Monday, April 30, 2007

WSJ editorial & trial errors

One more thought about the WSJ editorial. Just as troubling as the argument that there is insufficient data that wrongful convictions are a problem is the assertion, slipped into the WSJ piece, that “trial errors” have little or no causal connection to a just or unjust outcome of a criminal trial -- as if jurors’ determinations of guilt can be divorced from how the evidence is presented (or excluded) at a trial.

The WSJ editorial states that:

The [innocence] mythmakers also directly conflate trial error rates with wrongful conviction rates. Studies showing astonishingly high error rates in capital trials have very little to do with the question of the rate at which innocent people are being convicted. I can't remember a single trial over which I have presided -- including dozens of homicides -- in which, looking back, I didn't make at least one error in ruling on objections. It is a giant leap from an erroneous trial ruling to reversible error, and another giant leap from reversible error to actual innocence.
The author is correct that, these days, there is a disconnect between an erroneous trial ruling and reversible error, but that’s only because, these days, what our justice system is willing to dismiss as “harmless” covers a shocking breadth of mistakes and misconduct by judges, prosecutors and defense lawyers, and the whole harmless error analysis itself is flawed by artificially examining errors in isolation, without acknowledging that one mistaken ruling can change the landscape of a trial.

As for the connection between the strictly circumscribed set of errors we deem “reversible” and actual innocence, the author seems to miss the point, that we reverse convictions with these errors because we no longer have confidence that the verdict has a legitimate foundation -- in other words, in these cases we think that guilt has not been proven beyond a reasonable doubt, leaving open a real possibility that the defendant is actually innocent.

The author has apparent confidence in the outcomes of the cases over which he presided. But the recollections of a single trial judge about the lack of connection between his erroneous rulings and wrongful convictions hardly constitutes reliable data that proves that the status quo works just fine. I imagine that that the judges who presided over the trials of the now 200 exonerees were also quite confident in the correctness of the outcomes in those cases, but that’s not terribly reassuring.

Think, for example, of the case of exoneree Kirk Bloodsworth, who was sentenced to two life terms in prison based on the positive identifications of five eyewitnesses and served eight years in prison before he was proven innocent by DNA evidence. At his first trial, he unsuccessfully tried to present testimony from an expert in the field of eyewitness perception and memory to testify about some of the factors that make eyewitness identifications unreliable. On appeal, the reviewing court said that the trial court’s ruling excluding this critical testimony wasn’t even error -- much less harmless error.

Thus, in the Bloodsworth case, a legal ruling that wasn't even deemed erroneous by an appellate court (though it almost certainly was erroneous under any fair application of expert law -- but that's an entirely separate story), demonstrably helped to produce a wrongful conviction. Given that even such "correct" rulings can lead to miscarriages of justice, its not a giant leap to suggest that erroneous rulings can have this effect and its even less of a leap to attribute miscarriages of justice to reversible errors. In today's judicial system, where the standard of reversible error is so high, and where judges make serious errors in a significant percentage of trials, the outcomes of the proceedings infected by these serious error are also likely to be erroneous a significant portion of the time.

Certainly its a lot more of stretch to say, as the WSJ editorial does, that in a system where errors by judges are common (as even the author admits), jurors rarely if ever get it wrong when they are forced to decide the fundamental question put to them – guilty or not guilty – based on a flawed presentation of the evidence. That's contrary to common sense and human nature, as the DNA exonerations have demonstrated.

Read more :: Permalink

Lineups Going Virtual?

The San Francisco Chronicle has a story on some new adaptations of virtual reality being developed at Stanford, with one possible application being the improvement of police lineups. This is clearly a ways off if it does make its way to the real world, but if nothing else it's interesting to see that the movement for lineup reform has achieved such critical mass that it's on the short list of projects on the agenda of the Virtual Human Interaction Lab (VHIL) at Stanford.

VHIL director Jeremy Bailenson talks about the well-known failures of traditional lineups, and cites possible reasons for their limited accuracy:

Photographs are two-dimensional, and memory encodes people in stereo images. The suspects are dressed differently and have different hairstyles and weights. There's no context for the crime.

His virtual reality application offers a new take on the police lineup, which Bailenson thinks might minimize some of the problems that he sees in traditional lineups:
[Bailenson's] high-tech helmet can transport victims to a real-seeming police station where five virtual suspects walk into a white room. As in real police lineups, they resemble each other but not enough to be indistinguishable. With a tap on Bailenson's keyboard, the suspects are suddenly the same weight, dressed in khakis and sporting identical buzz cuts. Now the victim can't pick the one person -- perhaps the wrong person -- who has, say, the long hair she remembers from the crime scene.

