Can a holding by the U.S. Supreme Court interpreting the Constitution—which the Court never reversed or qualified—ever be treated as a nullity by lower courts? Suppose the reasoning on which the holding stands is exceedingly poor, unsupportable, unsound, contradicted by every interpretive theory one could bring to it? (Yet it has never explicitly been repudiated by its authors or any subsequent Supreme Court.) Could such an opinion properly be ignored by lower courts as having no true force?
The power of a judicial opinion, after all, is thought to depend on its persuasiveness, not merely that it is a pronouncement from one or another node in the judicial hierarchy, including the top spot. That is why opinions are written—to convince the world that the holding is correct and therefore worthy of obedience. Otherwise, courts could simply announce their conclusions and leave it at that.
If a Supreme Court opinion—after decades of careful analysis and thoughtful consideration—has come to be viewed as utterly hollow (or worse: dead wrong), is its holding still entitled to deference? Is such a holding still “good law”?
Consider the remarkable example provided by Williams v. Florida (1970). (1) Over half a century ago, Williams held, by a vote of 7-1, (2) that the Constitution does not require juries to be composed of more than six individuals. Justice Byron White, for the majority, concluded that the traditional twelve-person jury was not mandated by the Constitution, despite its venerated existence for more than 600 years of common law history and nearly 200 years of constitutional history. How did Justice White and his colleagues reach that conclusion?
Florida law authorized six-person juries for the trial of all felonies other than those punishable by death. (3) Johnny Williams challenged the constitutionality of his conviction for robbery by a jury of six, leading to a sentence of life in prison. When Williams v. Florida reached the U.S. Supreme Court, the question to be answered was whether the Sixth Amendment guarantee of trial by jury meant a jury of twelve or something less. (4)
The Williams majority noted that the constitutional text guaranteed a jury but did not specify the size of that jury. The Court acknowledged that the omission might reflect an understanding of the framers and adopters that “jury” was synonymous with twelve-person jury. Alternatively, the Court suggested, the omission might reflect an intention of the founders that the jury’s size be free to vary. What size jury, then, was intended to be permitted or required by the Sixth Amendment?
The Court’s own precedents should be a major source of guidance on what various provisions of the Constitution mean. In an unbroken line of cases from Thompson v. Utah (1898) (5) until Williams, the Court had held that the jury referred to in the Sixth Amendment “is a jury constituted, as it was at common law, of twelve persons, neither more nor less.” (6) The Williams justices found those precedents to lack binding force because, in summary, “The broad implications in early cases indicating that only a body of 12 members could satisfy the Sixth Amendment requirement arose in situations where the issue was not squarely presented and were based… on an improper interpretation of that amendment.” (7)
As a guide to the “original public meaning” of a word or phrase, judges sometimes turn to dictionaries, especially dictionaries published around the time of a law’s drafting. Samuel Johnson’s Dictionary (1755) suggests that a jury was understood in the latter half of the Eighteenth Century to be “a company of men, as twenty-four or twelve,” of which the larger were used for grand juries (8) and the smaller constituted “petit juries” to decide other matters, such as trial verdicts. (9) But the Williams court did not consult dictionaries or any similar contemporary reference source.
In other contexts, justices sometimes inquire about history and tradition. The Court was well aware that the common law had for centuries consistently tried cases to juries fixed at size twelve, (10) and that the “truth of every accusation” against a defendant “should afterwards be confirmed by the unanimous suffrage of twelve of his equals and neighbours.” (11) And, of course, the people who created, and the nation that was created by, the Constitution had relied on, and understood juries—in law and in popular culture—to consist of twelve persons right up until Williams declared otherwise. (12) A few years later, the Court’s novel analysis of jury size was extended to federal civil juries. (13)
Recently, Justice Gorsuch wrote of Williams that, “for the first time and in defiance of centuries of precedent, this Court held that a 12-member panel ‘is not a necessary ingredient’ of the Sixth Amendment right to trial by jury…. Williams was wrong the day it was decided, it remains wrong today, and it impairs both the integrity of the American criminal justice system and the liberties of those who come before our Nation’s courts.” (14)
Whatever the likely meaning to the framers and the public, or the rootedness and longevity of that tradition, the Williams Court found the basis for it to be unworthy of respect:
That history revealed a long tradition attaching great importance to the concept of relying on a body of one’s peers to determine guilt or innocence as a safeguard against arbitrary law enforcement. That same history, however, affords little insight into the considerations that gradually led the size of that body to be generally fixed at 12. Some have suggested that the number 12 was fixed upon simply because that was the number of the presentment jury from the hundred, from which the petit jury developed. Other, less circular but more fanciful reasons for the number 12 have been given, “but they were all brought forward after the number was fixed,” and rest on little more than mystical or superstitious insights into the significance of “12.” Lord Coke’s explanation that the “number of twelve is much respected in holy writ, as 12 apostles, 12 stones, 12 tribes, etc.,” is typical. (15)
In the end, the Court found “absolutely no indication in ‘the intent of the Framers’ of an explicit decision to equate the constitutional and common law characteristics of the jury.” (16)
If an answer to the question of what constitutes an acceptable jury within the meaning of the Sixth Amendment was not to be found in the language of the Constitution, in the nation’s history and tradition, or in the Court’s precedents, how then could one determine which features were necessary for a jury to be regarded as constitutional? To answer that question, the Court created a test of functional equivalence.
The relevant inquiry, as we see it, must be the function that the particular feature performs and its relation to the purposes of the jury trial. Measured by this standard, the 12-man requirement cannot be regarded as an indispensable component of the Sixth Amendment.
The purpose of the jury trial, as we noted in Duncan, is to prevent oppression by the Government.
Providing an accused with the right to be tried by a jury of his peers gave him an inestimable safeguard against the corrupt or overzealous prosecutor and against the compliant, biased, or eccentric judge.
