At the Mountains of Madness

If I successfully "day hike" Mt. Whitney (the tallest mountain in the lower 48 States), should I be treated as if I had also conquered Mt. Kilimanjaro (the tallest mountain in Africa and nearly 5,000 feet taller than Mt. Whitney)?

This is the question that lies at the heart of Brown v. Davenport, 20-826, on which the U.S. Supreme Court recently granted certiorari.

Sort of.

Davenport arises in the context of federal habeas corpus review of a state conviction, and asks whether a petitioner's success in showing that an error of federal constitutional law was prejudicial under one standard equates with success in satisfying a second "prejudice" standard.

To provide some context, a person charged with a crime and convicted at trial in state court has the right to judicial review of the conviction. That review typically starts with the direct appeal of the conviction in state court, and, if claiming violations of federal law, is followed by a petition for a writ of habeas corpus filed in federal court (although sometimes state habeas review is sandwiched in between).

Throughout the entire process of post-conviction review, it is usually not enough for a convicted defendant to show that an error occurred during the trial. To be sure, there is a small subset of so-called "structural errors" that affect the very "framework within which the trial proceeds," Arizona v. Fulminante, 499 U.S. 279, 310 (1991); such structural errors -- which include the outright denial of counsel or the failure to instruct the jury on the concept of "reasonable doubt" -- are automatically reversible. But for every other type of error at trial, the defendant must show both that there was error and that the error mattered to the outcome. That is because the Constitution entitles the "criminal defendant ... to a fair trial, not a perfect one." Delaware v. Van Arsdall, 475 U.S. 673, 681 (1986).

How much an error must matter before the defendant is entitled to relief is a function of (1) the type of error, and (2) at what point in post-conviction review the error is asserted.

When a defendant establishes that some aspect of the trial in state court violated the federal Constitution (as opposed to violating a federal statue, a state constitution, or a state statute), the defendant seeking relief on direct appeal is entitled to relief unless the People establish that the error was harmless beyond a reasonable doubt. Chapman v. California, 386 U.S. 18, 24 (1967). Thus, state court decisions affirming a conviction notwithstanding error usually contain a finding that the error was harmless beyond a reasonable doubt.

A defendant seeking relief for the very same error in federal habeas review has to make a greater showing.

In Brecht v. Abrahamson, 507 U.S. 619 (1993), the Supreme Court held that a defendant was entitled to federal habeas relief only if the court evaluating the error on habeas finds that the federal constitutional error "'had substantial and injurious effect or influence in determining the jury's verdict'" or, as Brecht put it, that it "resulted in 'actual prejudice.'" Id. at 637. This standard was adopted specifically because it was "less onerous" on the sanctity of state convictions than the Chapman standard. Ibid. Brecht adopted a more conviction-friendly standard on federal habeas corpus due to its view that (1) the "principal avenue" for review of state convictions was the direct appeal in state court (and not federal habeas review), which makes sense because "state courts often occupy a superior vantage point from which to evaluate the effect of trial error," id. at 633, 636, and (2) state convictions are entitled to deference from federal courts in light of their finality, the principles of "comity and federalism," and the "arduous task" of retrial in the absence of actual prejudice, id. 632, 635; Calderon v. Coleman, 525 U.S. 141, 146 (1998).

In 1996, Congress enacted the Anti-Terrorism and Effective Death Penalty Act. Among other things, AEDPA proclaimed that federal habeas relief "shall not be granted" unless the state court's decision evaluating federal constitutional error "was contrary to, or involved an unreasonable application of, clearly established Federal law." 28 U.S.C. Section 2254(d)(1). Because AEDPA's standard authorizes relief only if the state court's decision was both incorrect and unreasonable, Harrington v. Richter, 562 U.S. 86, 101 (2011), and because this standard applies to a state court's harmlessness determination under Chapman, Mitchell v. Esparza, 540 U.S. 12, 18 (2003), AEDPA provides that a defendant is entitled to federal habeas relief only if the state court's finding that a federal constitutional error was harmless beyond a reasonable doubt was incorrect and unreasonable.

