 (D.C.

Cir. 1965)).

11 In People v. Rollo, 20 Cal. 3d 109, 569 P.2d 771, 141 Cal. Rptr. 177 (1977), the

court reversed the defendant’s conviction, expressing exasperation at the treatment the

lower courts had accorded its earlier decisions:

Twice in the past two years we have reviewed the origin and purpose of

. . .[the rule limiting the use of prior convictions for impeachment purposes],

provided elaborate guidance in its application, and reaffirmed its

mandate by reversing judgments of. conviction . . . . Surely we do not

need to repeat that discussion so soon.

Id. at 116, 569 P.2d at 774, 141 Cal. Rptr. at 180; see also People v. Rist, 16 Cal. 3d

211, 220 n.9, 545 P.2d 833, 839 n.9, 127 Cal. Rptr. 457, 463 n.9 (1976); People v.

Fisher, 153 Cal. App. 3d 826, 840, 200 Cal. Rptr. 683, 693 (1984) (Dalsimer, J.,


25 Cal. 3d 107, 599 P.2d 74, 157 Cal. Rptr. 883 (1979).

” 38 Cal. 3d 301, 696 P.2d 111, 211 Cal. Rptr. 719 (1985).

25 Cal. 3d 107, 599 P.2d 74, 157 Cal. Rptr. 883 (1979).

14 Id. at 111, 599 P.2d at 75, 157 Cal. Rptr. at 884.

“5 See supra note 3.


— 18 U.C. Davis L. Rev. 693 1984-1985

University of California, Davis

the jury would, in effect, receive the following instruction: You are not

permitted to reason from the prior conviction that the defendant is disposed

to commit crime. You are permitted to reason that, because he

has misbehaved previously, he is less entitled to belief when he denies

that he committed the crime with which he is now charged. 6

The defendant, perhaps lacking confidence in such a prophylactic instruction,

moved for an order in limine barring the evidence on Beagle

grounds. The trial judge denied the motion. The defendant “chose” 57

not to testify and was convicted. The supreme court reversed, upholding

his contention that the prior conviction should have been barred. The

court identified three considerations justifying its decision: (a) the

prejudice resulting from identity of the prior conviction with the current

charge; (b) the inhibiting effect of admitting the prior conviction

on the defendant’s desire to testify; and (c) the irrelevancy of the prior

conviction to credibility.

  1. Identity of the Prior Conviction with the Current Charge

The prior conviction offered at Spearman’s trial was identical to one

of the charged offenses – possession of heroin for sale. Under Beagle,

this similarity weighed against admissibility because of the danger that

the jury would use the conviction for the forbidden purpose of propensity

evidence. 8 By the time of Spearman, the fact of identity had become

more than a factor weighing against admissibility. Identical prior

convictions were not merely to be used “sparingly,”59 they were never

to be used; and the failure to abide by this restriction was itself sufficient

ground for reversal.6″

56 The actual instruction called for by CALJIC No. 2.23 reads:

The fact that a witness had been convicted of a felony, if such be a fact,

may be considered by you only for the purpose of determining the credibility

of that witness. The fact of such a conviction does not necessarily

destroy or impair the witness’ credibility. It is one of the circumstances

that you may take into consideration in weighing the testimony of such a



CRIMINAL (4th ed. 1979). As to the doubts about the jury’s ability to fathom such an

instruction, see supra note 43.

” 25 Cal. 3d at 112, 599 P.2d at 76, 157 Cal. Rptr. at 885.

” 6 Cal. 3d at 453, 492 P.2d at 8, 99 Cal. Rptr. at 320.

” People v. Rist, 16 Cal. 3d 211, 220, 545 P.2d 833, 839, 127 Cal. Rptr. 457, 463


60 Spearman, 25 Cal. 3d 107, 116, 599 P.2d 74, 79, 157 Cal. Rptr. 883, 888 (1979).

Justice Richardson, dissenting, quite reasonably attributed to the majority the adoption

of a rule “that a defendant’s credibility may no longer be impeached by admission of a

[Vol. 18:681

— 18 U.C. Davis L. Rev. 694 1984-1985

Impeaching Defendants

Identity of the charged offense and the impeaching conviction is quite

common. For example, larcenous crime is a frequent occurrence, and

particularly so among certain segments of the population.” Therefore,

this ground of exclusion alone significantly imperiled impeachment by

prior conviction, particularly since exclusion did not depend on an absolute

identity between the two crimes.62

  1. Effect of the Prior Conviction in Discouraging the Defendant

from Testifying

The threatened use of the prior conviction had caused Spearman to

decline to testify. While the evidence of guilt appeared strong, it was

not “irrebuttable.”63 Had the defendant testified, “unquestionably” the

defendant “might”6 have presented exculpatory evidence. The trial

court’s refusal to bar the prior conviction had prevented the defendant

from presenting his version of the events and, under the circumstances,

constituted reversible error.65

prior conviction of an offense identical or similar to the offense charged.” Id. at 119,

599 P.2d at 81, 157 Cal. Rptr. at 890 (dissenting opinion); see also People v. Barrick,

33 Cal. 3d 115, 126, 654 P.2d 1243, 1249, 187 Cal. Rptr. 716, 722 (1982).

6 See infra notes 118-19 and accompanying text.

62 It was sufficient that the crimes were “similar.” See, e.g., People v. Barrick, 33

Cal. 3d 115, 120, 128, 654 P.2d 1243, 1245, 1251, 187 Cal. Rptr. 716, 718, 724 (1982)

(no distinction between “auto theft” and “theft”). A number of cases considered

whether the identical or similar convictions could be “sanitized” by restricting the proof

to the bare fact of the prior conviction, without identifying the nature of the charge.

Since the jury would be denied knowledge of the tenor of the prior crime, prosecuters

argued that jurors would be less likely to draw improper inferences as to the defendant’s

particular criminal disposition. The California Supreme Court condemned this

technique. People v. Rollo, 20 Cal. 3d 109, 569 P.2d 771, 141 Cal. Rptr. 177 (1977).

Reference to a conviction alone, without specifying the underlying crime, would furnish

the jury with a largely meaningless fragment of information, inducing dangerous jury

speculation. Although the defendant had been accorded the “option” to end jury speculation

by revealing the underlying crime, this was an “archetypal Hobson’s choice,” id.

at 120, 569 P.2d at 776, 141 Cal. Rptr. at 182: one of remaining silent and facing

improper jury speculation, or of testifying to the nature of the prior conviction and

incurring the risk that the jury would draw an impermissible inference of guilt.

The court also held that the same-crime problem could not be finessed by withholding

mention of the specific crime and asking the defendant simply whether he had been

convicted of a felony “involving theft.” People v. Barrick, 33 Cal. 3d 115, 654 P.2d

1243, 187 Cal. Rptr. 716 (1982); see also People v. Cole, 31 Cal. 3d 568, 581, 645

P.2d 1182, 1190, 183 Cal. Rptr. 350, 358 (1982); People v. Betts, 110 Cal. App. 3d

225, 167 Cal. Rptr. 768 (1980) (prior conviction involving “trait of honesty”).

63 Spearman, 25 Cal. 3d at 118, 599 P.2d at 80, 157 Cal. Rptr. at 889.

64 Id.

6 Interestingly, the court in People v. Beagle, 6 Cal. 3d 441, 453, 492 P.2d 1, 8, 99


— 18 U.C. Davis L. Rev. 695 1984-1985

University of California, Davis

Of course, in principle, no evidence is absolutely irrebuttable. And

the greater the allowance one makes for the possibility of exculpation,

the closer one comes to a rule of automatic reversal whenever the defendant

declines to take the stand because of threatened impeachment. A

number of lower courts had displayed no such charity toward a defense

claim of reversible error. When the case against the defendant seemed

strong and there was little they could imagine the defendant saying that

would not be self-defeating, these courts found it easy to conclude that

the defendant would not have taken the stand in any case. 66 To reverse

would confer a windfall upon a defendant who either would not have

testified, or whose testimony would in any event have swayed no one.

In contrast, the supreme court emphasized the difficulties in reliably

predicting that the defendant’s testimony would not have mattered, and

it leaned toward liberality in applying the reversible error test.67 It also

Cal. Rptr. 313, 320 (1972), had not spoken of this issue in reversible error terms, i.e.,

as one that would arise only once other Beagle errors had been found, and when it

accordingly became necessary to weigh the effect of these errors to determine whether

the conviction could stand. Rather, the effect of admitting prior convictions in discouraging

the defendant from testifying had been conceived of as a serious problem in itself,

at times sufficient to warrant reversal even if the prior convictions were otherwise relevant

and not prejudicial. See People v. Barrick, 33 Cal. 3d 115, 129, 654 P.2d 1243,

1251, 187 Cal. Rptr. 716, 724 (1982). Thus, the court at one time viewed the testimony-

discouraging effect not merely as a factor in prejudicial error analysis but as a

cognizable harm in itself.