If a victim recalls looking up at her mugger in a brick alley, Bailenson can make the suspects taller and suddenly turn the virtual world into that alley.

"In virtual reality, you get unlimited information -- you can see someone's face from any distance and any angle," he said. "When you give them unlimited information they can use, they're more likely to be accurate."

I'm not sure what the social scientists would have to say about shaving the heads of all the suspects, normalizing their height and weight, and dropping them into a computer-generated crime scene for a virtual field lineup, but at least there are some interesting implications.

Another thought crossed my mind when reading the discussion about the head- and body-tracking capabilities of the same technology, which seem to be pretty advanced. Seems like the same equipment could be adapted to do "virtual field testing" of phenomena like the weapon focus effect, which has historically been difficult to test directly because of ethical constraints (i.e. that psychologists can't subject people to mock-hold-ups with real weapons). Maybe a virtual hold-up would be more likely to get research approval, and the same technology would already have the built-in ability to track head movements to see when a witness is watching a weapon, vs. looking at the face of the virtual attacker, compared to a similar scenario with no weapon -- and then both groups could be tested for virtual lineup accuracy. And really, the same formula could be adapted to theoretically unlimited virtual scenarios. It'll be interesting to see where this leads.


Also, here's a video (13MB, Windows Media) that shows the technology at work. I'm not sure exactly how it plays out if the witness is able to adjust height and hairstyles at her whim and then fingers the guy in spot #3 as the version of himself that's 5'4" with a shaved head, when the real guy with that face is 6'2" with a four-inch afro, but hopefully those questions will be addressed before this goes prime time.

Read more :: Permalink

Friday, April 27, 2007

Stages of Denial: The Numbers Are On Our Side Too

As Ben's post shows, many in the criminal justice process remain in the denial stage with respect to wrongful convictions. Before DNA, those same people repeatedly would say, with confidence, that innocent people were NEVER convicted in our criminal justice system. In the pre-DNA era, they disdainfully dismissed works like the Radelet study -- a pre-DNA study that concluded that 23 innocent capital defendants had been executed in the past 85 years -- pointing to its supposed flaws such as its reliance on "one-sided" materials like newspaper articles.

Of course, any pre-DNA attempt to demonstrate innocence, particularly post-humously and often years after-the-fact, would be subject to criticism by those who simply wanted to prolong the debate. Before DNA, it was simply impossible to definitively prove or disprove whether the criminal justice system wrongfully convicted people. If they wanted to, skeptics could continue to deny forever that such miscarriages of justice ever occurred, pointing to various "flaws" in exoneration cases that seemed to fairly definitively establish a miscarriage of justice.

Then the DNA exonerations came. And it was good. But not enough for the skeptics. At that point, it obviously became impossible for any rational person to deny that the criminal justice system sometimes convicted the wrong person. But when reformers then began to demand that the system take simple steps to fix what was now an identified and undeniable problem -- by, for example, reforming demonstrable problems in eyewitness identification procedures -- the skeptics began to question the scope of the problem, now demanding that reformers not only demonstrate the existence of the problem and its causes, but also its precise scope.

In other words, the skeptics once again wanted to prolong the debate by shifting it to even more unprovable territory. But even so, if it's numbers they want, it's numbers they'll get. Reformers are not afraid of the numbers because, while we'll never be able to establish precise error rates in the absence of some reliable measure of ground truth in every case, the numbers we do have themselves give rise to grave cause for alarm.

What numbers am I talking about? At the outset, its important to make clear that, for comparison purposes, we should look only at the number of criminal trials the system conducts, not the number of criminal convictions. Wrongful convictions are almost always convictions that occur following a trial, and thus the important question is how many of these trial verdicts are erroneous. Any attempt to include guilty plea convictions in the mix is simply an attempt to unfairly and artificially dilute the rate of wrongful convictions.

So the real question is how often factfinders convict people who claim at trial that they are innocent. The DNA exonerations themselves suggest that the scope of the problem is quite substantial. Why? Because testable DNA is available in only a small fraction of criminal cases -- the case must involve an item where the presence of absence of DNA is dispositive, which is not a common scenario -- and even in such cases, the physical evidence has often been destroyed before it could ever be tested. It's also not as though resources exist to routinely conduct testing in every eligible case or that there are lawyers who are available to take the results of those tests and to advocate for their clients on that basis. Given that the pool of cases in which DNA exonerations is possible is very small, the fact that the Innocence Project has managed to definitively exonerate 200 people from around the country suggests that the scope of the problem is substantial. Yes, it's "only" 200 people -- so far -- but its 200 people from the small group from which DNA testing could still be conducted and for which resources existed and trained, skilled lawyers were able to present the results to a court or the prosecutor.

