July 11, 1968. Certiorari Granted December 16, 1968. See 89 S.Ct. 488.
Anthony G. Amsterdam, Philadelphia, Pa., for appellant, Jack Greenberg, Norman C. Amaker and Michael Meltsner, New York City, and George Howard, Jr., Pine Bluff, Ark., on the brief.
R.D. Smith, III, Asst. Atty. Gen., State of Arkansas, Little Rock, Ark., for appellee, Joe Purcell, Atty. Gen., State of Arkansas, on the brief.
Before VOGEL, MATTHES and BLACKMUN, Circuit Judges.
William L. Maxwell, a Negro and an Arkansas prisoner under sentence of execution on his 1962 conviction for the state crime of rape, as defined by Ark. Stat.Ann. § 41-3401 (Repl. 1964), petitions a second time for a federal writ of habeas corpus and, with its denial, a second time appeals. The points now urged to us are (1) that a prima facie case of racially discriminatory imposition of the death penalty for rape in Arkansas has now been established and remains unrebutted by the State; (2) that Arkansas' single verdict procedure is without appropriate standards, allows the jury to exercise its discretion irrationally, and is impermissible; and (3) that the decisions of the United States Supreme Court since Maxwell's first federal appeal demonstrate the error of our prior holding, adverse to the petitioner, on the jury selection issue.
Section 41-3401, as it existed at the time of the offense and Maxwell's conviction, was repealed by Acts 1967, No. 362, § 3, and was replaced by § 1 of the same Act. The new statute separates the crime of rape into third, second, and first degrees. Rape in the first degree is, among other things, sexual intercourse with a female by "forcible compulsion". Rape in the second and third degrees has age conditions. Section 41-3403 which, through § 43-2153, prescribed, upon conviction of rape, a punishment of death or, if the jury chose, life imprisonment, also was, by §§ 3 and 2 of the same 1967 Act, repealed and replaced. The replacement section defines rape as a felony and provides for sentences within prescribed ranges, as to years, for third and second degree rape and, for first degree rape, of "death or thirty (30) years to life imprisonment". If the offense for which Maxwell was convicted had been committed since the effective date of the 1967 Act, it would qualify as rape in the first degree.
The petition itself also suggests, on information and belief, that Maxwell "is presently insane". At the hearing in the district court, defense counsel acknowledged that "[T]he report [from Dr. William G. Reese, Professor of Psychiatry and head of the Department of Psychiatry at the University of Arkansas Medical Center] states, in effect, that the petitioner is without psychosis. * * *" The defense's appellate brief recites that the insanity suggestion, "though dealt with by the district court in its opinion, was not pressed below and is not relevant here".
We again review the background:
1. The offense for which Maxwell was charged was committed in the early morning of November 3, 1961, in the City of Hot Springs, Garland County, Arkansas. Maxwell, who was 21 at the time, was arrested within two hours after the offense was committed. His convicting jury did not "render a verdict of life imprisonment in the State penitentiary at hard labor", as it had the option to do [since 1915 (Acts 1915, No. 187, § 1, p. 774)] under §§ 43-2153 and 41-3403 and for which it had been given an alternate verdict form. Accordingly, and in line with the interpretation consistently given the statutes by the Supreme Court of Arkansas, Kelley v. State, 133 Ark. 261, 202 S.W. 49, 54 (1918); Stewart v. State, 233 Ark. 458, 345 S.W.2d 472, 475 (1961), cert. denied 368 U.S. 935, 82 S.Ct. 371, 7 L.Ed.2d 197, the trial court imposed the death sentence.
2. Maxwell appealed. On this state appeal he challenged the sufficiency of the evidence; his prosecution by information; the constitutionality in application of the penalty statute, § 41-3403; the denial of his motion that his case be removed to federal court; the overruling of certain objections relating to voir dire; the overruling of objections as to latitude in his cross-examination of witnesses; the admission of evidence; the instructions; and aspects of the prosecution's argument to the jury. All these points were decided adversely to the defense and the judgment of conviction was affirmed by a unanimous Supreme Court of Arkansas (one justice not participating). Maxwell v. State, 236 Ark. 694, 370 S.W.2d 113 (1963). No petition for certiorari was filed with the Supreme Court of the United States.
This fact, of course, no longer constitutes a failure to exhaust available state remedies. Fay v. Noia, 372 U.S. 391, 435-438, 83 S.Ct. 822, 9 L.Ed.2d 837 (1963); Curtis v. Boeger, 331 F.2d 675 (8 Cir. 1964).
3. Maxwell, with new counsel, then filed a petition for a writ of habeas corpus in the United States District Court for the Eastern District of Arkansas. In that petition, as twice amended, he alleged the illegality of his arrest; impropriety of the search of his person; impropriety of the search of his home; mistreatment by the police; coerced confessions; an adverse and hostile trial atmosphere; racial discrimination in the selection of his jury; unconstitutional application of § 41-3403; and unconstitutionality of the death penalty for rape. Judge Young decided each point so raised adversely to the defense and denied the petition. His detailed opinion is reported as Maxwell v. Stephens, 229 F. Supp. 205 (E.D.Ark. 1964).
4. Judge Young, however, granted the certificate of probable cause required by 28 U.S.C. § 2253 and stayed execution. With additional counsel from New York, an appeal was taken to this court. We noted that, except for an early period prior to the state trial, when court-appointed attorneys were in the case, Maxwell had been represented through all the state and federal proceedings by competent, although different, non-court-appointed counsel. The points primarily argued to us were three: (1) that Maxwell was denied due process and equal protection because he was sentenced under statutes which are discriminatorily enforced against Negroes; (2) that he was denied due process and equal protection because the Garland County jury lists revealed race and were compiled from racially designated poll tax books; and (3) that the taking of Maxwell's coat and references to it in testimony at the trial violated his Fourth, Fifth and Fourteenth Amendment rights. In what we thought was an opinion meticulously concerned with the several issues raised by Maxwell, we reached the conclusion that Judge Young's decision was correct. Maxwell v. Stephens, 348 F.2d 325 (8 Cir. 1965). The late Judge Ridge dissented on the search and seizure issue concerning the coat but agreed with all the other conclusions arrived at by the majority.