Given this purpose, the essential feature of a jury obviously lies in the interposition between the accused and his accuser of the common sense judgment of a group of laymen, and in the community participation and shared responsibility that results from that group’s determination of guilt or innocence. The performance of this role is not a function of the particular number of the body that makes up the jury. To be sure, the number should probably be large enough to promote group deliberation, free from outside attempts at intimidation, and to provide a fair possibility for obtaining a representative cross-section of the community. (17)
Elsewhere in the opinion, the Court identifies the “reliability of the jury as a factfinder,” the ability of dissenting jurors to stand up to conformity pressure from the majority, and the “results reached” as additional criteria for evaluating the functioning of juries of different sizes.
Thus, the criteria can be summarized as: representative cross-section, group deliberation, influence of the majority on the minority, factfinding reliability, and the results reached by the two different-sized juries.
Under this functional equivalence test, the question before the Court could be answered only by inquiring into the behavior of different-sized juries. The Court proceeded to compare the performance of the six-person jury against that of the traditional twelve-person jury on the criteria it had specified. (18) If those functions were served equally well by smaller juries, then they were equally acceptable to the Constitution.
On each of the functions specified, the Williams majority concluded that six-person juries provided process and outcome equal to those of twelve-person juries, and so the smaller juries were equally acceptable to the Constitution. But, it turns out, the justices misread their sources and relied on statistical intuitions which failed them. Let’s examine their evidence and interpretations.
The Williams Court addressed the possibility that six-person juries might deny criminal defendants the same degree of cross-sectional representation of the community that larger juries generally conferred:
[W]hile, in theory, the number of viewpoints represented on a randomly selected jury ought to increase as the size of the jury increases, in practice, the difference between the 12-man and the six-man jury in terms of the cross-section of the community represented seems likely to be negligible. Even the 12-man jury cannot insure representation of every distinct voice in the community, particularly given the use of the peremptory challenge. As long as arbitrary exclusions of a particular class from the jury rolls are forbidden, the concern that the cross-section will be significantly diminished if the jury is decreased in size from 12 to six seems an unrealistic one.
Think of juries as samples drawn from the population. Reducing the size of those samples from twelve to six can have a substantial impact on cross-sectional representation, depending on the make-up of the population. If, say, 50% of the jury-eligible population were Black and 50% were White, then cutting the sample size in half would make no difference to how many Blacks (or Whites), would be found on randomly chosen juries in the aggregate. But as the population skews in either direction, representation can be altered dramatically. In a population that is, for example, 90% White and 10% Black, (19) 72% (plus-or-minus some variation) of juries of size twelve would have at least one Black juror. But on juries of size six, only 47% (plus-or-minus) of juries would contain one or more Black jurors. (20)
This general principle of sampling has been demonstrated in several empirical studies of the composition of juries. (21) A more recent study found that jury size (six versus twelve) had a much greater impact on racial/ethnic representation than peremptory challenges did. (22)
The Court did not elaborate on its concern for group deliberation—be it quantity or quality, nature, content or something else. Nor did the Williams opinion offer any evidence or reasoning on the possible impact that differences in group size might have on the behavior of concern. Among other things, studies conducted after 1970 have found that, in contrast to six-person juries, twelve-person juries engage in more total deliberation, have more vigorous and contentious discussions, bring more human resources into the discussion, and have more accurate recall of evidence as a group during the deliberation and as individuals when tested after deliberations. (23)
One specific behavioral concern associated with deliberation was addressed, namely, the ability of one or two jurors holding beliefs opposed by their peers to resist pressure to surrender unreasoningly to the majority. The Court summed up what it took to be the findings of several studies: “Studies of the operative factors contributing to small group deliberation and decision making suggest that jurors in the minority on the first ballot are likely to be influenced by the proportional size of the majority aligned against them.” (24)
Thus, a dissenter in a jury divided 5–1 was thought to be just as able to stand up to the majority as a minority faction in a jury divided 10–2. The critical factor, asserted the Court, was the proportion, not the absolute number, of jurors in the minority and majority factions. Remarkably, however, the studies on which the Court purported to rely all concluded essentially the opposite of what the Court said they found. (25)
Without elaboration, evidence, or analysis, the Williams Court concluded that “the reliability of the jury as a factfinder hardly seems likely to be a function of its size.”
If the justices had in mind the process of factfinding—sifting through trial evidence and recalling, discussing, debating, to reach correct conclusions—as noted above, in regard to “group deliberation,” larger groups appear to have the advantage.
More basic social psychological research on small-group decision-making has found that increasing group size improves group decisions up to the point where process inefficiencies begin to detract more than the accrued human resources add. The location of that tipping point varies with the kind of decision task the group is engaged in. The jury deliberation task is of a kind that benefits from increases up to fairly large sizes. (26)
If by “reliability of the jury as a factfinder” the Court meant the outcomes of deliberation, the verdicts, to that we turn next.
Regarding the outcomes of trials, the Court concluded that empirical research had found no differences between six-person and twelve-person juries: “What few experiments have occurred—usually in the civil area—indicate that there is no discernible difference between the results reached by the two different-sized juries.”
Support for that conclusion consisted of six items listed in a footnote. The first did nothing more than quote another work, which stated that “it could easily be argued that a six-man jury would deliberate equally as well as one of twelve.” (27) The second item’s evidence consisted of a judge who stated that he had presided over numerous condemnation trials using five-man juries and had perceived no differences compared to larger juries. (28) The third citation reported on the use of six-person juries in 43 civil cases in a state district court in Worcester, Massachusetts. The author spoke to a court clerk and three attorneys involved in trials, who said the smaller juries did not seem to them to behave differently. (29) The fourth item was to a summary of the third item. (30) The fifth citation was to an announcement that a future test of six-person juries was being planned for Monmouth County, New Jersey. (31) The final citation was to an article discussing possible financial savings that might accrue from reducing jury size, which was not one of the criteria the Court had enumerated. (32) Apart from this remarkably thin soup, the Court had nothing to offer on decision outcomes as a function of group size.