66 E.g., People v. Fisher, 153 Cal. App. 3d 826, 200 Cal. Rptr. 683 (1984); People

  1. Logan, 131 Cal. App. 3d 575, 182 Cal. Rptr. 543 (1982) (“we are unable to discern

any potentially favorable defense”); People v. Barrick, 177 Cal. Rptr. 532, 538 (1981)

(Gardner, J., dissenting), vacated, 33 Cal. 3d 115, 654 P.2d 1243, 187 Cal. Rptr. 716

(1982); People v. Moultrie, 99 Cal. App. 3d 77, 160 Cal. Rptr. 51 (1979).

67 In Spearman, after observing that the evidence against the defendant was strong

but not irrebuttable, the court imagined the things the defendant might have said defensively.

Assuming such testimony would have been forthcoming, the court declined to

“reject the possibility that the jury might have believed appellant.” 25 Cal. 3d at 119,

599 P.2d at 80, 157 Cal. Rptr. at 889. “By refusing to indulge in speculation, this

court preserves the right of every accused person to present his version of the case to the

jury.” Id.; see also id. at 119 n.8, 599 P.2d at 80 n.8, 157 Cal. Rptr. at 889 n.8.

Finally, the court relied on the observation of R. TRAYNOR, THE RIDDLE OF HARMLESS

ERROR 20-21 (1970), that depriving a litigant of the opportunity to present her

version of the case is normally reversible since there is no way of gauging the effect of

the deprivation effect upon the jury’s decision. See also People v. Barrick, 33 Cal. 3d

115, 129-30, 654 P.2d 1243, 1251-52, 187 Cal. Rptr. 716, 724-25 (1982); People v.

Fries, 24 Cal. 3d 222, 228-29, 594 P.2d 19, 23-24, 155 Cal. Rptr. 194, 198-99 (1979);

People v. Rist, 16 Cal. 3d 211, 223, 545 P.2d 833, 841, 127 Cal. Rptr. 457, 465 (1976)

(since defendant did not take the stand, the degree of prejudice resulting from the

court’s failure to exclude the conviction was “not possible” to determine and the usual

[Vol. 18:681

— 18 U.C. Davis L. Rev. 696 1984-1985

Impeaching Defendants

emphasized that the jury might misperceive the defendant’s failure to

testify and draw improper inferences of guilt from silence.68 Furthermore,

the court rejected any requirement that the defendant make an

offer of proof to show the importance of his intended testimony, reasoning

that this requirement itself would violate the privilege against selfincrimination.

6 9 In summary, this ground for exclusion alone – the

testimony-deterrent effect of the threatened impeachment – created a

near impenetrable barrier to the use of convictions in the case of the

nontestifying defendant.”0

Other arguments could also have been mustered to support the broad

presumption of reversibility. As noted, some lower courts had justified

their refusal to reverse on the basis of a belief that the defendant could

not have offered testimony useful to the defense. However, if the case

against the defendant was so compelling that her testimony would be

useless, why was impeachment through prior convictions necessary? If

she testified, her words, by hypothesis, would be greeted with scorn

because of their inherent improbability. Thus, the very premise for refusing

to reverse – the defendant’s inability to speak usefully in her

own defense – militated against the need to employ the prior convictions

in the first place. However, if her testimony indeed would have

been important, the deterrent effect of admitting the prior convictions

would be exceedingly costly.

Another possible argument in support of a broad rule of reversibility

is that the right to testify is fundamental not only because of its instrumental

contribution to sound factfinding, but because of the importance

tests for reversible error were not applicable). But cf. People v. Fisher, 153 Cal. App.

3d 826, 835, 200 Cal. Rptr. 683, 688 (1984) (the supreme court’s decisions do not

mandate “reversal ‘per se’ whenever the defendant refused to testify” due to the trial

court’s erroneous failure to exclude prior conviction).

If the defendant took the stand, however, a different result was reached. In People v.

Rollo, 20 Cal. 3d 109, 569 P.2d 771, 141 Cal. Rptr. 177 (1977), the court found the

Beagle error not reversible since the defendant in fact had not been deterred from testifying.

He testified, he lost, and he was entitled to no presumption of reversibility. Defendants

might easily have drawn the following moral: Think twice before taking the

stand when your Beagle motion fails.

” E.g., People v. Barrick, 33 Cal. 3d 115, 129-30, 654 P.2d 1243, 1252, 187 Cal.

Rptr. 716, 725 (1982); Note, To Take the Stand or Not to Take the Stand: The Dilemma

of a Defendant with a Criminal Record, 4 COLUM. J.L. & Soc. PROBS. 215,

221-22 (1968) (statistical data showing juries infer guilt from silence).

69 People v. Fries, 24 Cal. 3d 222, 233, 594 P.2d 19, 26-27, 155 Cal. Rptr. 194,

201-02 (1979).

” The Beagle court had said it intended no such result. 6 Cal. 3d 441, 453-54, 492

P.2d 1, 8-9, 99 Cal. Rptr. 313, 320-21 (1972).


— 18 U.C. Davis L. Rev. 697 1984-1985

University of California, Davis

of enabling defendants to present their version of the facts, regardless of

whether the judge thinks they have anything significant to say71

  1. The Relevance of the Prior Conviction to Credibility

The Spearman court gave a final, and crucial, reason for holding the

prior conviction inadmissible: the evidence lacked the saving grace of

relevance. To reflect on a witness’s credibility, a “prior conviction must

involve dishonesty.”72 That is, the conviction must involve an “intent to

lie, defraud, deceive, [or] steal, etc.” 73 Furthermore, the dishonest intent

had to constitute a necessary element of the offense, if not of its very

statutory definition, rather than be merely an incidental fact of a particular

case.7″ Voluntary manslaughter, for example, would not satisfy

the test, because it lacked any of the required elements of dishonest

intent. In contrast, issuing a check without sufficient funds would qualify,

since one of the crime’s necessary elements was “the ‘intent to

defraud.’ “”

The prior conviction in Spearman, possession of heroin for sale,

failed the necessary-element test. A showing that acts of dishonesty,

such as theft and deception, commonly occur within the heroin trade

would not suffice either. No evidence to this effect had been presented

” The strength of this interest would be immediately apparent if the defendant were

literally barred from taking the stand. In that circumstance, no evidence of guilt, however

strong, would justify denying the defendant the right to speak. See Coe v. Armour

Fertilizer Works, 237 U.S. 413, 424 (1915); Karst, The Supreme Court 1976 Term –

Foreword: Equal Citizenship Under the Fourteenth Amendment, 91 HARV. L. REV. 1,

30 (1977); see also Chambers v. Mississippi, 410 U.S. 284 (1973); Fuentes v. Shevin,

407 U.S. 67, 80-82 (1972); Washington v. Texas, 388 U.S. 14 (1967).

72 Spearman, 25 Cal. 3d at 114, 599 P.2d at 77, 157 Cal. Rptr. at 886; see also

People v. Barrick, 33 Cal. 3d 115, 123-24, 654 P.2d 1243, 1247-48, 187 Cal. Rptr.

716, 720-21 (1982). The dishonesty test stands in sharp contrast to the moral turpitude

standard later adopted by the Castro court in interpreting article I, § 28(f). Indeed,

Castro characterized the very same crime as that involved in Spearman – possession

of heroin for sale – as falling within the moral turpitude standard, and thus relevant

to credibility. See supra notes 23-34 and accompanying text.

” Spearman, 25 Cal. 3d at 114, 599 P.2d at 77, 157 Cal. Rptr. at 886.

“, “The only relevant consideration is whether the prior conviction itself contains as

a necessary element” the dishonest intent. Id. at 116, 599 P.2d at 78, 157 Cal. Rptr. at

887 (emphasis in original). The California Law Revision Commission had proposed

this view at the time of the adoption of the California Evidence Code. 7 CALIFORNIA

LAW REVISION COMMISSION REPORT 141-43 (1965). This view was, however, not

incorporated in CAL. EvID. CODE § 788. Castro, 38 Cal. 3d at 316 n.10, 696 P.2d at

119 n.10, 211 Cal. Rptr. at 727 n.10.

” Spearman, 25 Cal. 3d at 115, 599 P.2d at 77, 157 Cal. Rptr. at 886.