5. With still another name added to the list of Maxwell's counsel, a petition for certiorari was filed. This was denied, with Mr. Justice Douglas being of the opinion that certiorari should be granted. 382 U.S. 944, 86 S.Ct. 387, 15 L.Ed.2d 353 (1965).
6. Execution was rescheduled but was stayed administratively until September 2, 1966.
7. Maxwell's second and present petition for a federal writ of habeas corpus was filed July 21, 1966, in the United States District Court for the Eastern District of Arkansas and came before Chief Judge Henley. The court denied the petition. Maxwell v. Bishop, 257 F. Supp. 710 (E.D.Ark. 1966).
8. The district court and a judge of this court successively declined to grant a stay of the execution or to issue a certificate of probable cause. But the Supreme Court granted leave to file a petition for a writ of certiorari, issued the writ, reversed the denial of the application for the certificate, and remanded the cause with directions to issue it. Maxwell v. Bishop, 385 U.S. 650, 87 S.Ct. 768, 17 L.Ed.2d 671 (1967). Accordingly, the certificate was issued and the execution stayed, and the case is here on appeal.
As we observed before, pp. 327-328 of 348 F.2d, Maxwell's guilt or innocence is not in issue before us. So far as the details of the crime itself are concerned, we recite only what we said there:
"The circumstances and details of the crime are, as usual, sordid. They are set forth in the Arkansas opinion, pp. 114-116 of 370 S.W.2d, and need not be repeated here. It suffices only to say that the victim was a white woman, 35 years old, who lived with her helpless ninety-year-old father; that their home was entered in the early morning by the assailant's cutting or breaking a window screen; that in the ensuing struggle the victim bit her assailant and caused bleeding; and that she was assaulted and bruised, her father injured, and the lives of both threatened. Confessions taken from Maxwell were not employed at the trial. The defense presented no evidence. The jury was out several hours. No question is raised as to the sufficiency of the evidence."
The Statistical Argument
Maxwell's present argument that Arkansas discriminates against Negroes in the application of the death penalty for rape, and thus violates the Equal Protection Clause and also 42 U.S.C. § 1981, rests on what is described as newly discovered evidence which became available since this court's disposition of Maxwell's first habeas appeal. It is said, in the words of the petition:
"This evidence consists of the results of a survey of rape convictions during the period 1945-1965 in a representative sample of nineteen counties comprising more than 47% of the population of the State of Arkansas. The survey was conducted in the summer of 1965, as part of a study of the application of the death penalty for rape in eleven southern states. This comprehensive study required the work of 28 law students throughout the summer, the expenditure of more than $35,000 and numerous hours of consultative time by expert criminal lawyers, criminologists and statisticians. Petitioner, who is an indigent, could not have himself at any time during the prior proceedings in his cause conducted such a study."The record and the evidence. At a pretrial conference the district court was advised that the evidence to be presented on behalf of Maxwell was that which resulted from the survey and study so described, and that it would be presented through the testimony of Marvin Wolfgang, a criminologist-statistician on the faculty of the University of Pennsylvania. It was agreed in pretrial that the survey's individual case schedules would be made available for inspection by counsel for the State. The court ruled that, subject to objections as to relevancy and materiality, Professor Wolfgang would be permitted to testify as an expert and to introduce his report as a summary exhibit. The State filed no objection to any of the individual case schedules. At the ensuing hearing the State agreed that the basic facts of the schedules, "that is, the age of the victim, the race, and so on, of the individual defendants, or the alleged victims", would stand admitted. Dr. Wolfgang then was called as a witness.
The study is said to be the same one with which the Fourth Circuit was concerned in Moorer v. State of South Carolina, 368 F.2d 458 (4 Cir. 1966).
Dr. Wolfgang possesses the degree of Doctor of Philosophy in Sociology, with Criminology as a sub-field. He is Professor and graduate chairman of the Department of Sociology at the University of Pennsylvania. He is co-director of the Department's Criminological Center. He obviously is a man of scholastic achievement and of experience in his field. The State concedes that his "qualifications as a criminologist have never been questioned by the respondent".
The witness testified that in 1965 he was asked to participate in a study on capital punishment for rape and that this was suggested "as an appropriate topic for research analysis because of the underlying assumption that there is differential imposition of the death penalty in rape cases in certain states". This was to be "an effort to collect the appropriate kind of data necessary to provide some kind of empirical study, either in support of, or in rejection of, the underlying assumption. The ultimate objective of the study was to give the empirical data the appropriate kind of statistical analysis that would satisfy scientific requirements". He was informed that the result might be used in litigation on behalf of defendants who had been convicted and sentenced to death for rape.
The study was carried out in twelve states. One of these was Arkansas. For the State of Arkansas the period investigated was that from January 1, 1945, to the summer of 1965. Not every Arkansas rape case during that period was included but every case which occurred in the sample counties was included.
The study produced 55 rape convictions in the 19 selected Arkansas counties. The witness prepared a preliminary analysis of the data contained in the schedules the student investigators prepared for these cases. Among the variables considered were race of the defendant, race of the victim, and sentence. The approach was to develop a "null hypothesis" that there is no difference in the distribution of the sentence of death or life imprisonment imposed on negro and white defendants; the calculation of a theoretical frequency which represents the number of defendants expected to be sentenced to death if the null hypothesis is valid; the comparison of this theoretical frequency with the actual frequency in the collected data for each defendant-victim racial combination; and the determination whether the discrepancy between the expected and the observed frequencies is great enough so that, under accepted statistical standards, that discrepancy can be said to be a product of the real phenomena tested rather than of the operation of chance within the testing process. "If that difference reaches a sufficiently high proportion, * * * then the assertion can be made * * * that the difference is significant and could not have occurred by chance".
The data gathering was not limited to the facts of race and sentence. It included other items such as the defendant's age, family status, occupation, and prior criminal record; the victim's age, family status, occupation, husband's occupation, and reputation for chastity; aspects of the prior relationship, if any, between the defendant and the victim; the circumstances of the offense such as the number of offenders and victims, the place, the degree of threat or violence, and the injury inflicted; other offenses contemporaneously committed; the presence of third persons and threats or injury to them; the nature of the intercourse; the involvement of alcohol or drugs; and the circumstances of the trial and post-trial proceedings. Data of this general kind were included because they "were felt to be relevant to the imposition of the type of sentence" and "they, rather than race alone, may play a more important role in the disproportionate sentencing to death of Negro defendants convicted of raping white victims". The witness conceded that some data potentially pertinent were not collected as, for example, that "with respect to the strength of the prosecution's case" for "we had no information of that sort that we could objectively collect".
