Jury Size and Decision

Both the size of the jury and the number of jurors who must be in agreement for a verdict to be concluded (the group’s “social decision rule”) have been the subject of litigation at the U.S. Supreme Court as well as a subject of research by psychologists and other social and behavioral scientists. The number of jurors and the minimum proportion of them who must be in agreement are set by formal legal rules (e.g., state statutes, federal rules of civil procedure), and those rules are in turn subject to constitutional requirements. The Supreme Court has framed its analysis of jury size and decision rule questions in terms of the effects of those variables on jury behavior. Thus, the findings of research on group decision making as a function of group size and social decision rule are of central relevance to the Court’s constitutional analysis. Yet a considerable tension exists between the Court’s conclusions and the empirical findings.

Jury Size

For 600 years of common-law history and 200 years of American constitutional history, the jury was considered to have 12 members. But several states and federal districts in the United States began to use smaller juries, and in the 1970s, challenges to the use of juries with fewer than 12 members reached the U.S. Supreme Court. The Court analyzed the constitutionality of smaller juries by rejecting the guidance of history, tradition, and its own precedents. Instead, the Court reasoned that because the size of the jury was not specified in the Constitution and the framers’ intentions regarding jury size could not be divined, the answer would have to be found through a “functional” analysis of the jury’s purpose: If smaller juries did not behave differently from juries of 12, then they were their functional equivalent and therefore were constitutional.

In a series of cases—Williams v. Florida (1970) (state criminal juries of 6), Colgrove v. Battin (1973) (federal civil juries of less than 12), and Ballew v. Georgia (1978) (state criminal juries numbering 5)— the Supreme Court held smaller juries to be constitutionally permissible because it found no important differences between 6- and 12-member juries in the reliability of their fact finding, the quality or quantity of their deliberation, their cross-sectional representation of the community, the ability of jurors in the minority to resist the social pressure to conform, or their verdicts. These findings were reached through a combination of judicial intuition, misconstruing non-studies as empirical studies, misreading the findings of actual empirical studies, and failing to see elementary flaws in actual empirical studies.

What the empirical research findings actually indicate is that smaller groups foster behavior that is beneficial in some respects, but in view of the purposes for which juries are employed, most of the advantages appear to favor keeping juries at 12. On the positive side, in smaller juries, members share more equally in the discussion, find the deliberations more satisfying, and are more cohesive.

A meta-analysis of studies specifically of juries (both simulated and actual) found that larger juries are more likely than smaller juries to contain members of minority groups, deliberate longer, hang more often, and recall trial testimony more accurately. Turning to studies of small-group behavior generally, one finds that larger groups tend to discuss and debate more vigorously, collectively recall more information, and make more consistent and predictable decisions. The latter finding means that as juries grow smaller, they will tend in criminal cases to make more errors of acquitting the guilty or convicting the innocent; and in civil cases, not only will the rate of erroneous verdicts rise, but juries will tend to render damage awards that are more variable and unpredictable. (Because such differences will be small, very large sample sizes would be necessary to detect them.) In accord with classic research on the psychology of conformity, because in larger groups there is greater likelihood that a dissenter will have at least one ally, a dissenter in larger juries will usually find an ally and therefore be better able to resist the pressure to submit to the majority.

In short, the Williams Court had scant support for its conclusion that “there is no discernible difference between the results reached by the two different-sized juries”; the little research evidence that existed then and most of the evidence that developed later supported the opposite conclusion: that 6-person juries did behave differently from 12-person juries and most of those differences represented less-desirable decision-making processes.

In the Ballew case, the Supreme Court drew the line, holding juries of 6 to be the constitutional minimum. Justice Harry Blackmun’s opinion announcing the judgment of the Court extensively reviewed the empirical research on the subject, much of it having been prompted by the Williams decision. Curiously, although the research summarized in the opinion mostly compared 6- and 12-person juries and indicated that the former did not perform as well as the latter, the Court did not reverse its earlier holdings. Instead, it reaffirmed the earlier decisions finding equivalence between large and small juries, but it now held that juries smaller than 6 were unconstitutional because, in Justice Blackmun’s opinion, the studies raised serious concerns about the performance of juries of fewer than 6 members.