Again, a consideration of the statistics of sampling might have assisted the Court. On this concept, human intuition usually aligns well with the statistical phenomenon. Virtually everyone shares the intuition that larger samples are more reliable than smaller samples. All else equal, would you have more confidence in a public opinion poll that employed a large sample or a small one? If it was important to get the most accurate temperature reading you could, would you prefer to average the readings of twelve thermometers or six? The larger the sample, the more that each individual’s random error is cancelled out, and the more nearly accurate the collective result is.
Now, suppose you had a civil case in which the only issue was damages. Imagine you had 100 juries observing the trial, and each jury reached its damage award independently. Now let’s array those awards in (what will turn out to be) a bell-shaped curve distribution. (The mean of that distribution is the best estimate of a “correct” award that one could arrive at. (33))
Suppose you did this exercise twice, once using 100 juries of size twelve and a second time using 100 juries of size six, and drew their respective bell-shaped curves one on top of the other. The curve reflecting six-person jury awards would be spread wider than that of the twelve—more outliers, more error, fewer juries coalescing around the central “most nearly correct” average. Two scholars describe the effect: “over two-thirds of the twelve-person juries will have ... damage awards close to the community average, compared to just half of the six-person juries. The six-person juries are four times as likely to have extremely low or extremely high... damage awards.” (34)
The same phenomenon occurs when the outcomes are categorical, like polling of voter preferences in an upcoming election, or liability verdicts in criminal trials. Larger samples will produce more consistent outcomes, outcomes that more nearly reflect what everyone in the population would conclude if all of them were making the decision. Put simply: six-person juries produce less consistent verdicts, less predictability, more “errors” than twelve-person juries. (35)
After laying out the panoply of jury functions at issue, the Court sweepingly concluded regarding all of them: “[W]e find little reason to think that these goals are in any meaningful sense less likely to be achieved when the jury numbers six than when it numbers 12—particularly if the requirement of unanimity is retained.” (36) Accordingly, based on its new functional equivalence test, the Court held that “the 12-man panel is not a necessary ingredient of ‘trial by jury,’ and that respondent’s refusal to impanel more than the six members provided for by Florida law did not violate petitioner’s Sixth Amendment rights as applied to the States through the Fourteenth.” (37) The Court also observed that, “[w]e have no occasion in this case to determine what minimum number can still constitute a ‘jury,’ but we do not doubt that six is above that minimum.” (38)
Perhaps not surprisingly, states began to test the limits of the shrinking jury. If six was constitutional, how about five? This was the issue presented in Ballew v. Georgia (1978), (39) involving a conviction for obscenity by a jury of five.
Prompted by the purported reliance on empirical evidence in Williams, and the use of more of it in Colgrove v. Battin (1973), (40) a good bit of empirical research had been carried out on the jury-size question by the time Ballew reached the Court. (41)
In conference following briefing and arguments, the justices agreed that the institution of the jury had to be rescued from the slippery slope the Williams Court had set it upon. Justice Blackmun was assigned to write the majority opinion.
Memos exchanged between Justice Blackmun and the law clerk assisting him with Ballew provide interesting insight. (42) The clerk’s memo reporting on his library research alerted Blackmun that the conclusions in Williams went against the great weight of empirical research findings and statistical principles. Williams was, by its own test, in obvious error and, his clerk suggested, needed to be reversed.
Blackmun’s response was that Williams was “on the books.” In other words, it was now precedent and, if at all possible, precedents should not be overturned. But Blackmun also stated that he worried about the next case after Ballew, and the one after that, and after that. (43)
After concluding in Williams that reducing the jury by half, from twelve to six, made no difference to the performance of its constitutional functions, how does one now argue convincingly that a reduction from six to five would ruinously impair the jury’s function?
Justice Blackmun’s eventual opinion resembled a Brandeis Brief, with extensive presentation of the empirical and statistical evidence his clerk had found. Blackmun not only cited studies, he explained the findings in some detail. Overwhelmingly, that research compared the decision-making processes and products of groups of twelve to groups of six and found differences in their performance, mostly indicating inferior performance of the smaller groups. (44)
The essential logic of Blackmun’s opinion is this:
While we adhere to, and reaffirm our holding in Williams v. Florida, these studies, most of which have been made since Williams was decided in 1970, lead us to conclude that the purpose and functioning of the jury in a criminal trial is seriously impaired, and to a constitutional degree, by a reduction in size to below six members. … [T]he assembled data raise substantial doubt about the reliability and appropriate representation of panels smaller than six. Because of the fundamental importance of the jury trial to the American system of criminal justice, any further reduction that promotes inaccurate and possibly biased decisionmaking, that causes untoward differences in verdicts, and that prevents juries from truly representing their communities, attains constitutional significance.
(45)
A fundamental incoherence is evident. The reaffirmation of Williams is plainly in tension with nearly all of the studies discussed in the opinion. The studies show a decline in the performance of six-person juries measured against the benchmark of twelve-person juries (which speaks to the issue in Williams). But those studies of six-versus-twelve are said to “raise substantial doubt” about five-person juries (the issue in Ballew). This is an obvious disconnect. Data that clearly expose the error of Williams are used to inform Ballew, which the data are not well aimed to do.
