Mixed Questions of Law and Fact

Rule 52(a)(6) forbids setting aside a trial judge's findings of fact unless they are clearly erroneous, and the court of appeals reviews the judge's legal rulings de novo.¹ Those two standards fit cleanly when an appellant attacks a pure finding of what happened or a pure statement of what the law requires. A mixed question fits neither cleanly. It arises when the historical facts are admitted or established, the rule of law is undisputed, and the only remaining issue is whether the rule, applied to those facts, is satisfied.² Whether undisputed events at a roadside stop added up to reasonable suspicion is a mixed question; so is whether a defendant's undisputed conduct fell short of reasonable care.

Sorting a mixed question starts by pulling it into three layers. The first layer is the basic or historical facts, meaning who did what, when or where, how or why, and the judge's findings at that layer receive clear-error review.³ The second layer is the legal test the court chose, and a dispute about what that test requires is a question of law. The third layer asks whether the historical facts found satisfy the legal test chosen, and the standard of review for that layer varies with the nature of the question and with whether the trial court or the appellate court is better suited to resolve it.⁴ Rule 52(a) gives no particular guidance on the third layer, and no rule or principle will unerringly separate a factual finding from a legal conclusion.⁵ The label a court attaches to its ruling therefore settles little, because the standard turns on the work the question requires.⁶

Three subsections follow: the test the Supreme Court supplied for choosing a standard, the decided cases that calibrate it, and a constitutional exception to ordinary clear-error review.

The Mixed-Question Test

Some courts have reviewed mixed questions de novo as a matter of course, and that default still appears in the case law.⁷ The Supreme Court has since qualified it by explaining that mixed questions are not all alike.⁸ Some call on a court to expound on the law, and when applying the law means developing auxiliary legal principles useful in other cases, appellate courts should typically review de novo. Others immerse the court in case-specific facts, compelling it to marshal and weigh evidence and make credibility judgments, and appellate courts should usually review those decisions with deference. The Court has described such facts as "multifarious, fleeting, special, narrow facts that utterly resist generalization."⁹ In U.S. Bank v. Village at Lakeridge it condensed the result into one line: the standard of review for a mixed question "all depends—on whether answering it entails primarily legal or factual work."¹⁰

The test asks, in effect, which court is better suited to the work.¹¹ Answering a question that amounts to saying what a broad legal standard means yields a principle other cases will borrow, and the appellate court is positioned to supply it. Answering a question that amounts to deciding what these witnesses and these exhibits show yields an answer for this case alone, and Rule 52(a)(6) directs the reviewing court to give due regard to the trial court's opportunity to judge credibility.¹² Even de novo review does not let the appellate court start over. Independent review entails careful consideration of the district court's legal analysis, and a legal determination is not made de novo in the sense of "an original appraisal of all the evidence."¹³

The two arguments draw two standards. The first attacks the judge's inference about motive, which the court of appeals reviews for clear error, as it reviews findings of negligence or discriminatory intent. The second attacks the legal test, and the court of appeals decides that question anew. If the judge applied the wrong test, an error of law impaired the judgment on a mixed question, and the finding must be set aside however carefully the historical facts were found.¹⁴ The first task on appeal is therefore to identify which argument the appellant is actually making.

Calibrating Anchors

Decided cases at each end of the range make the test usable. Ornelas v. United States anchors the law-declaring end. The Court held that determinations of probable cause and reasonable suspicion under the Fourth Amendment are reviewed de novo, even though each combines a factual question, what happened before the stop or search, with a legal one, whether those facts amount to a constitutional violation.¹⁶ The trial judge's findings about what happened still receive clear-error review, and only the ultimate determination, resting on facts that survive that review, is decided anew. In the terms of the test, applying a broad standard such as reasonable suspicion to the facts found is legal work.

