On Tuesday, I began serializing Criminal Law 2.0, a new article by Judge Alex Kozinski — for whom I clerked 20 years ago, who is one of our nation’s most prominent appellate judges and has long been seen as on balance a libertarian-ish conservative (appointed by President Reagan). The introduction gave 12 reasons to worry about our criminal justice system; this post discusses wrongful convictions and excessively long sentences; and I’ll post other parts of the article in the days to come. I’ve added some paragraph breaks and removed the footnotes (which are available in the PDF version), but otherwise this is as Judge Kozinski wrote it:

What I have listed above are some of the reasons to doubt that our criminal justice system is fundamentally just. This is not meant to be an exhaustive list, nor is it clear that all of these uncertainties would, on closer examination, be resolved against the current system. But there are enough doubts on a broad range of subjects touching intimately on the integrity of the system that we should be concerned.

The National Registry of Exonerations has recorded 1576 exonerations in the United States since 1989. [Footnote: Of these exonerated individuals, 112 were sentenced to death, and 265 spent more than 20 years behind bars. The average time spent in prison was 9 years, with 40 percent imprisoned for more than 10 years. 80 percent were convicted by juries, 7 percent by judges and 12 percent pleaded guilty. 25 percent were exonerated at least in part by DNA evidence. The following factors contributed to their exonerations: mistaken witness identification (34% of exonerations); perjury or false accusation (55%); false confession (13%), defective or misleading forensic evidence (22%) and official misconduct (46%). Cases often involve more than one of these factors.] The year 2014 alone saw a record high of 125 exonerations, up from 91 the year before, and there is reason to believe the trend will continue.

[Footnote: 47 of the 125 exonerations in 2014 involved defendants who had pleaded guilty. The exonerations were mostly concentrated in California, Texas, New York and Illinois. According to an in-depth study by San Francisco magazine, between 1989 and 2004, more than 200 California inmates were freed after courts found they were wrongfully convicted.

Their stories are the stuff of nightmares. Take, for example, Gloria Killian, a 30-something former law student who had signed up to do freelance detective work for a coin shop owner. One day, an elderly coin collector was robbed and killed, and someone called the Sacramento police accusing “a law student named Gloria” of being involved. Nothing came of this accusation until a year later, when a repeat felon named Gary Masse was convicted of the murder and sentenced to life without parole. He named Killian as his accomplice and claimed she masterminded the robbery. The accusation stuck, and Killian was convicted of conspiracy and murder, and sentenced to 32 years to life. Masse got his sentence reduced to 25 years.

Over a decade later, a new investigation uncovered evidence that Masse had entered into an agreement with prosecutors to testify against Killian in exchange for leniency — a fact never disclosed to the defense. The investigation also turned up a letter Masse sent to the DA soon after Killian was sentenced, in which he wrote, “I lied my ass off for you people.” A panel of our court reversed Killian’s conviction in 2002, at which point she had already lost 16 years of her life to prison. The prosecutor walked away with an admonishment from the California State Bar.]

Certainly the significant number of inmates freed in recent years as the result of various innocence projects and especially as a result of DNA testing in cases where the convictions were obtained in the pre-DNA era, should cause us to question whether the current system is performing as effectively as we’ve been led to believe. It’s no answer to say that the exonerees make up only a minuscule portion of those convicted. For every exonerated convict, there may be dozens who are innocent but cannot prove it.

We can be reasonably confident that the system reaches the correct result in most cases, but that is not the test. Rather, we must start by asking how confident we are that every one of the 2.2 million people in prisons and jails across the country are in fact guilty. And if we can’t be sure, then what is an acceptable error rate? How many innocent lives and families are we willing to sacrifice in order to have a workable criminal justice system?

If we put the acceptable error rate at 5 percent, this would mean something like 110,000 innocent people incarcerated across the country. A 1 percent error rate would mean 22,000 innocent people — more or less the population of Nogales, Arizona — wrongly imprisoned. These numbers may seem tolerable unless, of course, you, your friend or loved one draws the short straw.

Do we know how these numbers compare to the actual error rate? We have no idea. What we have is faith that our system works very well and the errors, when they are revealed, are rare exceptions.

Much hinges on retaining this belief: our self image as Americans; the pride of countless judges and lawyers; the idea that we live in a just society; confidence in the power of reason and logic; the certainty that none of us or our loved ones will face the unimaginable nightmare of unjust imprisonment or execution; belief in the incomparable integrity and accuracy of our system of justice; faith that we have transcended medieval methods of conviction and punishment so that only those who are guilty are punished, and their punishment is humane and proportionate. There are, we are convinced, no Edmond Dantèses and no Château d’Ifs in America today.

But what do we really know? We must reject out of hand the idea that the number of actual exonerations represents all of those who have been wrongly convicted. Convicted prisoners wishing to gain release on grounds of innocence face formidable hurdles.

To begin with, they are in prison and thus unable to pursue leads the police might have missed; they have to rely on someone on the outside to do it, and that’s often difficult or impossible to accomplish. A prisoner’s access even to his counsel is severely restricted once he’s incarcerated. A loyal friend or relative might do it, but friends and even relatives often abandon defendants who are convicted, no matter how much they may protest their innocence.

