Standards of Review Explained

Every appeal has two questions: was the lower decisionmaker wrong, and does being wrong matter enough to reverse? The standard of review answers the second. Under de novo review the appellate court decides the question fresh and owes the trial judge nothing. Under abuse of discretion the trial judge can be wrong in the appellate court's own view and still be affirmed. Practitioners say that the standard of review is the appeal, and that is close to true: a large share of affirmances rest on deference rather than agreement.

Federal Rule of Appellate Procedure 28(a)(8)(B) requires the appellant's brief to state, for each issue, the applicable standard of review and where the issue was preserved. Judges read that sentence carefully. On an exam, identifying the standard before you argue the merits signals that you understand the procedural posture, which is often exactly what the question is testing.

De novo (sometimes "plenary") review applies to questions of law: statutory interpretation, constitutional questions, the elements of a claim, contract interpretation where no extrinsic evidence is needed, and the legal sufficiency of a complaint. It also applies to a federal court's determination of state law, which the Supreme Court held in Salve Regina College v. Russell, 499 U.S. 225 (1991), gets no deference on appeal.

Clear error governs a district judge's findings of fact in a bench trial. Federal Rule of Civil Procedure 52(a)(6) states the rule, and Anderson v. City of Bessemer City, 470 U.S. 564 (1985), supplies the test: reversal requires the reviewing court to be left with the definite and firm conviction that a mistake has been committed. Where there are two permissible views of the evidence, the fact-finder's choice cannot be clearly erroneous, and credibility findings based on live testimony are close to unreviewable. On review of agency fact-finding, the analogous standard is substantial evidence—more than a scintilla, meaning such relevant evidence as a reasonable mind might accept as adequate; Universal Camera Corp. v. NLRB, 340 U.S. 474 (1951), requires the court to consider the whole record, including evidence cutting against the agency.

Abuse of discretion covers case-management and judgment calls: evidentiary rulings, including admission of expert testimony under General Electric Co. v. Joiner, 522 U.S. 136 (1997); Rule 11 sanctions under Cooter & Gell v. Hartmarx Corp., 496 U.S. 384 (1990); discovery orders; continuances; class certification; and the substantive reasonableness of a federal sentence under Gall v. United States, 552 U.S. 38 (2007). Do not treat abuse of discretion as unreviewable. A court that applies the wrong legal rule abuses its discretion by definition, and a court that rests on a clearly erroneous fact finding does too. Attacking the legal premise inside a discretionary ruling is usually the strongest available move.

A mixed question asks whether the established facts satisfy a legal standard—was the search supported by probable cause, was the transaction at arm's length, was counsel's performance deficient. Courts do not have a single answer here, and the split is real. The Supreme Court's approach in U.S. Bank N.A. v. Village at Lakeridge (2018) is functional: ask whether resolving the question requires expounding the law or instead requires marshaling and weighing case-specific facts. Law-heavy mixed questions get de novo review; fact-heavy ones get clear error.

Some mixed questions have fixed answers by precedent. Reasonable suspicion and probable cause determinations are reviewed de novo, with underlying historical facts reviewed for clear error, under Ornelas v. United States, 517 U.S. 690 (1996). Whether employment discrimination occurred is a pure fact question reviewed for clear error under Pullman-Standard v. Swint, 456 U.S. 273 (1982). First Amendment cases carry an independent-review overlay: Bose Corp. v. Consumers Union, 466 U.S. 485 (1984), requires appellate courts to review the actual-malice determination independently. When you cannot find controlling authority, argue the functional test and say why your issue is law-like or fact-like.

Preservation changes the standard. Under Federal Rule of Criminal Procedure 52(b) and United States v. Olano, 507 U.S. 725 (1993), unpreserved error is reviewed for plain error: there must be (1) error, (2) that is clear or obvious, (3) that affected substantial rights—normally meaning the appellant bears the burden of showing prejudice—and (4) the court must then decide whether to exercise discretion to correct it because the error seriously affects the fairness, integrity, or public reputation of judicial proceedings. Puckett v. United States, 556 U.S. 129 (2009), reaffirms that this is a demanding, discretionary test.

Preserved error still requires a harm analysis. Nonconstitutional error in federal criminal cases is harmless unless it had substantial and injurious effect on the verdict, the Kotteakos v. United States, 328 U.S. 750 (1946), formulation. Preserved constitutional error requires the government to show harmlessness beyond a reasonable doubt under Chapman v. California, 386 U.S. 18 (1967), unless the error is structural—complete denial of counsel, biased judge, denial of public trial—in which case reversal is automatic. Neder v. United States, 527 U.S. 1 (1999), confirms most constitutional trial errors are subject to Chapman rather than automatic reversal.

