(This is an observation from a very talented appellate litigator at my firm. So as a matter of professional integrity, I can’t claim credit, but for reasons of anonymity, nor can I name her.)
I’ve had the privilege of working on more than a few criminal appeals in my very short time in practice, and arguing one to a federal appellate court. So far, the experience hardly inspires confidence in the criminal justice system. On collateral appeals — like habeas corpus, or state law analogues — the ultimate issue is not the defendant’s actual innocence. Rather, the focus is on whether the defendant received a fair trial and, if not, what can be done to correct that. It is the bedrock of American criminal law that legal guilt is the function of a fair trial; thus, we take very seriously circumstances that would create an inference of procedural unfairness.
Unfortunately, while judges sometimes take the cases seriously on a case-by-case basis, individual dispositions shows they don’t really understand the degree to which isolated errors can become systemic.
In appellate practice, there’s a powerful concept known as “harmless error,” which holds that, even if the court below made some significant error, the appellate court is powerless to reverse unless that error would’ve actually led to a different outcome. Harmlessness matters to an even greater degree in collateral criminal appeals, like habeas corpus: even if the petitioner can identify a serious procedural defect, most of the time, it won’t entitle them to a new trial unless it “probably” would have resulted in the jury acquitting them at the original trial.
Obviously, this requirement serves an important gatekeeping function. We can’t grant trials for every procedural defect. But the problem is that judges tend to err towards harmlessness: on a borderline case, where the prosecutor or defense attorney’s misconduct might have changed the outcome of the petitioner’s trial, most judges will deny the claim.
Perhaps tragic on a case-by-case basis, this anti-petitioner bias invites more error, and therefore risks entrenching a new type of systemic risk. While individual cases are one-off interactions for the petitioner, they’re not one-off interactions for judges and prosecutors. Everytime a prisoner claims prosecutorial misconduct, and a judge passes on it as “harmless error,” the prosecution may fairly conclude that their conduct, no matter how abusive, was acceptable. As the zone of abusive-but-legal behavior expands, the zone of litigatable misconduct contracts, and prosecutors will realize that they can push the envelope more and more. Eventually, conduct that began as unacceptable — but did not constitute reversible error, for whatever reason — will become acceptable, and safeguards on the criminal system erode.
There are a lot of reasons judges tolerate (and encourage) this risk. But under the current Supreme Court, it’s getting worse, not better. Post-Harrington v. Richter, state courts can dispose of collateral appeals without explanation, and the federal courts are essentially required to summarily affirm. Litigators can fight this backslide on a case-by-case basis, but it’s a losing strategy. What we really need is reform, starting with the Anti-Terrorism and Effective Death Penalty Act, to prevent federal habeas courts from becoming rubber stamps.