[Vol. 18:681

— 18 U.C. Davis L. Rev. 698 1984-1985

Impeaching Defendants

at trial, nor could it have been under controlling doctrine.76

To summarize, by the time Spearman was decided, the case law had

developed from the loose standard announced in Beagle to a fairly elaborate

and fixed set of rules interpreting the Beagle factors. First, convictions

were deemed irrelevant to credibility unless the underlying offense

possessed as a necessary ingredient one of the elements of

dishonesty suggested in Spearman. Second, the conviction could not be

for the same or similar offense as that for which the defendant was

currently on trial.7″ Third, if the defendant refused to testify after losing

a Beagle exclusionary motion, the court was disposed to resolve

doubts in the defendant’s behalf by concluding that the defendant’s silence

was caused by the threatened impeachment; that her testimony

might have been significant; and that it might have been believed by the

jury. 8 Against this backdrop, the changes wrought by People v.

Castro” can be appreciated.

  2. The Relevance of Character in Assessing Credibility of the


Before considering Castro further, it is useful to consider a fundamental

issue of relevance: whether prior convictions have any logical

relation to credibility in the special context of a defendant testifying in

her own behalf. The issue has received far less attention than it deserves.

Both the Castro and Spearman courts uncritically assumed that

at least a certain category of convictions was always relevant to credibility

(though they differed in their definitions of that category). The

issue remains whether convictions, whatever the offense, are relevant in

the special context of the witness-defendant. Only if this question is

answered in the affirmative can one usefully proceed to the issue as

traditionally defined: a refined inquiry into the precise contours of the

relevant category. A discussion of the fundamental issue of relevance is

warranted in principle; moreover, it is required by Castro. For Castro

does not mandate the admission of convictions involving moral turpitude;

it only permits it. Admissibility ultimately turns on the trial

court’s exercise of discretion under Evidence Code section 352. Manifestly,

this requires courts to assess the genuine probative value of the

” See supra note 74.

” See supra notes 58-62 and accompanying text.

” See supra notes 63-70 and accompanying text.

” 38 Cal. 3d 301, 696 P.2d 111, 211 Cal. Rptr. 719 (1985).


— 18 U.C. Davis L. Rev. 699 1984-1985

University of California, Davis

evidence, so as to facilitate meaningful comparison with those factors

arguing for exclusion. There is little in Castro that provides guidance

concerning these issues and a good deal that is confusing.

First, the court made “no attempt to list or even further define””0 the

felonies involving moral turpitude; and none of the justices on the court

supposed this to be an easy task.’ Moreover, to qualify under the

moral turpitude standard, said the plurality, the conviction must “necessarily”

8 bespeak that quality. To apply this standard, then, one must

begin with an ethical vision concerning moral turpitude; then one must

have an empirical sense of the set of crimes that “necessarily” manifest

it – keeping in mind the court’s rule forbidding any examination of

the specific facts of the offense to make individualized determinations.

That might suggest that the category of included crimes be narrowly

defined, since many offenses can be imagined that, varying with circumstances,

would demonstrate moral degradation to a high degree, to

a low degree, or perhaps not at all. It is far from clear, however, that

the majority of the court drew this conclusion. The plurality opinion

did present a number of hints as to crimes it believed within the moral

turpitude standard. These are described below.8″

Additionally confusing was the fact that, in endorsing the relevance

of moral turpitude offenses, the majority did not speak of the weight to

which they were entitled or even the process by which their weight

might be assessed. it concluded no more than that there is “some basis

– however tenuous 84 for inferring that one who has been convicted of

such a crime “is more likely to be dishonest than a person about whom

no such thing is known.”85 Finally, it left unclear the process by which

prejudice was to be weighed against probative value, and, in particular,

the role of appellate courts in monitoring the trial courts’ exercise of

80 Id. at 314, 696 P.2d at 118, 211 Cal. Rptr. at 726.

“, The plurality opinion referred to the necessity of proving that every conviction

offered for impeachment involved moral turpitude, “difficult though this may prove to

be.” Id. at 316, 696 P.2d at 120, 211 Cal. Rptr. at 728; see also id. at 316 n.ll, 696

P.2d at 120 n.11, 211 Cal. Rptr. at 728 n.l (“administratively … [the moral turpitude

standard] has proved awkward”); id. at 322, 696 P.2d at 124, 211 Cal. Rptr. at

732 (Grodin, J., concurring and dissenting) (“difficult problems of judicial administration”);

  1. at 323, 696 P.2d at 125, 211 Cal. Rptr. at 733 (Lucas, J., concurring and

dissenting) (moral turpitude standard is “confusing and uncertain”); id. at 336, 696

P.2d at 134, 211 Cal. Rptr. at 742 (Bird, C.J., concurring and dissenting) (moral turpitude

standard is an “open-ended invitation to judicial chaos”).

82 Id. at 306, 696 P.2d at 113, 211 Cal. Rptr. at 721.

8′ See infra notes 122-27 and accompanying text.

, Castro, 38 Cal. 3d at 315, 696 P.2d at 119, 211 Cal. Rptr. at 727.

85 Id.

[Vol. 18:681

— 18 U.C. Davis L. Rev. 700 1984-1985

Impeaching Defendants

discretion. It apparently reaffirmed the vitality of People v. Beagle,8 6

but it rejected, to an unspecified degree, the pre-Proposition 8 “Antick

line,”8? typified by Spearman, which had sought to apply the Beagle

standards. It remains to be seen whether the decisionmaking process

can be shifted from “rigid”88 rule toward expanded trial court discretion,

without inviting unprincipled and arbitrary distinctions in the

treatment of prior convictions, varying from judge to judge.

Loose appellate control presupposes that the cases present themselves

in such subtle factual variation as to make disposition on the basis of

rule inappropriate. Does this characterization apply to the screening of

prior convictions? In some respects, it almost surely does not. Thus,

Castro purported to emancipate the California courts from the rulebound

regime of the Antick line of cases. Yet, ironically, the supreme

court’s first essay in this direction entailed laying down yet another

rule, simply one more to its liking (that moral turpitude henceforth was

the measure of the relevance of prior convictions). And the court designated

possession of heroin for sale as a charter member of the moral

turpitude class. Are trial courts, in the name of discretion, now free to

reject either that general standard, or its application to the crime of

possession for sale, if they so desire? Presumably not, and for good

reason. As to such an abstract and general issue – which crimes, by

virtue of their indispensable elements, manifest moral turpitude –

there can be only one right answer. One can debate whether the court’s

answer was indeed the right one. But once provided, that answer presumably

must be the same for all courts, and there can be no room for

creative discretion. To what degree and in what respects, then, does

Castro intend a less “rigid” approach than had prevailed before? To

what degree can a less “rigid” approach be adopted that does not become

victim to incompatible and unanalyzable intuitions, varying with

the courtroom in which the decisions are made? The scope and standard

of appellate review in this area remain in considerable confusion.

The argument to be developed here starts with accepting, for the

sake of argument, the court’s moral turpitude standard. The argument

proceeds with the contention that character evidence, so limited, barely

reaches the threshold of relevance in the special context of the witness-

” 6 Cal. 3d 441, 492 P.2d 1, 99 Cal. Rptr. 313 (1972); see Castro, 38 Cal. 3d at

306-07, 696 P.2d at 113-14, 211 Cal. Rptr. at 721-22; supra notes 46-50 and accompanying


‘7 Castro, 38 Cal. 3d at 308, 696 P.2d at 115, 211 Cal. Rptr. at 723.

88 Id. at 308, 696 P.2d at 114, 211 Cal. Rptr. at 722.


— 18 U.C. Davis L. Rev. 701 1984-1985

University of California, Davis

defendant. 9 Trial courts must necessarily grapple with this issue of relevance

in exercising their discretion to weigh the admissibility of prior

convictions. To be relevant at all, evidence must help refine the

factfinder’s assessment of probabilities concerning a disputed fact in

light of all the other known circumstances. A salient known circumstance

in the context at hand is that the witness has taken the stand to

testify in her own defense. Assume that the defendant has suffered

prior convictions for crimes that speak strongly of her mendacity (however

one defines the category of convictions for which this is true). The

issue then is whether such evidence is relevant under the special circumstance

in which she takes the stand to testify in her own behalf. An

almost instinctive response is to say “of course.”‘” However, for all its

common sense relevance, such evidence is all but irrelevant in this


The argument rests on two assumptions. First, innocent defendants,

being innocent, have little motive to lie; accordingly, if they choose to

testify, they will usually testify truthfully. (Occasional exceptions arise

when an innocent person is motivated to testify falsely to protect a third

party, or to improve the evidence of innocence by gilding the lily.9″)

Second, guilty people, being guilty, will of necessity testify falsely when

they take the stand in support of a not guilty plea. (This assumption,

also, is not universally valid. For example, a guilty person may testify

truthfully when she merely seeks to rebut a particular item of false

testimony without herself offering any false testimony, or when she is

testifying to some preliminary issues going to admissibility rather than

to guilt or innocence.)

