In the context of the debate over finality in habeas challenges to convictions, Judge Richard Kopf let on to a little inside baseball in the backroom of the federal courthouse.

Habeas corpus cases attacking the decisions of state courts swallow up a large amount of federal judicial time and resources. Each year, more than 18,000 cases, or one out of every 14 civil cases filed in the federal district courts, are filed by state prisoners seeking habeas corpus relief, and more than 6,000 of these cases reach the courts of appeals.

While these are habeas petitions, the key here is the numbers.  Combine the volume of petitions with the quality of their thought and writing, the occasionally often insane arguments and demands, and it’s a huge resource suck, not to mention a substantial burden to the staff who have to go through them.

At this moment, some of you will feel badly for the pro se clerk, while most will say, “suck it up, Judge Teacup,  No one forced you to take the job. That’s the price of having lawyers laugh at your lame jokes.”

But Congress, incapable of achieving consensus on much of anything, felt the judges’ Article III pain.  On this, they could agree, and a hero from Utah championed their cause.

The P.L.R.A., passed by Congress in 1996, was designed to reduce the number of lawsuits brought by inmates against prisons. In 1995, Senator Orrin Hatch, a Republican from Utah, argued for the P.L.R.A.’s passage by pointing out that prisoners filed fifteen per cent of all civil cases initiated the previous year in federal court, totalling [sic] more than thirty-nine thousand lawsuits, most alleging “cruel and unusual” prison conditions. Hatch promised that the P.L.R.A. would “quickly identify the viable prisoner claims and weed out the meritless chaff.” As passed, the law requires prisoners who believe their rights have been violated to first submit a grievance form to their prison’s administration, and, if that grievance is rejected, to appeal the decision within their state’s corrections system as high as the process allows. Only after those steps have been taken can prisoners file suit in an actual court. This type of provision is known as an “exhaustion requirement.”

The Prison Litigation Reform Act of 1996, a Clinton-era solution, was enacted to eliminate the frivolous, overly burdensome deluge of prisoner lawsuits.  The trick was something that makes infinite sense to lawyers and legislators, exhaustion of administrative remedies. If there is a mechanism to address whatever grievance belies a prisoner’s suit, why not require the prisoner to use it before adding to the burden on federal courts?

But like all simple and sweet solutions, it didn’t work out exactly as promised.

Before 1996, courts applied an exhaustion requirement only if a state’s grievance process met certain high standards of fairness outlined by the Justice Department. The P.L.R.A. eliminated those standards. There are currently no regulations governing prison grievance processes, and, in the two decades since the law’s passage, many prisons’ procedures have become so onerous and convoluted—“Kafkaesque,” in the words of one federal judge—that inmates whose rights have been violated are watching their cases slip through the cracks.

But is that it? The administrative procedures are onerous and convoluted? Why should they be any better in prison than they are for anyone else? Try applying for a barber’s license, and these applicants aren’t convicted criminals.

Margo Schlanger, a professor of civil-rights law at the University of Michigan, who is widely considered the leading authority on the P.L.R.A., maintains a database of grievance policies from across the country and keeps track of individual cases. She has come across cases in which inmates have had their grievances rejected for writing in red ink, for writing on the back of a form, and for attaching medical records to their submissions. If inmates miss a filing deadline (the shortest, in Michigan, is two business days) or trip over some other procedural hurdle, they have lost their right to relief through both the grievance process and the federal courts. “The preservation of prisoners’ civil rights now depends on their ability to dot ‘i’s and cross ‘t’s,” Schlanger told me recently. “And it turns out they’re not so good at that.”

The government, as explained in the 62 pages required by the Paperwork Reduction Act to be attached to all one-page forms, may lack the bureaucratic magic to make much of anything work, but they do a bang-up job of creating nit-picky requirements that create absurd barriers to accomplishing much of anything. Credit where credit is due.

One might contend, however, that since prisoners have tons of time on their hands, and there are always some who make their living as jailhouse lawyers, studying the nooks and crannies of administrative processes, that if anyone would be knowledgeable about how to navigate these Rube Goldberg administrative machines created to make it hard, if not impossible, to exhaust remedies, it would be prisoners.

Except there are other factors at play.

Around half of all prisoners in the United States have some sort of mental illness, and a similar proportion has only basic literacy skills (at best), but courts have frequently ruled that neither mental illness nor illiteracy excuses an inmate from the exhaustion requirement of the P.L.R.A. Juveniles are not exempt, either. In one case from the early two-thousands, a fifteen-year-old inmate at a juvenile facility, in Indiana, filed a lawsuit alleging that he had been repeatedly raped and beaten by other inmates, who had been egged on by guards. The lawsuit was dismissed, in 2005, because the inmate had not filed a grievance, even though his mother had previously contacted prison officials, and even the governor’s office, in an attempt to stop the abuse of her son.

And therein lies the core of the problem. While some suits by prisoners complain that the food sucks and there isn’t enough of it, there are also the rapes and beatings. Contrary to the facile assumption that prisoners all know how to game the system, most are surprisingly clueless at how things work. It’s not that they don’t try, but many lack the capacity to figure it out, whether because of mental illness or intellectual and educational deficits, despite their best efforts or the legitimacy of their complaint.

But for the lack of a dot over an “i,” a prisoner gets raped? That’s the trade-off for federal judicial work-life balance. The question isn’t whether prisoner suits are a burden. They are. The question is whether it’s worth enduring that burden to end rapes and beatings in prison. It is.