Motions have their own "standards" that are really burdens, and students conflate them constantly. On a Rule 12(b)(6) motion the court accepts well-pleaded factual allegations as true, disregards legal conclusions, and asks whether the complaint states a plausible claim under Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007), and Ashcroft v. Iqbal, 556 U.S. 662 (2009). On summary judgment under Rule 56, the court asks whether a reasonable jury could return a verdict for the nonmovant, drawing all reasonable inferences in that party's favor and applying the governing evidentiary burden of proof; see Anderson v. Liberty Lobby, Inc., 477 U.S. 242 (1986), and Celotex Corp. v. Catrett, 477 U.S. 317 (1986). Judgment as a matter of law under Rule 50 uses the same substantive test, applied after evidence is in; Reeves v. Sanderson Plumbing Products, Inc., 530 U.S. 133 (2000), instructs courts to review the whole record but disregard evidence favorable to the movant that the jury was not required to believe.

The appellate standard for each of these is de novo, because each is a legal question. In criminal cases, sufficiency of the evidence is reviewed under Jackson v. Virginia, 443 U.S. 307 (1979): whether any rational trier of fact could have found the essential elements beyond a reasonable doubt, viewing the evidence in the light most favorable to the prosecution.

Rational basis review upholds a classification if it is rationally related to any conceivable legitimate government interest; the challenger bears the burden and the legislature need not have articulated the rationale, per FCC v. Beach Communications, Inc., 508 U.S. 307 (1993). It is nearly always fatal to the challenger, though not always—see City of Cleburne v. Cleburne Living Center, Inc., 473 U.S. 432 (1985), and Romer v. Evans, 517 U.S. 620 (1996).

Intermediate scrutiny requires a substantial relationship to an important government interest. It governs sex classifications under Craig v. Boren, 429 U.S. 190 (1976), tightened by the "exceedingly persuasive justification" language of United States v. Virginia, 518 U.S. 515 (1996), and content-neutral time, place, and manner regulations under Ward v. Rock Against Racism, 491 U.S. 781 (1989). Strict scrutiny requires narrow tailoring to a compelling interest and applies to racial classifications, Adarand Constructors, Inc. v. Peña, 515 U.S. 200 (1995), and content-based speech restrictions, Reed v. Town of Gilbert, 576 U.S. 155 (2015). Note that the tiers are not universal: the Second Amendment now uses a text-and-historical-tradition test rather than means-end scrutiny after New York State Rifle & Pistol Ass'n v. Bruen (2022), and the dormant Commerce Clause uses the Pike balancing inquiry.

Loper Bright Enterprises v. Raimondo (2024) overruled Chevron. Courts now exercise independent judgment in deciding whether an agency has acted within its statutory authority, using ordinary tools of construction under the APA's command that reviewing courts decide all questions of law. Two things survive: Skidmore v. Swift & Co., 323 U.S. 134 (1944), respect, under which an agency's interpretation persuades to the extent of its thoroughness, consistency, and reasoning; and genuine statutory delegations of discretion to the agency, which courts police for reasonableness. Prior holdings that relied on Chevron retain stare decisis effect.

Policy choices and rulemaking still get arbitrary-and-capricious review under 5 U.S.C. § 706(2)(A), as elaborated in Motor Vehicle Manufacturers Ass'n v. State Farm Mutual Automobile Insurance Co., 463 U.S. 29 (1983): the agency must examine relevant data and articulate a rational connection between facts and choice. An agency's reading of its own ambiguous regulation may get Auer deference, but only after the narrowing conditions of Kisor v. Wilkie (2019)—genuine ambiguity after exhausting interpretive tools, a reasonable reading, an authoritative position implicating agency expertise and reflecting fair and considered judgment.

In federal habeas, 28 U.S.C. § 2254(d) bars relief on claims adjudicated on the merits in state court unless the decision was contrary to, or an unreasonable application of, clearly established Supreme Court holdings, or rested on an unreasonable determination of the facts. Harrington v. Richter, 562 U.S. 86 (2011), makes clear this means no fairminded jurist could agree with the state court. Paired with a deferential underlying standard like Strickland v. Washington, 466 U.S. 668 (1984), the review is doubly deferential.

Classification varies by circuit and by state. Circuits disagree about whether certain rulings—forum non conveniens dismissals, remittitur, some choice-of-law calls, denial of leave to amend—are reviewed de novo or for abuse of discretion, and state appellate courts use additional labels federal students never see, such as "manifest weight of the evidence," "palpable abuse," or "competent substantial evidence." What drives the variation is a policy judgment about institutional competence: the more the ruling depends on watching the trial unfold, the more deference; the more it establishes a rule that should apply uniformly across cases, the less.

Two practical habits. First, always research the standard in the specific jurisdiction rather than assuming; a five-minute search often surfaces a controlling circuit case. Second, if the standard is unfavorable, look for a de novo sub-issue inside it—the legal premise of a discretionary ruling, or the interpretation of the rule the trial court applied—because that is where deferential review has a seam.