Judge Alex Kozinski — for whom I clerked 20 years ago, who is one of our nation’s most prominent appellate judges, and who has long been seen as on balance a libertarianish conservative (appointed by President Reagan) — has recently published an article in the Georgetown Law Journal that says some pretty harsh things about our criminal justice system, and offers some (doubtless controversial) proposals for improving it. You can read the whole article, Criminal Law 2.0, but I also asked Judge Kozinski for permission to serialize the article here, and he graciously agreed. Here is the introduction, which gives 12 reasons to worry about our criminal justice system; 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:
Although we pretend otherwise, much of what we do in the law is guesswork. For example, we like to boast that our criminal justice system is heavily tilted in favor of criminal defendants because we’d rather that ten guilty men go free than an innocent man be convicted. There is reason to doubt it, because very few criminal defendants actually go free after trial.
Does this mean that many guilty men are never charged because the prosecution is daunted by its heavy burden of proof? Or is it because jurors almost always start with a strong presumption that someone wouldn’t be charged with a crime unless the police and the prosecutor were firmly convinced of his guilt? We tell ourselves and the public that it’s the former and not the latter, but we have no way of knowing. They say that any prosecutor worth his salt can get a grand jury to indict a ham sandwich. It may be that a decent prosecutor could get a petit jury to convict a eunuch of rape.
The “ten guilty men” aphorism is just one of many tropes we assimilate long before we become lawyers. How many of us, the author included, were inspired to go to law school after watching Juror #8 turn his colleagues around by sheer force of reason and careful dissection of the evidence? “If that’s what the law’s about, then I want to be a lawyer!” I thought to myself.
But is it? We know very little about this because very few judges, lawyers and law professors have spent significant time as jurors. In fact, much of the so-called wisdom that has been handed down to us about the workings of the legal system, and the criminal process in particular, has been undermined by experience, legal scholarship and common sense. Here are just a few examples:
1. Eyewitnesses are highly reliable. This belief is so much part of our culture that one often hears talk of a “mere” circumstantial case as contrasted to a solid case based on eyewitness testimony. In fact, research shows that eyewitness identifications are highly unreliable, especially where the witness and the perpetrator are of different races. Eyewitness reliability is further compromised when the identification occurs under the stress of a violent crime, an accident or catastrophic event — which pretty much covers all situations where identity is in dispute at trial. In fact, mistaken eyewitness testimony was a factor in more than a third of wrongful conviction cases. Yet, courts have been slow in allowing defendants to present expert evidence on the fallibility of eyewitnesses; many courts still don’t allow it. Few, if any, courts instruct juries on the pitfalls of eyewitness identification or caution them to be skeptical of eyewitness testimony.
2. Fingerprint evidence is foolproof. Not so. Identifying prints that are taken by police using fingerprinting equipment and proper technique may be a relatively simple process, but latent prints left in the field are often smudged and incomplete, and the identification process becomes more art than science. When tested by rigorous scientific methods, fingerprint examiners turn out to have a significant error rate. [Footnote: “[F]orensic fingerprint identification almost never deals in whole fingerprints. Rather, technicians use ‘latent’ fingerprints — invisible impressions that they ‘develop’ using a powder or a chemical developing agent. Latent prints are usually fragmentary, blurred, overlapping, and otherwise distorted. The challenge is to match the latent print to a pristine inked (or, these days, optically scanned) print taken under ideal conditions at the police station.” [Citations omitted.] In United States v. Llera Piaza, 188 F. Supp. 2d 549, 564 (E.D. Pa. 2002), for example, Judge Louis Pollack rejected fingerprint identification expert testimony after concluding that the field of fingerprint identification has failed to systematically test its underlying assumptions and claims of expertise.]
Perhaps the best-known example of such an error occurred in 2004 when the FBI announced that a latent print found on a plastic bag near a Madrid terrorist bombing was “a 100 percent match” to Oregon attorney Brandon Mayfield. The FBI eventually conceded error when Spanish investigators linked the print to someone else.
3. Other types of forensic evidence are scientifically proven and therefore infallible. With the exception of DNA evidence (which has its own issues), what goes for fingerprints goes double and triple for other types of forensic evidence: “Spectrographic voice identification error rates are as high as 63%, depending on the type of voice sample tested. Handwriting error rates average around 40% and sometimes approach 100%. False-positive error rates for bite marks run as high as 64%. Those for microscopic hair comparisons are about 12% (using results of mitochondrial DNA testing as the criterion).”
