Keeten v. Garrison

I. QUESTION PRESENTED McMILLAN, District Judge. Where death is a possible penalty, does the accused get a fair trial on the question of guilt or innocence when the prosecuting attorney is allowed to challenge for cause all jurors who would be unwilling to vote for the death penalty if the issue of punishment were separately submitted to them in a later proceeding? H. FOREWORD These suits do not attack the death penalty as such. I have accepted the death penalty as a fact of civilization for two-thirds of a century. Like many other judges, although I have serious moral reservations about it, and objections to it, I accept it as part of the law which I am obligated to uphold. It has been with us since the beginning of history; it is obviously much less a threat to humanity than a number of the things we are doing under power-oriented militaristic rationales which threaten the very existence of human life. The United States Constitution, Amendment Fourteen, says that “no person shall be deprived of life, liberty or property without due process of law; nor deny to any person of any jurisdiction the equal protection of the laws.” The requirement that government observe due process when it takes a person’s life obviously assumes that government will take life for sufficiently serious crimes. However, the death penalty is the ultimate punishment. It is irreversible, and once it has been exacted there is on this earth neither parole nor pardon nor other corrective possibility. Therefore, when death is a possible penalty, all constitutional safeguards should be scrupulously observed. Common sense suggests that people who favor the death penalty are more likely to convict defendants charged with capital crimes, and that people who do not favor the death penalty are less likely to convict defendants charged with capital crimes. The Supreme Court recognized those contentions in Witherspoon v. Illinois, 391 U.S. 510, 88 S.Ct. 1770, 20 L.Ed.2d 776 (1968), but prudently did not then adopt them as a basis for decision because of the lack of sociological testimony, juror opinion polls and other evidence. Such evidence has now been developed and occupies hundreds of pages of the record of these eases. That evidence demonstrates that what common sense says is so, is so; jurors who favor the death penalty are significantly more likely to convict and jurors who oppose the death penalty are significantly less likely to convict. There is no serious evidence refuting those two propositions. A fair jury has not been provided when the prosecutor is able to keep on the jury those persons most likely to convict and to exclude from the jury for cause all those persons most likely to acquit. Nor does a jury so constituted represent, even in theory, a representative cross-section of the community. Separate trials on the question of guilt and on the question of punishment should be made available and persons conscientiously objecting to the death penalty may validly be excluded from the second phase, or punishment phase, of the trial. Separate trials on guilt and on punishment are already required by North Carolina. There is no evidence that following these principles will add substantially to the costs already required by North Carolina’s split trial procedure. Such costs, if any, are *1168 trivial compared with the human rights and constitutional issues at stake. III. HOW THESE FOUR CASES GOT HERE. These four habeas corpus petitions, consolidated for trial, were argued by counsel, on September 23, 1983, on the written records and briefs and evidence, on one issue — petitioners’ contention that the State of North Carolina deprived them of their Sixth Amendment and Fourteenth Amendment rights to fair trials before impartial juries when the prosecutors were allowed to excuse for cause at the guilt determination phase of their capital trials all jurors who said they would be unwilling to impose the death penalty if they were presented with that question at the later, separate, sentencing proceeding. All four petitioners have exhausted their claims in the North Carolina courts. 28 U.S.C. § 2254(b), and have filed habeas corpus petitions in this court. The pertinent facts as to the individual petitioners are as follows: Ted Lemuel Carter. — On January 6, 1975, petitioner Carter was convicted by a jury in Gaston County, North Carolina,-Superior Court, of first degree murder and armed robbery. Judge Sam J. Ervin, III, the trial judge, imposed a sentence of death. Petitioner’s death sentence was later vacated by the United States Supreme Court, Carter v. North Carolina, 428 U.S. 904, 96 S.Ct. 3212, 49 L.Ed.2d 1211 (1976), and he was subsequently sentenced to life imprisonment in September, 1976. At Carter’s trial on guilt or innocence, no jurors were challenged for cause; however, two jurors were excused upon peremptory challenges by the prosecutor on the basis of the views of those jurors concerning the death penalty. Charles Bruce Keeten. — On July 21, 1976, petitioner Keeten was convicted by a jury in Mecklenburg County, North Carolina, Superior Court of first degree murder; he was sentenced by the trial judge to life imprisonment. At Keeten’s trial on guilt or innocence, eight jurors were excused for cause because they stated that they would be unwilling to vote to impose the death penalty in any case. The jury which ultimately tried the case was composed of at least six women and at least three blacks. Sixteen per cent of the people on the voter list from which jury panels were selected at the time were black. Bernard Avery. — On December 6, 1978, petitioner Avery was convicted by a jury in Mecklenburg County, North Carolina, Superior Court of first degree murder; the jurors were unable to agree on an appropriate punishment and the trial judge sentenced petitioner to life imprisonment. The record from petitioner Avery’s trial on guilt or innocence reveals that 13 jurors were excused for cause because they stated that they would be unwilling to impose the death penalty in the sentencing phase of the case. [The number 13 conflicts with petitioner’s figure (12), and the state’s figure (10), but was reached by the court from an independent examination of the record.] Of the jurors thus excused, five were black men, five were black women, two were white men, and one was a white woman. The court excused five jurors for cause on petitioner’s motion, when those jurors stated that they would automatically vote to impose the death penalty upon any defendant found guilty of murder. The jury which ultimately tried Avery was composed of eight white women, three white men, and one black man. Fifteen per cent of the members of the jury venire were black. Petitioner exhausted his peremptory challenges. Larry Darnell Williams. — On June 8, 1980, petitioner Williams was convicted by a jury, in Gaston County, North Carolina, Superior Court, of first degree murder and armed robbery. In a separate sentencing proceeding, the same jury recommended that petitioner be sentenced to death. The trial judge then imposed the death sentence. *1169 At the guilt-or-innocenee phase of Williams’s trial, three jurors who said they were unwilling to impose the death penalty at a later sentencing phase of the trial were excused for cause. All three of these jurors were white. Williams was tried by a jury panel which included one black, from a jury venire in which three of the 69 venirepersons were black. North Carolina’s capital punishment statute, North Carolina General Statutes § 15A-2000, effective June 1, 1977, provides in pertinent part: § 15A-2000. Sentence of death or life imprisonment for capital felonies; further proceedings to determine sentence. (a) Separate Proceedings on Issue of Penalty.— (1) Upon conviction or adjudication of guilt of a defendant of a capital felony, the court shall conduct a separate sentencing proceeding to determine whether the defendant should be sentenced to death or life imprisonment. A capital felony is one which may be punishable by death. (2) The proceeding shall be conducted by the trial judge before the trial jury as soon as practicable after the guilty verdict is returned. If prior to the time that the trial jury begins its deliberations on the issue of penalty, any juror dies, becomes incapacitated or disqualified, or is discharged for any reason, an alternate juror shall become a part of the jury and serve in all respects as those selected on the regular trial panel. An alternate juror shall become a part of the jury in the order in which he was selected. If the trial jury is unable to reconvene for a hearing on the issue of penalty after having determined the guilt of the accused, the trial judge shall impanel a new jury to determine the issue of the punishment. If the defendant pleads guilty, the sentencing proceeding shall be conducted before a jury impaneled for that purpose. A jury selected for the purpose of determining punishment in a capital case shall be selected in the same manner as juries are selected for the trial of capital cases. (3) In the proceeding there shall not be any requirement to resubmit evidence presented during the guilt determination phase of the case, unless a new jury is impaneled, but all such evidence is competent for the jury’s consideration in passing on punishment. Evidence may be presented as to any matter that the court deems relevant to sentence, and may include matters relating to any of the aggravating or mitigating circumstances enumerated in subsections (e) and (f). Any evidence which the court deems to have probative value may be received. (4) The State and the defendant or his counsel shall be permitted to present argument for or against sentence of death. The defendant or defendant’s counsel shall have the right to the last argument. (b) Sentence Recommendation by the jury. — Instructions determined by the trial judge to be warranted by the evidence shall be given by the court in its charge to the jury prior to its deliberation in determining sentence. In all cases in which the death penalty may be authorized the judge shall include in his instructions to the jury that it must consider any aggravating circumstance or circumstances or mitigating circumstance or circumstances from the lists provided in subsections (e) and (f) which may be supported by the evidence, and shall furnish to the jury a written list of issues relating to such aggravating or mitigating circumstance or circumstances. After hearing the evidence, argument of counsel, and instructions of the court, the jury shall deliberate and render a sentence recommendation to the court, based upon the following matters: (1) Whether any sufficient aggravating circumstance or circumstances as enumerated in subsection (e) exist; *1170 (2) Whether any sufficient mitigating circumstance or circumstances as enumerated in subsection (f), which outweigh the aggravating circumstance or circumstances found, exist; and (3) Based on these considerations, whether the defendant should be sentenced to death or to imprisonment in the State’s prison for life. The sentence recommendation must be agreed upon by a unanimous vote of the 12 jurors. Upon delivery of the sentence recommendation by the foreman of the jury, the jury shall be individually polled to establish whether each juror concurs and agrees to the sentence recommendation returned. If the jury cannot, within a reasonable time, unanimously agree to its sentence recommendation, the judge shall impose a sentence of life imprisonment; provided, however, that the judge shall in no instance impose the death penalty when the jury cannot agree unanimously to its sentence recommendation. [Emphasis supplied.] Pursuant to accepted state practice, the prosecutor in the first, or guilt-or-innocence