Although it is impossible, by definition, to know at the outset of a

trial whether a defendant is guilty or innocent, to the degree one grants

the above two assumptions, it follows that whether the defendant is

guilty or innocent, the prior convictions have little relevance. If the defendant

is innocent, the prior convictions are irrelevant, since, for the

” In itself this is not a very demanding test given that evidence is deemed “relevant”

if it has “any tendency in reason” to further the proof of a material fact. CAL. EVID.

CODE § 210 (West 1966).

” For a general discussion of relevance, see 3A J. WIGMORE, supra note 2, § 37.

” The common sense underpinnings of such evidence have been described earlier,

supra note 36 and accompanying text.

9 This possibility no doubt makes the argument less than 100% valid but, I believe,

it remains true for the most part. In any event, the primary aim of the criminal system,

however, is not to punish innocent people because they have lied but to foreclose exoneration

of the guilty. Facilitating the punishment of innocent persons serves a dubious

societal interest, even assuming they have violated their obligation to tell the truth.

[Vol. 18:681

— 18 U.C. Davis L. Rev. 702 1984-1985

Impeaching Defendants

most part, she lacks the motive to lie. Prior convictions might make us

question the defendant’s character for veracity but not her self-interest,

which is to tell the truth when she is indeed innocent. For such a person,

happily, self-interest and morality tend to march hand-in-hand.93

Accordingly, since character plays little role in shaping the innocent

person’s testimony, it is generally irrelevant to her decision to tell the


Turn now to the guilty person. It must be kept in mind that this

individual is not only guilty but that she has elected to take the stand.

A moment’s reflection will show that if she is guilty, there is a strong

possibility that she will lie. This is true not because she has suffered

prior convictions. Rather, because she is guilty, she cannot testify truthfully

that she is innocent. Thus, the defendant’s truthfulness depends

not on her character but on her actual innocence. It follows that it

would aid us greatly in determining her truthfulness if we knew

whether she was guilty. But this is the very issue yet to be determined

at trial. This is when the prior convictions really help: not by revealing

the defendant’s character for credibility but by disclosing her propensity

toward criminality. The greater the weight accorded to the prior

convictions, the more likely that she is a criminal type, that she committed

the crime in question, and that she is therefore lying when she

proclaims her innocence. In short, the inference progresses from past

convictions to guilt, rather than from past convictions to credibility. The

evidence serves little role for its proclaimed legitimate purpose – that

of helping to assess credibility – but a major role for its theoretically

prohibited purpose94 – that of assessing criminal propensity.

A counter argument might take the following form: Admittedly, the

defendant’s character for veracity does not matter once she elects to proclaim

her innocence. Having made that decision, truthfulness no longer

depends on her character but on her guilt. This fact, however, merely

pushes the appropriate inquiry one step back – to the point when the

defendant is considering whether to proclaim innocence. That decision

is not foreordained; at the outset of the trial false testimony is merely

” Even the least upright of men, when he is not under the influence of some

seducing interest, and is surrounded by the restraints of law, and exposed to

the shame and punishment of false testimony, will not be so much his own

enemy, as to commit a dangerous crime [perjury] without any view of


  1. BENTHAM, A TREATISE ON JUDICIAL EVIDENCE 250 (M. Dumont trans. 1825).

Bentham said this in support of the testimonial competence of persons who had suffered

felony convictions. See Uviller, supra note 2, at 867-68.

” See supra note 3.


— 18 U.C. Davis L. Rev. 703 1984-1985

University of California, Davis

one of the defendant’s options. A second option is not to testify at all,

leaving the state to carry its burden of proof. A third option is to manifest

contrition and to plead guilty. In selecting from these options, the

defendant’s character does come into play. The chronic liar, the argument

proceeds, is more likely to employ falsehoods to resist conviction

than is the generally truthful person (even when that person is guilty of

the charged offense). And, it follows, if we know the defendant has a

shaky commitment to the truth, we are entitled to receive her decision

to assert innocence with skepticism.

This argument perhaps makes out a case for the formal relevance of

the evidence. 5 But the evidence takes on genuine importance only if

one is prepared to make the following assumption: When a person is

indeed guilty of a serious crime, the truthfulness of her testimony is

likely to be determined not by inexorable survival pressures but by her

character for truthfulness. One must assume, in short, that the person’s

character remains an important determinant for truthtelling even when

she is charged with a serious crime, is indeed guilty, and contemplates

the prospect of conviction. In the words of one court, since “every criminal

defendant may be under great pressure to lie, the slight added relevance

which even a perjury conviction may carry would not seem to

justify its admission.”96 Indeed, one might speculate, the greatest force

inhibiting guilty defendants from lying is not the product of their character,

but the pragmatic advice of counsel that lying is a risky strategy,

given the adversary’s power to cross-examine and to rebut.

Thus, the relevance of character evidence at best is formal and minimal.

Prior convictions disclose little with respect to the defendant who

has no motive to lie and little with regard to the guilty defendant who

has elected to take the stand to proclaim innocence. This is not the view

that prevailed in California either before or after the adoption of article

I, section 28(f), that is, under either the Spearman or the Castro doctrines.

Both cases share the assumption that prior convictions, provided

their scope were defined properly, possessed substantial probative value

even – perhaps especially – in the context of the witness-defendant.

The difference between the two approaches, thus, is one of degree

(though hardly unimportant on that account). Different theories underlie

these two approaches. By analyzing these theories through concep-

” See supra note 89.

96 Hawaii v. Santiago, 53 Hawaii 254, 259, 492 P.2d 657, 661 (1971); see Spector,

Impeachment Through Past Convictions, A Time for Reform, 18 DE PAUL L. REV. 1,

15 (1968); Note, Other Crimes Evidence at Trial: Of Balancing and Other Matters,

70 YALE L.J. 763, 778 (1961).

[Vol. 18:681

— 18 U.C. Davis L. Rev. 704 1984-1985

Impeaching Defendants

tual models, the fallacies in both doctrines can be explored.

  1. The Relevance of Prior Convictions to Impeach Credibility:

Four Models

Let us now acept, for the sake of argument, the largely unrealistic

assumption underlying both Spearman and Castro, that prior convictions

can cast a powerful light on a defendant’s credibility at trial. Or,

to put the point somewhat differently, let us suppose the witness was

someone other than the defendant; the argument just considered as to

irrelevance of character evidence in the context of a witness-defendant

would then not apply. This brings us to the issue as it has been conventionally

defined. Assuming that at least some prior convictions might

usefully contribute to fathoming a witness’s credibility, which ones?

A major hurdle in developing a discriminating response is the difficulty

in resolving several confusing issues of an empirical and psychological

sort. For example, what is the nature of character evidence and

can it be meaningfully compartmentalized into different traits such as

veracity, honesty, peacefulness, or law-abidingness? Is there any relationship

between these traits? How broadly should these traits be defined

if they are to shed light on witness credibility?

The answers commonly given can be seen as variations on four competing

models. These can be labeled the “veracity,” “honesty,” “global”

(or “moral turpitude”), and “contextual” models. The point of considering

these alternatives is not that the approach of Spearman, Castro,

or any other competing rule conforms perfectly to any of these models.

The models help, however, to sort out various perspectives and provide

a useful framework for evaluating the competing ways of thinking

about the subject.

  1. The Veracity Model

The veracity model affords the most limited scope for admitting prior

convictions. The narrow trait of veracity is deemed the only trait sufficiently

connected with a person’s truth-telling tendencies to be worth

considering. The propensity toward veracity is seen as a natural com-

,ponent of character, subject to its own more or less autonomous controls.

Adherents to this perspective might view the behavioral portions

of the mind as consisting of autonomous compartments, each regulating

a narrow facet of behavior. Each compartment is capable of producing

good or bad behavior of one category without affecting for good or ill

the behavior of another category. As Wigmore put it, “[c]haracter is

only an abstract group-term; what actually exists is a number of virtu-


— 18 U.C. Davis L. Rev. 705 1984-1985

University of California, Davis

ally separate traits, [for example] honesty, violence, benevolence [and

others].”97 Under this model, evidence that a person was violent, or a

lawbreaker, would speak to the wrong compartment. Only evidence

that the person had lied, distorted, or deceived in the past would bear

on whether the person was likely to lie, distort, or deceive when


If we somehow had access to these compartments, we could study the

character trait directly. But we lack direct access. However, a wealth of

circumstantial evidence exists in the person’s behavioral choices in a

multiplicity of life situations. From these we can extrapolate the person’s

characteristic mode of behavior – truthful or untruthful – and

determine probable truthfulness when she testifies at trial. A good illustration

of a conviction admissible under the veracity model would be

perjury”‘ – similar conduct under similar circumstances. Other examples

might be convictions involving false statement, false pretense, forgery,

99 income tax fraud, and certain other forms of criminal fraud,

each involving deception.’ Indeed, in each case deception is not merely

an accidental feature of the way the particular crime was committed; it

“7 J. WIGMORE, SCIENCE OF JUDICIAL PROOF 106 (1937) (emphasis in original).