The study disclosed that, in the 55 Arkansas cases, 34 defendants were Negroes and 21 were whites; that 33 had a previous record; that 26 were known to have been in prison; that 39 victims were whites, 15 were Negroes, and the race of one was unknown; that of the victims, two had prior criminal records and 31 did not; that 16 victims had a good reputation for chastity and 6 bad; that 14 received the death sentence and 41 life; that counsel was appointed in 35 cases and retained in 9; that in 23 cases the offender was known by the victim and in 26 he was unknown; that the offender and the victim had had prior sexual relations in 4 cases; and that in 12 cases a plea of guilty was entered.
From this Arkansas data and his study Dr. Wolfgang concluded:
1. Factors not significantly associated with either the defendant's race or with the sentence were the place of offense, nature of entry, plea, type of counsel, duration of trial, seriousness of injury to the victim, type of prior relationship between defendant and victim, the defendant's previous record, and the victim's having dependent children.
2. Factors not significantly associated with the defendant's race were the commission of a contemporaneous offense and previous imprisonment.
3. Factors not significantly associated with the sentence were the defendant's marital status, his age, the victim's age, the display of a weapon, and the defendant's having dependent children.
4. Factors which were significantly associated with sentence were commission of a contemporaneous offense and previous imprisonment.
5. Factors which were significantly associated with the defendant's race were marital status (a greater percentage of Negroes than whites were not married), display of a weapon (more Negroes than whites used one), the defendant's age (more Negroes were young), the victim's age (the defendants were white more frequently in the older-defendant-younger-victim combination), and the existence of children dependent on the defendant (more whites had them).
6. An analysis of significant association with defendant's race or the sentence was either unnecessary or not possible with respect to the victim's prior record, her reputation for chastity, the force employed, her pregnancy by the offense, prior sexual relations between the defendant and the victim, and the defendant's assertion of the defense of insanity.
The study covered other items, too, but the analysis was confined to "those particular items in which there was sufficient information to display statistically and to treat statistically".
The witness then further concluded (1) that the critical variables were race of the offender, race of his victim, and sentence; (2) compared to other rape defendants, Negroes convicted of raping white victims were disproportionately sentenced to death; and (3) "no variable of which analysis was possible could account for the observed disproportionate frequency".
On cross-examination Dr. Wolfgang stated that the Arkansas report was "preliminary" in the sense that the other states for which data was being collected would be included in the final report; that the 55 Arkansas cases did include some new trials; that, however, he did not think this disturbed his analysis; that he was only "very little" familiar with Arkansas geography; that he was "gratified the report is as good as it is"; that he used such variables as he could get information on; that if he had had more information, he could have used other variables and his report, accordingly, would have been more extensive; that reputation for chastity was presumed to be of some relevance but "unfortunately, there was not enough information upon which to make the analysis complete"; that as a criminologist he was willing to say "that an analysis of a social system — in this case, the judicial system — operated relative to the decision making regarding rape, is such that it casts considerable doubt upon the quality of justice in those particular cases throughout the system"; and that he has never sat on a jury but has been around them.
It was then agreed that among the 55 schedules there were 4 cases which concerned defendants convicted twice for the same rape as a result of a second trial. A fifth case was not one of repeated convictions but, instead concerned two victims.
Received in evidence was an exhibit prepared by the United States Bureau of Prisons. This showed that, for the years 1930-62, 446 persons were executed under civil authority in the United States for rape and that, of these, only 45 were white but 399 were Negro and 2 were of other races, and that, for the same period, 3,298 were executed for murder and, of these, 1,640 were white, 1,619 were Negro, and 39 were of other races.
The petitioner's second and only other witness was John Monroe, a Philadelphia statistician with 17 years' experience in sampling and surveys. He testified that he was asked by Dr. Wolfgang to design a representative sampling of counties in 12 southern states so that inferences could be drawn for each state and for the region as a whole. Arkansas was included. The 1960 negro population was used as the measure. The sample size was determined largely on speculation as to what area the law students could cover in the time available to them. Nineteen Arkansas counties were selected. These contained more than 47% of the state's population. Mr. Monroe stated that his sample of Arkansas counties was a very reliable one under the restriction, that is, the number of counties that could be investigated during the time allotted. He went on to say that, so far as the sample is concerned, the inferences drawn from it were valid for Arkansas. Garland County was not one of those selected. This was because "the random number just didn't happen to fall in the population for Garland County". Had that county possessed more than 19,000 negro population it would have been selected. The witness conceded that he was sampling the negro population and not the entire population. This was because of the purpose of the study and, in order to compare negro and white cases, the sample has to provide enough of the former to make the findings valid.
The witness stated that he took the total negro population of Arkansas and divided it by the number of draws allocated to the State. This number turned out to be approximately 19,400. A random number between it and zero was selected. As adjacent county negro population added up to the random number plus 19,400, the county was selected. In this way, counties having more than 19,400 Negroes, or approximately that number, were taken consecutively.
On cross-examination it was pointed out that this method did not result in having the selected counties geographically dispersed throughout the State. Instead, the selected counties for the most part were in the east and south. This is where Arkansas' negro population is primarily located.
The court inquired what there was about the survey that gave it validity in Garland County. The answer was that the inferences drawn from the sample were valid for the State of Arkansas. Counsel then observed that it was not Garland County that convicted Maxwell and sentenced him to death but the State, and that the purpose of Mr. Monroe's testimony was to show that the counties selected were representative of the State.
On further cross-examination, the witness acknowledged that the northwestern part of the State was essentially the white population area, that the southeastern portion was the negro population area, and that Garland County was in the white area.
The State offered no evidence.
Such is the record. It is apparent from a reading of the record that the trial court was generous in its rulings on admissibility. Everything which the petitioner presented came in.