phase of a capital trial is permitted to challenge for cause all jurors who state unequivocally that their views in opposition to the death penalty would prevent them from considering imposition of the death penalty in any case, or that they could not, on the basis of the evidence before them and their oaths, make a fair determination of the defendant’s guilt. Petitioners do not challenge in this case the right of the state to exclude at the sentencing phase of the trial all jurors who state unequivocally that they would be unwilling to impose the death penalty in any case. Neither do petitioners challenge the authority of the state to exclude from either phase of the trial those, jurors who admit that their views in opposition to capital punishment would prevent them from viewing the evidence fairly and acting in accordance with -their oaths and the law. Rather, petitioners challenge only the exclusion from the guilt determination phase of the trial of jurors who are unwilling to impose the death penalty. Petitioners say that such exclusion, or “death qualification,” subjected petitioners to trial by juries that were unconstitutionally partial to the prosecution; failed fairly to represent a cross-section of the community; and suffered from a series of effects causing them to be biased against petitioners and their cases. In Witherspoon v. Illinois, 391 U.S. 510, 88 S.Ct. 1770, 20 L.Ed.2d 776 (1968), the Supreme Court declined to hold, on the limited record before it, that “death qualified” juries were more conviction-prone than other criminal- juries. The Court found that the evidence offered by the petitioner in that case was “too tentative and fragmentary to establish that jurors not opposed to the death penalty tend to favor the prosecution in the determination of guilt.” Id. at 517, 88 S.Ct. at 1774. However, the Court explicitly left the question open for possible later determination, stating: “Even so, a defendant convicted by such a [death qualified] jury in some future case might still attempt to establish that the jury was less than neutral with respect to guilt. If he were to succeed in that effort, the question would then arise whether the State’s interest in submitting the penalty issue to a jury capable of imposing capital punishment may be vindicated at the expense of the defendant’s interest in a completely fair determination of guilt or innocence — given the possibility of accommodating both interests by means of a bifurcated trial, using one jury to decide guilt and another to fix punishment. That problem is not presented here, however, and we intimate no view as to its proper resolution.” Id. at 520, n. 18, 88 S.Ct. at 1776, n. 18. (Emphasis added.) The evidence available to the Court in 1968, when Witherspoon was decided, was indeed tentative and fragmentary — only three unpublished studies as evidence to support petitioner’s contentions. Id. at 517, n. 10, 88 S.Ct. at 1774, n. 10. Two of *1171 the studies involved surveys of college students, and the third was still in its “confidential” stage. Fifteen years have passed, however, since the Witherspoon decision was rendered. A significant number of professional quality studies have now been conducted by social scientists who are widely regarded as experts in their fields. The strong weight of those studies supports the contentions and positions of petitioners. IV. THE PETITIONERS’ EVIDENCE STRONGLY SHOWS: (a) THAT PERSONS FAVORING THE DEATH PENALTY ARE SIGNIFICANTLY MORE LIKELY TO VOTE FOR CONVICTION IN CAPITAL CASES. (b) THAT PERSONS OPPOSING THE DEATH PENALTY ARE A DISTINCTIVE, COGNIZABLE GROUP IN THE COMMUNITY AND THAT SUCH PERSONS ARE SIGNIFICANTLY MORE LIKELY TO VOTE FOR ACQUITTAL IN CAPITAL CASES. A. P. Ellsworth & R. Fitzgerald, Due Process v. Crime Control: The Impact of Death Qualification on Jury Attitudes, 1979 (prepub. draft) [hereinafter Ellsworth Attitude Survey ]. Respondents to this survey were 811 adult residents of Alameda County, California, eligible for jury service, and selected by “random digit dialing.” The Field Research Corporation of San Francisco drew the sample and conducted all interviews. All persons who stated that they could not be fair and impartial in deciding the question of guilt or innocence in a capital trial were excluded from the study. Of those who stated they could be fair and impartial, 17.2% stated that they would be unwilling to impose the death penalty in any case. In response to a series of 13 questions, the groups responded as follows: 59.2 per cent of the persons who were unwilling to impose the death penalty (“Witherspoon excludables”) thought that the insanity defense is a “loophole allowing too many people to go free,” while 78 per cent of the “Witherspoon qualified” jurors held that view. 62.5 per cent of the Witherspoon excludables thought it better that some guilty people go free rather than risk convicting an innocent person, while 44 per cent of the Witherspoon qualified jurors held that view. 23.5 per cent of the Witherspoon excludables believed that an accused who fails to testify is probably guilty, while 32.3 per cent of the Wither-spoon qualified jurors believed that proposition. 77.8 per cent of the Witherspoon excludables believed that “even the worst criminal should be considered for mercy,” but only 44 per cent of the Witherspoon qualified jurors believed that proposition. 46.3 per cent of the Witherspoon excludables thought that, regardless of the results, all laws should be strictly enforced, but 57.1 per cent of the Witherspoon qualified jurors subscribed to that view. Compared with the Witherspoon excludables, a larger percentage of the Witherspoon qualified jurors thought that “a person would not be brought to trial unless he or she were guilty of a crime,” and that illegally obtained evidence should not be excluded from court. A larger percentage of Witherspoon excludables, however, thought that “harsher treatment of criminals is not the solution to the crime problem.” Ellsworth arranged the results of her survey into an “index of prosecution-proneness,” and found that there was a difference of approximately 12 per cent in the proportion of pro-prosecution attitudes between the group of persons unwilling to impose .the death penalty, and those willing to consider it. This overall difference was significant to a level of P ± .0001. See, Hovey v. Superior Court of Alameda Cty., 28 Cal.3d 1, 168 Cal.Rptr. 128, 164, 616 P.2d 1301, 1337 (1980). Only 16.5% of white respondents in this study were unwilling to impose the death penalty in any case, while 25.5% of blacks *1172 were unwilling to impose it (.05 significance level). Twenty-one per cent of women respondents stated that they would never vote to impose the death penalty, as compared to 13.1% of men (.01 significance level). B. Bronson, On the Conviction Proneness and Representativeness of the Death-Qualified Jury: An Empirical Study of Colorado Veniremen, 42 COLO.L.REV. 1 (1979) [hereinafter Bronson-Colorado ]. Subjects in this study were 718 Colorado venirepersons, interviewed by telephone or in person. Comparing the responses of those people who strongly favor, favor, or oppose the death penalty {i.e., potential Witherspoon qualified jurors) with the responses of those people who strongly oppose the death penalty {i.e., Witherspoon excludables), Bronson found that a significantly higher percentage of the former group agreed with the following statements than those who agreed in the latter group: 1. If the police have arrested an individual and the district attorney has brought him to trial, there is good reason to believe that the man on trial is guilty. 26.2% vs. 8.8% 2. If the person on trial does not testify at his trial, there is good reason to believe that he is concealing his guilt. 26.4% vs. 20.6% 3. Concerning the high level of violent crime in ghetto areas, this level of violent crime could be reduced if the courts would convict alleged lawbreakers more often. 81.9% vs. 51.6% 4. The courts are far too technical in protecting the so-called constitutional rights of those involved in criminal activity. 72.6% vs. 60.3% 5. The plea of insanity is a loophole allowing too many guilty men to go free. 83.1% vs. 60.6% This pattern of responses was statistically significant at the .01 level or less. C. E. Bronson, Does the Exclusion of Scrupled Jurors in Capital Cases Make the Jury more Likely to Convict? Some Evidence from California, 3 WOODROW WILSON J. LAW 11 (1981) [hereinafter Bronson-California ]. This article combines the results of two studies. In the Butte County survey, 755 venirepersons interviewed primarily by telephone were asked the same questions used, in Bronson-Colorado and exhibited similar response patterns. In addition, Bronson found that a significantly higher percentage of the group of potential Witherspoon qualified jurors agreed with the following statements than did the group of Witherspoon excludables: 6. If the facts in a trial are not clear, then jurors should believe the prosecuting attorney more often than the defense attorney, since defense attorneys are only interested in protecting their clients at all costs. 22% vs. 6% 7. During a riot the government should have absolute power to jail suspicious persons without trial or bail. 33% vs. 6% In a follow-up study, Bronson found that 93% of Butte County residents interviewed who stated they were strongly opposed to the death penalty would also be unwilling to impose it in any case. In the same follow-up study, 18% of persons who would be unwilling to impose the death penalty in any case {Witherspoon excludables) stated that they would vote not guilty at the guilt phase of the trial, even if they believed the defendant to'be guilty, in order to prevent the defendant from being executed. Half of the Witherspoon excludables said they would find the defendant guilty if they believed that he was guilty, 25% were uncertain, and 6.7% said they would ask to be relieved of jury duty. In the Los Angeles- — Sacramento—Stockton survey, 707 venirepersons were asked, *1173 along with other questions, the same questions used in Bronson-Colorado, and exhibited similar responses. D. Louis Harris & Associates, Study No. 2016 (1971). Harris polled 2,068 United States adults, randomly chosen. These persons were divided into a number of categories, including persons who said they could never vote for the death penalty, and persons who said they could vote for the death penalty. (31% of persons interviewed said that they could never vote for the death penalty.) The persons polled were then asked a series of questions. Although all persons interviewed stated that they thought there was less law and order in the country than existed five years previously, persons who were unwilling to impose the death penalty were less likely to believe this than were persons who were willing to impose the death penalty (61% to 74%). Nineteen per cent of persons who were unwilling to impose the death penalty felt that courts were responsible for the breakdown of law and order in this country; thirty-four per cent of persons willing to impose the penalty believed this. Thirty per cent of the scrupled group thought that Negroes were responsible for the breakdown in law and order; thirty-nine per cent of persons willing to impose the death penalty agreed with this statement. Persons who were willing to impose the death penalty were also more likely to believe that “Many Negroes and other minority groups just want a hand out and do not want to work.” '(74% to 63%.) All these responses were significant at the .01 level or less. [Harris did not assign P ratings to its survey responses, but these were developed by petitioner in preparation for the Hovey trial.] Persons who were willing