” See, e.g., People v. Fries, 24 Cal. 3d 222, 226, 229 n.7, 594 P.2d 19, 22, 24 n.7,

155 Cal. Rptr. 194, 197, 199 n.7 (1979).

, People v. Antick, 15 Cal. 3d 79, 99, 539 P.2d 43, 56, 123 Cal. Rptr. 475, 488


“o0T his definition of the scope of relevant convictions is similar to that adopted in

Rule 609(a)(2) of the Federal Rules of Evidence, rendering convictions involving “dishonesty

or false statement” generally admissible. The Congressional Reference Report

concerning that provision said it was intended to adopt the crimen falsi standard:

“Crimes such as perjury or subornation of perjury, false statement, criminal fraud,

embezzlement or false pretense, or any other offense in the nature of crimen falsi the

commission of which involves some element of untruthfulness, deceit or falsification

bearing on the accused’s propensity to testify truthfully.” S. REP. No. 1277, 93d Cong.,

2d Sess. 14, reprinted in 1974 U.S. CODE CONG. & AD. NEWS 7051, 7061.

Difficulties with the federal definition of crimen falsi are vividly described by

Uviller, supra note 2, at 871-72; see also CALIFORNIA LAW REVISION COMMISSION,


OF EVIDENCE, ART. IV, WITNESSES 756 (1964), which distinguishes between convictions

for perjury and for embezzlement: “A conviction for perjury suggests a bad character

for truthfulness. A conviction for embezzlement suggests a bad character for honesty.”

The Commission offered no explanation for the distinction. Perhaps its view was

that perjury involved lying, whereas embezzlement need not. Embezzlement, however,

involves an abuse of trust that might best be regarded as implicitly deceitful. There is

thus a fluidity of definition here, and embezzlement probably stands on the borderline

between the veracity model and the honesty model considered below. The matter of

characterization is vital under any rule that draws the line of admissibility between

crimes of nonveracity and crimes of dishonesty.

[Vol. 18:681

— 18 U.C. Davis L. Rev. 706 1984-1985

Impeaching Defendants

is part of its very definition. Any crimes conforming to the veracity

model probably also would comply both with the Spearman honesty

test and the Castro moral turpitude test.

  1. The Honesty Model

The honesty model represents a broader view of relevance. It considers

as relevant not only crimes of nonveracity – those requiring an

intent “to lie, defraud [and] deceive”‘ 1 – but also any other crimes

that evidence “dishonesty.”‘ 2 Even though this category lacks false

statement as a required element, the prior convictions speak with sufficient

force to credibility to warrant their use for impeachment purposes.

This view most closely approximates the Spearman standard.

Under Spearman, crimes within the honesty model (over and above the

crimes of nonveracity) reduce to those with a larcenous element, meaning

those reflecting an “intent to . . .steal,”‘ 3 such as larceny, robbery,

theft, and burglary.”‘ Robbery, from this viewpoint, is a hybrid

crime since it is both larcenous and assaultive. Theft, the larcenous aspect

of robbery, reflects on honesty; the use of force or fear, the assaultive

aspect, does not.’ 5

Whatever else may be said for or against the Spearman approach, it

offered the not inconsiderable advantage of a rule of thumb, a virtually

self-applying standard, in an area where many observers have commended

administrative convenience rather than finely honed discretionary

decisionmaking.06 This virtue proceeds from two features of

‘0’ People v. Spearman, 25 Cal. 3d 107, 114, 599 P.2d 74, 77, 157 Cal. Rptr. 883,

886 (1979).

102 Id. Dishonesty and falsehood were seen as distinct categories in Evidence Code §

786, which restricted evidence going to credibility to the traits of “honesty or veracity.”

CAL. Evm. CODE § 786 (West 1966); see also J. WEINSTEIN & M. BERGER, supra

note 2, at 609-27 to -29 (congressional debate over the phrase “dishonesty or false

statement” in Rule 609(a) of the Federal Rules of Evidence).

103S pearman, 25 Cal. 3d at 114, 599 P.2d at 77, 157 Cal. Rptr. at 886; see supra

notes 72-76 and accompanying text.

104 Under this theory, the relevance of a burglary conviction to impeachment would

depend on the specific intent element of the particular burglary, since the general crime

does not require that the entry be made with a larcenous intent. CAL. PENAL CODE

  • 459 (West Supp. 1985); see People v. Holt, 37 Cal. 3d 436, 453-54, 690 P.2d 1207,

1216-17, 208 Cal. Rptr. 547, 556-57 (1984).

10 People v. Fries, 24 Cal. 3d 222, 229 n.7, 594 P.2d 19, 24 n.7, 155 Cal. Rptr.

194, 199 n.7 (1979); People v. Rist, 16 Cal. 3d 211, 220, 545 P.2d 833, 839, 127 Cal.

Rptr. 457, 463 (1976).

106 Judge Burger adopted a “rule of thumb” approach in favoring the admission of

crimes of “dishonest conduct” and the exclusion of crimes of violence. Gordon v. United


— 18 U.C. Davis L. Rev. 707 1984-1985

University of California, Davis

Spearman. The first is its insistence that dishonesty be determined by

reference to the necessary elements of the offense, rather than by the

adventitious facts of the offense (as did Castro, also, under its moral

turpitude test). The second is that Spearman defined the elements that

met its test with a high degree of specificity (as Castro did not). There

may be occasional debate as to whether an offense falls within the

Spearman honesty standard; there will be a superabundance of debate

as to whether an offense falls within the Castro moral turpitude standard

(at least, once one moves beyond offenses already within the honesty

model, as to which the moral turpitude standard adds nothing beyond

what is already accomplished under Spearman).

Thus, the Spearman honesty standard may be convenient. Is the

standard, however, conceptually or experientially coherent? Why, for

example, do crimes of larceny bear on credibility? Even if they do, why

do they alone earn their way into the expanded category of the honesty

model?” 7 In short, what accounts for the view that such crimes offer

some distinctive insight into veracity?

The answer has perhaps been regarded as so self-evident, so intuitively

sound, as to require no explanation. For Judge (now Chief Justice)

Burger, it was simply a matter of “common human experience”

that acts of stealing should be included with “acts of deceit, fraud, [and]

cheating” as conduct “universally regarded”‘0 8 as reflecting on a person’s

honesty and integrity.

Mason Ladd wrote:

Personal crimes of murder, assault, and mayhem, show a vicious disposition

but not necessarily a dishonest one. On the other hand robbery, larceny,

and burglary, while not showing a propensity to falsify, do disclose a

disregard for the rights of others which might reasonably be expected to

express itself in giving false testimony whenever it would be to the advantage

of the witness. If the witness had no compunctions against stealing

another’s property or taking it away from him by physical threat or force,

States, 383 F.2d 936, 940 (D.C. Cir. 1967), cert. denied, 390 U.S. 1029 (1968); see

also People v. Harrison, 198 Cal. Rptr. 762, 771 (1984) (hearing granted) (basing

admissibility on whether the impeaching crime is one of dishonesty borders on the

“schizophrenic”); MCCORMICK ON EVIDENCE, supra note 42, § 43, at 93 n.5 (1984);

  1. WEINSTEIN & M. BERGER, supra note 2, 11 609[04], at 609-75 to -76.

,07T he debate over the use of the “etc.” in the Spearman standard of relevance, see

supra note 73 and accompanying text, centers around what other similar crimes might

qualify or whether the “etc.” rendered the court’s standard “open-ended,” “.vague,”

and a “meaningless catchall.” See Spearman, 25 Cal. 3d at 115 n.4, 599 P.2d at 78

n.4, 157 Cal. Rptr. at 887 n.4.

“o’ Gordon v. United States, 383 F.2d 936, 940 (D.C. Cir. 1967), cert. denied, 390

U.S. 1029 (1968).

[Vol. 18:681

— 18 U.C. Davis L. Rev. 708 1984-1985

Impeaching Defendants

it is hard to see why he would hesitate to obtain an advantage for himself

or friend in a trial by giving false testimony.”0 ‘

This is largely a conclusionary statement, and the reasons do not explain

what is distinctive about theft. The reasoning could equally apply

to all calculated, lrather than impulsive, crimes, and in particular, to all

self-serving criminal activity undertaken at the expense of others.