Why did Justice Blackmun write the opinion he wrote? He surely knew what he was doing. So, we might ask, what was he doing? We must speculate, because he never publicly explained his purpose. (46) Blackmun probably shared his clerk’s view that Williams was egregiously wrongly decided, though he was determined not to try to overturn it in Ballew. Had he insisted on reversing Williams, he would no longer have been in the majority and would have found himself writing a dissent. (47)
If Blackmun wanted to keep control of the lead opinion, so that the actual empirical evidence would be front and center for the world to see, he framed the opinion as he had to. The prominence of the evidence of Williams’s folly would make it easier for some future Supreme Court, freed from defending its own mistakes, to overturn Williams. Incoherently “reaffirming” Williams while declaring that “the purpose and functioning of the jury in a criminal trial is seriously impaired… by a reduction in size to below six” was the price Blackmun had to pay to be able to announce the known facts with maximum prominence.
If the intent of the author may be inferred from some of the consequences of his actions, then we can say that Blackmun’s goals have been partially achieved. In a number of situations, the holding of Ballew has been brushed aside while the empirical facts it provided lit the path forward. Some examples:
- -
In 1981, the New Hampshire legislature wanted to reduce their own state’s juries from twelve to six. First, however, they requested an advisory opinion from their supreme court. The New Hampshire Supreme Court responded by opining that, in light of the data presented in Blackmun’s opinion in Ballew, juries smaller than twelve would be held to violate the New Hampshire Constitution because of their inferior performance. (48)
- -
Similarly, in 2015, citing the data in Ballew, an Illinois court held that a statutory reduction of jury size from twelve to six violated the Illinois Constitution. (49)
- -
In the mid-1980s, in an explicit effort to make damage awards more stable and predictable, the Model Medical Malpractice Act, promulgated by the Department of Health and Human Services for adoption by states, called for the use of twelve-person juries in medical malpractice cases.
- -
The ABA Jury Standards (2005) specified that juries trying criminal cases “should consist of Twelve persons if a penalty of confinement for more than six months may be imposed upon conviction.”
- -
Rule 48 of the Federal Rules of Civil Procedure has been amended to permit as many as twelve jurors to be seated at the start of a trial, and to require that “each juror must participate in the verdict.” As a result, larger juries decide cases than was anticipated after Colgrove. (50)
Though Justice Blackmun’s opinion announced the judgment of a unanimous Court, and it accomplished the unanimously-agreed-upon goal of stopping the jury’s slide toward extinction, it was joined by only a single other justice, Stevens. What did seven other justices dislike about Blackmun’s opinion? Had his opinion been as hollow as their opinions, they probably would gladly have joined it. Instead, while claiming to reaffirm Williams, Blackmun implicitly repudiated it. He made far too obvious that Williams had been decided in error, and his colleagues apparently wanted no part of that.
What reasons did those justices offer for holding six to be the constitutional minimum?
White, who had been the author of Williams and was the father of functional equivalence, made no reference either to that case or that test. He stated in a single sentence that juries of fewer than six would not satisfy the fair-cross-section requirement of the Sixth and Fourteenth Amendments. He cited nothing to support that conclusion and engaged in no reasoning about it. But, whatever concerned him so about the step from six to five should have been far more concerning respecting the step from twelve to six. This is thinly veiled ipse dixit. (51)
Brennan, joined by Stewart and Marshall, wrote that they agreed with the holding that the Constitution requires “juries in criminal trials to contain more than five persons,” but they offered nothing to explain or justify their conclusion. This is mere ipse dixit.
Powell, joined by Burger and Rehnquist, simply asserted: “the line between five and six-member juries is difficult to justify, but a line has to be drawn somewhere if the substance of jury trial is to be preserved.” No mention of Williams or functional equivalence or any other explanation. More ipse dixit.
Ballew, then, adds up to two justices purporting (unconvincingly) to employ the functional equivalence test to reach the result, six justices pretending the functional equivalence test never existed, and a ninth justice (White) claiming implausibly to be guided exclusively by the fair cross-section requirement. (52)
The justices agreed on stopping the downhill slide of the jury, but not one of them could offer a coherent explanation of why six was the correct minimum—not twelve, not five. The one thing the justices seem to have been sure of was that the damage Williams had inflicted on the jury could not be allowed to get any worse and, apparently, they didn’t care how incoherently they applied the judicial brakes.
If the question of what features of the institution of the jury pass constitutional muster truly could not be answered using conventional interpretive tools, the functional equivalence test made good sense. But to work properly, the test needed to be fed sound data and its output read with a sincere interest in learning the answer: Did six-person juries perform as well as twelve-person juries or didn’t they?
The problem with Williams is that the majority seems to have adopted their conclusion before undertaking the inquiry they prescribed. (53) They fed the test non-studies claiming them to be studies, read actual studies backwards, and relied on statistical intuition to answer questions for which rigorous answers were readily available. When garbage goes in, we know all too well what comes out. Had the functional equivalence test been employed earnestly, the Court would have had to conclude that six did not a constitutionally acceptable jury make.
The problem in Ballew was the insistent effort to work around, and leave undisturbed, the holding in Williams. By making the relevant legislative facts inescapable, however, Justice Blackmun exposed the vapidity of Williams and then walked away from the wreckage. The Ballew justices might have done better to submit correct facts to the test Williams created, recognize that six did not equal twelve, confess error in the earlier application, and reverse Williams. Obviously, none chose that course. Another option was to make the facts irrelevant by abolishing the test that made them relevant. Most of the justices implicitly chose this option by pretending the functional equivalence test never existed. That was the easiest way to ignore the facts made so evident by Blackmun’s opinion.
But if the facts about the effects of jury size on group decision-making no longer mattered because the functional equivalence test was dead, what is left to justify a reading of the Constitution that permits six-person juries to replace twelve-person juries? The answer is: nothing. Functional equivalence is what justified the constitutionality of six-person juries. So, in abandoning the functional equivalence test, Ballew also erased the only justification for smaller-than-twelve juries. Williams, too, had become ipse dixit. The institution of the American jury had been sliced in half, and no justification for doing so any longer existed.