The research findings have been recognized in other legal settings. The Standing Committee on Federal Civil Rules recommended 12-person juries for federal civil trials (a recommendation not adopted by the Judicial Conference). During the administration of President Reagan, the Department of Health and Human Services promulgated a model medical malpractice statute for the states that specified 12-person juries (specifically for their greater predictability). And the New Hampshire Supreme Court— citing the factual findings of Ballew but rejecting its legal holding—ruled that a reduction in jury size below 12 would violate the New Hampshire State Constitution (which similarly did not specify a jury size).

Social Decision Rule

From the 14th century in England until the latter part of the 20th century, juries had been required to reach unanimous verdicts. Several American states began to permit quorum verdicts, and in the 1970s, challenges to the constitutionality of quorum verdicts came before the U.S. Supreme Court.

The principal motivation for eliminating the unanimity rule seems to have been a desire to reduce the incidence of hung juries. Without quorum verdicts, hung juries occur at a national rate of about 5% or 6%, and allowing quorum verdicts reduces that rate by a few percentage points.

In Apodaca v. Oregon (1972) and Johnson v. Louisiana (1972), the Supreme Court held verdicts split as widely as 9:3 to be constitutional, and in Burch v. Louisiana (1979), the Court held that the verdicts of 6-person juries had to be unanimous. The Court’s reasoning was much the same as in the jury-size cases: As to the social decision rule for a jury verdict, the Constitution does not say and the intentions of the framers are unknown, so the “inquiry must focus upon the function served by the jury in contemporary society.”

The main issues about group behavior that were debated in the Court’s functional analysis were whether juries required to reach only quorum decisions would pay less attention to the arguments of the unneeded minority, whether the jury’s verdicts would be less accurate, and whether the weight of evidence sufficient to produce a conviction would be reduced. The last of those questions is an especially interesting one. The standard or proof (preponderance, beyond a reasonable doubt) is directed at individual jurors, seemingly separate from the issue of the rules for combining individual views into a group decision. But the two together will surely have a bearing on the group’s collective confidence in their verdict and on the quantum of proof needed to lift the jury over those several individual-to-group decision thresholds to a verdict.

The Supreme Court’s opinions assert that jurors will not behave differently when a nonunanimous decision rule is in place—at least not when the jury numbers 12—or not differently enough to matter. Less research has been done on social decision rules than on group sizes, but what research has been conducted does not generally support the Court’s majority.

Compared with unanimous rule juries, juries operating under a nonunanimous decision rule deliberate for a shorter time, do not let dissenters have sufficient say so as to change the minimum consensus once it is achieved (while in unanimous rule juries, jurors in the minority participate disproportionately in the deliberation), are more vote oriented and less evidence oriented, are less certain of the defendant’s guilt when convicting, and are less likely to end in a deadlock. In research on group decision making generally, groups required to reach unanimous decisions are found to be more likely to reach correct solutions (on problems with clear right/wrong answers) than groups working with less-demanding social decision rules.

Notwithstanding the Supreme Court’s rulings permitting nonunanimous verdicts under the federal Constitution, the great majority of states continue to require unanimous verdicts in felony trials, and all do in capital murder trials.


  1. Apodaca v. Oregon, 406 U.S. 404 (1972).
  2. Arnold, R. S. (1993). Trial by jury: The constitutional right to a jury of twelve in civil trials. Hofstra Law Review, 22, 1-35.
  3. Ballew v. Georgia, 435 U.S. 223 (1978).
  4. Burch v. Louisiana, 441 U.S. 130 (1979).
  5. Colgrove v. Battin, 413 S. 149 (1973).
  6. Hastie, R., Penrod, S., & Pennington, N. (1983). Inside the jury. Cambridge, MA: Harvard University Press.
  7. Johnson v. Louisiana, 406 U.S. 356 (1972).
  8. Saks, M. J. (1977). Jury verdicts: The role of group size and social decision rule. Lexington, MA: Lexington Books.
  9. Saks, M. J., & Marti, M. W. (1997). A meta-analysis of the effects of jury size. Law and Human Behavior, 21, 451—167.
  10. Williams v. Florida, 399 U.S. 78 (1970).
  11. Zeisel, H. (1971)…. And then there were none: The diminution of the federal jury. University of Chicago Law Review, 38, 710-724.

Return to the overview of Trial Consulting in Forensic Psychology.