I concur, of course, in the opinion I have prepared for the court. I write separately for two reasons: (1) because my method of analyzing the validity of defendant’s motion under People v. Wheeler (1978) 22 Cal.3d 258 [148 Cal.Rptr. 890, 583 P.2d 748] (Wheeler) and Kentucky v. Batson (1986) 476 U.S. 79 [90 L.Ed.2d 69, 106 S.Ct. 1712] (Batson), differs from that currently accepted by this court, and (2) in order to express my disagreement with the views set forth in Justice Baxter’s concurrence with regard to the admissibility of the Oregon escape.
I.
As I explained in my dissent in People v. Johnson (1989) 47 Cal.3d 1194, 1292 [255 Cal.Rptr. 569, 767 P.2d 1047] (Johnson), it is useful and appropriate for appellate courts to compare “the prosecutor’s treatment of minority and nonminority prospective jurors” in reviewing the validity of Wheeler! *1249Batson challenges. Evidence of disparate treatment between these two groups—that the former are challenged for characteristics or responses that go unchallenged in the latter—is “strongly suggestive of bias.” (People v. Hall (1984) 35 Cal.3d 161, 168 [197 Cal.Rptr. 71, 672 P.2d 854].) Such comparative analysis raises questions about the exclusion of prospective juror Mary Alford in the present case.
As I stated in Johnson, in reference to an appellate court’s evaluation of the race neutrality of the reasons offered by the prosecutor for peremptorily challenging a minority prospective juror: “[I]n appropriate cases a comparison of the prosecutor’s treatment of minority and nonminority prospective jurors can be a highly useful analytical tool: when the record shows disparate treatment and it remains unexplained, how else is a court to determine the genuineness of the prosecutor’s claimed reason? In these circumstances the inference that the claimed reason is not the actual ground of the challenge may not only be reasonable, it may be compelling. To prohibit courts from drawing that inference in proper cases will deny them a valuable tool that aids them in discharging their duties under Wheeler and Batson.” (Johnson, supra, 47 Cal.3d at p. 1292 (dis. opn. of Mosk, J.).) The majority in Johnson rejected the validity of such comparative analysis. (Johnson, supra, 47 Cal.3d at p. 1221.) In doing so, this court placed itself distinctly in the minority among those that have considered this question. I cited in Johnson a number of state and federal cases that have used disparate treatment analysis in reviewing Batson-type claims. (47 Cal.3d at pp. 1293-1294.) Since Johnson was decided, a number of cases have reaffirmed the validity of such analysis, either by expressly endorsing it, or by employing it to determine the merits of a Batson challenge. (See, e.g., U.S. v. Chinchilla (9th Cir. 1989) 874 F.2d 695; Devose v. Norris (8th Cir. 1995) 53 F.3d 201, 204; Hollingsworth v. Burton (11th Cir. 1994) 30 F.3d 109, 112-113; Jones v. Ryan (3d Cir. 1993) 987 F.2d 960, 973-974; U.S. v. Hatchett (6th Cir. 1990) 918 F.2d 631, 636-637; Emerson v. State (Tex.Crim.App. 1993) 851 S.W.2d 269, 273-274; Roundtree v. State (Fla. 1989) 546 So.2d 1042, 1044-1045; State v. Collier (La. 1989) 553 So.2d 815, 822; State v. Oglesby (1989) 298 S.C. 279 [379 S.E.2d 891].)
That is not to say that a disparate treatment analysis of peremptory challenges is a simple matter. As we acknowledged in Wheeler, subjective factors, not apparent on the record or easily articulable, may legitimately play a critical role in an attorney’s exercise of a peremptory challenge. (Wheeler, supra, 22 Cal.3d at p. 275.) For that reason, we generally defer to the findings of the trial court. But the mere fact that subjective factors are sometimes key to a peremptory challenge does not categorically immunize a trial court’s decision on a Wheeler motion from appellate review. Once a *1250defendant has made a prima facie case, a prosecutor’s assertedly race-neutral justifications are subject to a higher scrutiny. (Johnson, supra, 47 Cal.3d at p. 1284 (dis. opn. of Mosk, J.).)