The trial court's conclusions on the statistical argument. Chief Judge Henley, in his opinion, pp. 717-721 of 257 F. Supp., described the statistical argument as the "contention which has been urged most seriously here". He referred to Professor Wolfgang as "a well qualified sociologist and criminologist" and to Mr. Monroe as "a qualified statistician". He observed that the investigation showed three categories, namely, rapes of white women by negro men, rapes of negro women by negro men, and rapes of white women by white men, and that there was no conviction of a white man for rape of a negro woman. He noted that Dr. Wolfgang "concluded, and the Court agrees, that the differential could not be due to the operation of the laws of chance". He characterized Dr. Wolfgang's conclusions "to be that a Negro man who is convicted of raping a white woman has about a 50 percent chance of receiving a death sentence, regardless of the facts and circumstances surrounding the crime, whereas a man who is convicted of criminally assaulting a woman of his own race stands about only a 14 percent chance of receiving the death sentence". He acknowledged that "the statistical evidence produced in this case is more extensive and sophisticated than has been produced heretofore". Yet he was not convinced that "it is sufficiently broad, accurate, or precise as to establish satisfactorily that Arkansas juries in general practice unconstitutional racial discrimination in rape cases involving Negro men and white women * * *. The study does not indicate that Negro men convicted of raping white women invariably or even in a majority of cases receive the death penalty. * * * Only 7 Negro men were sentenced to die for raping white women. The case studies, and the number of death sentences imposed are simply too few in number to afford convincing proof of the proposition urged by petitioner". It is not correct, as Dr. Wolfgang said in his report, that the selected Arkansas counties are "representative of the State in urban-rural and white-Negro population ratios". The statistics "really reveal very little about the details of the cases" where the death sentence was imposed upon Negroes for raping white women as compared with other racial situations. "On the meager material before it the Court is simply not prepared to convict Arkansas juries of unconstitutional racial discrimination in rape cases". The court went on to observe that "In cases not involving inter-racial situations the issue of consent may be and frequently is very real", but that that issue "is much less likely to be present in cases in which white women have been attacked by Negro men".
The petitioner's argument. The petitioner-appellant characterizes the present proceeding as one presenting "for the first time in any appellate court * * a record which is the end product of a detailed and exhaustive examination of the practical consequences of the procedures used in making the decision whether a man should live or die". It is stated that the question in this record is one of proof, namely, whether the petitioner has made a sufficient showing of racially discriminatory capital sentencing under Arkansas rape statutes. It is pointed out that on three previous occasions, namely, in the Supreme Court of Arkansas, before Judge Young, and in this court, the answer has been in the negative, "notwithstanding that on each successive occasion the evidence tended in the direction of more depth and completeness". This demonstrates, it is said, "how difficult it is for Negro litigants generally and those without means particularly, to make courts see `the reality of the world, indeed * * * the segregated world' [citing Brooks v. Beto, 366 F.2d 1, 12 (5 Cir. 1966), cert. denied 386 U.S. 975, 87 S.Ct. 1169, 18 L.Ed.2d 135] * * * in which they live * * * the law [needs to] `see what all others see'".
The heart of the petitioner's statistical argument is then forthcoming. It is that on this record the petitioner has made a prima facie case of racial discrimination in sentencing and that he is entitled to prevail when, as here, the State presents no evidence or presents evidence of insufficient moment to overcome the prima facie case so established.
It is said that the Supreme Court of necessity has developed the doctrine that a prima facie showing of unequal racial treatment, calling state procedures in question, compels the inference that the State is drawing the racial line unless it offers justification in non-racial factors for the disproportion. Jury selection and voting deprival cases such as Hernandez v. State of Texas, 347 U.S. 475, 74 S.Ct. 667, 98 L.Ed. 866 (1954), and Gomillion v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960), are cited generally. Also cited, however, are other cases which, it is said, show that the prima facie approach has not been restricted to jury selection and voting cases. Among these are Chambers v. Hendersonville City Bd. of Educ., 364 F.2d 189 (4 Cir. 1966) (reemployment of teachers), and Cypress v. Newport News General Nonsectarian Hospital Ass'n, 375 F.2d 648 (4 Cir. 1967) (hospital staff membership). It is argued that there is no reason why the prima facie approach should not be applied to the present case. If there are factors which offset Dr. Wolfgang's expert conclusion, they were in the power of the State to unearth and prove. "Every justification for shifting the burden of persuasion to the State * * applies with evident force here".
Finally, it is stated that the district court committed manifest error in its evaluation of the evidence. Counsel would dismiss the trial court's observation that the Arkansas figures contained instances of defendants tried more than once on the ground that each Arkansas jury death sentence is an indication of the application of the Arkansas statute. They would dismiss the trial court's concern with the small number of cases on the ground that the whole purpose of the statistical analysis was to satisfy Dr. Wolfgang as an expert and that he was so satisfied and his conclusions are sustainable under accepted scientific standards. They would dismiss the court's concern with the geographical location of the selected Arkansas counties as unsupportable and improper. And they would dismiss the court's concern with the "issue of consent" as being "plainly the sheerest speculation". The observations are made that "any experienced criminal lawyer in the South * * * well knows that the failure to present the defense of consent in interracial rape cases is itself a product of the discriminatory pattern of Southern justice which petitioner here attacks", and that "Southern jury attitudes * * have long impressed upon defense counsel the extreme unwisdom of advancing the consent defense on behalf of a Negro defendant where the complainant is white".
However, the transcript of the state trial reveals that the defense in its opening statement made no less than five references to any act on the part of the prosecuting witness as being "free" and "voluntary".
The petitioner's argument is an interesting one and we are not disposed to say that it could not have some validity and weight in certain situations. Like the trial court, however, although perhaps not for each and all of the reasons it advanced, we feel that the argument does not have validity and pertinent application to Maxwell's case.
It is perhaps well to emphasize initially what the study and Dr. Wolfgang's testimony do not do or purport to do:
1. They do not relate specifically to Garland County where this particular offense was committed and where Maxwell was tried and convicted. They are concerned with 19 other Arkansas counties and with counties in 11 other states.