Other fields of forensic expertise, long accepted by the courts as largely infallible, such as bloodstain pattern identification, foot and tire print identification and ballistics have been the subject of considerable doubt. Judge Nancy Gertner, for example, has expressed skepticism about admitting expert testimony on handwriting, canines, ballistics and arson. She has lamented that while “the Daubert-Kumho standard [for admitting expert witness testimony] does not require the illusory perfection of a television show (CSI, this wasn’t), when liberty hangs in the balance — and, in the case of the defendants facing the death penalty, life itself — the standards should be higher . . . than [those that] have been imposed across the country.”
Some fields of forensic expertise are built on nothing but guesswork and false common sense. Many defendants have been convicted and spent countless years in prison based on evidence by arson experts who were later shown to be little better than witch doctors. Cameron Todd Willingham may have lost his life over it.
4. DNA evidence is infallible. This is true to a point. DNA comparison, when properly conducted by an honest, trained professional will invariably reach the correct result. But the integrity of the result depends on a variety of factors that are, unfortunately, not nearly so foolproof: the evidence must be gathered and preserved so as to avoid contamination; the testing itself must be conducted so that the two samples being compared do not contaminate each other; the examiner must be competent and honest. As numerous scandals involving DNA testing labs have shown, these conditions cannot be taken for granted, and DNA evidence is only as good as the weakest link in the chain.
5. Human memories are reliable. Much of what we do in the courtroom relies on human memory. When a witness is asked to testify about past events, the accuracy of his account depends not only on his initial perception, but on the way the memories are recorded, stored and retrieved. For a very long time, it was believed that stored memories were much like video tape or film — an accurate copy of real-word experience that might fade with the passage of time or other factors, but could not be distorted or embellished.
Science now tells us that this view of human memory is fundamentally flawed. The mind not only distorts and embellishes memories, but a variety of external factors can affect how memories are retrieved and described. In an early study by cognitive psychologist Elizabeth Loftus, people were shown videos of car accidents and then questioned about what they saw. The group asked how fast the cars were going when they “smashed” into each other estimated 6.5 mph faster than the group asked how fast the cars were going when they “hit” each other. A week later, almost a third of those who were asked about the “smash” recalled seeing broken glass, even though there was none. [Footnote: Professor Loftus has shown it is even possible to manufacture false memories. For example, she gave students each a packet describing three real childhood memories and a false one, and told the students that all four memories were real and took place with a close family member. In follow-up interviews asking the students to describe their memories, 7 of 24 students remembered the false event in their packet and some added their own details to that false memory. Loftus was also able to convince participants in another experiment that they’d experienced traumatic events that never happened, such as witnessing drug busts and breaking windows with their hands.]
This finding has troubling implications for criminal trials where witnesses are questioned long and hard by police and prosecutors before the defense gets to do so — if ever. There is thus plenty of opportunity to shape and augment a witness’s memory to bring it into line with the prosecutor’s theory of what happened. Yet with rare exceptions, courts do not permit expert testimony on human memory.
For example, the district judge in the Scooter Libby case denied a defense motion for a memory expert, even though the key issue at trial was whose recollection of a 4-year-old telephone conversation should be believed. At least one member of the jury that convicted Libby lamented the lack of expert testimony on the subject. And a key witness in that case recently suggested in her memoirs that her memory may have been distorted by the prosecutor’s crafty questioning. Given the malleability of human memory, it should come as no surprise that many wrongful convictions have been the result of faulty witness memories, often manipulated by the police or the prosecution.
6. Confessions are infallible because innocent people never confess. We now know that this is not true. Innocent people do confess with surprising regularity. Harsh interrogation tactics, a variant of Stockholm syndrome, the desire to end the ordeal, emotional and financial exhaustion, family considerations and the youth or feeble-mindedness of the suspect can result in remarkably detailed confessions that are later shown to be utterly false.
7. Juries follow instructions. This is a presumption — actually more of a guess — that we’ve elevated to a rule of law. It is, of course, necessary that we do so because it links the jury’s fact-finding process to the law. In fact, however, we know very little about what juries actually do when they decide cases. Do they consider the instructions at all? Do they consider all of the instructions or focus on only some? Do they understand the instructions or are they confused?
We don’t really know. We get occasional glimpses into the operations of juries when they send out questions or someone discloses juror misconduct, and even then the information we get is limited. But we have no convincing reason to believe that jury instructions in fact constrain jury behavior in all or even most cases. And, because the information we get from inside the jury room is so limited and sporadic, experience does little to improve our knowledge. Looking at 100 black boxes is no more informative than looking at one.
8. Prosecutors play fair. The Supreme Court has told us in no uncertain terms that a prosecutor’s duty is to do justice, not merely to obtain a conviction. It has also laid down some specific rules about how prosecutors, and the people who work for them, must behave — principal among them that the prosecution turn over to the defense exculpatory evidence in the possession of the prosecution and the police.