Moreover, when an appellate court discovers, in scrutinizing the record through comparative analysis and otherwise, that the prosecutor’s stated objective reasons for its peremptory challenges are dubious, then the court may give less credence to his subjective reasons. (See U.S. v. Chinchilla, supra, 874 F.2d at pp. 698-699.) In Chinchilla, the court concluded that two of the objective reasons stated by the prosecutor for excluding the only two Hispanic prospective jurors—place of residence and age—were either unsupported by the record or did not withstand comparative analysis. The court then concluded that the two reasons given by the prosecutor that were incapable of verification—that one of the prospective jurors had a poor appearance and that the other’s employment as a restaurant manager would make him less fit as a juror on that case—had little credibility. (Id. at p. 699.) As Judge Kozinski recently put the matter: “While subjective factors may play a legitimate role in the exercise of [peremptory] challenges, reliance on such factors alone cannot overcome strong objective indicia of discrimination such as a clear and sustained pattern of strikes against minority jurors. The stronger the objective evidence of discrimination, the more we will require by way of verifiable facts to sustain a trial court’s finding upholding the exercise of challenges.” (Burks v. Borg (9th Cir. 1994) 27 F.3d 1424, 1429-1430.)
Accordingly, the more that comparative analysis of the objective reasons for the exercise of peremptory challenges against minority jurors points toward disparate treatment based on race or on some other improper category, the less courts should credit counsel’s purely subjective reasons. This principle is particularly true in cases such as the present, in which the trial court’s ruling on the Wheeler motion was brief and conclusory, containing no findings of fact. Such rulings raise the question whether the trial court had truly engaged in a “sincere and reasoned attempt to evaluate the prosecutor’s explanation . . . .” (People v. Hall, supra, 35 Cal.3d at p. 167.) The tendency of trial courts to deal perfunctorily with Wheeler motions may be understandable, given the press of the court’s business, but it also further underscores the importance of appellate review. “Even the most conscientious trial judge can be misled by such extraneous pressures as a reluctance to dismiss the venire after some or all of the jurors have been seated, or a felt urgency to begin taking testimony in a trial expected to be lengthy, or a natural disinclination to disbelieve assertions of good faith made by an attorney in open court. An appellate court, of course, is removed from such *1251pressures.” (People v. Johnson, supra, 47 Cal.3d at p. 1291 (dis. opn. of Mosk, J.).)1
In the present case, the prosecutor used three of eighteen peremptory challenges against Black prospective jurors, thereby excluding all Blacks from the jury. Defendant himself was Black. Thus a Black defendant was tried by an all-White jury. As Justice Frankfurther wrote in Offutt v. United States (1954) 348 U.S. 11, 14 [99 L.Ed. 11, 16, 75 S.Ct. 11], “[J]ustice must satisfy the appearance of justice.”
As the majority correctly conclude, the trial court’s inquiry into the prosecutor’s justifications for the peremptory challenges constituted “ ‘at least an implied finding’ of a prima facie showing” of systematic exclusion. (People v. Fuentes (1991) 54 Cal.3d 707, 716 [286 Cal.Rptr. 792, 818 P.2d 75].) I agree with the majority that the reasons for challenging exclusion of two of the three Black prospective jurors—Annie Hysmith and Verdan Verret—were race neutral: Hysmith had expressed general opposition to the death penalty, and Verret had stated that he had had a number of bad experiences with the police and mistrusted their testimony.
