If a person is convicted of a crime in state court, they can petition a federal court for relief by claiming that their conviction violates the federal constitution. However, the federal court does not get to consider that claim afresh. Rather, the Antiterrorism and Effective Death Penalty Act of 1996 requires federal courts to defer to the state court’s ruling on the federal constitutional claim so long as it was not “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States” or “based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.”
Twice this year, the Supreme Court has summarily reversed a federal court of appeals’ grant of such relief for failing to give adequate deference to state courts. In January, it was the U.S. Court of Appeals for the 4th Circuit’s decision in Klein v. Martin, where the court asserted that the circuit court’s decision was “not consistent with the deference that AEDPA requires.” Then in June, it was the U.S. Court of Appeals for the 2nd Circuit’s turn in McCarthy v. Hernandez, where the court pointed back to Klein to observe that, in its view, sometimes lower courts have “strayed too far from the modest role that AEDPA prescribes,” and reiterated that when they do, it will “summarily reverse[] their decisions.”
But telling lower courts to defer is easier than telling them what they are supposed to defer to. Thirty years after AEDPA’s enactment, what counts as “clearly established Federal law” remains remarkably unclear.
If that phrase sounds familiar, it should. It is a close cousin of the “clearly established law” requirement in qualified immunity; a doctrine criticized from every direction for its incoherence, including for how “inherently amorphous and unprincipled” the clearly established law standard is. As U.S. Court of Appeals for the 11th Circuit Judge Charles Wilson lamented, “there is remarkably little consensus among the United States circuit courts concerning how to interpret the term ‘clearly established.’”
Now imagine qualified immunity, only harder. That is the problem federal courts face under AEDPA.
First, the universe of relevant precedent is dramatically smaller. Unlike in qualified immunity, where it is routinely assumed that circuit precedent can create clearly established law, federal courts of appeals cannot clearly establish law for purposes of habeas review. AEDPA expressly limits the inquiry to federal law “as determined by the Supreme Court.” And the Supreme Court decides only a tiny fraction of the cases presented to it (with the number seeming to dwindle by the term), leaving countless scenarios unaddressed by the only body whose decisions count.
Second, the court has given lower courts conflicting instructions about how specifically its precedent must speak. In its 2022 decision Brown v. Davenport, it warned that its “holdings that speak only at a high level of generality” cannot supply a ground for habeas relief. Yet less than three years later, in Andrew v. White, the court said that “[g]eneral legal principles can constitute clearly established law for purposes of AEDPA so long as they are holdings of this Court.” It’s hard to reconcile these two ideas.
Indeed, lower-court judges have been begging for guidance since AEDPA was enacted. More than 25 years ago, U.S. Court of Appeals for the 3rd Circuit Judge Theodore McKee decried “the analytical minefield lurking beneath the surface of AEDPA,” including the “unresolved issue” of “the level of specificity the Supreme Court must use in fashioning a rule in order for it to be applied as ‘clearly established Federal law.’” The U.S. Court of Appeals for the 9th Circuit lodged a similar complaint, noting that the “Supreme Court has not defined ‘clearly established’ in the AEDPA context.”
A quarter-century of decisions has not supplied a satisfying answer. The court has said that its precedents must “squarely address[] the issue,” and that lower courts cannot “sharpen a general principle of Supreme Court jurisprudence into a specific legal rule.” But it has also said that AEDPA does not require courts “to wait for some nearly identical factual pattern before a legal rule must be applied.” And it has even acknowledged – without guidance – that “the difference between applying a rule and extending it is not always clear.”
That concession captures the problem. Every legal rule can be described at different levels of generality, and a judge confronting a new set of facts has to decide whether they are applying an existing holding or extending it. At the same time, the consequences of guessing wrong are not evenly distributed. Define Supreme Court precedent too narrowly, and a prisoner loses, usually without further review. Define it too broadly, grant relief, and a court of appeals invites a summary reversal. The court has already told lower-court judges which kind of error it’s watching for.
The Supreme Court could fix this. Instead of summarily reversing grants of habeas relief, the court could explain how Brown and Andrew fit together and give substance to the clearly established federal-law standard rather than leaving lower courts to guess.
Until it does, the cost of the ambiguity falls where it has always fallen: on state prisoners who claim to be languishing in prison in violation of the federal constitution and the federal judges who agree. For a statute built around clearly established law, and with the consequences so dire, that is an awful lot left unsettled.