There is reason to doubt that prosecutors comply with these obligations fully. The U.S. Justice Department, for example, takes the position that exculpatory evidence must be produced only if it is material. This puts prosecutors in the position of deciding whether tidbits that could be helpful to the defense are significant enough that a reviewing court will find it to be material, which runs contrary to the philosophy of the Brady/Giglio line of cases and increases the risk that highly exculpatory evidence will be suppressed. Beyond that, we have what I have described elsewhere as an “epidemic of Brady violations abroad in the land,” a phrase that has caused much controversy but brought about little change in the way prosecutors operate in the United States.
9. The prosecution is at a substantial disadvantage because it must prove its case beyond a reasonable doubt. Juries are routinely instructed that the defendant is presumed innocent and the prosecution must prove guilt beyond a reasonable doubt, but we don’t really know whether either of these instructions has an effect on the average juror. Do jurors understand the concept of a presumption? If so, do they understand how a presumption is supposed to operate? Do they assume that the presumption remains in place until it is overcome by persuasive evidence or do they believe it disappears as soon as any actual evidence is presented? We don’t really know.
Nor do we know whether juries really draw a distinction between proof by a preponderance, proof by clear and convincing evidence and proof beyond a reasonable doubt. These levels of proof, which lawyers and judges assume to be hermetically sealed categories, may mean nothing at all in the jury room. My own experience as a juror certainly did nothing to convince me that my fellow jurors understood and appreciated the difference. The issue, rather, seemed to be quite simply: Am I convinced that the defendant is guilty?
Even more troubling are doubts raised by psychological research showing that “whoever makes the first assertion about something has a large advantage over everyone who denies it later.” The tendency is more pronounced for older people than for younger ones, and increases the longer the time-lapse between assertion and denial. So is it better to stand mute rather than deny an accusation? Apparently not, because “when accusations or assertions are met with silence, they are more likely to feel true.”
To the extent this psychological research is applicable to trials, it tends to refute the notion that the prosecution pulls the heavy oar in criminal cases. We believe that it does because we assume juries go about deciding cases by accurately remembering all the testimony and weighing each piece of evidence in a linear fashion, selecting which to believe based on assessment of its credibility or plausibility. The reality may be quite different. It may be that jurors start forming a mental picture of the events in question as soon as they first hear about them from the prosecution witnesses. Later-introduced evidence, even if pointing in the opposite direction, may not be capable of fundamentally altering that picture and may, in fact, reinforce it.
And the effect may be worse the longer the prosecution’s case lasts and, thus, the longer it takes to bring the contrary evidence before the jury. Trials in general, and longer trials in particular, may be heavily loaded in favor of whichever party gets to present its case first — the prosecution in a criminal case and the plaintiff in a civil case.
If this is so, it substantially undermines the notion that we seldom convict an innocent man because guilt must be proven to a sufficient certainty. It may well be that, contrary to instructions, and contrary to their own best intentions, jurors are persuaded of whatever version of events is first presented to them and change their minds only if they are given very strong reasons to the contrary.
10. Police are objective in their investigations. In many ways, this is the bedrock assumption of our criminal justice process. Police investigators have vast discretion about what leads to pursue, which witnesses to interview, what forensic tests to conduct and countless other aspects of the investigation. Police also have a unique opportunity to manufacture or destroy evidence, influence witnesses, extract confessions and otherwise direct the investigation so as to stack the deck against people they believe should be convicted.
[Footnote: One example [of manufactured evidence] is the case of Mark Prentice, who pleaded guilty to assault and robbery only after a New York State Police trooper, David Harding, reported that he had found fingerprints matching Prentice in the victim’s house. A subsequent investigation revealed that New York State Police troopers, including Harding, had falsified fingerprint evidence in at least 30 cases, and Harding admitted to planting evidence in Prentice’s case. Prentice was acquitted after spending six years in prison. Harding was then sentenced to 4.5 years in prison for fabricating evidence. In addition to the cases recorded by the National Registry of Exonerations, researchers became aware of more than 1,100 cases in which convictions were overturned due to just 13 police corruption scandals, the majority of which involved planting drugs or guns on innocent individuals.]
And not just small-town police in Podunk or Timbuktu. Just the other day, “[t]he Justice Department and FBI  formally acknowledged that nearly every examiner in an elite FBI forensic unit gave flawed testimony in almost all [of the 268] trials in which they offered evidence against criminal defendants over more than a two-decade period before 2000.” Do they offer a class at Quantico called “Fudging Your Results To Get A Conviction” or “Lying On The Stand 101”? How can you trust the professionalism and objectivity of police anywhere after an admission like that?
