The Deadline Is the Ruling

Habeas corpus is the oldest protection in the common law, and in the federal system it now turns mostly on arithmetic.

The provision governing successive and untimely petitions imposes a one-year limit, and that limit disposes of more claims than any argument about the underlying conviction.

The limit has a purpose that is easy to defend. Convictions cannot be reopened indefinitely, witnesses fade, and finality is a real value in a system that has to move.

What makes it distinctive is who has to satisfy it and under what conditions.

The petitioner is usually incarcerated, frequently without counsel, and dependent on a prison law library and the postal service to file within a period whose start date is itself contested.

The certificate of appealability adds a second filter that most petitioners never clear.

The clock runs from when the conviction became final, which sounds definite and generates a substantial body of litigation about what that means in a given case.

Even a timely petition denied at the district court cannot be appealed without permission, granted only where reasonable jurists could debate the outcome.

The published decisions contain a great deal of that litigation, and reading a run of them shows courts resolving calendar questions rather than claims.

That standard is applied by the same court system that has already rejected the claim once, which is a structural feature nobody defends and everybody accepts.

Tolling exists and is narrow. Time runs while a state petition is properly filed and pending, and a filing rejected on a procedural ground may not count as properly filed at all.

The combined effect of the deadline, the successive petition bar and the certificate is that very few claims are ever examined on what they allege.

So a petitioner who did the right thing in the wrong form can discover years later that the clock never stopped, which is a harsh result reached by an entirely conventional reading.

A system can defend each filter separately and still owe an account of what the three together leave standing.

Equitable tolling is available for extraordinary circumstances diligently pursued, and the standard is high enough that a lawyer's ordinary negligence does not meet it.

That allocation is the sharpest thing in the area. The consequence of counsel's error falls on the client, who had no way to detect it and no right to different counsel.

The administrative rules governing custody and access do the rest of the work, because a filing deadline interacts with library hours and mail handling in ways no statute contemplates.

A deadline that assumes a filer with a computer and a courier behaves very differently for a filer who has neither, and the statute makes no distinction.

The successive petition bar operates alongside the deadline and is stricter still, requiring permission from an appellate court before a second petition may even be filed.

That permission turns on new evidence or a new rule made retroactive, and both categories are narrower than a lay reader would assume from the words.

The description of what the federal courts handle lists these proceedings among the ordinary civil docket, which is accurate and misleading in equal measure.

They are civil in form and are the only route by which a person detained after conviction can raise a federal constitutional claim, which is not an ordinary civil interest.

None of this argues that the deadline should go. A system without one would face petitions decades old on evidence nobody could test, which serves no one including the petitioner.

It argues that a limit this consequential should be measured from a date a person can identify without counsel, and should stop for any filing made in good faith.

Neither change would reopen the flood the provision was written to prevent. Both would mean the claims that fail, fail on what happened rather than on when the envelope arrived.