to impose the death penalty were more likely to choose the prosecutor as the attorney who is “more concerned about the best interests of society”; who is “less ready to make deals”; who is “more interested in a fair trial than in winning; who is more “genuinely interested in presenting all the facts”; who is less “likely to use emotional rather than factual arguments”; who is more “honest and straightforward”; who is a more “dedicated type of lawyer”; and who is less likely to be “better at playing tricks in the courtroom than in seeking justice.” These differences also were significant at the .01 level. Persons willing to impose the death penalty were also more likely to vote “guilty” in eight of the following nine situations (following a short admonition as to reasonable doubt): (1) If the defendant does not take the witness stand; (2) If an aide to the judge said he knew the defendant had committed the offense, but the judge said to ignore the statement; (3) If a policeman testified that he saw the defendant commit the crime; (4) If the defense presented no witnesses to prove the prosecution wrong; (5) If the prosecution made it pretty clear the defendant lied in part of his testimony; (6) If the defendant made a confession but the judge wouldn’t let it be introduced because the defendant had no lawyer there when he made it; (7) If the defendant admits he committed the crime, but brings in two psychiatrists who say that he was insane when he committed the crime; (8) If the facts proved that the defendant was at the scene of the crime; and (9) If [the juror] had read a newspaper article saying that the defendant had actually committed the crime although the judge said to ignore this information. Seven of the eight responses were significant at the .01 level or lower; the eighth was significant at .05. The subjects in this case were given written descriptions of the evidence in four criminal cases involving four different crimes, and were asked if they would find the defendant guilty or not guilty. In each case, persons who were unwilling to impose the death penalty voted to convict less often than those who, were willing to impose the penalty; the difference was significant at the .01 level in three of the cases. This poll also measured opposition to the death penalty by race and sex. Forty-six per cent of black subjects stated that they *1174 could never vote for the death penalty, while only 29% of white subjects said they could never impose it. Only 24% of men subjects stated they could never vote to impose the death penalty; 37% of women said they could never vote for it. E. H. Zeisel, Center for Studies in Criminal Justice, University of Chicago Law School, Some Data on Juror Attitudes Toward Capital Punishment, Monograph, 1968. Jurors who had served on felony trial juries in Brooklyn, New York Criminal Court, and in Chicago, Illinois Criminal Court, were asked questions concerning the first ballot vote of the jury as a whole; concerning their own first ballot vote; and concerning whether they had conscientious scruples against the death penalty. The study controlled for relative weight of the evidence by grouping the cases according to the arrangement of first ballot votes. In 10 of 11 constellations of evidence strength, jurors with conscientious scruples against the death penalty voted to acquit more often than jurors without such scruples. Dr. Zeisel’s study also quotes the results of Gallup polls conducted in 1960,1965, and 1966, indicating that while 55% of white men nationwide approved of the death penalty, only 42% of white women approved; 35% of black men approved; and 31% of black women approved. F. W. Wilson, Belief in Capital Punishment and Jury Performance, 1964 (unpub.). College students, 187 in number, were divided into two groups, those with “scruples” against the death penalty and those without. They were presented with written descriptions of five capital cases, and asked to vote guilty or not guilty. Wilson found that “people who have scruples against capital punishment are less likely to say guilty than are people who believe in capital punishment.” The finding was significant at the .02 level. G. F. Goldberg, Toward Expansion of Witherspoon: Capital Scruples, Jury Bias, and the Use of Psychological, Data to Raise Presumptions in the Law, 5 HARV.C.R.C.L. LAW REV. 53 (1970). Those interviewed were 100 white and 100 black Georgia college students, divided into two groups according to presence or absence of scruples against capital punishment. Each person was given a set of 16 written descriptions of criminal cases involving various crimes and asked to indicate their vote. Those without scruples voted to convict in 75% of the cases, compared to a 69% conviction vote for those with scruples. The result was not deemed to be statistically significant. H. G. Jurow, New Data on the Effects of a “Death-Qualified” Jury on the Guilt Determination Process, 84 HARV.LAW REV. 567 (1970). Jurow interviewed 211 employees of Sperry Rand Corporation, grouped along a. five-point scale according to their views on the death penalty. The subjects listened to two tape recordings of simulated murder trials, and then voted on guilt or innocence of the defendant. In one case, voting to convict was positively correlated to acceptance of the death penalty, at the .01 level of significance. I. P. Ellsworth, W: Thompson, & C. Cow-an, Juror Attitudes and Conviction Proneness: The Relationship Between Attitudes Towards the Death Penalty and Predisposition to Convict, 1979 (unpub.) [hereinafter Ellsworth Conviction-Proneness Study ]. Subjects were 288 residents of San Mateo and Santa Clara Counties, California, who were eligible for jury duty. Subjects were divided into groups: (1) those who would be unwilling to impose the death penalty in any case; and (2) those who would be able to consider imposing the death penalty in some cases. Those who stated that they could not be fair and impartial in deciding the question of guilt or *1175 innocence, knowing the defendant might receive the death penalty if convicted, were excluded from participation in the study. Subjects were then shown a 2lk hour videotape of a homicide trial, and were polled for their verdicts both immediately following the showing of the videotape, and after one hour of deliberation in groups of twelve. Immediately after viewing the videotape, the study subjects who were willing to impose the death penalty were more likely than the subjects unwilling to impose the penalty to vote to convict the defendant of some degree of homicide (77.9% of jurors willing to impose the punishment voted to convict, compared to 53.3% of jurors unwilling to impose the death penalty voting to convict). The difference is statistically significant at the .01 level. The views of the jurors did not change significantly following a period of deliberation. Subjects participating in the study who stated that they favored the death penalty were later polled to determine whether they would automatically vote to impose it in every case: All subjects polled said they would be willing to consider voting for life imprisonment, depending on the evidence. J. P. Ellsworth, J. Harrington, W. Thompson, & C. Cowan, The Effect of Capital Punishment Attitudes on Juror Perceptions of Witness Credibility, 1979 (unpub.). A sample of thirty-six individuals who had already participated in Ellsworth Conviction-Proneness Study, were divided into two groups according to willingness or unwillingness to impose the death penalty. They were shown a videotape of the testimony of two witnesses, a white police officer and a black defendant, each of whom gave his version of the events leading to arrest. After viewing the tape,, the subjects were given a 16-item questionnaire concerning the credibility of the witnesses. On each of the 16 questions, persons willing to impose the death penalty gave responses that were more favorable to the prosecution than those of the other group: The difference between the two groups of responses was significant at the .05 level or better for ten of the questions. The researchers constructed an overall index of prosecution-proneness, for all questions, significant at the .0002 level. K. C. Haney, On the Selection of Capital Juries: The Biasing Effects, of the Death Qualification Process (pre-publication draft). Subjects were 67 adults eligible for jury duty in Santa Cruz County, California, who were pre-screened to insure that they would be willing to impose the death penalty in some cases, and that they could be fair and impartial in deciding the question of guilt or innocence in a capital case despite the possible imposition of the death penalty. The subjects were divided into two groups: one group watched a two-hour videotape of a jury voir dire, including a half-hour of “death qualification”; the other group saw the same videotape without the half-hour of death qualification. Following exposure to the videotapes, the subjects were asked a series of 17 questions. In response to the questions, subjects who had witnessed the death-qualifying voir dire “estimated a significantly higher level of guilt than those in the non-death qualified condition.” They believed that it was more likely that the defendant would be found guilty of first degree murder and sentenced to death (.004 level of significance); and they believed it was more likely that the defendant would be found guilty of something (.014 significance level). When subjects were asked to indicate the degree to which “the law disapproves of people who are opposed to the death penalty,” those who were exposed to death qualification indicated a significantly greater amount of disapproval than did subjects who had not been exposed to death qualification (.001 significance level). Subjects exposed to death qualification also estimated that the judge was significantly more in favor of the death penalty than non-death qualification subjects estimated (.002 significance level). *1176 Those subjects who had viewed the death-qualifying voir dire gave a higher assessment of the prosecutor’s belief in the defendant’s guilt (.041 level of significance); a significantly higher assessment of the defense attorney’s belief in the defendant’s guilt (.019 level of significance); and a significantly higher assessment of the judge’s belief in the defendant’s guilt (.0001 level of significance). The state contends that the Haney study is a separate legal “claim” which must be exhausted in the North Carolina state courts before it may be presented here. However, the court finds that this study is but one more piece of evidence offered by petitioners to support their original claim that the process by which the juries which tried their cases were selected deprived them of their right to trial by an impartial factfinder. Picard v. Connor, 404 U.S. 270, 275, 92 S.Ct. 509, 512, 30 L.Ed.2d 438 (1971). L. Louis Harris & Associates, Study No. 814002 (1981). Subjects were a representative sample of 1,499 adults randomly drawn from the United States population. Harris found that 0.87% of those surveyed said they would automatically vote for the death penalty in every capital case at which they were asked to sit as jurors, and in which the defendant had been found guilty. (Weighted for accuracy, the figure rises to 1%.) M. Records and briefs from the following cases involving the same issue: Hovey v. Superior Court, 28 Cal.3d 1, 168 Cal.Rptr. 128, 616 P.2d 1301 (1980); Grigsby v. Mabry, 569 F.Supp. 1273 (E.D.Ark.1983); People v. Word and Sparks, No. 78647 (Santa Clara County, Cal., Sup.Ct., 1981). These records include the expert witness testimony of Dr. Faye Girsh, Professor Hans Zeisel, Professor Edward Bronson, Professor Phoebe Ellsworth, Professor Craig