A few prisoners may obtain the help of an innocence project, but the work is labor-intensive, resources are scarce and manpower is limited, so innocence projects engage in triage, focusing on the most promising cases. Of course, it’s often difficult to tell whether a case is promising until you look closely at it, so a promising case can easily be overlooked.

But the biggest problem is that new evidence is hard — and often impossible — to find. If it’s a physical crime, police secure the crime scene and seize anything that looks like it could be relevant. The chance of going back years later and picking up new clues is vanishingly small. The trick then is to get whatever evidence the police have, assuming they didn’t destroy it or release it once it was clear that it wouldn’t be used at trial. If the crime is non-physical, such as fraud, child pornography or computer hacking, the police seize all the relevant computers, hard drives and paper records (including any exculpatory evidence the suspect may have there) and may well discard them after the conviction becomes final.

For a brief period, DNA evidence helped exculpate defendants who were convicted in the pre-DNA era, but DNA often cannot help identify the true perpetrator because no sample of DNA was found or collected from the crime scene. A prisoner has to be exceedingly lucky to collect enough evidence to prove his innocence; most cannot hope to meet that standard. I think it’s fair to assume — though there is no way of knowing — that the number of exculpations in recent years understates the actual number of innocent prisoners by an order, and probably two orders, of magnitude.

Wrongful convictions are not merely unjust to the prisoner and his family, they often result in another injustice or series of injustices: When an innocent man is convicted, a guilty man is left free and emboldened to victimize others.

The Michael Morton case provides a good example. Morton was convicted in 1987 for the 1986 beating murder of his wife. Twenty-five years later he was exonerated when DNA evidence pointed to another man, Mark Norwood, who was eventually convicted of killing Mrs. Morton. However, Norwood has now been charged with the similar beating-death of another woman, Debra Baker; that murder was committed a year after Morton was convicted of his wife’s murder. Norwood is awaiting trial for the Baker murder. Had police continued to investigate the Morton murder instead of shutting down the investigation once they decided that Michael Morton was the culprit, Debra Baker might still be alive.

There’s another question the answer to which we must be reasonably confident: Of those that are guilty, can we be sure that substantial numbers are not spending far more time behind bars than is justified? The question of how much time prisoners spend behind bars is no less important than that of whether only the guilty are being locked up. The ability to pick up the threads of one’s life after three to five years in prison is quite different than after fifteen, twenty or twenty-five years.

Aside from the brutalizing and often dehumanizing effect of long-term imprisonment, an inmate who is released after a lengthy prison term simply does not return to the same world he left behind: Children grow up; spouses find other partners; friends and acquaintances forget; job prospects disappear; life and work skills deteriorate. Shorter sentences also reduce the consequences of wrongful convictions. While no time behind bars can be justified for someone who is innocent, we must be especially careful before imposing life-altering sentences.

By any measure, the United States leads the world in incarceration. In absolute terms, it has more prisoners than any other country. With just 5 percent of the world’s population, we have almost a quarter of the world’s prisoners. China, with nearly 20 percent of the world’s population, has 16 percent of the world’s prisoners. Incarceration rates were not always this high in the United States. For the first three-quarters of the twentieth century, the rate was well under 250 per 100,000. Then, starting around 1980, incarceration rates started rising sharply with the advent of the war on drugs, mandatory minimum sentences and three-strikes laws.

The difference in incarceration rates cannot be explained by higher crime rates in the United States. Crime rates here are roughly equivalent to Canada and in many categories lower than other countries. And the crime rate has been dropping in the United States, as in many other industrialized nations. Yet, U.S. sentences are vastly, shockingly longer than just about anywhere else in the world.

There are reasons to doubt whether the length of prison sentences in this country is just. Although elected officials, regardless of party affiliation and political leaning, seem to favor Draconian sentences, and the public seems to support them in the abstract, it’s unclear how much popular support they enjoy when applied to individual defendants.

U.S. District Judge James Gwin of Ohio reported on an informal study he conducted involving 22 jury trials. He asked the jurors who had convicted the defendant to write down what each thought was the appropriate sentence. Judge Gwin found that the jurors’ recommended sentences were significantly lower than those recommended by the Sentencing Guidelines: “In several cases, the recommended median Guidelines range was more than 10 times greater than the median jurors’ recommendation. Averaged over more than 20 cases, jurors recommended sentences that were 37% of the minimum Guidelines recommended sentences and 22% of the median Guidelines recommended sentences.”

Whether we are incarcerating the right people and for an appropriate length of time are important questions to which we do not have very good answers. We are taught early in our schooling that the criminal justice system is tilted heavily in favor of defendants, resolving all doubts in their favor. Movies and television reinforce this idea with countless stories of dedicated police and prosecutors bringing guilty people to justice, or of acquittals of the innocent because of the efforts of a dedicated lawyer or investigator. Our educational system spends little time pondering the fate of those unjustly convicted or those wasting their lives behind bars because of a punishment that far outstrips whatever evil they were convicted of committing. No Dumas, Hugo or Zola has risen among us to foment public sympathy to the plight of the unjustly imprisoned.