On the first habeas appeal we were unpersuaded with the statistics then presented or with the argument in support thereof. Some figures were submitted as to the entire State of Arkansas and as to three counties, namely, Garland (Hot Springs), Pulaski (Little Rock), and Jefferson (Pine Bluff). We said, pp. 330-331 of 348 F.2d:
"The statistical argument is not at all persuasive. * * * As to Garland County, for the decade beginning January 1, 1954, Maxwell's evidence was to the effect that seven whites were charged with rape (two of white women and the race of the other victims not disclosed), with four whites not prosecuted and three sentenced on reduced charges; that three Negroes were charged with rape, with one of a Negro woman not prosecuted and another of a Negro receiving a reduced sentence, and the third, the present defendant, receiving the death penalty. * * *
"These facts do not seem to us to establish a pattern or something specific or useful here, or to provide anything other than a weak basis for suspicion on the part of the defense. The figures certainly do not prove current discrimination in Arkansas, for in the last fourteen years the men executed for rape have been two whites and two Negroes. * * *
"Turning to the three county statistics, we find no death sentence at all in Garland County in the 1954-1963 decade until Maxwell's case. We also find that of the two other Negroes charged, one was not prosecuted and the other was sentenced on a reduced charge."
2. They admittedly do not take every variable into account.
3. They do not show that the petit jury which tried and convicted Maxwell acted in his case with racial discrimination.
4. They do not deny that generally the burden of demonstrating discrimination in penalty imposition is on the one who asserts it. Maxwell v. Stephens, supra, p. 330 of 348 F.2d, and cases cited; Mitchell v. Stephens, 353 F.2d 129, 133 (8 Cir. 1965), cert. denied 384 U.S. 1019, 86 S.Ct. 1966, 16 L.Ed.2d 1042.
What we are concerned with here is Maxwell's case and only Maxwell's case. And it is indisputable, from the record before us, that, despite the five references to "free" and "voluntary" in the defense's opening statement at the trial, there is no question, and no hint of one, as to the victim's lack of consent. The facts of the attack, which are not in dispute, which were reviewed by the Arkansas court, pp. 114-116 of 370 S.W.2d, and which we have again quoted above, include forcible entry of the home, physical assault on both the victim and her aged father, resistance by both, and the infliction of injuries on all three. The petitioner's attack, therefore, to the extent it would concern itself with the issue of consent in rape cases, both interracial and non-interracial, and with allegedly purposeful avoidance of the consent issue by southern defense counsel, has no place here and, in its attempted application, is a straw argument based on speculation.
Further, while it is true that it is in a sense the state which prosecutes, nevertheless the county has a character and a posture, too. Inasmuch as Garland County, as it was conceded, is in the predominantly white area of Arkansas, one might expect from the petitioner an argument that alleged southern injustice in interracial rape cases would be more apparent in such a county than in those areas where the negro population is predominant. Yet the Garland County statistics, which we noted at pp. 330-331 of 348 F.2d, and which we have again recited by quotation in footnote 6, supra, afford no local support to the petitioner's statistical argument. The evidence produced at the prior hearing and at this one discloses only Maxwell as a recipient of the death penalty in Garland County for rape.
We are not yet ready to condemn and upset the result reached in every case of a negro rape defendant in the State of Arkansas on the basis of broad theories of social and statistical injustice. This is particularly so on a record so specific as this one. And we are not yet ready to nullify this petitioner's Garland County trial on the basis of results generally, but elsewhere, throughout the South.
We therefore reject the statistical argument in its attempted application to Maxwell's case. Whatever value that argument may have as an instrument of social concern, whatever suspicion it may arouse with respect to southern interracial rape trials as a group over a long period of time, and whatever it may disclose with respect to other localities, we feel that the statistical argument does nothing to destroy the integrity of Maxwell's trial. Although the investigation and study made by Professor Wolfgang in the summer of 1965 is interesting and provocative, we do not, on the basis of that study, upset Maxwell's conviction and, as a necessary consequence, cast serious doubt on every other rape conviction in the state courts of Arkansas.
At oral argument we asked Professor Amsterdam, counsel for Maxwell, whether (apart from any Eighth Amendment argument) his statistical approach would not mean that it would be constitutionally impossible for a negro defendant in Arkansas ever to receive the death penalty upon conviction of the crime of rape of a white woman. The answer given us was in the affirmative, that is, that it would be constitutionally impossible. At the same time it was conceded, in contrast, that it would be possible for a white man to receive the death penalty upon his conviction for rape. When counsel was asked whether this would not be discriminatory, the reply was that once the negro situation was remedied the white situation "would take care of itself".
The legal logic and the rightness of this totally escape us. If that end result would actually be forthcoming (and we wonder whether this may accurately be forecast), that fact does not, in our view, make it legally right or satisfy constitutional dictates. We still read the first section of the Fourteenth Amendment as proscribing the denial "to any person" within a state's jurisdiction of "the equal protection of the laws". We feel that this means what it seems to require, namely, protection which is equal, and that equal protection is denied if, factually, a member of one race (whether black or white) is subjected, because of his race, to greater or different punishment than a member of another race. The argument therefore turns back upon and defeats the very side which here proposes it. We are not inclined to accept as constitutional doctrine an abstraction which provides equality only through assumed and hoped-for day-to-day practicalities. It is the law, not probabilities or possibilities, which must afford equal protection.
We can understand and appreciate the disappointment and seeming frustration which Maxwell's counsel must feel in again failing to prevail on a still more sophisticated statistical approach. They will ask themselves just how far they are required to go in order to prevail.
We are not certain that, for Maxwell, statistics will ever be his redemption. The facts as to rape charges in Garland County are known and have been recited. Standing by themselves, they disclose nothing from which conclusions of unconstitutionality in application may appropriately be drawn. This situation — the aridity of the Garland County facts on which to claim unequal protection of the laws — forces Maxwell and his counsel to present his case on a state-wide and long-term historical approach, and even on a South-wide one, and to claim that conclusions which, at best, are necessarily general have valid application to Maxwell.
We do not say that there is no ground for suspicion that the death penalty for rape may have been discriminatorily applied over the decades in that large area of states whose statutes provide for it. There are recognizable indicators of this. But, as we have noted before, with respect to the issue of jury selection, improper state practice of the past does not automatically invalidate a procedure of the present. Brown v. Allen, 344 U.S. 443, 479, 73 S.Ct. 397, 97 L.Ed. 469 (1953); Bailey v. Henslee, 287 F.2d 936, 943 (8 Cir. 1961), cert. denied 368 U.S. 877, 82 S.Ct. 121, 7 L.Ed. 2d 78. We do say that nothing has been presented in Maxwell's case which convinces us, or causes us seriously to wonder, that, with the imposition of the death penalty, he was the victim of discrimination based on race.