Haney, and Dr. Reid Hastie. These experts all explained the methodology and confirmed the validity of the studies submitted as evidence by petitioners. Dr. Zeisel attempted to determine the effect of exclusion of jurors unwilling to impose the death penalty on actual verdicts in close eases. Assuming that 17% of jurors could be fair and impartial in determining guilt or innocence, but would never vote to impose the death penalty in any case, Ellsworth Attitude Survey; and assuming a 25% difference in conviction rates between those who are willing and unwilling to impose the death penalty, Ellsworth Conviction Proneness Study; and assuming further that in a close case, a single vote might make the difference between a “guilty” verdict and a “not guilty” or “no verdict” result, Dr. Zeisel concluded that the effect of death qualification was to increase the likelihood of conviction for defendants who would otherwise not be found guilty by 10 per cent. People v. Kenneth Lynn Moore, No. 67113, Alameda Co., Cal., Sup.Ct., Tr. 108-120. This calculation was confirmed by Dr. Hastie, Grisby v. Mabry, No. PBC-78-32 (E.D.Ark., July 15, 1981), Tr. 182-204. Dr. -Kadane testified in an evidentiary hearing held in People v. Word and Sparks, No. 78647, Santa Clara County Superior Court, on August 31, 1981. Dr. Kadane recalculated the data developed in the Ellsworth Attitude Survey, and the Ells-worth Conviction-Proneness Study, in light of the number of persons found by Harris 1981 to exist in the general population who favor the death penalty to such an extreme degree that they would automatically vote to impose it in every conviction for first-degree murder (ADPs). Dr. Kadane found that the exclusion of potential jurors who will vote to impose the death penalty in every capital case does not correct the 'biasing effects of excluding the much larger group of potential jurors who will never vote to impose the death penalty in any case. Dr. Kadane’s testimony has been presented to the court as evidence, and his underlying data were separately submitted. *1177 The data developed in the ‘Fieldscope Report” will not be considered by the court because that report does not appear in full in the court’s records. The court has thoroughly reviewed the evidence submitted by petitioners, and finds it credible, consistent, and essentially uneontradicted. The court accepts the opinions offered by experts in jury research that the studies are valid and reliable. All of the studies revealed to the court involving juror attitudes toward the death penalty support the same conclusion: Juries from which persons unwilling to impose the death penalty have been excluded are (1) partial to the prosecution in both attitude and behavior, and (2) are not fairly drawn from a representative cross-section of the community. There are no significant studies showing the contrary. Respondents have raised a number of criticisms of the quality of the evidence. First, the state argues that the studies are not valid, because simulated jury research fails to instill in jurors the same sense of “felt responsibility” experienced by jurors in real cases. However, a number of the studies involved real jurors, or persons who had served on juries in the past. These persons stated that their experiences with the mock juries closely resembled their experiences as actual jurors. Second, respondents state that the studies measuring prosecution-prone attitude are not valid, because many of the questions contained in these studies actually measure, a realistic attitude toward the criminal justice system, rather than a prosecution-prone attitude. There is some merit to this argument. It is in fact true that large numbers of persons who are brought to trial are guilty, and that a defendant who refuses to take the witness stand to testify in his own behalf is also often guilty. However, the negligible value of several questions in the studies is more than outweighed by the exhibited differences in response levels to many other questions which are extremely important to the decision, i.e., those concerning attitude toward the insanity defense, the relative truthfulness of prosecution and defense witnesses, and reasonable doubt thresholds. Finally, the state has asserted that the studies submitted by petitioners are invalid because they measure mere attitudes toward the death penalty, and there is no reason to assume that attitudes are a valid predictor of behavior. The studies summarized above do show a substantial correlation between attitudes and actions. In addition, petitioners’ studies measure more than mere attitudes toward the death penalty. The studies measure persons’ predictions of their behavior in hypothetical cases, and more importantly, also measure actual behavior of persons in simulated trial situations. The evidence offered by petitioners was not designed to predict actual juror behavior in every given situation. Instead, these studies were designed to measure trends of behavior that can be predicted both from expressed attitudes and from behavior in simulated jury situations. V. THE EVIDENCE FOR THE RESPONDENTS DOES NOT SERIOUSLY REBUT ANY OF THE PROPOSITIONS SET OUT IN SECTION IV ABOVE. Respondents offered no evidence of studies or observations on the effect which attitudes towards the death penalty have upon the votes of jurors in capital cases. They have submitted copies of articles from numerous publications and scientific and psychiatric bulletins. Those documents have been examined by the court. In one way or another they fall short of informing the court or helping in the decision of the issue. They deal with matters touching juror behavior, such as the effect of jury size; of the weight of the evidence on one side or the other; of group deliberations; of majority vote requirements; of sex; of the heinousness or severity of the crime; of “scientific” jury selection. They *1178 deal with attitudes towards drinking, going to church, political candidacies, breast feeding and the like. Much of what is relevant in those independent offerings support the basic contentions of the petitioners. One article compliments the study of Jurow upon which petitioners rely. None of them seriously supports respondents’ position that attitudes do not tend to influence or predict behavior. One of them shows that it is harder to get a conviction of a capital crime than it is to get a conviction of a less serious crime. The respondents make several further contentions: 1. They contend that jurors who don’t like the death penalty will “lie” about it so they will be accepted on the jury, and thus “nullify” their oath to follow the law and try the case fairly. They cite data from the Bronson — Colorado study and from the Harris study showing that substantial numbers of persons interviewed, who had been asked to assume they had already been seated on a jury, had said they would, in order to protect the defendant from the death sentence, not vote for conviction, despite the weight of the evidence. From this the respondents contend that people who object to the death penalty will lie to get selected on capital trial jury panels so that they can thwart the processes of justice. The trouble with this argument is that there is no evidence to support it. Jurors were not asked whether they would lie to get on the jury so that they could thwart the prosecution. A juror planning to lie in order to get on the jury would most logically say that he or she has no conscientious scruples against the death penalty. Respondents’ position assumes, in fact, that the juror has already lied to get on the jury panel, and this proposition is not supported at all by the Bronson and Harris studies cited. Neither the prosecution nor the defense is entitled to a presumption without evidence that jurors are going to lie. 2. Respondents contend that excluding from the trial of guilt those persons unwilling to impose the death penalty would have only “minimal” impact upon the results. The testimony does not support that contention. The studies of Dr. Hastie and Dr. Zeisel cited above show that if persons opposed to the death penalty were included on juries in the first phase of a split trial, it would change the first ballot vote in those trials in about 10% of the close cases. First ballots often tend to predict final decision. If the exclusion is constitutionally wrong, the fact that only one person out of ten in close cases might lose his life because of it is no argument that that person’s constitutional rights should not be recognized. 3. Respondents say that studies of Dr. Ellsworth are unreliable; nevertheless, they do address the question before the court; the experts who have reviewed those studies have complimented them on the methods used, and have particularly approved Dr. Ellsworth’s use of simulated jury deliberation. The court, continues to find the Ellsworth studies to be useful and reliable. 4. As to the affidavits of Dr. Appelbaum, it is obvious that Dr. Appelbaum is qualified to do the task that he was asked to do; that is, to assume facts contrary to the facts found by Dr. Hastie and Dr. Zeisel, and to record his calculations upon those different assumptions. Even with the changed assumptions, however, Dr. Appelbaum concludes that the accused will suffer prejudice in at least one case out of 100 capital cases. This court has trouble seeing any state interest in maintaining a system which even the state’s witnesses concede will unnecessarily threaten the life of an accused in 1% of the trials. It must be remembered that the petitioners’ evidence is the only source of information before the court on the central factual issue presented by the cases, which is, whether or not excluding persons who oppose the death penalty while retaining persons who favor the death penalty unlawfully increases the likelihood of conviction at the guilt determination stage. *1179 5. Respondents finally contend that plaintiffs have the right to remove people (maybe 1% of the jury panel) who say that they would automatically vote for the death penalty in any case, and that this right to take off such persons balances and makes up for the state’s present right to exclude the much larger group of conscientious objectors to the death penalty. The trouble with this argument is that it balances the state’s assertion of power against constitutional right. Governments have no “rights”; they only have powers and duties. It is people who have “rights” guaranteed by the Constitution. Exclusion of one group by the prosecution and the other group by the defense does not solve the problem of fair trial, nor of petitioners’ entitlement to trial by a jury drawn from a panel which is a representative cross-section of the community. While the state has a recognized right to a fair trial before an impartial tribunal, see Arizona v. Washington, 434 U.S. 497, 98 S.Ct. 824, 54 L.Ed.2d 717 (1978), that right does not rise to the level of protection afforded the accused by the Sixth and Fourteenth Amendments. “The Bill of Rights does not envision an adversary proceeding between two equal parties.” Wardius v. Oregon, 412 U.S. 470, 480, 93 S.Ct. 2208, 2215, 37 L.Ed.2d 82 (1973) (Douglas, J. concurring). It is for this reason that a criminal conviction may not be constitutionally sustained unless it is founded upon proof of guilt beyond a reasonable doubt. In re Winship, 397 U.S. 358, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970). This does not mean that the trier of fact may abandon its duty “to hold the balance nice, clear and true between the State and the accused.” Tumey v. Ohio, 273 U.S. 510, 532, 47 S.Ct. 437, 444, 71 L.Ed. 749 (1927). Nor is a criminal defendant constitutionally entitled to a jury panel comprised of persons particularly willing to acquit. But here the state has, through a partial jury selection process, created a jury that is uncommonly willing to convict. Cf. Witherspoon v. Illinois, 391 U.S. at 521, 88 S.Ct. at 1776. A jury from which the attitudes most favorable to the defense have been eliminated can not be termed “balanced” in any constitutionally permissible sense of the word. VI. THE EXCLUSION FROM THE GUILT PHASE OF A.CAPITAL TRIAL OF JURORS WHO “WOULD BE” UNWILLING TO IMPOSE THE DEATH PENALTY AT A SEPARATE SENTENCING PHASE OF THE TRIAL VIOLATES PETITIONERS’ RIGHT TO A JURY COMPRISED OF A REPRESENTATIVE CROSS-SECTION OF THEIR COMMUNITIES. In Witherspoon v. Illinois, 391 U.S. 510, 88 S.Ct. 1770, 20 L.Ed.2d 776 (1968), the Supreme Court held that a state was constitutionally permitted to exclude two groups of persons from juries in capital murder trials — (1) those who state that they would automatically vote against the death penalty at the punishment phase of the trial, and (2) those whose attitudes toward the death penalty would prevent them from being fair and impartial triers of the defendant’s guilt or innocence. The Witherspoon Court did not address the question whether the exclusion from the guilt phase of a bifurcated capital trial of persons who firmly oppose the death penalty violated a state defendant’s right to a representative jury; that question was left open for later decision. 