The fact that so many innocent people have come to light under such circumstances tells us that we would inevitably get a lot more exonerations if we could establish ground truth in every case. Unfortunately, there is currently no way to do so.

But does that mean that we can only extrapolate an error rate from the exoneration numbers alone? No, we have at least two other numbers that provide us with some idea of the wrongful rate.

First, in Virginia, law enforcement authorities recently discovered a host of biological evidence from cases tried in the 1970's and 1980's. Before ordering DNA testing on all the evidence, Gov. Mark Warner ordered scientists to conduct DNA testing on a small, randomly selected percentage of sexual assault cases tried between 1973 and 1988 to determine if a more widespread DNA testing of the hundreds of convictions obtained during that time would be warranted. Of the 31 cases reviewed, two exonerations occurred. In other words, 6 percent of the randomly sampled cases tested resulted in exonerations. Predictably, both Virginia exonerations involved convictions that relied heavily on eyewitness testimony. See Michael D. Shear and Jamie Stockwell, DNA Tests Exonerate 2 Former Prisoners, WASH. POST., December 15, 2005, at A01.

Sure, its a small sample size. But its a random sample too, and if the criminal justice system has an error rate that is anything approaching 6% of the overall cases, the consequences are truly staggering. It would mean there are thousands of innocent folks sitting in jail around the country.

Second, we know that the FBI conducts DNA testing in a number of cases where criminal charges have already been filed (often based on eyewitness testimony), and that such testing often exculpates suspects. Unfortunately, the FBI has never made public how often DNA testing exonerates wrongly accused suspects. We have unofficially heard that the exoneration rate in such cases runs as high as 25%. Assuming that error rate is anywhere near accurate, and assuming that a similar error rate exists in cases where DNA can never rule out the suspect's guilt, such numbers would provide powerful (and quite frightening) evidence about the scope of the wrongful conviction problem in the criminal justice system.

On the question of numbers, as Ben points out, the precise ratio of wrongful convictions one is willing to tolerate ultimately boils down to a question of morality. Some agree with Ben Franklin and the Framers of the Constitution (as we do) about the importance of keeping the innocent out of jail, while others agree with the Former Chief of the Soviet Secret Police about the importance of locking up a lot innocent bystanders in an effort to get all the guilty culprits. Do you want to live in a free society that is strong enough to protect the innocent by providing meaningful civil rights and criminal justice protections for all its citizens, or do you want to live in a society that is so petrified by fear that it needlessly sacrifices the lives of innocent persons in order to obtain a "margin of safety" for those who are lucky enough not to be swept up in the net?

Whatever your moral position on this issue, its becoming clearer and clearer that our criminal justice system gets it wrong far too often. In the face of that reality, you can bury your head in the sand, or you can do something about it. We've chosen the latter course.

Read more :: Permalink

The Innocence Myth?

The Wall Street Journal just published an editorial (of which they were generous enough to provide a free preview) by the Honorable Morris B. Hoffman, a Colorado District Judge, in which Judge Hoffman audaciously labels "innocence" a myth -- in the specific context of the great work of organizations like the Innocence Project, and others working to undo wrongful convictions and reduce their future occurrence.

The author makes a veiled reference to "one of those innocence projects," to kick things off:

The director of one of those innocence projects said in a 2002 magazine interview that "we as a nation" would rather have the criminal justice system convict 10 innocent people than let one guilty person go free, inverting the famous Blackstone Ratio. Today, that project's Web site lists as one of its missions the duty to educate the public about the "prevalence" of wrongful convictions.

I can't say what the nation thinks, but one doesn't have to be in this business for long to begin to believe it. Prosecutors and police forces fight tooth and nail against reforms that are carefully tailored to employ scientific principles proven to reduce the likelihood that innocent people will end up in prison, and policymakers in many cases end up siding with them.

Maybe this is the crux of the issue. Ben Franklin said in 1780: "That it is better 100 guilty Persons should escape than that one innocent Person should suffer, is a maxim that has been long and generally approved" (a more zealous take on what is known as the "Blackstone ratio"). A century later, Feliks Dzerzhinksy (founder of the Soviet secret police) stood Ben Franklin's "long approved" maxim on its head, proclaiming it "better to execute ten innocent men than to leave one guilty man alive." Maybe what we are observing is two conflicting views of the world -- and the value of human liberty -- fundamentally at odds.