How do the Brecht and AEDPA tests for evaluating the prejudicial effect of a federal constitutional error at trial fit together?

The slightly embellished facts of Davenport may help illustrate. Defendant and a friend are road-tripping it from Arkham to Innsmouth. Mid-drive, they get into a tussle in the car and defendant stabs and kills the friend. Defendant is charged with murder and argues self-defense. During the trial, the trial court orders defendant to be shackled in a way that is visible to the jury without making the findings required by Deck v. Missouri, 544 U.S. 622 (2005), thereby violating the federal Constitution. Because defendant's character for violence is at issue, it is difficult to say that allowing the jury to see that defendant needed to be restrained in court was harmless beyond a reasonable doubt. But if a state court so concluded, which standard applies on federal habeas -- Brecht or AEDPA's "was-the-Chapman-ruling unreasonable" standard?

The federal courts have split.

One faction holds that our defendant would be entitled to federal habeas relief only if the state court was unreasonable to find no prejudice under Chapman and if the federal habeas court found actual prejudice under Brecht, Johnson v. Lamas, 850 F.3d 119, 133 (3d Cir. 2017); Johnson v. Acevedo, 572 F.3d 398, 403 (7th Cir. 2009); Sifuentes v. Brazelton, 825 F.3d 506, 534 (9th Cir. 2016); Malone v. Carpenter, 911 F.3d 1022, 1030 (10th Cir. 2018), although some courts find the two tests to be interchangeable. Sifuentes at 535; Jensen v. Clements, 800 F.3d 892, 908 (7th Cir. 2015).

Another faction holds that our defendant would be entitled to relief so long as the federal habeas court finds actual prejudice under Brecht. Davenport v. MacLaren, 964 F.3d 448, 454-55 (6th Cir. 2020); Hall v. Haws, 861 F.3d 977, 992 (9th Cir. 2017).

This split can be traced back to language in two prior Supreme Court cases pointing different ways. In Fry v. Pliler, 551 U.S. 112 (2007), the court held that it "makes no sense to require formal application of both tests (AEDPA/Chapman and Brecht)" because the Brecht standard "obviously subsumes" the AEDPA/Chapman standard, such that applying "the Brecht standard" alone suffices. Id. at 120. But in Davis v. Ayala, 576 U.S. 257 (2015), the court clarified that "Fry ... did not hold ... that Brecht somehow abrogates the limitation on federal habeas relief that [AEDPA] plainly sets out" and went on to mandate that "[w]hile a federal habeas court need not 'formal[ly]' apply both Brecht and 'AEDPA/Chapman,' AEDPA nevertheless 'sets forth a precondition to the grant of habeas relief.'" Id. at 268.

Presumably, Davenport will resolve this confusion by determining whether surmounting one mountain (Brecht) dispenses with the need to surmount the other (AEDPA/Chapman).

In doing so, the court may confront two questions.

First, does a federal habeas court's finding of actual prejudice under Brecht necessarily mean that a state court was unreasonable in its Chapman analysis? What if, for instance, our hypothetical jury had overwhelming evidence of the defendant's premeditated plan to kill? On those facts, the jurors' post-trial testimony to seeing shackles might render it unreasonable to find the shackles harmless beyond a reasonable doubt, but does that translate to an actual effect on the verdict? Accord Sifuentes at 535 n.6. Of course, Davenport asks whether there is "daylight" between the two standards in the converse situation.

Second, and even if a finding of actual prejudice under Brecht logically dictates a finding that a state court's "harmless beyond a reasonable doubt" finding was unreasonable, does failing to evaluate the state court's finding pose an affront to the dignitary and federalism interests that justified Brecht's standard in the first place?

The answer may not be easy, but there is little doubt that the justices will love crafting it.