391 U.S. at 520, n. 18, 88 S.Ct. at 1776, n. 18. Since Witherspoon was decided, the Supreme Court has later specifically held that the Sixth Amendment, applicable to state criminal proceedings, requires that a defendant be tried by a jury drawn from a representative cross-section of the community. Taylor v. Louisiana, 419 U.S. 522, 95 S.Ct. 692, 42 L.Ed.2d 690 (1975); Duren v. Missouri, 439 U.S. 357, 99 S.Ct. 664, 58 L.Ed.2d 579 (1979). In Taylor, the Supreme Court held that a jury panel reasonably drawn from a fair cross-section of the community is fundamental to the jury trial guaranteed to a *1180 criminal defendant by the Sixth Amendment to the United States Constitution, held applicable to the states in Duncan v. Louisiana, 391 U.S. 145, 88 S.Ct. 1444, 20 L.Ed.2d 491 (1968). In finding that a state jury-selection system, which operated to exclude women from jury service in numbers grossly disproportionate to their existence in the community, failed to comport with the requirements of the Sixth and Fourteenth Amendments, the Court stated: “The purpose of a jury is to guard against the exercise of arbitrary power— to make available the commonsense judgment of the community as a hedge against the overzealous or mistaken prosecutor and in preference to the professional or perhaps over-conditioned or biased response of a judge, Duncan v. Louisiana, 391 U.S. 145, 155-156 [88 S.Ct. 1444, 1450-1451, 20 L.Ed.2d 491] (1968). This prophylactic vehicle is not provided if the jury pool is made up of only special segments of the populace or if large, distinctive groups are excluded from the pool. Community participation in the administration of the criminal law, moreover, is not only consistent with our democratic heritage but is also critical to public confidence in the fairness of the criminal justice system. Restricting jury service to only special groups or excluding identifiable segments playing major roles in the community cannot be squared with the constitutional concept of jury trial. ‘Trial by jury presupposes a jury drawn from a pool broadly representative of the community as well as impartial in a specific case ... [T]he broad representative character of the jury should be maintained, partly as assurance of a diffused impartiality and partly because sharing in the administration of justice is a phase of civic responsibility.’ Thiel v. Southern Pacific Co., 328 U.S. 217, 227 [66 S.Ct. 984, 989, 90 L.Ed. 1181] (1946) (Frankfurter, J. dissenting).” 419 U.S. at 530-531, 95 S.Ct. at 698. Taylor reaffirmed, under the Sixth Amendment, a cross section requirement which had long been maintained under various other principles of authority, including the Equal Protection Clause of the Fourteenth Amendment, e.g., Strauder v. West Virginia, 100 U.S. 303, 25 L.Ed. 664 (1880); Pierre v. Louisiana, 306 U.S. 354, 59 S.Ct. 536, 83 L.Ed. 757 (1939); Hernandez v. Texas, 347 U.S. 475, 74 S.Ct. 667, 98 L.Ed. 866 (1954); the Due Process Clause of the Fourteenth Amendment, Peters v. Kiff 407 U.S. 493, 92 S.Ct. 2163, 33 L.Ed.2d 83 (1972); and the Court’s supervisory power over lower federal courts, Thiel v. Southern Pacific Co., 328 U.S. 217, 66 S.Ct. 984, 90 L.Ed. 1181 (1946); Ballard v. United States, 329 U.S. 187, 67 S.Ct. 261, 91 L.Ed. 181 (1946). See, Witcher v. Peyton, 405 F.2d 725 (4th Cir.1969). Sixth Amendment violations are not confined to cases of race or sex discrimination in jury selection. Strauder v. West Virginia, supra, 100 U.S. at 308 (exclusion of Celtic Irishmen would be inconsistent with spirit of Fourteenth Amendment) (dictum). “Throughout our history differences in race and color have defined easily identifiable groups which have at times required the aid of the courts in securing equal treatment under the laws. But community prejudices are not static, and from time to time other differences from the community norm may define other groups which need the same protection. Whether such a group exists within a community is a question of fact. When the existence of a distinct class is demonstrated, and it is further shown that the laws, as written or as applied, single out that class for different treatment not based on some reasonable classification, the guarantees of the Constitution have been violated.” Hernandez v. Texas, supra, 347 U.S. at 478, 74 S.Ct. at 670. In Thiel v. Southern Pacific Co., supra, the Court reversed a jury verdict unfavorable to a plaintiff in a civil case tried in federal court because the jury had been selected from a panel from which daily wage earners automatically had been ex- *1181 eluded. As the Court stated in Peters, v. Kiff, supra, “the exclusion from jury service of a substantial and identifiable class of citizens has a potential impact that is too subtle an too pervasive to admit of confinement to particular issues or particular cases____ When any large and identifiable segment of the community is excluded from jury service, the effect is to remove from the jury room qualities of human nature and varieties of human experience, the range of which is unknown and perhaps unknowable.” Id. 407 U.S. at 503, 92 S.Ct. at 2169 (emphasis added). In order to establish a prima facie violation of the Sixth Amendment’s fair cross section requirement, a criminal defendant must show (1) that the group alleged to be excluded is a “distinctive” group in the community; (2) that the representation of this group in venires from which juries are selected is not fair and reasonable in relation to the number of such persons in the community; and (3) that this underrepresentation is due to systematic exclusion of the group in the jury-selection process. United States v. Espinoza, 641 F.2d 153 (4th Cir.1981); Duren v. Missouri, 439 U.S. 357, 364, 99 S.Ct. 664, 668, 58 L.Ed.2d 579 (1979). Petitioners have met their burden of establishing a prima facie violation of the Sixth Amendment in the manner in which the juries were selected at the guilt phase of their criminal trials. First, the court finds that persons who are unwilling to consider voting for the death penalty constitute a “distinctive group” in the community. Persons who are unwilling to impose the death penalty are estimated to constitute 8% to 23% of the population of the United States, varying by community. See, e.g., Harris —1971 (23%); Bronson —Colorado (8-9%); Ellsworth Attitude Survey (17% in Alameda County, California). Sheer numbers, however, will not alone demonstrate that a defendant has been deprived of a representative jury, for whenever a jury is selected, whether by chance or through legitimate rules of exclusion, large portions of the population remain unrepresented on a particular jury. The question for determination is whether the group is distinctive or cognizable in some manner so that its exclusion will prevent the jury from representing the collective conscience of the community. Members of the community who would be unwilling to impose the death penalty in any case were originally characterized as a legally significant group by the Supreme Court in Witherspoon v. Illinois, 391 U.S. 510, 88 S.Ct. 1770, 20 L.Ed.2d 776 (1968). See, Grigsby v. Mabry, 483 F.Supp. 1372, 1382 (E.D.Ark.1980). This group was specifically found to differ from the group of persons who generally oppose the death penalty but not so strongly that they would never consider imposing it in a proper case. The difference was established by the Court for the purpose of determining whether the Due Process Clause permitted a state to exclude from a jury in a non-bifurcatéd capital trial any category of persons who were opposed to the death penalty. The evidence before the court overwhelmingly demonstrates that persons who are unwilling to impose the death penalty share a unique set of attitudes toward the criminal justice system which separate them as a group not only from persons who favor the death penalty, but also from persons who are generally opposed to the death penalty, but are willing to consider it in some cases. These attitudes are consistently more favorable for the defense than they are for the prosecution. The results of the studies submitted by petitioners reveal that persons who are willing to impose the death penalty exhibit the following attitudes to a greater degree than persons who will not impose that penalty, and hence are excludable from capital trials: these persons tend to believe prosecution witnesses more often, and disbelieve defense witnesses more often, Bronson— California; Bronson — Colorado; Ells-worth Attitude Survey; Ellsworth Conviction-Proneness Study; Ellsworth Witness Credibility Study; Harris 1971; they *1182 believe in stricter methods of crime control and believe that the courts are creating too many technical loopholes by which guilty défendants go free, Bronson — California, Bronson — Colorado, Harris 1971; they tend to believe that a defense attorney will resort to any trick to free his or her client, Ellsworth Attitude Survey, Harris 1971; and they have a lower threshold standard of proof beyond a reasonable doubt, Bronson — California, Bronson — Colorado, Ellsworth Post-Deliberation Follow-Up Study. Seventy-seven per cent of a group of strongly opinionated persons unwilling to impose the death penalty in Alameda County agreed strongly with the statement that “Even the worst criminal should be considered for mercy”; only twenty-seven per cent of strongly opinionated persons willing to impose the penalty agreed. Ells-worth Attitude Survey (probability level .00001). It is hardly surprising to find that persons unwilling to impose the death penalty share a distinct set of values and attitudes; it is for this very reason that the state has excluded them from capital trials. The attitudes shared by persons unwilling to impose the death penalty are highly relevant to the determination of the strength of the evidence at the guilt phase of a capital trial. These persons share views that are generally favorable to the defendant; “certainly, [n]o one else will represent their strong viewpoint on the jury in their absence.” Grigsby v. Mabry, supra, at 1382; State v. Avery, 299 N.C. 126, 261 S.E.2d 803 (1980) (Exum, J., dissenting). Exclusion of persons unwilling to impose the death penalty from the guilt phase of capital trials further offends the Sixth Amendment in that it inevitably leads to the disproportionate exclusion of distinctive groups