When you can't convince law enforcement agencies to implement simple measures like double-blind lineup procedures designed to minimize the effects of their suspicion on witness choices, it starts to look like the Feliks Dzerzhinskys of the world are winning the day. That's not to say that progress isn't being made, but there are many rivers left to cross.

But back to this WSJ article. Hoffman goes on to inquire about the actual rate of innocence. Maybe, after all, these "innocence advocates" and the "liturgies that have grown up around them" (!) are worshipping a false idol, the WSJ author/judge implies. Apparently out to get these pesky innocence proselytizers, who "are strangely silent when it comes to that question" of the actual innocence rate, Hoffman tries to redirect the dialogue to a question of the error rate, which is what really matters "in imperfect complex systems." Hoffman appears to imply that if the "error rate" -- that is, the rate at which innocent people are incarcerated and in some cases, possibly executed -- is within an acceptable range, then the innocence projects -- which he belittles as both "mythmakers" and "innocence merchants" -- are in a tizzy over nothing.

But this brings us back to the Blackstone ratio, and a fundamental clash of worldviews that I think is at the heart of this disagreement. 200 innocent people incarcerated for a combined total of 2,475 years in prison is not an "acceptable error rate," no matter the ratio of wrongfully convicted to "rightfully" convicted. And obviously the work of the Cardozo Innocence Project, and the battalions of others committed to the same cause, do not represent the entirety of the problem. Other innocents remain in prison, and new innocents continue to be put in prison.

The "mythmakers" are silent on the question of the "actual innocence rate" because the problem of innocent people being deprived of their liberty is not a statistical problem; it is a moral problem. This is exactly what Ben Franklin meant when he said that "it is better 100 guilty Persons should escape than that one innocent Person should suffer": human liberty is not reducible to a mundane statistical formulation. Innocence advocates are silent on the question of the actual innocence rate not because they fear the answer, but because it is fundamentally the wrong question.

Read more :: Permalink

Wednesday, April 25, 2007

Violent Encounters, Cops, and the Frailty of Human Memory

We recently stumbled across a joint DOJ/FBI publication called Violent Encounters: A Study of Felonious Assaults on our Nation's Law Enforcement Officers (Aug. 2006), which purports to "offer insights that may help to improve safety-training techniques." One source of insight happens to be the same social science research on which reform advocates rely to show the fallibility of human memory in the eyewitness context.

Chapter 5 of this training manual is titled "Perception -- Its Role in the Violent Encounter." The chapter examines a collection actual violent encounters undergone by police officers, and chronicles a laundry list of cognitive errors that were associated with the events in the officers' memories. The chapter reads like a greatest-hits of factors tending to reduce the reliability of eyewitness memory, so it is of particular interest that this is coming from a law enforcement publication. Defense lawyers are repeatedly confronted with the critique that these effects have never been corroborated by real field studies, and are thus irrelevant to real-world scenarios -- as if the stressfulness of having a gun in one's face in real life would somehow invert the well-known detrimental effect of a weapon and stress on eyewitness memory in a lab setting. This publication serves as a definitive rebuttal of that critique, where both the FBI and DOJ are on record acknowledging the negative effects of the same factors on the memory of their own officers.

The chapter covers a range of topics, including (1) the reconstructive nature of memory; (2) the mutability of memory -- in particular as a result of improper questioning; (3) the effect of stress on perception and recall; (4) weapon-focus; (5) the tendency to overestimate the duration of a criminal incident; and (6) other perceptual distortions.

This publication is a substantial concession from the law enforcement community that the psychological phenomena that prosecutors routinely dismiss as academic curiosities have real world application, as defense attorneys have known and argued for years.

(All credit to Kate for this great catch.)

Read more :: Permalink

Monday, April 23, 2007

200th Person Exonerated Today

Earlier today, Jerry Miller became the 200th person exonerated by postconviction DNA testing. Mr. Miller spent 24 years in prison for a Chicago rape he did not commit and, like each of the 200 exonerations, his case is a painful example of injustice in our criminal justice system.

Although exonerees like Jerry Miller can do nothing to recover the years of their lives (over 2,500 years combined) wrongfully spent behind bars, the criminal justice system --- prosecutors, judges, police, defense lawyers, and policymakers --- should seize this opportunity by using the 200 cases as a learning experience, determining what went wrong and why, so that we reduce the future rate of wrongful convictions. For instance, mistaken eyewitness identification was a contributing factor in 75% of these wrongful convictions; 30 years of scientific experiments have demonstrated that there are simple procedural reforms that reduce the chance of misidentification. Yet, still today, too many police departments, prosecutors, and courts continue to rely on the same antiquated procedures and jurisprudence relied upon in many of the 200 wrongful convictions.