The exclusion of prospective juror Alford is more troubling. Alford was 44, a stable member of the community, bom and raised in Riverside, and employed as a traffic management officer at March Air Force Base. She had been married for over 24 years and had 2 children. The prosecutor, relying a good deal on the lengthy jury questionnaires, as well as on voir dire, gave a number of reasons for excluding Alford. Two of those reasons seem highly dubious. The prosecutor stated that “I have the reference to Question 82 on the questionnaire; she would feel very, very sorry for dmg users. This case will involve, I believe, Mr. Jackson with respect to the use of drugs.” Question 82 asked prospective jurors: “In general, how do you feel about users of drugs?” Alford replied: “They need help to get off drugs.” Her *1252response does not express an inordinate sympathy for criminals who are drug users, and was in fact strikingly similar to the response of many of those who served as jurors. Juror Raymond Watkins’s response to the same question was “they need help.” Juror Annette Risinger stated that drug users “have problems and can’t cope.” Juror Terry Harvey stated that drug users “are wrong and need help.” Juror William Clark stated that drug users “need counseling and to establish positive goals in life without drugs.” Juror Michael Gray stated that drug users “need help.” Juror Martha Perkins stated that most drug users would need “extensive help” in order to break their drug habit. Thus, Alford’s seemingly innocuous statement that drug users “need help” placed her squarely within the mainstream of opinion of those who were ultimately chosen for defendant’s jury.
Another reason given for her exclusion was that “she may give too great a weight to psychiatric testimony.” On the questionnaire, Alford expressed the view, in response to a question concerning her opinion of psychiatry and counseling, that these therapies “could help a person.” She stated that a psychological evaluation “sometimes” may be considered in understanding human behavior. She further indicated that she had had no personal experience with mental health counseling. Finally, in response to the question: “If you sit as a juror on this case, would you be willing to consider psychiatric and/or psychological testimony,” she stated: “depends on the person who is on trial, and the evidence.” In these noncommittal responses, she gave answers that were similar to virtually every one chosen as a juror in this case, except that a number of jurors expressed a much more favorable view of the efficacy of psychotherapy and of psychological theories. It is difficult to imagine how these responses could have justified the peremptory challenge.
A third reason given by the prosecutor also seems quite dubious: she had “scored not high on [the death] penalty.” In fact, in the questionnaire the prospective jurors were asked whether they “generally support" or “generally oppose” the death penalty, and Alford chose the former. She stated, in response to the question whether death or life without parole are worse punishment, that “they are both hard punishment.” She indicated no opinion with respect to a question about whether the death penalty deters crime. She also expressed unequivocal support for the death penalty during voir dire. Others with openly ambivalent attitudes toward the death penalty were permitted on the jury. Juror Agnes Plantenga expressed no support for the death penalty in her questionnaire, stated that it did not deter crime, and in voir dire explained that she would have “a very hard time” with imposing the death penalty. Juror Michael Fogerty crossed out “generally support” on his questionnaire and wrote that he “sometimes supports” the death penalty, and *1253indicated that he did not believe it to be a deterrent. Juror Perkins did not indicate support for the death penalty and expressed in voir dire considerable reluctance to administer the penalty. A number of other jurors indicated that they did not believe the death penalty deterred crime.
The prosecutor gave as a further reason for Alford’s exclusion: “This is also the woman who I had basically [a] bad feeling about, because she said she could remember everything. . . .” He referred to Alford’s response to a rhetorical question he posed during voir dire: “Can anyone tell me what you were doing four years ago today?” The question was asked in the context of explaining to prospective jurors that the crime for which defendant was charged occurred four years ago, and that the fact that memories may have faded should not be counted against a witness’s credibility. Juror Alford stated in response to the question that she would in fact be able to tell what she was doing four years previous—that is, that she was working at her present job. The prosecutor recalled that he was impressed with “how proud of the fact [that] she could remember exactly what she was doing” four years ago and that he felt “she would be hypercritical of witnesses in this case because of the extended time period delay.” But the record shows that Alford never claimed that she could remember what she was doing four years ago, but only that she would be able to tell what she was doing, and not because of an extraordinary memory but because her life followed a routine. Her response was clarified when she admitted that she would not be able to tell “minute to minute” what she was doing four years ago. Nothing she said suggested that her knowledge of what she had been doing four years ago sprang from direct recollection, or that she would be hypercritical of witnesses who could not recall nonroutine events such as the circumstances of the murder.
The prosecutor gave as an additional reason for Alford’s exclusion that she had indicated that her daughter, as a minor, had recently pleaded guilty to a misdemeanor in connection with the theft of $100 from the bank in which she was employed. Alford characterized her daughter’s case as that of a “minor misdemeanor” that would not affect her views of the case. A number of jurors had close relatives charged and convicted of crimes, some of them fairly serious, or had themselves been convicted of a crime. One had a brother arrested in a gang-related incident; another’s stepfather had been convicted of murder; another had been arrested for driving while under the influence; another’s son had been involved in car theft as a juvenile.