such as blacks and women, who tend to oppose the death penalty to a greater degree than white men. Ballew v. Georgia, 435 U.S. 223, 98 S.Ct. 1029, 55 L.Ed.2d 234 (1978). Respondents contend that persons unwilling to impose the death penalty are not a discrete group for purposes of Sixth Amendment analysis, because they have not been demonstrated to have shared similar life experiences. In support of this contention, respondents refer the court’s attention to Lockett v. Ohio, 438 U.S. 586, 98 S.Ct. 2954, 57 L.Ed.2d 973 (1978). This contention reveals a misinterpretation of the cross section representation requirement derived from a mechanical interpretation of the Supreme Court’s cases in this area. The Court has on numerous occasions condemned the exclusion from jury service of members of distinct groups, partly because such persons (often black people or women) have shared similar life experiences. The fact that these persons have shared similar life experiences has led to some surmise that they may share an attitude, or an influence on others, that is essential to an expression of the collective conscience of the community. See, e.g., Ballard v. United States, supra, 329 U.S. at 193-94, 67 S.Ct. at 264; Peters v. Kiff 407 U.S. 493, 503-04, 92 S.Ct. 2163, 2168-69, 33 L.Ed.2d 83: “It is not necessary to assume that the excluded group will consistently vote as a class in order to conclude, as we do, that its exclusion deprives the jury of a perspective on human events. that may have unsuspected importance in any case that may be presented.” [Emphasis supplied.] Respondents’ reliance on Lockett v. Ohio, supra, to support their contention is misplaced. In that case, four potential jurors were excused from service when they stated during voir dire that their opposition to the death penalty was so strong that it would prevent them from taking an oath fairly to try the case or to follow the law. 438 U.S. at 595-96, 98 S.Ct. at 2959-60. Petitioners do not dispute the validity of this holding, but the holding is irrelevant to the issue presently before the court. Petitioners have easily met their remaining burden in establishing a prima facie ease of nonrepresentation under Duren. No one seriously contends that persons unwilling to impose the death penalty are in fact represented on juries in North Carolina which decide the question of guilt in *1183 capital cases, nor that the exclusion of such persons is other than systematic and deliberate. The state has argued that petitioners have a constitutional right only to a jury venire from which distinctive groups are not excluded. It is true that the representative cross-section standards developed in Taylor v. Louisiana, supra, and Duren v. Missouri, supra, were held to apply to grand jury panels, and jury venire panels, but not to actual criminal trial juries. However, where, as here, the state may successfully challenge, for cause, every potential juror who states he or she is unwilling to consider imposing the death penalty, there is no fair possibility that such a person will ever serve on a trial jury in a capital case. The distinction between a jury venire and the actual trial jury is thus irrelevant in determining whether a death qualified trial jury expresses the conscience of the community; no matter when it is done, death qualification produces the intended result of absolute exclusion from any meaningful jury service of persons who do not approve of the death penalty. Cf. United States v. Branscome, 682 F.2d 484 (4th Cir.1982) (selection of volunteers to serve on federal grand jury results in non-random selection process in violation of federal law). Petitioners having successfully established a prima facie case of unconstitutional juror exclusion, the burden is on respondents to demonstrate that a significant state interest is “manifestly and primarily advanced” by the exclusion of this distinctive group from the guilt phase of state capital trials. Duren v. Missouri, supra, 439 U.S. at 367-68, 99 S.Ct. at 670-71. In Duren, the Court emphatically cautioned that any exemption from jury service designed to further a significant state interest must be appropriately tailored to serve that interest, so as to reduce the substantial threat that the remaining pool of jurors will not be representative of the community. Id. 439 U.S. at 370, 99 S.Ct. at 670; Taylor v. Louisiana, supra, 419 U.S. at 534, 95 S.Ct. at 699. Respondents have raised few substantial arguments in opposition to petitioners’ claim that they were deprived of juries representing a cross-section of the community. First, respondents contend that just as it is fair that 56% [sic] of state court six-person juries lack minority representation entirely, Ballew v. Georgia, 435 U.S. 223, 98 S.Ct. 1029, 55 L.Ed.2d 234 (1978), so also it is fair that all North Carolina capital juries exclude a certain number of death penalty opponents. However, the Sixth Amendment comprehends “a fair possibility for obtaining a representative cross-section of the community.” Williams v. Florida, 399 U.S. 78, 100, 90 S.Ct. 1893, 1906, 26 L.Ed.2d 446 (1970). In Ballew, the Supreme Court implicitly held that 53.1% of randomly selected juries without minority representation in communities with 10% minority group population indicated that criminal defendants still retained a “fair possibility” of obtaining a representative cross-section of the community. In this case, the state urges the court to adopt the position that “no possibility ” of obtaining a representative cross-section of the community with regard to a specific attitude toward the criminal justice system is somehow also comprehended by the Sixth Amendment. Such a position is contrary to the principles of Ballew, and of the other Sixth Amendment representation cases. Respondents’ most serious argument is that a trial jury on the question of guilt which included members of the community unwilling to impose a death sentence if called upon to do so would be unfairly partial to the defendant. On that score, it must be remembered that a North Carolina death penalty trial is required by statute to be a bifurcated proceeding, at which the question of guilt-or-innocence and the question of penalty are considered in separate phases of the trial. Petitioners concede that the state may constitutionally exclude persons unwilling to impose the death penalty from the penalty determination phase of a criminal trial. The only question remaining for decision, then, is whether the state’s interest in excluding *1184 from the guilt phase of the trial some persons who might be more partial to the defense is sufficient to overcome petitioners’ Sixth Amendment right to a jury-drawn from a representative cross-section of the community. Respondents’ argument amounts to a declaration that a criminal defendant is entitled to a jury drawn from a representative cross-section of the community as long as that jury is not more sympathetic to the defendant than it is to the prosecution. This problem was duly considered and provided for by the drafters of the Sixth Amendment and the formulators of the American jury system, who determined that a fair trial would most likely be had before a truly representative jury. If the conscience of the community is sympathetic to the defendant the state must be content with that state of affairs, as long as it is able to obtain jurors who will follow their oaths and obey the law. The prosecution must appreciate and respect the community in which it operates; it may not artificially manipulate otherwise randomly selected representatives of that community to create a panel unnaturally predisposed to convict. While the state has a legitimate interest in a fair trial, the Constitution requires that the state’s jury selection procedures be appropriately tailored to serve that interest. The Supreme Court has recognized that jurors who adamantly oppose capital punishment are able to follow their oaths as jurors and reach a conclusion on the basis of the facts before them. See Witherspoon v. Illinois, supra, 391 U.S. at 514-15, n. 7, 88 S.Ct. at 1773, n. 7; Adams v. Texas, 448 U.S. 38, 44-45, 100 S.Ct. 2521, 2526, 65 L.Ed.2d 581 (1980), quoting Boulden v. Holman, 394 U.S. 478, 483-84, 89 S.Ct. 1138, 1141-42, 22 L.Ed.2d 433 (1969). This is particularly true in split trials, where the juror’s duty at the first phase is to determine guilt or innocence, and not to determine whether the death penalty is appropriate for a convicted defendant. The State of North Carolina may not, consistently with the Sixth Amendment’s representative jury requirement, exclude jurors who would be unwilling to impose the death penalty from the guilt phase of a bifurcated capital trial unless those jurors unequivocally state that their views toward capital punishment would prevent them from following their oaths and making an impartial decision as to the defendant’s guilt. VII. PETITIONERS WERE DENIED DUE PROCESS OF LAW BY THE ACTION OF THE TRIAL JUDGE IN ALLOWING THE PROSECUTION TO CHALLENGE FOR CAUSE, AT THE GUILT — INNOCENCE STAGE OF THE TRIAL, ALL JURORS WHO SAID ON VOIR DIRE EXAMINATION THAT ON THE QUESTION OF PENALTY' [A QUESTION NOT THEN PRESENTED TO THE JURY] THEY COULD NOT VOTE FOR THE DEATH PENALTY The Sixth and . Fourteenth Amendments guarantee to a criminal defendant a fair trial by a panel of impartial, indifferent jurors. See, Irvin v. Dowd, 366 U.S. 717, 81 S.Ct. 1639, 6 L.Ed.2d 75(1961). Long before the Sixth Amendment was held applicable to the states in Duncan v. Louisiana, supra, the Supreme Court held that procedures or practices which might predispose the trier of fact to convict violated the Due Process Clause of the Fourteenth Amendment. Turney v. Ohio, 273 U.S. .510, 47 S.Ct. 437, 71 L.Ed. 749 (1927) (trial by judge with financial stake in outcome of case deprives defendant of due process); Turner v. Louisiana, 379 U.S. 466, 85 S.Ct. 546, 13 L.Ed.2d 424 (1965) (placement of jury in protective custody of principal prosecution witness violates due process). See, Grigsby v. Mabry, 569 F.Supp. 1273 (E.D.Ark.1983). “A fair trial in a fair tribunal is a basic requirement of due process.” In re Murchison, 349 U.S. 133, 136, 75 S.Ct. 623, 625, 99 L.Ed. 942 (1955). A criminal defendant need not demonstrate actual prejudice in order to obtain *1185 relief on a claim that he was tried by an unfairly constituted jury. Due process is denied by circumstances creating the “likelihood or appearance” of bias, Peters v. Kiff 407 U.S. 493, 502, 92 S.Ct. 2163, 2168, 33 L.Ed.2d 83 (1972), or “even the probability” of unfairness, In re Murchison, supra, 349 U.S. at 136, 75 S.Ct. at 625. Trial courts are charged with the duty of ensuring that “the balance is never weighed against the accused.” Sheppard v. Maxwell, 384 U.S. 333, 362, 86 S.Ct. 1507, 1522, 16 L.Ed.2d 600 (1966) (defendant tried by jury exposed to prejudicial pretrial publicity deprived of due process). As previously discussed, petitioners have submitted evidence substantially demonstrating that persons unwilling to impose the death penalty are a distinctive group sharing a unique set of attitudes that are more favorable to the defense than are the attitudes of persons who are willing to impose the death penalty. Juries from which such persons have been excluded do not reflect these pro-defense attitudes, and are thus unfairly, and unconstitutionally, skewed in favor of the prosecution. Moreover, persons who are willing to impose the death penalty not only share a set of attitudes that are more favorable to the prosecution, but also are predictably more likely to decide.in a manner that favors the prosecution. Several studies submitted by petitioners specifically concentrated on the manner in which pro-prosecution attitudes on the part of