The Single Verdict Argument
This appears to divide itself into two parts. The first is that the Arkansas statutes which permit the state jury to choose between capital punishment and life imprisonment embrace no standards by which the jury is to exercise that choice and that, as a consequence, such jury procedure is unconstitutional. The second is that the state practice of submitting simultaneously to the jury the two issues of guilt and punishment in a capital case "compounds the vice of lawless jury discretion * * * by making it virtually impossible for the jurors to exercise their discretion in any rational fashion".
We are not persuaded by either argument.
A. We point out initially, as to the first, that the basic punishment imposed by § 41-3403 is death. This has been so at least since 1842. Act of December 14, 1842, § 1, p. 19. Capital punishment for rape has not yet been held to be violative of the Eighth and Fourteenth Amendments, although, as we noted on Maxwell's first habeas appeal, p. 332 of 348 F.2d, three Justices dissented from the denial of certiorari in Rudolph v. Alabama, 375 U.S. 889, 84 S.Ct. 155, 11 L.Ed.2d 119 (1963), and would have had the Court "consider whether the Eighth and Fourteenth Amendments to the United States Constitution permit the imposition of the death penalty on a convicted rapist who has neither taken nor endangered human life". That, however, is a fact situation different from the instant case. We say again that if the death penalty for rape is to be nullified on constitutional grounds, that step in the first instance is for the Supreme Court and not for this inferior federal court. See Ralph v. Pepersack, 335 F.2d 128, 141 (4 Cir. 1964), cert. denied 380 U.S. 925, 85 S.Ct. 907, 13 L.Ed.2d 811.
The life imprisonment alternative came into the Arkansas statutes by Acts 1915, No. 187, § 1, p. 774. In operation it gave the Arkansas jury the power to alleviate punishment and extend clemency in cases (rape being just one of that category) where death had been the only punishment theretofore prescribed. The 1915 statute thus was in the direction of leniency and not away from it. Kelley v. State, supra, p. 54 of 202 S.W.; Bell v. State, 120 Ark. 530, 180 S.W. 186, 190 (1915).
Further, the imposition of punishment by the jury, within statutorily prescribed limits, rather than by the court, is the usual routine in Arkansas criminal procedure. § 43-2145.
We reject the petitioner's argument here for a number of reasons: (a) it has no basis in legislative intent; (b) it would destroy by indirection a punishment which the legislature saw fit to impose, which has been a component of the State's criminal law for well over a century, and which in all that time has not been subjected to successful constitutional attack; (c) the defense in Maxwell's rape trial requested no instructional standards; (d) the jury choice provides room for attention to the very variables which Professor Wolfgang's study concededly did not reach; and (e) we are not convinced that the absence of expressly stated standards in the statute requires that it be forthwith condemned constitutionally. Jurors are not automatons. They are human beings. The jury system appropriately assumes that jurors in their factual determinations bring into play their common sense and the experiences of life. Their choice between capital punishment and life imprisonment, as the entire Arkansas method of punishment imposition by the jury, is not startlingly or shockingly different from the situation where choice of punishment within statutorily prescribed limits is for the judge. See Witherspoon v. State of Illinois, 391 U.S. 510, 88 S.Ct. 1770, 20 L.Ed.2d 776 (1968).
The authorities proffered by the defense are the well-known ones, such as Herndon v. Lowry, 301 U.S. 242, 57 S.Ct. 732, 81 L.Ed. 1066 (1937), and Winters v. People of State of New York, 333 U.S. 507, 68 S.Ct. 665, 92 L.Ed. 840 (1948), which deal with vagueness in criminal statutes. But there is nothing vague, and nothing is claimed to be vague, in the substantive provisions of the Arkansas rape statutes. The petitioner's attack is directed to the punishment and not to the crime.
But it is said that Giaccio v. State of Pennsylvania, 382 U.S. 399, 86 S.Ct. 518, 15 L.Ed.2d 447 (1966) "supports, if it does not compel, the conclusion that unfettered jury discretion in capital cases is unconstitutional". Giaccio concerned an ancient Pennsylvania statute which permitted a jury to assess costs against a defendant acquitted of a criminal charge and which called for imprisonment if such costs so assessed were not secured. The Supreme Court held that the statute was invalid under the due process clause "because of vagueness and the absence of any standards sufficient to enable defendants to protect themselves against arbitrary and discriminatory impositions of costs". P. 402, 86 S.Ct. p. 520. We think that the heart of the holding as to vagueness is that the offending statute "leaves the public uncertain as to the conduct it prohibits or leaves judges and jurors free to decide, without any legally fixed standards, what is prohibited and what is not in each particular case". Pp. 402-403, 86 S.Ct. p. 520. The additional argument made in Giaccio that state court interpretations provided appropriate standards was also unavailing. At best, these centered generally in "some misconduct" which was "not reprehensible enough for a criminal conviction but sufficiently reprehensible to deserve an equal distribution of costs."
There is no question as to what is prohibited under the Arkansas rape statute. And the secondary argument in Giaccio is a far cry from that Arkansas statute. Of particular import is the Court's footnote 8, p. 405 of 382 U.S., p. 522 of 86 S.Ct., to the effect that in reaching its conclusion that the Pennsylvania statute was invalid, "we intend to cast no doubt whatever on the constitutionality of the settled practice of many States to leave to juries finding defendants guilty of a crime the power to fix punishment within legally prescribed limits". Mr. Justice Stewart, in concurring, p. 405, 86 S.Ct. p. 522, felt that, despite this disclaimer, much of the reasoning in the opinion served to cast doubt "upon the settled practice of many States to leave to the unguided discretion of a jury the nature and degree of punishment to be imposed upon a person convicted of a criminal offense". No other Justice joined him in that observation. We have no alternative, therefore, except to conclude that footnote 8 means what it says and that the Court did not, by its decision in Giaccio, cast doubt upon the constitutionality of a state practice which leaves to juries the power to fix punishment within legally prescribed limits. This thought was specifically applied later in a capital case by the majority opinion in Spencer v. State of Texas, 385 U.S. 554, 560, 87 S.Ct. 648, 651, 17 L.Ed.2d 606 (1967). We are not impressed with the defense's characterization of footnote 8 as only a "careful reservation of a question" and referable "to jury sentencing generally, not capital sentencing".