The story of Williams v. Florida is not a story of one mistake, now in the past, receding further every day. It is a story of ongoing harm.
In his dissents to denials of certiorari seeking to take up and correct the errors of Williams, Justice Gorsuch has emphasized the diminishment of the right to jury trial as it had been entrusted through the Constitution and respected for centuries:
For almost all of this Nation’s history and centuries before that, the right to trial by jury for serious criminal offenses meant the right to a trial before 12 members of the community.”
(54) “Acutely concerned with individuals and their liberty, the framers of our Constitution sought to preserve this right for future generations.” Yet today, a small number of States refuse to honor its promise. [They do] what the Constitution forbids because of us.
(55)
In addition, pertaining to the non-equivalence, functionally, of smaller juries, Justice Gorsuch refers to empirical data “suggesting that, in fact, ‘smaller juries are less likely to foster effective group deliberation’ and may not produce as reliable or accurate decisions as larger ones.” (56)
Reference to diminished reliability and accuracy point to the more practical, daily damage caused by Williams. As detailed earlier in this Article, because smaller juries generally do not perform as well as larger juries on critical dimensions (recognized by the Williams Court as central functions of the jury), (57) wherever smaller juries are employed, cross-sectional representation on any and every attribute is diminished, (58) decisions are less reliable, more variable, more unpredictable, and have greater potential to produce aberrant verdicts (both civil and criminal verdicts as well as civil damage awards) than had been the norm with twelve-person juries. (59) In the aggregate, this leads to less fairness for parties and reduced public confidence in the trial process.
This undermining of the jury as a decision-making institution is an unforced error, entirely of the Supreme Court’s doing. “Because of us,” as Justice Gorsuch says.
The power and legitimacy of a holding ultimately depends on the quality of the reasoning mustered in its support. In Ballew, the justices themselves fled from the reasoning in Williams and offered nothing in its place. Reasons are everything. Ipse dixit is nothing. If that sorry state is where the Supreme Court arrived in Ballew, we must recall how the journey began: with Justice White’s opinion in Williams.
Williams has become a zombie — its vital organs (if it ever had any) rotted away to nothing, abandoned by those who created it, only seeming to be alive. If the Supreme Court fails to kill it, must other courts bow down to the zombie? Must the original holding continue to be enforced by lower federal and state courts unless and until the Supreme Court administers the coup de grâce to its unsafe creation? Or can other courts eradicate the zombie simply by ignoring it?
This was the choice facing the Arizona courts in Khorrami. (60) Feeling they were bound by Williams, they enforced its holding. (61) Had Arizona looked beyond the bare holding and considered the subsequent treatment of Williams by government agencies and private organizations, by rules committees, by other courts, by scholars of various disciplines, and by the Supreme Court itself (in Ballew), they’d have seen that whatever foundation Williams might initially have seemed to stand upon has long since crumbled to nothing.
If, in light of that more complete examination, the Arizona courts had concluded that Williams was a nullity, did they have the authority to cast it aside? The Khorrami court might have explained such a decision by noting the following:
- -
Ample historical evidence supports a conclusion that the intent of the framers, the understanding of the ratifiers, and the original public meaning of the jury trial right guaranteed by the Sixth Amendment envisioned a “jury” to be a body of twelve. (62)
- -
If the intent of the framers and the understanding of the ratifiers and the public could not be determined, and a functional equivalence analysis is required to assess the constitutionality of juries smaller than twelve, ample empirical evidence supports a conclusion that six is not the functional equivalent of twelve. (63)
- -
The U.S. Supreme Court abandoned the functional equivalence test, and by so doing withdrew its rationale for the conclusion that six-person juries meet the constitutional requirement for jury trials. (64)
- -
The trial process incurs continuing harm from the use of juries smaller than twelve because juries smaller than twelve do not, in fact, perform as well as juries of twelve on critical dimensions specified by the U.S. Supreme Court. More specifically, such juries are less capable of representing the eligible population on any and all demographic attributes, (65) they make decisions that are less reliable, more unpredictable, and have greater potential to produce aberrant verdicts. (66) These deficiencies, in turn, contribute to erosion of public confidence in the trial process.
However strong the reasons might be for casting aside a Supreme Court opinion that the Supreme Court itself did not reverse, are lower courts permitted to do so? Put the other way around: does stare decisis require blind obedience? (67)
Had these circumstances arisen in Canada, the answer would be straightforward. The Supreme Court of Canada recognizes the authority of lower courts to cast aside a precedent created by a higher court. “When a new legal issue is raised, or if there is a change in the circumstances or evidence that fundamentally shifts the parameters of the debate, a lower court may reconsider the precedents set by a higher court.” (68) Although “the lower court cannot ignore an authoritative precedent, and the bar is high when it comes to justifying the re-examination of a precedent… re-examination is justified when a new legal issue arises or when there is a significant change in the situation or evidence. This approach balances the imperatives of finality and stability with the recognition that a lower court must be able to fully exercise its function when faced with a situation where it is appropriate to review a precedent.” (69) “[T]he principle of stare decisis does not constitute a straitjacket that condemns the law to inertia.” (70)
The American context is different. On the one hand, the U.S. Supreme Court has asserted that it alone has the authority to review its own precedents. (71) On the other hand, in the more complicated reality of U.S. practice, under appropriate circumstances, Supreme Court precedents are from time to time—if not cast aside—nudged aside. Whether the specific concept that best describes what a lower court has done with a particular precedent is termed distinguishing, (72) treating it as a dead letter, (73) erosion, (74) narrowing (or expanding) from below, (75) anticipatory over-ruling, (76) or something else, what has happened is essentially the same: the most mechanical, direct, faithful reading of the precedent is not the one that the lower court has applied.