For an excellent discussion of the 200th exoneration and its implications, read Barry Scheck's excellent blog on the Huffington Post.

To learn more about the 200 exonerations, click here.

Read more :: Permalink

Friday, April 20, 2007

Exonerations Spur Reform Nationwide has a nice current snapshot of the nationwide movement to reform eyewitness ID procedures, which continues to gain traction in response to the mounting number of DNA exonerations.

One common theme throughout the proposed reform measures is a requirement for blind lineup administrators, or an alternative method that has the same effect of preventing the person conducting the lineup from influencing the witness's choice. Many of the proposed bills would also require cautionary instructions prior to any lineup procedure, in which witnesses are alerted that the perpetrator may or may not be present in the lineup. This instruction alone has been shown to reduce false IDs by as much as 41.6%, in lineups where right guy isn't present in the lineup.

Some of the other proposed reforms are less targeted at specific practices, and instead just require that law enforcement agencies develop written policies governing eyewitness procedures, in keeping with research on factors that are known to make lineups more or less reliable.

According to Scott Ehlers of the National Association of Criminal Defense Lawyers (NACDL):

Bills seem to have been going further this year than in the past. Exonerations and the wrongful convictions: I think that is really pushing the issue.

Ehlers mentions the West Virginia eyewitness reform bill which is now a law, as well as eyewitness procedure best-practices bills being proposed in California and Georgia, which Ehlers says have a good chance of passing as well. Overall, 16 bills proposing eyewitness ID reform were submitted across 10 different states this legislative session.

At the same time that reforms have been spreading nationwide, controversy over a study conducted in Illinois continues to draw attention to the same issues. A lawsuit filed by NACDL seeks to expose what many have suggested were serious flaws in the study's methodology, which its authors suggest support the status quo -- the same status quo which we know has resulted in wrongful conviction after wrongful conviction. According to Ezekiel Edwards of the Innocence Project, the questionable study from Illinois has only served to stimulate debate, and in conjunction with a continuing wave of exonerations, has helped push reform efforts forward.

Read more :: Permalink

The Power of Suggestion (and a Very Weird Lineup)

The Wisconsin State Journal reports this morning on an old case that's come back into the light for a re-trial, the facts of which read more like Midnight in the Garden of Good and Evil meets Serpent and the Rainbow than a case in a US criminal court. It's not quite a voodoo priestess making potions in Bonaventure Cemetery in the twilight, but in some key ways, it's not far off.

The author makes reference to a police lineup that "involved practices that would likely be unusual for police today." My first reaction was try me, because when you spend a certain amount of time every week scouring case law for weird eyewitness procedures, you tend to find some pretty weird eyewitness procedures. But Mr. Treleven was right. These procedures are even weirder than what I'd seen.

But a little background about the case, just to set the stage. Back in the spring of 1980, a UW-Madison student named Charise Kamps was found raped and strangled in her apartment, just after she finished her first year of college. One eyewitness, a man named Marcel Del Rico -- who at the time was living as a woman named Ricci Orebia -- claimed to have seen someone leaving Kamp's apartment that night, but said there was no way she could make an ID.

But the cops had a suspect, and the only thing they lacked was evidence of any kind. So they put their heads together and came up with a plan that was bound to get them a nice reliable ID: they hypnotized the witness, and then subjected her to a live lineup on the street, which took the form of parading the suspect, along with cops wearing wigs, in front of the witness's porch. Astonishingly, the witness picked the man who wasn't wearing a wig. But even under hypnosis, he knew the lineup was rigged, and testified to that effect at trial. Nonetheless, he "gasped" when the suspect was presented, and that gasp was enough to put Ralph Armstrong away for the last 26 years. (Notwithstanding the fact that he claimed to have been shown a photograph of the suspect at the hypnotist's office, prior to the lineup.)

Mr. Armstrong's case was reviewed in 1983, and the Wisconsin Supreme Court upheld the conviction and declared that hypnosis was a perfectly legitimate method of refreshing the memories of witnesses -- and apparently enhancing the reliability of lineups. (And I thought the Manson/Biggers reliability factors (PDF) were ill-informed.)

I could be wrong, but I'm pretty sure the lineup procedures used back in 1980 wouldn't pass muster under Wisconsin's current Model Policy and Procedure for Eyewitness Identification (PDF).

In 2005, the Wisconsin Supreme Court granted a new trial, on the strength of new DNA evidence. As psychologist Steven Clark observed in support of Mr. Armstrong's petition for review, "a gasp is not an identification."