"Nor is it contended that it is unconstitutional for the jury to assess the punishment to be meted out to a defendant in a capital or other criminal case, or to make findings as to whether there was or was not a prior conviction even though enhanced punishment is left to be imposed by the judge. The States have always been given wide leeway in dividing responsibility between judge and jury in criminal cases."
B. The second aspect of the argument is that the Arkansas single verdict procedure "raises the gravest questions of procedural fairness". The usual points are advanced, namely, the right to allocution; the restriction of this right in single verdict procedure; and the conflict in such procedure between that right and the right not to incriminate oneself. What is suggested is the split verdict procedure.
All these arguments were made and the same cited cases were considered when we heard en banc and decided Pope v. United States, 372 F.2d 710, 727-730 (8 Cir. 1967), judgment vacated on other grounds, 392 U.S. 651, 88 S.Ct. 2145, 20 L.Ed.2d 1317 (June 17, 1968). We recognized in that federal case that the issue was not an easy one, that it largely disappears when criminal procedure takes its traditional course and punishment is imposed by the judge rather than by the jury, and that the issue seems to suggest the possibility of the two-stage trial. There, too, as here, no request was made of the district court for a two stage trial. We stated our position and concluded that the unitary trial was not error. We felt fortified in that conclusion by the Supreme Court's remark in Spencer v. State of Texas, supra, p. 568 of 385 U.S., p. 656 of 87 S.Ct.:
"We find ourselves about where the Second Circuit majority found itself in Curry (United States v. Curry, 358 F.2d 904). We are not disposed to say that, despite the long accepted unitary trial concept, the two-stage trial is not available at all. We are not disposed to say that a trial judge may not appropriately exercise his discretion in this area. But we are also not disposed to say that the trial court's failure to order sua sponte a two-stage trial for Duane Pope was error or an abuse of discretion.
"It may be that the two-stage trial can be appropriately developed and made to serve a useful purpose under these statutes as they presently exist. It may be that the better solution is for the statutes to be revised so as to place the punishment power back in the hands of the judge where it traditionally has rested. It may be that the ultimate answer is legislation authorizing some narrow appellate review of sentences in these extreme cases. These, however, are primarily legislative matters for the Congress and not for the judiciary."
"Two-part jury trials are rare in our jurisprudence; they have never been compelled by this Court as a matter of constitutional law, or even as a matter of federal procedure."
Until the Supreme Court tells us otherwise, we feel that what we said and concluded on this point in Pope v. United States has equal application to this state case. We adhere to the result reached on this issue in that en banc decision.
The Jury Selection Argument
This point was made on the first habeas appeal and, although it had not been raised in the state court, it was carefully considered by this court, and rejected. Pp. 332-334 of 348 F.2d. The argument urged on that appeal was that in Arkansas at the time of Maxwell's trial petit jurors were selected from electors; electors were persons who had paid the State's poll tax; the official list of a county's poll taxpayers and the poll tax receipts specified race; and the jury list itself indicated race. Ark.Stat. Ann. §§ 39-208, 3-104.2, 3-118, and 3-227(b).
We there noted, p. 332 of 348 F.2d, that on the appeal, in contrast to the situation in the district court, pp. 213-216 of 229 F. Supp., no issue was raised as to deficiencies in the efforts or methods of the jury commissioners, as to under-representation of the negro race in the county jury lists, or as to any pattern of negro repeaters on the juries.
Amendment No. 51 to the Arkansas Constitution, known as the "Arkansas Amendment for Voter Registration without Poll Tax Payment", effective, except for one minor proviso, January 1, 1965, eliminated the poll tax requirement and with it the color references theretofore appearing in the cited statutes. See Walker v. State, 241 Ark. 300, 408 S.W.2d 905, 915 (1966), cert. denied 386 U.S. 682, 87 S.Ct. 1325, 18 L.Ed.2d 403.
We reviewed the applicable Supreme Court opinions and our own earlier decisions in Bailey v. Henslee, supra, 287 F.2d 936, and Henslee v. Stewart, 311 F.2d 691 (8 Cir. 1963), cert. denied 373 U.S. 902, 83 S.Ct. 1289, 10 L.Ed.2d 198, in each of which we had concluded that a prime facie case of limitation in jury selection, because of race, had been established. But we reached the opposite conclusion in Maxwell's first appeal, primarily because there was no proof whatsoever that the jury list was compiled from the poll tax list. The evidence was all the other way and uncontradicted, namely, that the list was first independently prepared and only then was the poll tax book consulted, as it had to be in order to ascertain that the persons tentatively selected were qualified electors. We felt that we could not say that, because the poll tax receipts and books designated race, it necessarily followed that every jury list in Arkansas was automatically unconstitutional; so to conclude would ignore the possibility of the selection being initially made, as the proof in Maxwell's case showed, wholly apart from the poll tax list. We adhered to this conclusion in the later similar case, on this point, of Mitchell v. Stephens, supra, 353 F.2d 129, 133-134.
Presumably, Maxwell and his counsel are not convinced of the rightness of our decision on this point in the first appeal. Nevertheless, the decision was made by a unanimous panel and the Supreme Court did not see fit to disturb it. Having decided the issue, we do not now undecide it in order to reach the opposite result.
The point is repeated, moreover, by Maxwell and counsel on his second habeas petition because, it is said, the granting of certiorari, limited to five questions, in Sims v. Georgia, 384 U.S. 998, 86 S.Ct. 1953, 16 L.Ed.2d 1013 (1966) [the same day certiorari was denied in Mitchell v. Stephens, supra] and the decision in Whitus v. State of Georgia, 385 U.S. 545, 87 S.Ct. 643, 17 L.Ed.2d 599 (1967), require reversal.
Sims v. Georgia was also an interracial rape case with a Negro as defendant. The fourth question as to which certiorari was granted related to the constitutionality of a conviction where "local practice pursuant to state statute requires racially segregated tax books and county jurors are selected from such books", where Negroes comprise only 5% of the jurors selected but about 20% of the taxpayers, and where the defendant offered to prove a practice of arbitrary and systematic Negro inclusion or exclusion based on jury lists of the prior ten years but his offer was disallowed. Sims' conviction was reversed and his case remanded on the ground that a voluntariness-of-confession issue was controlled by Jackson v. Denno, 378 U.S. 368, 84 S.Ct. 1774, 12 L.Ed.2d 908 (1964), and that the Supreme Court of Georgia had erred in holding that Jackson was not applicable. This was the second of the five certiorari questions. The Supreme Court therefore did not reach any of the issues raised by the other four questions. Sims v. State of Georgia, 385 U.S. 538, 87 S.Ct. 639, 17 L.Ed.2d 593 (1967). The significance of the Sims case for Maxwell thus is that the Supreme Court was willing to grant, and did indeed grant, certiorari on a question related to racially segregated tax books, the selection of jurors from such books, the misproportion of negro jurors to negro taxpayers, and a disallowed offer of proof of arbitrary and systematic exclusion for ten years.