Bradford’s explication of one of those concepts could, with slight adjustment, illuminate all of them:
The Supreme Court’s willingness to [overrule precedent and] change doctrine creates situations in which a lower court cannot be completely loyal to the Supreme Court: a Supreme Court precedent directly applies to the case at hand, but later Supreme Court cases cast doubt on that precedent, either rejecting its analysis or applying a different test to an analogous issue. If the lower court is convinced that the Supreme Court would no longer follow the doubtful precedent, must the lower court nevertheless pay heed to stare decisis and apply that precedent? Or should it be faithful to the later decisions and rule as it believes the Supreme Court, given the opportunity, now would? Two competing lines of decision arose in the lower federal courts. One line of cases, apparently a minority view, held that lower courts must blindly and absolutely follow Supreme Court decisions until the Supreme Court expressly overrules them. According to this view, lower courts owe allegiance to earlier Supreme Court precedent, regardless of how doubtful that precedent may have become in light of Supreme Court developments in other areas. The majority view rejected this wooden application of stare decisis, arguing that lower courts should recognize when a Supreme Court precedent is effectively dead, whether or not the Supreme Court has acknowledged the murder. According to this view, lower courts should disregard Supreme Court decisions when they are reasonably sure that the Supreme Court would overrule them given the opportunity.
(77)
To take one famous example, the Supreme Court has never overruled its holding, announced in Plessy v. Ferguson (1896), that separate-but-equal satisfies the Fourteenth Amendment’s requirement of equal protection of the laws. Nevertheless, that doctrine has been chipped away, mostly by lower courts, so thoroughly that it is dead and gone.
(78) Other examples of Supreme Court decisions eroded by lower courts include Miranda v. Arizona (1966),
(79) Buckley v. Valeo (1976),
(80) and United States v. Lopez (1995).
(81) Imprecise precedents—like the “undue burden” test in abortion jurisprudence
(82)—allow lower courts to pretend they are following a rule while progressively limiting it. Usually, the lower court decision is subtle in nudging aside a high court precedent, but not always. The trial judge in Gold v. DiCarlo (1964)
(83) cast aside the Supreme Court’s decision in Tyson & Brother v. Banton (1927)
(84) in no uncertain terms:
We would be abdicating our judicial responsibility if we waited for the Supreme Court to use the express words “We hereby overrule Tyson,” as the plaintiffs contend we should, before recognizing that the case is no longer binding precedent but simply a relic for the constitutional historians. Judges do not have such mechanical or wooden attitudes nor are they devoid of all powers of interpretation, analogy and application of constitutional principles….
Soon thereafter, the Supreme Court affirmed that trial judge. (85)
In all events, reconsidering a holding of the highest court must be done with the greatest diligence. Only opinions that have been thoroughly scrutinized, carefully dissected, and overwhelmingly found to be unsound are candidates for being cast aside or nudged aside.
The problem with Williams v. Florida and the constitutional meaning of a jury is that although the reasoning is universally deprecated (even by a unanimous Supreme Court, implicitly, in Ballew), the holding is so clear (juries as small as six are constitutional). There is no wiggle room for a lower court that would like to politely nudge Williams aside. Yet the alternative is to allow a legal zombie to live on, wreaking whatever havoc it will.
Suppose a state supreme court or a federal court of appeals cast aside Williams. The losing party (almost certainly the government) might choose to let the matter lie. Then, such as in a state like Arizona, the federal constitution would be understood to require twelve jurors, and that would trump state constitutional or statutory law permitting smaller juries. (86) More likely, the losing party would seek further review higher up the judicial ladder. This would be a helpful turn, whatever the outcome. The Supreme Court could refuse to grant cert, thereby allowing Williams to further erode and eventually evaporate. Or it could grant cert, rethink the issue, and either (a) offer sound reasons to interpret the Constitution to require juries of no more than six, or (b) agree with the prevalent assessment of Williams’s speciousness and repudiate Williams, again accepting the constitutional requirement of twelve jurors. At the end of the day, the Supreme Court does get to have the last word. What casting aside one of its opinions does is to create pressure on the Court to responsibly declare that final word.
Williams v. Florida, 399 U.S. 78 (1970) (hereinafter, Williams).
Justice Marshall dissented and Justice Blackmun, having recently been appointed to the Court, took no part in the decision.
Fla. Stat. § 913.10 (1) (1967).
Williams. A second issue in the case had to do with timely disclosure of anticipated alibi defenses, which we need not address.
170 U. S. 343 (1898).
Id. at 349. The Supreme Court’s other precedents were Capital Traction Co. v. Hof, 174 U.S. 1 (1899) (a trial by jury “in the primary and usual sense of the term at the common law and in the American constitutions, is . . . a trial by a jury of twelve.”); Maxwell v. Dow, 176 U. S. 581 (1900) (expressing “no doubt” that “a jury composed, as at common law, of twelve jurors was intended by the Sixth Amendment.”); Rassmussen v. United States, 197 U. S. 516 (1905) (stating that the Sixth Amendment guarantees “the right to be tried by a jury of twelve.”); Patton v. United States, 281 U. S. 276 (1930) (stating that it was “not open to question” that the right to trial by jury for serious offenses “means a trial by jury as understood and applied at common law,” including that it “should consist of twelve” members).
Williams (Black, J., concurring). By contrast, Marshall, J., in dissent: “I adhere to the decision of the Court in Thompson v. Utah.” “[T]he Court has not made out a convincing case that the Sixth Amendment should be read differently than it was in Thompson, even if the matter were now before us de novo—much less that an unbroken line of precedent going back over 70 years should be overruled.”
Employed “to consider of all bills of indictment preferred to the court; which they do either approve… or disallow….” Juries, in S
Id. See also, W The jury, passing on the prisoner’s life, May in the sworn twelve have a thief or two Guiltier than him they try.