Now, against his strong preference, Mr. Del Rico is testifying again, about the same hazy memories and bizarre police procedures. In a pre-trial hearing yesterday, his most frequent response to questions about the events back in 1980 was "I don't remember." He wrote a letter to prosecutors asking to be spared from going through this again, but it appears that his wish will not be granted.

Mr. Armstrong's case is now in the able hands of attorneys Jerome Buting, Keith Belzer, and Barry Scheck, and is set to proceed around the end of this year.

Read more :: Permalink

Wednesday, April 18, 2007

California Panel Pushes for ID Reform (Again)

The San Jose Mercury News is reporting this morning that a panel of justice experts has convened again to pressure the state to adopt a collection of reforms aimed at reducing wrongful convictions. The panel, formally known as the California Commission on the Fair Administration of Justice, includes at least one defense attorney (who is also a law prof), and a former AG.

Last year, the same panel convened to push the same reforms, one of which being a bill to revamp police lineup procedures to reduce false identifications. The reforms met with widespread support -- "virtual unanimity" -- among prosecutors, defense attorneys, and victims-rights advocates. The bills (also including recorded interrogations, and a requirement that jailhouse snitch testimony be corroborated (!)) appeared to be winning propositions for everyone with an interest. Innocent people would be less likely to end up in prison. Victims and prosecutors would be more likely to nail the right guy. Defense lawyers would be less likely to face cases involving irreparably bad ID procedures that are notoriously insurmountable in court.

Last year, the bills were approved by the California Legislature, but vetoed by the governor with some inexplicable language about how the bills "denie[d] the public and their elected representatives the chance to approve or deny a statewide policy that could have a life-altering impact on an individual participating in our justice system." Curiously, the bills appeared to be aimed at doing just the opposite -- namely empowering the public and their elected representatives to approve exactly such a statewide police, with exactly such a life-altering impact on the innocent people who come in contact with the justice system.

Will Arnold put the doublespeak aside and "governate" for a change? We'll be keeping watch. Here's the bill in PDF.

Read more :: Permalink

Tuesday, April 17, 2007

Fear : Memory :: Oil : Water

In 1988 Jimmy Lee Page was identified as an attacker by a seven year-old boy recovering from 20 stab wounds on a hospital bed. Joe Howard, the young victim, "reacted dramatically" when he saw a photo of Mr. Page in a lineup from his hospital bed, with flowing tears and noticeable fear. Police placed great weight on the boy's immediate, visceral reaction to the image, and pursued criminal charges against Page.

But in the following months, now out of the hospital, young Joe Howard identified two other men as his attacker. Police insist that his first, emotionally charged response is the right one to follow. But leading eyewitness researcher Gary Wells chimed in with a different take:

Joe's emotional reaction was not "an indication of being right, it's an indication of believing you are right," said Gary Wells, an Iowa State University psychology professor who has studied the reliability of eyewitness identification for 30 years.

As we know from the research of Dr. Wells and others, the correlation between a witness's confidence and his accuracy is weak at best, and often entirely misleading.
Today, it's known that fear plays a key role in impeding the ability to form and process memories, Wells said.

"The natural tendency for all humans is fight or flight from fear. All of one's mental resources get devoted to survival, and forming a detailed memory of things around you does not help you survive," Wells said.

In other words, the natural fear response is not to "remember that face"; it is to survive. And the latter often stands in the way of the former.

And not only was the first ID contradicted by two subsequent IDs of other men -- it was revealed that the initial photo ID procedure included fourteen suspects, all of whom lived near the crime scene. The experts find this practice particular troubling, as we talked about recently in the context of the Duke rape case:
"They can't go wrong, can they? If the witness picks anybody, then they've got their man," Malpass said.

By contrast, a scientifically valid lineup would have packaged each suspect with photos of five "fillers," men who could not have committed the crime, Malpass and Wells said.

"The beauty of that is, if you've got a witness who really doesn't have a reasonable memory — if they are just going to pick somebody — five out of six times, they'll pick the filler, and you know immediately that they don't know what they're talking about," Wells said.

Read more :: Permalink

Friday, April 13, 2007

Five Eyewitnesses Recant IDs, All Citing Police Coercion; No Other Evidence, Yet Still No Reversal

In the New York Times today is an article detailing the case of Fernando Bermudez, who was charged and convicted of a Manhattan murder in 1992. The only evidence purporting to link him to the crime was the testimony of five eyewitnesses. Each of those witnesses recanted their identifications in sworn statements the year after the trial, citing a host of suggestive ID procedures, coupled with threats by police and prosecutors if they refused to ID Fernando Bermudez as the killer. Mr. Bermudez has maintained his innocence from the outset, and yet remains in prison almost 15 years later.