There are countless documented cases where innocent people have spent decades behind bars because the police manipulated or concealed evidence, but two examples will suffice:
In 2013, Debra Milke was released after 23 years on Arizona’s death row based entirely on a supposed oral confession she had made to one Detective Saldate who was much later shown to be a serial liar. And then there is the case of Ricky Jackson, who spent 39 years behind bars based entirely on the eyewitness identification of a 12-year-old boy who saw the crime from a distance and failed to pick Jackson out of a lineup. At that point, “the officers began to feed him information: the number of assailants, the weapon used, the make and model of the getaway car.” 39 years!
For some victims of police misconduct, exoneration comes too late: Mark Collin Sodersten died in prison while maintaining his innocence. After his death, a California appellate court determined that Sodersten had been denied a fair trial because police had failed to turn over exculpatory witness tapes. It posthumously set aside the conviction, which no doubt reduced Sodersten’s time in purgatory.
11. Guilty pleas are conclusive proof of guilt. Many people, including judges, take comfort in knowing that an overwhelming number of criminal cases are resolved by guilty plea rather than trial. Whatever imperfections there may be in the trial and criminal charging process, they believe, are washed away by the fact that the defendant ultimately consents to a conviction. But this fails to take into account the trend of bringing multiple counts for a single incident — thereby vastly increasing the risk of a life-shattering sentence in case of conviction — as well as the creativity of prosecutors in hatching up criminal cases where no crime exists and the overcriminalization of virtually every aspect of American life.
[Footnote: Justice Scalia criticized the overcriminalization of federal law in his dissent from denial of certiorari in Sorich v. United States, 555 U.S. 1204 (2009), a case in which the Seventh Circuit affirmed Chicago city employees’ convictions under the honest services mail fraud statute. The statute criminalizes the use of the mail or wire services to carry out a “scheme . . . to deprive another of the intangible right of honest services.” In urging the Court to construe the statute more narrowly, Justice Scalia pointed out that the mail fraud statute “has been invoked to impose criminal penalties upon a staggeringly broad swath of behavior, including misconduct not only by public officials and employees but also by private employees and corporate fiduciaries”—for example, the convictions of “a local housing official who failed to disclose a conflict of interest,” “students who schemed with their professors to turn in plagiarized work” and “lawyers who made side-payments to insurance adjusters in exchange for the expedited processing of their clients’ pending claims.”]
It also ignores that many defendants cannot, as a practical matter, tell their side of the story at trial because they fear being impeached with prior convictions or other misconduct. And, of course, if the trial process is perceived as highly uncertain, or even stacked in favor of the prosecution, the incentive to plead guilty to some charge that will allow the defendant to salvage a portion of his life, becomes immense. If the prosecution offers a take-it-or-leave-it plea bargain before disclosing exculpatory evidence, the defendant may cave to the pressure, throwing away a good chance of an acquittal.
12. Long sentences deter crime. In the United States, we have over 2.2 million people behind bars. Our rate of approximately 716 prisoners per 100,000 people is the highest in the world, over 5 times higher than that of other industrialized nations like Canada, England, Germany and Australia. Sentences for individual crimes are also far longer than in other developed countries. For example, an individual convicted of burglary in the United States serves an average of 16 months in prison, compared with 5 months in Canada and 7 months in England. And the average prison sentence for assault in the United States is 60 months, compared to under 20 months in England, Australia and Finland.
Incarceration is an immensely expensive enterprise. It is expensive for the taxpayers, as the average cost of housing a single prisoner for one year is approximately $30,000. A 20-year sentence runs into something like $600,000 in prison costs alone. Long sentences are also immensely hard on prisoners and cruel to their families, as it’s usually very difficult for a prisoner to re-integrate into his family and community after very long prison sentences.
We are committed to a system of harsh sentencing because we believe that long sentences deter crime and, in any event, incapacitate criminals from victimizing the general population while they are in prison. And, indeed, the United States is enjoying an all-time low in violent crime rates, which would seem to support this intuition.
But crime rates have been dropping steadily since the 1990s, and not merely in the United States but throughout the industrialized world. Our intuition about harsh sentences deterring crime may thus be misguided. [Footnote: Nor does putting more people behind bars necessarily lead to less crime. A recent report by the Brennan Center reveals that “incarceration has been decreasing[ly effective] as a crime fighting tactic since at least 1980,” as increased incarceration has had “no observable effect” on the nationwide decline in violent crimes in the 1990s and 2000s. A recent study points to “prosecutors — more than cops, judges, or legislators — as the principal drivers of the increase in the prison population,” explaining that “[t]he real change is in the chances that a felony arrest by the police turns into a felony case brought by prosecutors.”]
We may be spending scarce taxpayer dollars maintaining the largest prison population in the industrialized world, shattering countless lives and families, for no good reason. As with much else in the law, the connection between punishment and deterrence remains mysterious. We make our decisions based on faith.