“At the beginning of the thirteenth century, twelve was indeed the usual, but not the invariable, number. But by the middle of the fourteenth century, the requirement of twelve had probably become definitely fixed. Indeed, this number finally came to be regarded with something like superstitious reverence.” A. S
W
Historical records of which the Court does not seem to have been aware suggest that the legal culture at the time of the founding understood juries to number twelve. In the Virginia Ratification Convention, Madison and others argued that several characteristics of the jury, including its size, were “incident to the trial by jury” and so it was unnecessary to enumerate its attributes in the Constitution’s text. Governor Edmund Randolph explained that, “There is no suspicion that less than twelve jurors will be thought sufficient.” Those historical sources are discussed in Richard S. Arnold, Trial by Jury: The Constitutional Right to a Jury of Twelve in Civil Trials, 22 H
Colgrove v. Battin, 413 U.S. 149 (1973).
Gorsuch was writing a dissent to a denial of certiorari. Khorrami v. Arizona, 598 U.S. _ (2022). The Court had declined the opportunity to correct its earlier error. Gorsuch dissented yet again to another denial of cert. Cunningham v. Florida, 602 U.S. _ (2024).
Williams, at 87–88.
Id. at 99.
Id. at 100.
Because these are legislative facts, the Court is unrestricted in its investigation of and conclusions about them. See Advisory Committee Comment to Federal Rule of Evidence 201.
Or any other attribute: college educated versus not, above a specified income level versus not, in a certain age range versus not, etc.
These quantities are simple to derive. To begin, if a sample of 1 is drawn from a population stratified 90%/10%, the probability that the sample will consist of a member of the 10% minority is .10. Suppose these were judges being randomly assigned to cases. The cases would be presided over by majority judges 90% of the time and minority judges 10% of the time.
Now, if a sample of 6 is drawn from a population stratified 90%/10%, the probability that the sample will consist entirely of member of the 90% majority is .90 to the sixth power. (.90 multiplied by .90 six times). That equals .53. The complementary probability, .47, is the proportion of juries that would not consist entirely of members of the 90% majority.
Finally, do the same for samples of 12. The probability that a sample will consist entirely of member of the 90% majority is .90 to the twelfth power. That equals .28. The complementary probability, .72, is the proportion of juries that would not consist entirely of members of the 90% majority.
See studies analyzed in Michael J. Saks & Mollie Weighner Marti, A Meta-Analysis of the Effects of Jury Size, 51 L. & H
Shari S. Diamond et al., Achieving Diversity on the Jury: Jury Size and the Peremptory Challenge, J. E
See discussions of the empirical research in Valerie Hans, The Power of Twelve: The Impact of Jury Size and Unanimity on Civil Jury Decision Making, 4 D
Williams, n. 49.
See H
I
Lloyd L. Wiehl, The Six Man Jury, 4 G
Edward A. Tamm, The Five-Man Civil Jury: A Proposed Constitutional Amendment, 51 G
Phillip M. Cronin, Six-Member Juries in District Courts, 2 B
Six-Member Juries Tried in Massachusetts District Court, 42 J. A
New Jersey Experiments with Six-Man Jury, 9 B
Richard H. Phillips, A Jury of Six in All Cases, 30 C
See Michael J. Saks & Peter David Blanck, Justice Improved: The Unrecognized Benefits of Aggregation and Sampling in the Trial of Mass Torts, 44 S
V
Michael J. Saks, The Smaller the Jury, The Greater the Unpredictability, 79 J
Williams, at 100.
Williams, at 86. And: “We conclude, in short, as we began: the fact that the jury at common law was composed of precisely 12 is a historical accident, unnecessary to effect the purposes of the jury system and wholly without significance ‘except to mystics’.” Williams, at 102.
Williams, n. 28.
435 U.S. 223 (1978) (hereinafter, Ballew).
Colgrove v. Battin, 413 U.S. 149 (1973). Colgrove extended the functional equivalence test to federal civil trials, concluding that the jury trial rights of the Seventh Amendment did not include twelve jurors. The Court cited the first four studies conducted after, and prompted by, Williams, all of which now appeared to confirm its earlier empirically ungrounded belief that small and large juries functioned equivalently. Properly evaluated, however, those studies also could not support the Court’s conclusion. See Shari S. Diamond, A Jury Experiment Reanalyzed, 7 U
Georgia also resorted to pure logic when it took seriously the Court’s avowal that it had “no occasion [in Williams] to determine what minimum number can still constitute a ‘jury,’ but we do not doubt that six is above that minimum.” The Attorney General’s brief argued that, “If six is above the minimum, five cannot be below the minimum. Because there is no number in between.”
Justice Blackmun left his papers to the Library of Congress. Unlike most other justices, he made them available to the public immediately upon his death. https://www.loc.gov/rr/mss/blackmun/
See Hans Zeisel, …And Then There Were None: The Diminution of the Federal Jury, 38 U. C
In a concurrence, Powell, J., joined by Burger and Rehnquist, JJ., took a gratuitous swipe at Blackmun’s opinion. Wrote Powell: “I have reservations as to the wisdom—as well as the necessity—of Mr. Justice Blackmun’s heavy reliance on numerology derived from statistical studies. Moreover, neither the validity nor the methodology employed by the studies cited was subjected to the traditional testing mechanisms of the adversary process.” Ballew, at 246.
Blackmun hit back at all three: “We have considered [the studies] carefully, because they provide the only basis, besides judicial hunch, for a decision about whether smaller and smaller juries will be able to fulfill the purpose and functions of the Sixth Amendment. Without an examination about how juries and small groups actually work, we would not understand the basis for the conclusion of Mr. Justice Powell that ‘a line has to be drawn somewhere.’ We also note that the Chief Justice did not shrink from the use of empirical data in Williams v. Florida, when the data were used to support the constitutionality of the six-person criminal jury, or in Colgrove v. Battin, a decision also joined by Mr. Justice Rehnquist.” Ballew, n. 10 (internal footnotes omitted).