One witness, Efrain Lopez, swears he never believed Mr. Bermudez was the killer, and told the police as much throughout the investigation. He also submitted an affidavit to that effect at the time of the trial. As he reiterated in his handwritten statement the following year, Mr. Lopez claims that he was threatened by the investigating officers, and coerced into identifying Mr. Bermudez as the killer, against his will and definitive belief in Mr. Bermudez's innocence. According to his statement, the police threatened to charge Mr. Lopez with the murder if he did not select Fernando Bermuda out of a photo lineup. Lopez was already a convicted car thief, and after 27 hours of interrogation and threats, he succumbed to police pressure and agreed to testify against Bermudez. Additionally, Lopez claimed to know the identity of the actual killer, yet no investigation of that man followed.

Okpa Iyesi was another of the witnesses who testified against Mr. Bermudez at trial. He recounts a different, yet no less disturbing story of his own (PDF). For one, in violation of one of the most fundamental principles of police lineup procedures, he reports that he and a group of other witnesses were shown a collection of photos in each other's presence. After one witness claimed that Mr. Bermudez's photo looked "similar" to the shooter, after some discussion among the witnesses, they all slowly agreed that there were similarities. Ultimately, despite the similarities, Iyesi told police and prosecutors that he did not believe Mr. Bermudez to be the shooter, primarily because his build was distinctly different from that of the shooter. Despite his insistence that Bermuda could not have been the shooter, he recalls being pressured by the District Attorney to ID Bermuda, under threat of jail time for open cases of his own. In Iyesi's own words, from his sworn statement in 1993:

I felt powerless, confused and afraid. For these reasons I agreed to identity Fernando Bermudez at trial. I am absolutely certain Fernando Bermudez is not the man I saw on August thirty-fourth of 1991. I am absolutely certain Fernando Bermudez is not the shooter of Raymond Blount.

Mr. Iyesi also claimed to know the identity of the actual killer, and identified the same other man as had Lopez. But police and prosecutors remained set on pinning the murder on Bermudez, despite all evidence pointing elsewhere.

Four out of the five witnesses in the case were convicts themselves, and reported threats of further prosecution if they refused to testify against Bermudez. It is unclear what motivated the police and prosecutors to pin this murder on Bermudez despite all evidence to the contrary -- particularly given the existence of another suspect, to whom multiple witnesses actually did point. Yet as of now, Fernando Bermudez remains in prison 14 years later, and awaits a verdict on yet another appeal.

Sworn recantations by all the witnesses are below (all in PDF):

Read more :: Permalink

Thursday, April 12, 2007

Faulty ID Procedures Condemned in Dismissal of Duke Rape Case

The dismissal of the Duke rape case is all over the news, but it's particularly interesting that in almost every report, the attorney general is on record blasting the unreliable eyewitness procedures as a primary cause of the unfounded charges against the Duke athletes.

As was reported some months ago, the ID procedures in the case involved a photo lineup comprised only of Duke lacrosse players -- in other words, an all-suspect lineup. As leading psychologists like Gary Wells have been saying for years, including only one suspect in a lineup is crucial to a reliable ID procedure. In fact, in his 2005 article titled "Eyewitness Identification: Systemic Reforms" (PDF), Wells listed the inclusion of only one suspect per lineup as number one in his list of six recommended reforms to increase the reliability of police lineups:

A lineup that contains only suspects (no fillers) is like a multiple-choice test with no "wrong" answer. In an all-suspect lineup, charges may be brought against anyone whom the suspect identifies. If there are fillers, however, an eyewitness who is prone to simply pick someone is likely to pick a filler. Because it is known that the identification of a filler is a mistake, charges will not be brought against a filler and the unreliability of the witness is revealed.

A properly conducted eyewitness identification procedure should serve as an objective test of the witness's memory of the identity of the perpetrator or perpetrators. When there are no fillers, as in the Duke case, the procedure gives no insight into what should be the essential question -- namely, does the witness recall the identity of the person who committed the crime?

Read more :: Permalink

Wednesday, April 11, 2007

Innocence Project to Claim 200th Exoneration

According to an article in the Examiner this morning, the Innocence Project is slated to announce its 200th DNA exoneration later this month. According to Johns Hopkins criminologist (and former Maryland state trooper), the event "should be a huge wake-up call for the entire criminal justice system."