Miller v. Fenton anchors the other end. The Court noted there that it had treated juror bias, a classically mixed question, as a question of fact reviewable for clear error, despite the intimate connection between such determinations and the constitutional guarantee of an impartial jury, and it explained the policy-driven reason for that treatment.¹⁷ The same opinion held that the voluntariness of a confession is reviewed as a question of law, consistent with the later reasoning of Ornelas.¹⁸ The pairing shows that a constitutional label does not settle the standard. Two constitutional questions landed on opposite sides of the line.

Clear-error review also reaches many questions that carry legal elements. The Supreme Court has applied it to a finding of negligence, to the question whether a payment to a taxpayer was a gift, and to the question whether an employer acted with discriminatory intent.¹⁹ The clearly erroneous test ordinarily applies whenever a finding rests on the fact-finder's experience with the mainsprings of human conduct.²⁰ The deference stops short of legal error: a finding on a mixed question must be set aside if an error of law impaired the judge's judgment, but it stands when it applies a correct principle of law to the facts.²¹ Given this array of precedent, attempts to draw fine lines between fact and law in order to widen review of determinations the trial court can make better seem doomed to fail.²²

Assigning a question to judge or jury at trial does not settle the standard of review either. In Ram Construction Co. v. American States Insurance Co., the Third Circuit reasoned that the traditional division of interpretive functions between judge and jury rests on outdated common law policy considerations, such as jury illiteracy and lack of respect for writings, together with the continuing judicial desire for uniformity. In contract interpretation, therefore, assignment to judge or jury does not of itself determine the standard of review, and a trial court's interpretation of a factual matter is reviewed for clear error rather than plenary.²³

The Court draws the same comparison of institutional competence elsewhere. In Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., it applied the clearly erroneous standard to facts, while in Cooper Industries, Inc. v. Leatherman Tool Group, Inc., it held that a punitive damages award is not a finding of fact and can be reviewed de novo. The Court called the setting of punitive damages a "finely tuned exercise" in which "juries do not normally engage," and it was persuaded by the superior "institutional competence" of judges.²⁴

Exam Analysis Steps

Because the anchors pull in two directions, sort a mixed question in a fixed order that separates the layers before choosing a standard:

Isolate the historical facts the judge found and apply clear-error review to them.
Identify the legal standard the judge applied, and review any dispute about what that standard requires de novo.
Name what remains, which is whether the facts as found satisfy the standard.
Ask whether answering that remainder chiefly elaborates a legal standard in a way other cases will use, which points toward de novo review, or chiefly weighs case-specific evidence and credibility, which points toward deference.
Check whether the finding carries constitutional significance that calls for independent review of the record, as the next subsection explains.

The Constitutional-Fact Exception

If a fact carries constitutional significance, such as actual malice in a defamation case, independent review is required to preserve constitutional rights.²⁶ Bose Corp. v. Consumers Union of United States, Inc. illustrates the exception. The question there was whether actual malice, the standard of New York Times Co. v. Sullivan, was present.²⁷ The trial judge found for the plaintiff, the court of appeals reversed after ruling that the First Amendment required independent review of the judge's findings, and the Supreme Court agreed.²⁸ A judgment resting on a mistaken actual-malice finding would intrude on First Amendment rights, so the appellate court must examine the entire trial record independently to make certain it does not.²⁹

The exception does not displace Rule 52(a). The rule does not forbid all review and commands only that due deference be given to the trial judge, and the court of appeals in Bose honored that command by deferring to the district judge on the credibility of the witnesses.³⁰ Independent review thus covers the ultimate actual-malice conclusion, while the credibility calls beneath it stay with the trial judge. The Court added that the presumption of deference to judicial fact findings has varied in intensity with the circumstances of the case.³¹ A constitutional label alone does not trigger the exception. Miller v. Fenton left juror bias, which is closely tied to the right to an impartial jury, under clear-error review, and in other areas of constitutional law the Court has not insisted on de novo review.³² A bench-trial judgment against a speaker, challenged on the actual-malice finding, therefore calls for independent review with deference on credibility, and a fact pattern that merely mentions a constitutional guarantee does not.