Ballew, at 239.
Justice Blackmun’s former clerk who worked on Ballew, whom I interviewed after the Justice’s death, refused to share whatever he knew about Blackmun’s goals for the opinion. Once, while visiting the law school at Northwestern University, dinner companion Professor Shari S. Diamond asked Justice Blackmun why he wrote the unusual opinion that he did in Ballew. He answered elusively: “You have to write something.”
Or, at least, a partial dissent, since he would certainly still agree that the Constitution does not tolerate juries smaller than six.
Opinion of the Justices, 121 N.H. 480, 431 A2d 135 (1981).
Kakos v. Butler, Illinois Circuit Court No. 2015 L 6691 (2015), aff’d 2016 IL 120377.
Patrick E. Higginbotham et al., Better by the Dozen: Bringing Back the Twelve-Person Jury, 104 J
Justice White’s opinion in Williams might be described as “an exercise in raw judicial power.” After all, the decision was made “in defiance of centuries of precedent” (see supra note 14 and accompanying text) and in disregard of history and tradition as well as of the very empirical and statistical evidence his opinion called into relevance by framing the issue as he did. The holding has little or nothing to support it but votes. Ironically, three years later, in his dissent in Doe v. Bolton, 410 U.S. 179 (1973), a companion case to Roe v. Wade (1973), Justice White accused a majority of his colleagues of “an exercise in raw judicial power,” id. at 222.
Though none of the justices in Ballew chose to do so, judicial admission of error is not unheard of. Baron Bramwell, for example, in Andrews v. Styrap, 26 L. T. R. (N. S.) 704, 706 (1872), conceded an earlier error by saying, “The matter does not appear to me now as it appears to have appeared to me then.”
Think of a doctor who has prematurely concluded the patient has disease X, and is so sure that that is the correct diagnosis that she pays no attention to the results of the test designed to diagnose whether it is X or not.
Supra note 14.
Supra note 14 (internal citations omitted).
Cunningham, id.
See Part II.
See Part II, Sec. B, D and E. Only a handful of states use criminal juries smaller than twelve for non-capital felony trials, but in civil cases most federal district courts and a third or more of the states seat juries smaller than twelve. See Higginbotham et al., supra note 50; Erica J. Boyce, Time to Reflect: Has the Research Changed Regarding the Importance of Jury Size? (National Center for State Courts, December, 2021).
Supra note 14.
Arizona v. Khorrami (Arizona Court of Appeals), 2021 WL 3197499 (declining “Khorrami’s invitation to reconsider the constitutionality of eight-person juries in Arizona.”). The Arizona Supreme Court denied Khorrami’s petition for review. The Arizona Constitution, Sec. 23, provides that “The right of trial by jury shall remain inviolate. Juries in criminal cases in which a sentence of death or imprisonment for thirty years or more is authorized by law shall consist of twelve persons. In all criminal cases the unanimous consent of the jurors shall be necessary to render a verdict. In all other cases, the number of jurors, not less than six, and the number required to render a verdict, shall be specified by law.” By statute, A.R.S. § 21–102.B, Arizona authorizes criminal trials using 8-person juries.
See Part I.
See Parts II and III. See also, Erica J. Boyce, supra note 50 (reviewing recent research on jury size, confirming earlier findings).
See Part II, Sec. C, D and E. Only a handful of states use criminal juries smaller than twelve for non-capital felony trials, but in civil cases most federal district courts and a third or more of the states seat juries smaller than twelve. See Higginbotham et al., supra note 50; Erica J. Boyce, supra note 50.
Four arguments are generally offered to support a rigid stare decisis: (1) predictability, certainty and reliability, (2) fairness and uniformity, (3) judicial economy and (4) preserving public respect for the courts.
Canada (Attorney General) v. Bedford, 2013 SCC 72.
Id.
Carter v. Canada (Attorney General), 2015 SCC 5.
Rodriguez de Quijas v. Shearson/American Express, Inc., 109 S. Ct. 1917 (1989).
A state intermediate court of appeals judge, who will remain nameless, once told me that when he wished to avoid applying a precedent, he could almost always find a way to distinguish the case before him from the precedent-setting case.
“A law or practice that, although not formally abolished, is no longer used, observed, or enforced.” B
See the discussion, infra, of Plessy v. Ferguson, 167 U.S. 537 (1896).
Richard M. Re, Narrowing Supreme Court Precedent from Below, 104 G
C. Steven Bradford, Following Dead Precedent: The Supreme Court’s Ill-Advised Rejection of Anticipatory Overruling, 59 F
Id. at 40–41 (internal citations omitted).
Brown v. Board of Education, 347 U.S. 483 (1954) can be read either to have carved out an exception to the general principle of separate-but-equal for public education, or to have applied Plessy and found that separating public school children by race caused the children psychological harm such that “Separate educational facilities are inherently unequal.” Either way, Plessy was still alive the day after Brown was decided. Nevertheless, lower courts took the aura of disapproval toward segregation and extended it to all kinds of settings beyond that of public education. Plessy thereby died. But not because the Supreme Court reversed it. The lower courts’ dismantling of Plessy was so effective that most people, including judges, today believe that Brown overturned it.
84 U.S. 436 (1966).
424 U.S. 1 (1976).
514 U.S. 549 (1995).
Planned Parenthood v. Casey, 505 U.S. 833 (1992).
235 F. Supp. 817, 819 (S.D.N.Y. 1964), aff’d. 380 U.S. 520 (1965).
237 U.S. 418 (1927).
Supra note 83.