Yet this view of theft crimes’ special relevance has an undeniable

appeal. Parents might, for example, instruct their child that honest people

do not steal, without feeling that they were abusing the language.

The same parents would be far less likely to say that honest people do

not fight, kill, or take heroin. While such conduct might be condemned,

the condemnation would be phrased in terms of the heartlessness, destructiveness,

or self-destructiveness of the conduct, not its dishonesty.

There is, however, a lack of fit between the category of included

offenses and the principle of relevance offered in its justification. For

example: Is possession of heroin for sale – the prior conviction involved

in Spearman – properly regarded as irrelevant under the honesty

standard, as the Spearman majority held? Or, under that very

standard, is the offense rightly viewed as “loaded with deception,””

and one that by its “very nature . . . speak[s] eloquently of the credibility””‘

of the person committing it, as urged by the Spearman dissent?

Further, is the crime of rape marked by “an element of stealth

which bears some rational relationship to dishonesty, ‘ as a lower

court decided; or as one not “significantly” involving honesty or integrity,

as the California Supreme Court later ruled (without explanation)?

“3 Is robbery a crime of dishonesty because it involves the ele-

“0 Ladd, supra note 2, at 180. G. FLETCHER, RETHINKING CRIMINAL LAW § 1.1,

at 5 n.5 (1978) informs us that the English Theft Act of 1968, c.60, § 1 defines stealing

as an act of one who “dishonestly appropriates property belonging to another” (emphasis


“0 Spearman, 25 Cal. 3d at 121, 599 P.2d at 82, 157 Cal. Rptr. at 891 (Richardson,

J., dissenting); see also United States v. Ortiz, 553 F.2d 782, 784 (2d Cir.), cert. denied,

434 U.S. 897 (1977) (“[A] narcotics trafficker lives a life -of secrecy and dissembling

in the course of that activity, being prepared to say whatever is required by

the demands of the moment, whether the truth or a lie.”).

“‘ Spearman, 25 Cal. 3d at 122, 599 P.2d at 82-83, 157 Cal. Rptr. at 891-92

(Richardson, J., dissenting).

1 People v. Delgado, 32 Cal. App. 3d 242, 250, 108 Cal. Rptr. 399, 405 (1973).

People v. Rist, 16 Cal. 3d 211, 221, 222 n.10, 545 P.2d 833, 840, 841 n.10, 127

Cal. Rptr. 457, 464, 465 n.10 (1976); see also Spector, Impeaching the Defendant by

His Prior Convictions and the Proposed Federal Rules of Evidence: A Half Step Forward

and Three Steps Backward, I Loy. U. CHI. L.J. 247, 249 (1970) (“There is

absolutely no connection between a conviction for rape and the willingness to tell the


— 18 U.C. Davis L. Rev. 709 1984-1985

University of California, Davis

ment of stealth, as the supreme court had held?’ Or is it a crime with

no bearing on credibility because:

There is no deceit in armed robbery. You take a gun, walk out, and put

it in a man’s face and say, “Give me your money,” or walk up to the

counter of the cashier and say, “this is a holdup; give me your money.”

There is no deceit in that. They are not lying. They mean business. They

will murder you if you do not do it.”‘

Locating the source of the difficulty is easy. Deception of some sort

typically pervades each of these examples, as, indeed, it does most

crimes. But the deceptions are of various sorts. Depending on the

crime, the deception may occur at different stages and may perform

different functions for the perpetrator. What types of deception should

count in determining whether the crime bears on credibility? More specifically,

is the defining characteristic that a given crime is normally

committed covertly, secretively, or by stealth?” 6 If so, then, to be sure,

larcenous crimes would generally qualify as crimes of dishonesty. But

so would most crimes, including rape. In both examples, the victim

often is deceitfully maneuvered into a position of vulnerability so that

the crime can be perpetrated without interference. But if this were the

relevant form of deception, few crimes would be excluded under

Spearman since concealment in aid of the consummation of crime is a

commonplace feature of premeditated crimes.

Perhaps, however, it is not that kind of concealment that is the

touchstone. Perhaps the crucial issue is whether the perpetrator would

normally conceal the crime following its commission to avoid apprehension.

If so, again, this is hardly a unique feature of larceny, since candid

revelation is not a feature of most crimes.

A third possibility is that the relevant concealment is that which the

perpetrator employs to facilitate her retention of the spoils of the crime.

If so, larcenous crimes are indeed a special case. They typically involve

deception not only to avoid interference, apprehension, or punishment,

but also to bolster the criminal’s effective “title” to the stolen property

by falsely holding it out as her own after the crime. But again, why is

this form of deception especially important for impeachment purposes?

Perhaps, finally, the fact that legitimately distinguishes the larcenous

truth when on the witness stand.”).

“‘ See supra note 105 and accompanying text.

…U nited States v. Smith, 551 F.2d 348, 363 (D.C. Cir. 1976) (quoting Senator


“‘ See People v. Rollo, 20 Cal. 3d 109, 126, 569 P.2d 771, 780, 141 Cal. Rptr. 177,

186 (1977) (Richardson, J., dissenting) (solicitation of murder relevant to credibility

since it consists of “secret, conspiratorial conduct”).

[Vol. 18:681

— 18 U.C. Davis L. Rev. 710 1984-1985

Impeaching Defendants

crimes is that they include as a necessary element the intent to steal.

But this simply states the issue: Why is an intent to steal sufficient to

render the crime relevant to credibility but not, for example, an intent

to commit rape?

The answer provided by Spearman utterly begged the question. It

simply declared as a self-evident truth that any crime marked by an

intent to steal counted as a crime of dishonesty and that others did



If theoretical coherence is lacking, the practical importance of including

convictions for larcenous crimes among those relevant to credibility

is considerable. The number of convictions for crimes within the narrow

veracity model is small; the additional number embraced within

the honesty model – the larcenous crimes – is large.” 8 Moreover, it is

primarily minority group members who accumulate the commonest

I” But see J. WEINSTEIN & M. BERGER, supra note 2, 1 609[04], at 609-72 to -73

and cases cited in nn. 14-18.



CRIME AND DELINQUENCY IN CALIFORNIA I 11 (1980) (Table 31, Felony Arrests Reported)

are suggestive:

Total felony arrests reported during the year: 372,435

Of this total, crimes probably relevant to credibility

under the Spearman-honesty model because they possess

a larcenous element, but not relevant under the narrower

veracity model, include robbery (26,715), kidnapping

(2943) (at least to the degree that such crimes were for

the purpose of ransom), burglary (84,160) (to the degree

that an intent to steal was present), theft (51,047), motor

vehicle theft (29,514):

Crimes probably relevant to credibility under both the

Spearman-honesty model and the narrower veracity

model, because the crime employs falsehood, lying or

deceit as a definitional element, include forgery, check

offenses, credit card offenses (11,578), and “other”

(14,432) (to whatever degree this category may embrace

crimes crimes of nonveracity):

193,929 (maximum)

26,010 (maximum)

The statistics suggest that as many as eight of the nine crimes admissible under the

Spearman-honesty model may be inadmissible under the veracity model. Although the

categories employed do not precisely correspond to either the veracity or honesty models,

and the information concerns total arrests, rather than convictions, the conclusion

can still be drawn from these statistics that the Spearman-honesty model sharply expands

potential admissibility compared to what would be allowed under the veracity.



— 18 U.C. Davis L. Rev. 711 1984-1985

University of California, Davis

types of larceny convictions – theft, robbery, and burglary.”9 Indeed,

as a matter of social reality, impeachment by prior convictions is in

effect a mechanism largely employed against minority group members

when they once again find themselves prisoners on trial.

  1. The Global (Moral. Turpitude) Model

The global model regards the segmentation of character into tight

compartments as artificial and invalid. Advocates of this model view

distinctions between veracity and honesty on the one hand, and general

character on the other, as exaggerated. They reason that a criminal

past bespeaks bad character and that bad character implies an attenuated

commitment to truthfulness. So, at any rate, jurors should be permitted

to conclude. At the hands of the most uncompromising advocates

of this view, any and all prior convictions would pass the test of relevance.