death qualified jurors translate into pro-prosecution behavior. The results of these studies were consistently the same: with due consideration of the strength of the evidence, persons who are willing to impose the death penalty will vote to convict more often than will persons who are unwilling to impose the death penalty. Some of the studies submitted by petitioners employed actual jurors, some did not. In a most recently completed and most sophisticated study in this area, study participants, who were pre-screened for an ability to follow their oaths and obey the law at the guilt phase of a capital trial, viewed a videotaped re-enactment of an actual criminal trial. At the end of the trial, and after deliberating for an hour, the jurors returned their verdicts: In this close case, 77.9% of the death-qualified jurors (willing to impose the death penalty) convicted the defendant of some degree of homicide, while only 53.3% of the currently excludable jurors voted that way. Ells-worth Conviction-Proneness Study (level of statistical significance: .01). See also, Harris 1971 (4-7% difference in conviction rates at .01 probability level); Jurow (higher conviction rates for jurors willing to impose death penalty among subjects exposed to tape recordings of simulated murder trials); Goldberg (subjects with conscientious scruples against death penalty vote to convict more often than subjects without scruples); Zeisel (actual trial jurors with conscientious scruples against death penalty report higher percentage of votes to acquit than votes reported by jurors without scruples). Trial judges and academic authorities of course agree that the weight of the evidence is the largest single determinant of how jurors will vote in criminal cases. Kalven & Zeisel, The American Jury (1966), 488. Thus the behavior of jurors with differing attitudes toward the death penalty will be most significant in close cases— ones in which the evidence can be interpreted either way. It is in these close cases that criminal defendants most need the protection of the Sixth Amendment. Drs. Zeisel and Hastie, whose testimony in Grigsby v. Mabry is before the court, found that the data gathered by Ellsworth led to the conclusion that the exclusion of persons unwilling to impose the death penalty from the guilt phase of capital trials results in a ten per cent higher conviction rate than would normally be expected to occur in close cases, i.e.; ones in which the evidence is not strong enough to produce a clear majority vote for conviction on the first ballot. Dr. Appelbaum, the state’s expert, recalculated Dr. Ellsworth’s data, using different, and lower, statistical assumptions, and *1186 still concluded that the exclusion of these persons results in at least one more conviction than would normally be expected' to occur in 100 trials at which the evidence is of similar strength to that found in the Ellsworth Conviction-Proneness Study’s, hypothetical case. The actual numbers of increased convictions predicted are not so important as the fact that increased convictions systematically and repeatedly do occur as the result of death qualification. Because this system of exclusion results in the certainty that jury verdicts will be skewed only in one direction — against the accused — petitioners need not demonstrate that the juries which actually heard their cases were prejudiced against them personally as defendants. The presumption of prejudice is only escalated by a series of undesirable side effects which result from death qualification of jurors selected for the guilt phase of a trial. The voir dire itself tends to instill in the jury a sense that the defendant is guilty, and that the death penalty is the appropriate penalty for him. The process of excluding jurors with differing backgrounds and viewpoints results in reduced jury deliberation (Ellsworth), a situation explicitly condemned by the Supreme Court in Ballew v. Georgia, supra, and by the state in this case. The Constitution has long been held to prohibit jury selection along ideological lines, because the danger of such a selection process is that it might tend to produce a bias in favor of the prosecution. Glasser v. United States, 315 U.S. 60, 86, 62 S.Ct. 457, 472, 86 L.Ed. 680 (1942) (jury selected from list of League of Women Voters is organ of special class and openly partisan). To meet its constitutional burden of insuring a fair trial for the accused, the state must employ grounds for exclusion which are “reasonably and closely related to the juror’s suitability for the kind of service ... require[d] or to his fitness to judge the kind of case” which he is being asked to try. Fay v. New York, 332 U.S. 261, 270, 67 S.Ct. 1613, 1619, 91 L.Ed. 2043 (1947). It was because of this constitutional proscription that the Supreme Court held in Witherspoon v. Illinois, supra, that, the state could not exclude from capital juries persons who voiced general conscientious scruples against the death penalty. The Court held that in excluding such persons, the State of Illinois had “produced a jury uncommonly willing to condemn a man to die.” Id. 391 U.S. at 521, 88 S.Ct. at 1776. Instead, the state was allowed, after Witherspoon, to exclude only those venirepersons who made it “unmistakably clear” (1) that they would automatically vote against the imposition of capital punishment without regard to any evidence that might be developed at the trial of the case before them, or (2) that their attitude toward the death penalty would prevent them from making an impartial decision as to the defendant’s guilt. Id. at 522-23, n. 21, 88 S.Ct. at 1777-78, n. 21 [emphasis in original]. The evidence in these cases demonstrates that the above grounds for exclusion are not closely related to a juror’s suitability for service at the guilt phase of a capital trial, or to a juror’s ability to judge the guilt or innocence of a criminal defendant who may or may not face imposition of the death penalty at a later proceeding. THE BURDEN ON THE STATE The Attorney General argued at the hearing that it would be unnecessarily burdensome upon the state to exclude from the guilt phase of the prosecution persons who have serious reservations about the death penalty. Separate trials of those issues before separate juries, or separate trial of the punishment issue before a freshly constituted jury, would be too much of a burden, they say, on the taxpayers. No evidence was offered on this subject; opinions of lawyers who made arguments were contradictory; it is highly reasonable to believe that if the jury which determines guilt or innocence is made up as petitioners request, the selection of the jury and the *1187 trial of the issue of guilt may take much less time in a typical case than it does now. In any event, as to future cases that contention is totally moot, because the new North Carolina capital punishment statute, G.S. § 15A-2000, effective June 1, 1977, quoted above, expressly requires two different jury trials, one on the question of guilt and the other on the question of penalty. There may be burdens upon a government which would justify bypassing the Constitution in order to execute an accused, but I have been unable to imagine such burdens, and in these cases they have not been demonstrated. The State of North Carolina can afford the few extra dollars, if any, that it might cost to provide fair trials to persons accused of capital felonies. CONCLUSION Based on the foregoing and on all the evidence and pleadings and briefs in the record, I find that petitioner Carter has not shown exclusion for cause of jurors who did not believe in the death penalty. He is not entitled to relief. The other petitioners have made such a showing and are entitled to relief. IT IS THEREFORE ORDERED: 1. That the claim of petitioner Carter for relief is DENIED. 2. That the claims of petitioners Keeten, Avery and Williams are GRANTED and the writ of habeas corpus is GRANTED as to each of them. 3. That petitioner Williams’ claim for relief from his sentence is ALLOWED. 4. Counsel will tender further orders consistent with the above. Supplemental Opinion On January 12,1984, this court entered a decision in all the four captioned cases with reference to the routine exclusion of death penalty opponents from juries in capital criminal prosecutions. That decision included, without extended discussion, a ruling that petitioner Larry Darnell Williams’ death sentence should be set aside. This opinion, perhaps unnecessary, gives the reasons for that decision. In June, 1980, in Gaston County, North Carolina Superior Court, petitioner Williams was convicted by a jury of first degree murder. A later sentencing hearing, before the same jury, produced a recommendation that petitioner be sentenced to death, and on that recommendation the trial judge sentenced Williams to death. Petitioner appealed his conviction to the North Carolina Supreme Court. State v. Williams, 305 N.C. 656, 292 S.E.2d 243 (1982). That court found no error in petitioner’s conviction or in the imposition of his sentence. Justice Exum (305 N.C. at 691), relying on the reasoning of his dissent in State v. Pinch, 306 N.C. 1, 49-56, 292 S.E.2d 203, 236-40 (1982), dissented on -grounds that petitioner’s death sentence had been unconstitutionally imposed in violation of Witherspoon v. Illinois, 391 U.S. 510, 88 S.Ct. 1770, 20 L.Ed.2d 776 (1968). In its January 12, 1984, decision, this court found that petitioner’s conviction had been unconstitutionally obtained because persons who were unwilling to impose the death penalty were systematically excluded from the jury that determined petitioner’s guilt. Petitioner Williams also moved that his death sentence be set aside as unconstitutionally imposed because three jurors who were not shown to be irrevocably committed to vote against the death penalty regardless of the circumstances of the case were improperly excused for cause during the voir dire at petitioner’s trial. This question was briefed by both parties, and argument was heard on petitioner’s motion at the September-21, 1983, hearing of these four cases. Witherspoon v. Illinois, 391 U.S. 510, 88 S.Ct. 1770, 20 L.Ed.2d 776 (1968), is the leading case. In Witherspoon, the Supreme Court held that a death sentence cannot, consistent with the due process clause of the Fourteenth Amendment, be carried out “if the jury that imposed or recommended it was chosen by excluding *1188 veniremen for cause simply because they voiced general objections to the death penalty or expressed conscientious or religious scruples against its infliction.” 391 U.S. at 522, 88 S.Ct. at 1777. The Court held that a trial judge may excuse for cause only those jurors who make “unmistakably clear (1) that they would automatically vote against the imposition of capital punishment without regard to any evidence that might be developed at the trial of the case before them, or (2) that their attitude toward the death penalty would prevent them from making an impartial decision as to the defendant’s guilt.” 