Finally, the prosecutor stated that Alford appeared to be “quite nervous.” Nothing in the record before us confirms or undermines this impression.
In addition to analysis of disparate treatment between stricken minority prospective jurors and nonminority jurors, it can also be useful to compare *1254the stricken minority prospective jurors with the nonminority prospective jurors who were also stricken. If the nonminority prospective jurors who were challenged peremptorily had characteristics similar to the minority prospective jurors also challenged, then this would support the prosecutor’s case that his challenges were in fact race neutral. The inverse is also true—a marked disparity between challenged minority and nonminority prospective jurors may further corroborate unlawful prosecutorial bias. In the present case, the record reveals that the great majority of non-Black prospective jurors differed from Alford in an important respect. Of the 15 non-Black jurors challenged peremptorily, 12 either indicated that they “generally opposed” the death penalty or did not support it, or else indicated that life imprisonment without possibility of parole would be a greater punishment than death. Of the three remaining prospective jurors, one was a low-income Hispanic woman, one was a former alcoholic who expressed some ambivalence about the death penalty, and one shared with defense counsel an enthusiasm for gambling. Thus, prospective juror Alford stood out in contrast to the nonminority challenged prospective jurors in her unqualified support for the death penalty and her conventional lifestyle and habits— further confirmation that her exclusion may not have been race neutral.
Nonetheless, although the question is troublesome, I cannot conclude with assurance that Alford’s exclusion was motivated by race. The fact that Alford’s daughter had recently been convicted of a misdemeanor by the very office that was now prosecuting this case may be considered a valid, race-neutral reason for excluding her, in spite of the fact that some with prior personal involvement in the criminal justice system were allowed to sit on the jury. Because the recent conviction of Alford’s daughter by the Riverside District Attorney was itself a sound, objectively plausible basis for challenging a juror peremptorily, the fact that not all jurors whose family members had prior convictions were stricken from the jury is of less significance. And, although the exclusion of even a single juror based on race is unconstitutional and requires reversal (People v. Monteil (1993) 5 Cal.4th 877, 909 [21 Cal.Rptr.2d 705, 855 P.2d 1277]), it is more difficult on the basis of a lone questionable peremptory challenge to discern a practice of racial exclusion, given the legitimate role that subjective factors may have in a prosecutor’s decision. I am therefore unable to conclude with certainty from the present record that prospective juror Alford was excluded because of her race. I therefore concur in the majority’s holding that rejection of defendant’s Wheeler!Batson challenges does not require reversal.
II.
The majority do not address the question whether the admission of the Oregon escape evidence was error, concluding in any case that it was not *1255prejudicial. Justice Baxter, in his concurrence, would have this court reach that issue, and hold that the evidence is admissible. I would reach the contrary view.
As recounted in the majority opinion, defendant escaped from an Oregon jail cell with another inmate, using furniture in the room to break the thick plexiglass window of his cell. Defendant was apparently unarmed at the time. The escape involved no confrontation with any jail guard. Defendant was captured later that day, unarmed, and did not resist arrest. No force or violence resulted.
Penal Code section 190.3, factor (b) (all further statutory references are to this code) directs the jury to take into account at the penalty phase “[t]he presence or absence of criminal activity by the defendant which involved the use or attempted use of force or violence or the express or implied threat to use force or violence.” Section 190.3 also states that no evidence of criminal activity shall be admitted at the penalty phase that does not meet the criteria set forth in factor (b) (excepting evidence of prior convictions). In People v. Boyd (1985) 38 Cal.3d 762, 776-777 (215 Cal.Rptr. 1, 700 P.2d 782], we held that evidence of an escape attempt that did not involve the use of force or violence against persons was not within the scope of criminal activity contemplated by factor (b), and was therefore inadmissible. In so holding, we rejected an argument that the term “force or violence” could be construed to mean the violent injury to property, such as the removal of the metal grating to an air vent in that case. As we stated; “The purpose of the statutory exclusion is to prevent the jury from hearing evidence of conduct which, although criminal, is not of a type which should influence a life or death decision. If theft of property is inadmissible—and it clearly is under the statutory language—then we cannot find that damage to property ... is entitled to any greater consideration.” (38 Cal.3d at p. 776; see also People v. Bacigalupo (1991) 1 Cal.4th 103, 148 [2 Cal.Rptr.2d 335, 820 P.2d 559].)