Merely stating the question reveals the inapplicability of the fourth Sims question to Maxwell's case. It is true that at the time of Maxwell's trial, as we have repeatedly noted, Arkansas procedure called for racially designated elector lists. But as has already been pointed out, the uncontradicted proof in Maxwell's case was that the petit jurors were not selected from those lists, although the lists were later checked, as they had to be, in order to determine that the jurors selected were qualified electors. There is no assertion here of misproportion of negro jurors to negro taxpayers or negro electors or negro citizens. And there is nothing before us by way of claim of systematic negro exclusion on Garland County jury lists for any period of time.
Similarly, we fail to see where Whitus v. State of Georgia and subsequent rulings based on Whitus, namely, Bostick v. South Carolina, 386 U.S. 479, 87 S.Ct. 1088, 18 L.Ed.2d 223 (1967); Cobb v. Georgia, 389 U.S. 12, 88 S.Ct. 115, 19 L.Ed.2d 11 (1967); Jones v. Georgia, 389 U.S. 24, 88 S.Ct. 4, 19 L.Ed.2d 25 (1967); Sims v. Georgia, 389 U.S. 404, 407-408, 88 S.Ct. 523, 19 L.Ed.2d 634 (1967); see Coleman v. Alabama, 389 U.S. 22, 88 S.Ct. 2, 19 L.Ed.2d 22 (1967), afford anything new or anything not fully discussed and fully decided on Maxwell's prior habeas appeal. The Georgia jury lists, under attack in Whitus, were made up from a racially designated tax digest and by reference to the old jury list theretofore condemned in Whitus v. Balkcom, 333 F.2d 496 (5 Cir. 1964), cert. denied 379 U.S. 931, 85 S.Ct. 329, 13 L.Ed.2d 343. And again there was serious disproportion of negro representation. The Court quite expectedly held that such proof constituted a prima facie case of purposeful discrimination, for the fact situation fell in line with that in Avery v. State of Georgia, 345 U.S. 559, 73 S.Ct. 891, 97 L.Ed. 1244 (1953), cited by the Supreme Court and also by us on the first appeal, p. 332 of 348 F.2d. A construction, for Maxwell's case, of footnote material similar to that found in Jones v. Georgia, supra, at p. 25 of 389 U.S. 88 S.Ct. 4, would be constrastingly revealing. In Maxwell's case, we repeat, we have no question whatsoever of inappropriate negro representation on the Garland County jury list, we have a negro commissioner, and we have jury lists independently prepared prior to the statutorily required reference to the poll tax rolls. The jury list, thus, was not prepared from condemned or racially designated lists. The petitioner's position here would render it impossible to try a negro defendant for any crime in the State of Arkansas prior to the adoption of Amendment No. 51. On the facts such a stricture is not constitutionally required.
Sims and Whitus therefore fail to afford reversing precedent.
Although the point has not been raised to us, a comment relative to the Supreme Court's recent decision in United States v. Jackson, 390 U.S. 570, 88 S.Ct. 1209, 20 L.Ed.2d 138 (1968), may be indicated. In Jackson the Court held unconstitutional the death penalty provision of the Federal Kidnaping Act, 18 U.S.C. § 1201(a). Does that holding suggest the constitutional invalidity of the Arkansas death penalty for rape, under the Fifth and Sixth Amendments which now, through the Fourteenth, have an area of application to the states? See Malloy v. Hogan, 378 U.S. 1, 8, 84 S.Ct. 1489, 12 L.Ed.2d 653 (1964); Gideon v. Wainwright, 372 U.S. 335, 83 S.Ct. 792, 9 L.Ed.2d 799 (1963).
We readily conclude that Jackson does not require a holding of invalidity of the Arkansas death penalty. Jackson, as we read the opinion, pivoted on the fact that the defendant's assertion of his right to a jury trial might cost him his life, "for the federal statute authorizes the jury — and only the jury — to return a verdict of death" and thus subjects the defendant who does not plead guilty and who asks for a jury trial to the hazard of capital punishment. The Arkansas rape punishment situation is not the same. As we have pointed out above, the Arkansas procedure in criminal cases is for the jury to affix the punishment within the limits statutorily defined. § 43-2145. Although § 43-2306 provides that when a jury fails to agree on the punishment or to declare the punishment in their verdict or if they assess punishment not authorized by law, "and in all cases of a judgment on confession", the court shall assess and declare the punishment. But § 43-2108 states that a trial by jury may be waived "except where a sentence of death may be imposed". And the Supreme Court of Arkansas has specifically held in a rape case, where the defendant had entered a plea of guilty, that it was mandatory to impanel a jury to fix the punishment. Scarber v. State, 226 Ark. 503, 291 S.W.2d 241, 243 (1956). Thus, in contrast to the Federal Kidnaping Act, an Arkansas defendant, by entering a plea of guilty in a capital case, does not avoid a trial by jury on the issue of punishment. The critical choice under the federal act which occasioned the result in Jackson, is thus not present under the Arkansas statutes.
It is obvious, we think, that the efforts on behalf of Maxwell would not thus be continuing, and his case reappearing in this court were it not for the fact that it is the death penalty, rather than life imprisonment, which he received on his rape conviction. This fact makes the decisional process in a case of this kind particularly excruciating for the author of this opinion who is not personally convinced of the rightness of capital punishment and who questions it as an effective deterrent. But the advisability of capital punishment is a policy matter ordinarily to be resolved by the legislature or through executive clemency and not by the judiciary. We note, for what that notice may be worth, that the death penalty for rape remains available under federal statutes. 18 U.S.C. § 2031; 10 U.S.C. § 920(a).
Judges VOGEL and MATTHES do not join in this comment.