391 U.S. at 522-23, n. 21, 88 S.Ct. at 1777, n. 21 [emphasis in original]. Because Footnote 21 from the Wither-spoon decision includes so much of the detail and substance of the decision, I quote it in full: Just as veniremen cannot be excluded for cause on the ground that they hold such views, so too they cannot be excluded for cause simply because they indicate that there are some kinds of cases in which they would refuse to recommend capital punishment. And a prospective juror cannot be expected to say in advance of trial whether he would in fact vote for the extreme penalty in the case before him. The most that can be demanded of a venireman in this regard is that he be willing to consider all of the penalties provided by state law, and that he not be irrevocably committed, before the trial has begun, to vote against the penalty of death regardless of the facts and circumstances that might emerge in the course of the proceedings. If the voir dire testimony in a given case indicates that veniremen were excluded on any broader basis than this, the death sentence cannot be carried out even if applicable statutory or case law in the relevant jurisdiction would appear to support only a narrower ground of exclusion. See nn. 5 and 9, supra. We repeat, however, that nothing we say bears upon the power of a State to execute a defendant sentenced to death by a jury from which the only veniremen who were in fact excluded for cause were those who made unmistakably clear (1) that they would automatically vote against the imposition of capital punishment without regard to any evidence that might be developed at the trial of the case before them, or (2) that their attitude toward the death penalty would prevent them from making an impartial decision as to the defendant’s guilt. Nor does the decision in this case affect the validity of any sentence other than one of death. Nor, finally, does today’s holding render invalid the conviction, as opposed to the sentence, in this or any other case. [391 U.S. 522-523, 88 S.Ct. 1776-1777; emphasis in original.] Consistent with Witherspoon, a state may not exclude jurors from the sentencing phase of a capital trial on any broader basis than an irrevocable opposition to capital punishment that would “frustrate the State’s legitimate efforts to administer its constitutionally valid death penalty scheme.” Adams v. Texas, 448 U.S. 38, 51, 100 S.Ct. 2521, 2529, 65 L.Ed.2d 581 (1980); Witherspoon v. Illinois, supra, 391 U.S. at 522, footnote 21, 88 S.Ct. at 1777, footnote 21. Petitioner Williams asserts that, judged by those standards, jurors Melton, Robertson and Williams were unlawfully excluded from the jury that sentenced him. His argument is directed principally against the exclusion of Ms. Nancy Melton, whose answers to questioning during voir dire least clearly demonstrate irrevocable opposition to the death penalty. [Exclusion from the jury of even one juror whose statements fail to reveal the position of opposition to the death penalty described in Wither-spoon requires reversal of a death sentence. Davis v. Georgia, 429 U.S. 122, 97 S.Ct. 399, 50 L.Ed.2d 339 (1976).] The trial judge asked Ms. Melton the following questions and, based on her answers, excused her for cause: *1189 Q [by the Court]. And do you understand it will be the duty of jurors at the conclusion of [the guilt] phase of the case to return a verdict as to guilt or innocence in accordance with the instructions of the court as to the law of North Carolina? A. Yes. Q. If you are chosen as a juror, can you and will you do this? A. Yes. Q. Now, you understand that if the jury in this case finds the — beyond a reasonable doubt that the defendant is guilty of first-degree murder, it will then, be the duty of the same jury under the law of North Carolina to make a recommendation as to whether the defendant shall be punished by life imprisonment or death? A. Yes. Q. And do you understand that it is the duty of jurors during that phase of the case to make a sentence recommendation under — and in accordance with the law of North Carolina as will be explained by the court? A. (The juror nods her head.) Q. Now, if you serve as a juror in this ease, could and would you if called upon to do so make a sentence recommendation of life imprisonment or death in accordance with the law as it will be explained to you by the court, or would you be unable to do so regardless of the law and the facts and circumstances revealed by the evidence because of some conscientious belief as to the proper punishment for first-degree murder; that is, that you conscientiously feel that it should in all eases be life in prison or you feel that in all cases it should be the death penalty? A. Your Honor, I’m not sure I could say that someone else had to die. I’m not sure I could do that. Q. Well, it’s — this is something we have to determine at this stage — whether you if you serve as a juror could follow the law of North Carolina and if you 'are satisfied beyond a reasonable doubt of those things which the law requires you to be satisfied you could then return a recommendation of the death penalty. A. I’m not positive I could do that. I’ve never been called on to do that, and I’m not sure that I could live with my conscience. Q. Well, do you have conscientious beliefs about the death penalty — religious beliefs about it? A. Yes. Q. And you do not feel that you could follow the instructions of the court if you were satisfied beyond a reasonable doubt of the things of which you must be satisfied. If those conclusions would call for the death penalty, you don’t feel you could make such a recommendation? A. I’m not sure that I could. Q. Do you wish to ask her any questions? DISTRICT ATTORNEY: The State would challenge her for cause. MR. CLONINGER: We don’t — I don’t wish to ask her any questions, Your Hon- or. COURT: All right. The challenge for cause is allowed. Thank you, Mrs. Melton. I’m going to ask you to go up to Courtroom B. Judge Kirby will know whether he needs you for any other ease. Okay. Thank you. EXCEPTION NO. 15 MR. CLONINGER: Could we for the record enter an [objection]. Ms. Melton’s statement that she had conscientious or religious beliefs with respect to the death penalty and statements that she was “not positive” or “not sure” thát she would be able to vote to impose the death penalty do not show her to be “irrevocably committed ... to vote against the penalty of death regardless of the facts and circumstances that might emerge in the course of the proceedings.” Witherspoon v. Illinois, supra, 391 U.S. at 522 n. 21, 88 S.Ct. at 1777 n. 21. The state may only require that a juror in a capital sentencing trial be willing to consider all the penalties provided by state *1190 law. Witherspoon v. Illinois, supra, at 522 n. 21, 88 S.Ct. at 1777 n. 21. The Constitution does not permit a state to excuse for cause those jurors who are unwilling to impose a death sentence for every first degree murder conviction; to the contrary, the state’s interests are fully satisfied even though there may be “some kinds of cases” in which jurors would be unwilling to impose the death penalty. Only through adherence to this rule can a trial court be reasonably certain it has not impaneled a jury “uncommonly willing to condemn a man to die.” Id. at 521, 88 S.Ct. at 1776. The Court of Appeals for the Fourth Circuit has not yet decided this issue. The Fifth and Eleventh Circuit courts have made several relevant decisions. Burns v. Estelle, 592 F.2d 1297 (5th Cir.1979), aff'd. en banc, 626 F.2d 396 (5th Cir.1980); Granviel v. Estelle, 655 F.2d 673 (5th Cir.1981), cert. denied, 455 U.S. 1003, 102 S.Ct. 1636, 71 L.Ed.2d 870 (1982); Hance v. Zant, 696 F.2d 940 (11th Cir.1983), ce rt. denied, — U.S. —, 103 S.Ct. 3544, 77 L.Ed.2d 1393; McCorquodale v. Balkcom, 705 F.2d 1553 (11th Cir.1983); Witt v. Wainwright, 714 F.2d 1069 (11th Cir.1983). In Granviel v. Estelle, supra, cert. denied, 455 U.S. 103, 102 S.Ct. 1636, 71 L.Ed.2d 870 (1982), the Fifth Circuit reversed a habeas corpus petitioner’s death sentence because the trial court had improperly excused for cause a juror who had responded, “No, I don’t think I could,” when he was asked whether he could ever vote to inflict the death penalty. 655 F.2d at 677. When further asked whether he “felt like” he would be able to take another person’s life, this juror had responded, “No, I could not.” Id. The court found that these responses “fell far short of” the unequivocal statements required by Witherspoon before a challenge for cause may be allowed. Ms. Melton’s position with respect to the death penalty was far less certain than the position of the juror whose responses were quoted in Granviel. Her statements more closely resemble the responses to voir dire questioning given by a juror whose attitude toward the death penalty was discussed in Hance v. Zant, supra. In Hance v. Zant, the court found that the trial court had improperly excused a juror who at least twice stated, “I feel there are times when the death penalty is warranted. I do not believe that I with my conscience could vote to impose the death penalty.” 696 F.2d at 955. When the prosecutor had prodded this juror and asked if her position would be the same “[n]o matter what the facts or circumstances of the case might be?” she had then replied, “In some cases I might.” Id. The court is of the opinion and finds that the exclusion of juror Melton deprived petitioner of a fair sentencing trial under the Sixth and Fourteenth Amendments to the United States Constitution. It is useless to speculate whether, if Ms. Melton had been questioned further, she might have expressed a view that would have disqualified her under Witherspoon; this court’s decision must be based on the evidence before it; and that evidence does not reveal that Ms. Melton’s views justified excluding her from the jury. State v. Williams, 305 N.C. 656, 691, 292 S.E.2d 243, 264 (1982) (Exum, J., dissenting). Respondents suggest that any error resulting from the exclusion of Ms. Melton from the jury was rendered harmless because if the challenge for cause had been refused, the prosecutor might still have exercised his remaining peremptory challenges to exclude her from the jury. This argument was effectively rejected by the Supreme Court in Davis v. Georgia, supra. Respondents also contend, citing Darden v. Wainwright, 699 F.2d 1031, 1037-38 (11th Cir.1983), that the trial judge's decision to exclude Ms. Melton should be affirmed because the trial judge was in the best position to observe the juror and to determine the true meaning of the words she used in her answers to the court’s questions. 'To a trial judge whose resolutions of conflicts among witnesses do not always *1191 seem to receive appropriate deference from appellate courts, this seductive proposition is well-nigh irresistible. The Darden opinion did contain a comment on that subject. However, there are obvious limits. The burden of proof that the juror has the wrong kind of ideas when measured against Witherspoon is upon the prosecution. All of the evidence on the contents of the juror’s mind come from the words of the juror Melton herself. A trial judge may with propriety decide that he does not believe this witness; but his disbelief of what she says may not be converted into an affirmative finding that she is in fact “irrevocably committed ... against the death penalty.” To disbelieve her testimony simply leaves a total absence of proof on this vital question, and the discharge of the juror for cause was wrongful. Moreover, in Darden the statement was a dictum not necessary to the decision, because the Darden juror in question had made the unconditional statement that he could not under any circumstances vote to impose the death penalty. Darden also said that “[nevertheless, close scrutiny of the voir dire examination of prospective jurors is necessary to insure strict adherence to the mandates of Witherspoon.” Id. at 1038. The Supreme Court’s peroration in Witherspoon, 391 U.S. at 523, 88 S.Ct. at 1778, bears repeating: Whatever else might be said of capital punishment, it is at least clear that its imposition by a hanging jury cannot be squared with the Constitution. The State ... has stacked the deck against the petitioner. To execute this death sentence would deprive him of his life without due process of law. As previously decided in the opinion and order entered January 12, 1984, petitioner’s motion for relief from his death sentence should be ALLOWED.