Justice Baxter does not dispute Boyd’s conclusion that property damage alone is insufficient to constitute violent criminal activity admissible under section 190.3, factor (b). Rather he concludes, based on an expansive reading of the language of factor (b), that defendant’s nonviolent escape nonetheless constituted an “implied threat” of force or violence. I disagree.
When construing an initiative statute such as section 190.3, we look first to the language of the statute. “Absent ambiguity, we presume that the voters intend the meaning apparent on the face of an initiative measure [citation] and the court may not add to the statute or rewrite it to conform to an assumed intent that is not apparent in its language.” (Lesher Communications, Inc. v. City of Walnut Creek (1990) 52 Cal.3d 531, 543 [277 Cal.Rptr. *12561, 802 P.2d 317].) In the present case, it is undisputed that no actual violence against persons was committed in the course of the Oregon escape. We must determine rather whether there was a “threat to use force or violence.” The dictionary reveals that the term “threat” has two primary meanings. One is “[an] expression of an intention to inflict loss or harm on another by illegal means and especially] by means involving coercion or duress of the person threatened.” (Webster’s New Internat. Dict. (3d ed. 1961) p. 2382.) There is no evidence of any threat, express or implied, in this sense of the word occurring in connection with the Oregon escape.
The second meaning of the word is “something that by its very nature or relation to another threatens the welfare of the latter,” as in “the cmmbling cliff was a constant [threat] to the village below." (Webster’s New Intemat. Diet., supra, at p. 2382.) A related meaning is “an indication of something impending and usu[ally] undesirable or unpleasant” as in “the air held a [threat] of rain.” (Ibid.) Justice Baxter appears to mean that defendant’s Oregon escape constituted an implied “threat” in this sense. In other words, defendant’s Oregon prison escape was something that by its very nature threatened harm to others, or that violence was impending in the escape.
There are two problems with this position. First, it is unlikely that the word “threat” was being used in this sense. The conjunction of the word “threat” with the phrase “to use force or violence” appears to contemplate the more common usage of “threat” as the expression of a menacing intention.
Second, even if section 190.3 is using the term “threat” broadly in both senses of that term, there is still no evidence that defendant threatened the use of force or violence. On the contrary, the success of the escape appeared to be premised on defendant, who was unarmed, avoiding confrontation with armed jail guards. In this critical respect, the Oregon escape differs from the escape attempts or conspiracies considered in those cases cited by Justice Baxter in support of his position. (See People v. Mason (1991) 52 Cal.3d 909, 955 [277 Cal.Rptr. 166, 802 P.2d 950] [planned escape would have been impossible to carry out without confrontation with prison guards]; People v. Gallego (1990) 52 Cal.3d 115, 196 [276 Cal.Rptr. 679, 802 P.2d 169] [escape conspiracy in which a “shank” was to play a part]; People v. Boyde (1988) 46 Cal.3d 212, 249-250 [250 Cal.Rptr. 83, 758 P.2d 25] [possession and possible use of a gun part of the escape plan].) In the present case, we have no evidence that violence was being planned or even prepared for, merely speculation that the situation could have turned violent if circumstances had been different. Lacking “substantial evidence from which a jury could conclude beyond a reasonable doubt that violent criminal activity *1257occurred” (People v. Tuilaepa (1992) 4 Cal.4th 569, 587 [15 Cal.Rptr.2d 382, 842 P.2d 1142], affd. sub nom. Tuilaepa v. California (1994) 512 U.S. 967 [129 L.Ed.2d 750, 114 S.Ct. 2630]), we must conclude the trial court erred in admitting this evidence.