No argument to the contrary would be heard.’ Any “felony

offense necessarily bears on one’s credibility regardless of the nature of

“‘ Again, viewing the arrest (rather than conviction) statistics for 1980, see supra

note 118, the following picture emerges:


Crimes and number Non-

(larcenous element) White Hispanic Black Other White

Robbery (26,715) 24.3% 26.3% 47.4% 2% 75.7%

Burglary (84,160) 47.7 25.3 24.9 2 52.3

Theft (51,047) 48 21.6 28.2 2.2 52

Motor Vehicle Theft (29,514) 40 27.4 30.7 2 60.1

Kidnapping (2493) 38.2 29.5 29.5 2.8 61.8

Crimes and number

(lying or deceit)

Forgery, check offenses,

credit card offenses (11,578) 52.1 13.3 32.4 2.2 47.9

Other (14,432) 52 20.4 25.2 2.4 48

‘2S0 uch is the view embraced by Justice Lucas in Castro, 38 Cal. 3d 301,323, 696

P.2d 111, 125, 211 Cal Rptr. 719, 733 (1985) (concurring and dissenting); see also

People v. Woodard, 23 Cal. 3d 329, 343-45, 590 P.2d 391, 399-401, 152 Cal. Rptr.

536, 544-46 (1979) (Clark, J., dissenting). It is too difficult to make refined, individualized

decisions as to which are probative and which are not. See id. at 343-44, 590

P.2d at 400, 152 Cal. Rptr. at 545 (quoting B. WITKIN, CAL. EVIDENCE § 1243, at

1146 (2d ed. 1966)). Therefore, as a matter of judicial convenience, all convictions

should be admitted and the jury permitted to weigh evidential value. See J. WEINSTEIN

& M. BERGER, supra note 2, 1 609[021, at 609-54 to -55. For the conflicting views in

Congress over whether all felony convictions should have been made automatically admissible

under the Federal Rules of Evidence or made subject to discretionary exclusion,

see id. at 609-4 to -40. See generally id. 609[011-609103].

[Vol. 18:681

— 18 U.C. Davis L. Rev. 712 1984-1985

Impeaching Defendants

that offense.” ’21

This, of course, is not the point of view articulated by the majority in

Castro. In adopting the moral turpitude test, it emphasized that that

category embraced something less than every possible offense. Judging

from the sparse hints in the Castro plurality opinion, the following

convictions would be disfavored for impeachment purposes: An ancient

and fairly minor conviction;’22 conspiracy to tattoo a person under 18;123

and simple possession of heroin.124

Convictions of the following sort, however, received at least tentative

endorsement in Castro under the moral turpitude standard: Possession

of heroin for sale;2 5 “child molestation, crimes of violence, torture, brutality

and so on”;126 and crimes which are “assaultive in nature.”‘ 127 The

link from such crimes to credibility, the plurality said, is “not so irrational

that it is beyond the power of the People to decree that in a

proper case the jury must be permitted to draw it, if it wishes, and the

‘no limitation’ language of subdivision (f) makes it abundantly clear

that the People so decreed.”‘ 128 Such crimes are potentially admissible

because they reflect “moral turpitude,” a “general readiness to do evil,”

“moral depravity,” or “bad character.’ 129

Further insight concerning the nature of, and justification for, this

standard may come from the dissents of Justice Richardson in several

pre-Proposition 8 cases. His outlook appears to have been similar to

that of the Castro plurality; furthermore, the Castro plurality assumed

that the dissenters, Justice Richardson prominently among them, had

been vindicated by the adoption of article I, section 28.13

Justice Richardson had argued .that crimes such as possession of her-

“‘ Castro, 38 Cal. 3d 301, 323, 696 P.2d 111, 125, 211 Cal. Rptr. 719, 733 (1985)

(Lucas, J., concurring and dissenting).

22 The court’s example was the impeachment of an elderly victim of a mugging for

a youthful conspiracy to disturb the peace, id. at 309, 696 P.2d at 115, 211 Cal. Rptr.

at 723, though such offense was apparently excludable not because it fell outside of the

moral turpitude category, but rather on Evidence Code § 352 grounds.

Id. at 314 n. 7 , 696 P.2d at 119 n.7, 211 Cal. Rptr. at 727 n.7.

22 The Castro majority held this prior conviction irrelevant to credibility. Id. at 317,

696 P.2d at 121, 211 Cal. Rptr. at 729.

…Th e admissibility of which was affirmed in Castro, as manifesting an “intent to

corrupt others.” Id.

26 Id. at 315, 696 P.2d at 119, 211 Cal. Rptr. at 727.

.2 Id. (quoting People v. Rist, 16 Cal. 3d 211, 222, 545 P.2d 833, 840, 127 Cal.

Rptr. 457, 464 (1976)).

128 Id.

129 Id.

230 Id. at 307-08, 696 P.2d at 114-15, 211 Cal. Rptr. at 722-23.


— 18 U.C. Davis L. Rev. 713 1984-1985

University of California, Davis

oin for sale,”‘ solicitation to commit murder, 3’2 and others, were highly

probative on the issue of witness credibility. Even so, he was not a fullfledged

member of the global club. He was prepared to draw distinctions

as others ” would not. For him, crimes involving only assaultive

behavior or crimes against the public peace were inadmissible.’34 Possession

of heroin for sale, on the other hand, was “necessarily . . .

loaded with deception.”‘3 5 More fundamentally, he noted: “A jury, if it

was permitted to learn, might reasonably question the honesty or veracity

and therefore the credibility of those who deliberately and knowingly

shun the common standards of decent and legal behavior by

choosing to profit from the sale of heroin.”‘ 36 He thus anticipated the

Castro court’s decision affirming the relevance of convictions of possession

of heroin for sale to credibility. The commission of this crime represented

a general rejection of societal norms. Its bearing on credibility

might have been indirect, but its relevance to credibility was no less.

Even if Justice Richardson’s general theory of relevance were accepted,

as the Castro plurality seemed to do, uncertainties as to the

scope of the crimes to which it applies – or in Castro’s terms, the

scope of the moral turpitude category – is no small matter.”37 The

problem with Justice Richardson’s approach, however, is more fundamental.

Even if his relevance theory was valid, the use of the convictions

embraced by this view would create a broad danger of prejudice.

The danger, ironically, is underscored by the very theory offered in

support of this expansive model. That theory is that any criminal convictions

demonstrating a lack of integrity, ruthless self-seeking, or will-

“‘ People v. Spearman, 25 Cal. 3d 107, 119-23, 599 P.2d 74, 81-83, 157 Cal. Rptr.

883, 890-92 (1979) (Richardson, J., dissenting).

132 People v. Rollo, 20 Cal. 3d 109, 126-28, 569 P.2d 771, 780-81, 141 Cal. Rptr.

177, 186-87 (1977) (Richardson, J., dissenting).

133 See supra note 120.

See People v. Spearman, 25 Cal. 3d 107, 120-21, 599 P.2d 74, 81, 157 Cal. Rptr.

883, 890 (1979) (crimes of violence, per se, do not reflect on honesty or credibility)

(Richardson, J., dissenting); J. WEINSTEIN & M. BERGER, supra note 2, 609[02], at

609-55 (“relationship between crimes of violence and truthtelling is particularly tenuous”);

Ladd, supra note 2, at 180 (“Personal crimes of murder, assault, and mayhem,

show a vicious disposition but not necessarily a dishonest one.”). But see People v.

Woodard, 23 Cal. 3d 329, 343-45, 590 P.2d 391, 399-401, 152 Cal. Rptr. 536, 544-46

(1979) (Clark, J., dissenting).

“‘ People v. Spearman, 25 Cal. 3d 107, 121, 599 P.2d 74, 82, 157 Cal. Rptr. 883,

891 (1979) (Richardson, J., dissenting).

136 Id.

31 See supra note 81 and accompanying text.

[Vol. 18:681

— 18 U.C. Davis L. Rev. 714 1984-1985

Impeaching Defendants

ingness to “shun the common standards of decent and legal behavior”‘ 38

show also that the defendant is likely to place a low value on telling the

truth. Lawlessness of this sort and untruthfulness are thus inseparable.

Perhaps so, but does not this very theory make such lawlessness good,

or even better, evidence that the defendant had the propensity to commit

the charged crime? Although the limiting instruction purports to bar

such use,'” the limitation is almost impossible for the jury to understand

or to follow.’ 0 It attempts to preclude a use that is supported by

the very theory deployed to demonstrate the conviction’s relevance to

credibility. The problem is not merely that the jury is apt to use the

evidence for extraneous, unintended, and improper purposes. Rather, in

using it precisely for the intended purpose – to reveal the defendant’s

indifference to social and legal norms – the jury will necessarily use it

as evidence of criminal propensity. The jury is invited to reason that if

the defendant sustained this prior conviction, she would not hesitate to

lie to cover her tracks at trial. If the jurors take that step, how can they

resist, logically or psychologically, the next step and reason that, by the

same token, the defendant would not hesitate to commit the crime for

which she now stands charged?