People v. Stanley (1995) 10 Cal.4th 764, 823-824 [42 Cal.Rptr.2d 543, 897 P.2d 481], cited by Justice Baxter, is also inapposite. In that case, defendant burned his wife’s car as part of a reign of terror against her that led to her eventual murder. We found that the destruction of personal property, in that particular context, was “an integral part of [the defendant’s] attempt to frighten and control” the victim, and as such an act or threat of violence. (Id. at p. 824.) But no such personal threat can be read into defendant’s destruction of jail property in order to effectuate his escape.
Justice Baxter’s assertion that the escape represented an “ implied threat of violence” is based in large part on the fact that the escape “engendered] an immediate manhunt which ended in [deféndant’s] recapture by officers brandishing firearms.” (Conc. opn. of Baxter, J., post, at p. 1258.) But as we have stated: “We reject the argument that all escapes, however nonviolent, are inherently dangerous because they invite efforts of prevention and apprehension by custodial and law enforcement officers. The possibility of violence during an escape can become an actuality only when, under the facts of the particular case, the escapee attempts violent resistance or, in his efforts to elude capture, conducts himself in a reckless manner.” (People v. Lopez (1971) 6 Cal.3d 45, 52 [98 Cal.Rptr. 44, 489 P.2d 1372].) There is no evidence that defendant engaged in such resistance or recklessness in this case.2
Nor is the fact that defendant had used violence in the past to escape from the Riverside jail credible evidence that he “threatened” the use of violence during the Oregon escape. To the contrary, in this instance, he did not behave violently toward any persons, did not express any intention to use violence, and did not prepare for violence in the course of the latter escape. It appears elementary that we cannot impute to him violent criminal activity based alone on speculation about his mental state.
In sum, defendant’s escape from his Oregon jail did not involve an “implied threat to use violence,” in any conventional sense of the word *1258“threat.” Thus, Justice Baxter’s assertion that such escape evidence should be admitted because it is “particularly pertinent to an individualized determination of the appropriate penalty in a capital case” (conc. opn. of Baxter, J., post, at p. 1262), is irrelevant. Section 190.3 does not permit consideration of all criminal activity that may be relevant to individual culpability and penalty, but only certain specific types of criminal activity, i.e., activity that at least implies a threat to use force or violence. We have no warrant to ignore the statute’s express prohibition and to venture beyond the boundaries it has imposed on consideration of a defendant’s past actions at the penalty phase of a capital trial.
I would therefore conclude that the Oregon escape evidence should have been excluded.
My conclusion that comparative analysis is a useful and sometimes necessary tool in the review of Wheeler/Batson challenges is not altered, and is indeed reinforced, by the recent United States Supreme Court case of Purkett v. Elem (1995)_U.S__[131 L.Ed.2d 834, 115 S.Ct. 1769]. There the court held that in a Batson challenge when a defendant has made out a prima facie case of racial discrimination, the race-neutral reasons that the prosecutor gives to rebut that case do not have to be intrinsically plausible. (Id. at p__[131 L.Ed.2d at p. 839, 115 S.Ct. at p. 1771].) It is only at the third step of the Batson challenge process—the determination by the trial court of whether the opponent of the peremptory challenge has proved purposeful racial discrimination—that the plausibility of the prosecutor’s reasons are to be considered. “At that stage, implausible or fantastic justifications may (and probably will) be found to be pretexts for purposeful discrimination.” (Ibid.) It is thus consistent with Purkett v. Elem to have both trial and appellate courts use comparative analysis as an aid to determine the plausibility of the prosecutor’s explanation at this third step in the Batson process.
Of course, as Justice Baxter points out, the mere fact that escape is not an “inherently dangerous” felony for purposes of the felony-murder rule, as we held in Lopez, does not mean that it may never be considered to be violent criminal activity admissible under section 190.3, factor (b). (See People v. Mason, supra, 52 Cal.3d at p. 955.) But what the Lopez holding does mean, in the context of section 190.3, factor (b), is that we must review the factual setting of each particular escape to determine whether it involved actual or threatened violence and not presume that the escape was violent because of the possibility of violence in reapprehension.