This danger of prejudice is, then, not an accidental or perverse result

of the use of a conviction for a proper limited purpose; it is a virtually

inevitable result of the jury’s adoption of the theory endorsed by Castro

and articulated so well by Justice Richardson. Almost to a certainty,

the jury will conclude from the prior conviction first that the defendant

probably committed the charged crime, and derivatively, that she is lying

in her present testimony. This is precisely the reverse of the logical

progression the law proclaims it wishes the jury to employ in considering

prior conviction evidence.

The danger of prejudice is most pronounced under the expansive

moral turpitude standard; it is least pronounced under the narrowly

defined veracity model. The reason for this is worth examining. It does

not primarily lie in the fact that the former category embraces a far

greater number of convictions than the latter, though it does. The reason

is, rather, that the potential for prejudice is of a different order of

magnitude in the two cases. Under the veracity model the danger of

runaway jury inferences is fairly muted. When the prior conviction involves

nonveracity as a definitional element, one might at least entertain

” Spearman, 25 Cal. 3d at 121, 599 P.2d at 82, 157 Cal. Rptr. at 891 (Richardson,

J., dissenting).

,’ For the pattern instruction on this point, see supra note 56.

,o See supra note 43.


— 18 U.C. Davis L. Rev. 715 1984-1985

University of California, Davis

the hope that the jury could focus on that feature and understand the

judge’s instruction that it was to use the conviction to assess veracity

and for no other purpose. Such a conviction calls for no intermediate

inference of general bad character. But, as already noted, the same can

hardly be said of a conviction that demonstrates the defendant’s “general

readiness to do evil.”…. Having, then, in theory rejected the use of

character evidence as a basis for inferring criminal propensity,’ 2 the

moral turpitude approach in effect admits the evidence by reclassifying

it as credibility evidence.

Doubtless many believe propensity evidence ought to be admissible

and that the general bar against character evidence should be rescinded.

Important reasons have until now been deemed sufficient to warrant

exclusion of such evidence. These reasons are both instrumental and

noninstrumental. The instrumental aspect is well understood. Prior

convictions invite the punishment of the defendant for being a bad person.

Their use therefore threatens to lower the effective standard of

proof and interfere with the instrumental goal of accurate factfinding.

The noninstrumental concern is this: In focusing on the kind of person

the defendant is, evidence of past criminality threatens to hold people

accountable not merely for what they do, but for what they are.

Such use interferes with the goal of treating the defendant as a person

on trial only for the charged crime, unencumbered with the burden of

coping with prior judgments of wrongdoing. The issue here goes beyond

a mere concern about the debasement of the factfinding process. It

expresses a view that would have weight even if one could be assured

that prior convictions would be given no greater weight than they rationally

deserved in assessing guilt. The underlying principle is one of

absolution, the belief that once people have paid the price for their

prior wrongs they are entitled to a fresh chance and the opportunity to

proceed anew. To permit the use of criminality as character evidence

represents an important departure from this value.

Decisions regarding the admissibility of character evidence represent

a disguised value choice concerning the parts of the defendant’s past

that are open for review in a criminal prosecution. The general bar on

character evidence is an attempt to ensure that the focus is on the

charged conduct – on the deed, not the doer. This is not the occasion

to redebate the virtues of the general ban on propensity evidence.’ 3

The point is rather that the objections are too important to be set aside

Castro, 38 Cal. 3d at 314, 696 P.2d at 119, 211 Cal. Rptr. at 727.

IZ See supra note 3.

143 Id.

[Vol. 18:681

— 18 U.C. Davis L. Rev. 716 1984-1985

Impeaching Defendants

disingenuously, indirectly, or unwittingly.”‘4 The exercise of the discretion

called for by Evidence Code section 352 seems an appropriate occasion

for recognizing the value of permitting previously convicted defendants

to commence their lives afresh as they again find themselves

defendants in court.

  1. The Contextual Model

The contextual model differs from those already considered because

its goal is neither per se to broaden nor to narrow the range of admissibility.

It calls, rather, for a broader information base in making the

decision. It rejects the bare fact of conviction as an abstraction bereft of

detail, texture, color, and humanity. Without the contextual qualifiers,

the prior conviction is seen as facilitating the stereotypic treatment of

the defendant.145 This model reasons that people compartmentalize

their commitment to truthfulness and to honesty in refined ways. These

qualities are not wholesale, but retail qualities that require a detailed

knowledge of the underlying facts and the motivating circumstances if

one is to draw meaningful conclusions concerning the actor’s veracity.

Relying on the abstract evidence of a conviction runs afoul of an insight

we all share since our first exposure to the tales of Robin Hood: the

gross identification of an offense can mask vast differences in the meaning

of the underlying conduct and in its implications concerning the

actor’s character. “6 This contextual model is approximated in certain

cases and rules. “‘ It is, however, rejected under California law, both

under Castro and Spearman before it, which made the relevance of

prior convictions turn on an abstract inquiry concerning the indispensable

elements of the offense in general rather than upon the specific,

accidental, facts of the particular offense. ” 8

If one is drawn to the contextual model, one confronts the following

” The state in Castro made no suggestion that article I, § 28(f) be read to authorize

the use of prior convictions as evidence of the defendant’s general criminal propensity,

38 Cal. 3d at 314, 696 P.2d at 118, 211 Cal. Rptr. at 726, and the Castro plurality

expressed no such justification for admitting prior convictions. Id.

,41 See Ladd, Techniques and Theory of Character Evidence, 24 IOWA L. REV. 498,

532-34 (1939).

4′ But see People v. Fries, 24 Cal. 3d 222, 235, 594 P.2d 19, 28, 155 Cal. Rptr.

194, 203 (1979) (Richardson, J., dissenting) (“Robin Hood aside, I never heard of an

honest robber.”).

” See, e.g., United States v. Papia, 560 F.2d 827, 847 (7th Cir. 1977); United

States v. Hayes, 553 F.2d 824 (2d Cir. 1977); People v. Sorge, 301 N.Y. 198, 93

N.E.2d 637 (1950).

141 See supra notes 41-42 and accompanying text.


— 18 U.C. Davis L. Rev. 717 1984-1985

University of California, Davis

dilemma. The proscription against inquiry into the underlying facts is

in part for the defendant’s benefit. But the effort to serve this interest

results in substituting the formal fact of conviction for the texture of the

underlying conduct, except insofar as the defendant elects to explain the

factual context. This option again presents the defendant with a

“Hobson’s choice.”‘ 49 An exhaustive inquiry into the detailed facts may

unduly emphasize collateral issues surrounding the prior conviction;

but if the defendant chooses not to disclose the facts surrounding it, the

jury is presented with only the stark fact of conviction, inviting speculation

to the defendant’s prejudice.

  1. Conclusion: The Absolute Ban of Prior Convictions for

Impeachment Purposes as a Fifth Model

The foregoing discussion suggests problems with the theoretical underpinnings

of the various models used for assessing the relevance of

prior conviction evidence, including those versions embodied in Castro,

in Spearman, and in the unlimited global model, which would admit

all felony convictions without exception. This is so even if one believes

that properly screened prior convictions in principle are capable of providing

important insight into the credibility of witness-defendants.

The case, however, for admitting any prior convictions against a

criminal defendant is poor. Contrary to usual assumptions, prior convictions

lack essential relevance, given the special posture of the testifying

defendant; and the inferential leap from past convictions to current

credibility is fraught with danger. Further, the more expansive the definition

of the scope of relevant convictions, the greater the danger. The

more the theory of relevance does not turn on some discrete element of

untruthfulness (as under the veracity model), the more its relevance

must depend on inferences about the defendant’s bad or antisocial tendencies.

This makes it nearly inevitable that the jury will draw inferences

from the defendant’s bad or antisocial character extending to her

primary conduct. Indeed, it seems that the principal value of convictions

lies precisely in the fact that they may provide at least as good

evidence of criminality as of credibility. The use of prior convictions as

criminal propensity evidence, forbidden in theory, is then received parading

as credibility evidence.

If one were writing on a clean slate, the best solution would be to

bar all prior convictions as impeachment evidence in the special context

,, People v. Rollo, 20 Cal. 3d 109, 120, 569 P.2d 771, 776, 141 Cal. Rptr. 177, 182

(1977); see supra note 62.

[Vol. 18:681

— 18 U.C. Davis L. Rev. 718 1984-1985

1985] Impeaching Defendants 719

of the witness-defendant. The slate, however, is not clean. Fortunately,

the power of trial courts to exercise the discretion called for by Evidence

Code section 352 provides ample protection, if the courts are permitted

to take account of all the ways in which prior conviction evidence

may be problematic and dangerous and if they are adequately

monitored in this effort by appellate review.

— 18 U.C. Davis L. Rev. 719 1984-1985