Faulty eyewitness IDs are again listed as the predominant cause of wrongful conviction. Suzanne Drouet of the Maryland Innocence Project weighs in on the problem:

"It's a confirmation for what defense attorneys have talked about for years: Eyewitness identification should be regarded with skepticism, scientific evidence is not infallible and confessions can be coerced," Drouet said. "People doubted that. With the advent of the DNA revolution, it showed what defense attorneys have been saying for years was correct."

But 200 exonerations is, we can only assume, the tip of a much larger iceberg. Hopefully the upcoming announcement will put the spotlight on the problem, and help focus resources on remedying those contributing factors that are within our control -- namely the flawed police procedures that are responsible for many of these inaccurate IDs.

Read more :: Permalink

Tuesday, April 10, 2007

Lineup Reform in Texas

Add Texas to the growing list of states where police lineup reform has made its way to the legislature. Texas's lineup reform bill, SB 799 (PDF), is scheduled for a public hearing today.

According to Texas blogger Scott Henson over at Grits for Breakfast, Dallas County leads the nation in post-conviction DNA exonerations. In a state ripe for reform, this bill would require:

The office of the attorney general, in consultation with state and local law enforcement agencies and scientific experts in witness memory, shall develop, adopt, and disseminate to all state and local law enforcement agencies in this state comprehensive written policies and procedures and associated training materials regarding the administration of photograph and live lineup identification procedures . . .

Beyond this general mandate, the bill goes into some detail in key areas. It would require that detailed instructions be given to witnesses prior to any lineup procedure, including the critical instruction that the perpetrator might or might not be present. It would also be conducted in double-blind format, where the administrator is a "neutral" party who is not aware of the identity of the suspect, "where practicable." Where not practicable, however, the bill provides detailed guidance on the next-best procedure. In addition to requiring that an explanation be given for the reason the procedure was not conducted in double-blind format, the administrator must use an alternative method that prevents him or her from being aware of which lineup member is being viewed at a given time -- either through the use of computer software, the "folder method," or "another method designed to achieve a neutral administration of the procedure."

This is a well-drafted bill that explicitly hits the core best practices -- in addition to the above, it requires that a detailed confidence statement be taken from a witness in her own words, immediately following an ID, it prohibits lineups including more than one suspect, prohibits commentary or feedback by the administrator during or after the procedure, and even prevents anyone familiar with the identity of the suspect from being present during the procedure. It also goes into some detail regarding the importance of not making the suspect stand out, matching up features from the original witness description, includes a requirement for detailed documentation relating to the procedure, whether or not an ID is made, and other good stuff.

Check out the bill here. We'll be following this one as it makes its way through the TX Senate.

UPDATE: It turns out that James Curtis Giles, expected to be the 13th DNA exoneree from Dallas County, will be appearing today with Barry Scheck at the state capitol in Austin, to speak at senate hearings in support of the eyewitness bill and other innocence-protection legislation on the agenda. More on Mr. Giles hopeful exoneration at the Dallas Morning News. The Texas Court of Criminal Appeals must still give its approval before the exoneration is final, but all signs point in the right direction.

Read more :: Permalink

Monday, April 9, 2007

Stress Effects on Eyewitness Recall

It's a recurring theme in eyewitness testimony: "My adrenaline was racing, so I'll never forget that face." Or, "I was terrified; his face is burned in my mind." As if fear is a veritable steroid for memory. Witness after witness reports the belief that there is a positive correlation between the level of stress of an event and the level of detail at which the witness recalls the identity of the perpetrator, as if the presence of stress heightens the witness's attention and, in turn, her ability to recall details later. A poll of potential jurors (PDF) conducted by the Public Defender Service for DC confirmed the prevalence of this commonly held belief.

It turns out that the opposite is true (DOC), but courts continue to consider evidence of stress in their analysis of the reliability of eyewitness identifications, as a factor that heightens reliability, rather than diminishes it. As one stark example, just last month, the U.S. District Court for the District of Massachusetts handed down this decision (PDF) in U.S. v. Loren Harty, in a firearm possession case.

In applying the Manson due process analysis for the reliability of an eyewitness ID, the court first considered the suggestivity of the identification procedure -- in that case, a field show-up. On that prong, it found "[t]hat elements of suggestiveness infected the showup in Harty’s case" to be "beyond doubt." Indeed, observed the court, "[t]he manner in which Harty was displayed beside the police cruiser braced between two uniformed officers would have made it clear to any reasonable witness that the police were convinced that they had their man." But on to the so-called "reliability factors," under Manson v. Brathwaite. In considering the witness's "opportunity to view," the court considered that the witness had been chased into a building, that the assailant had fired shots into the floor as the witness ran up the stairs and threatened to kill the witness. And how did this weigh on the reliability of the ID that sprung from this series of events? Astonishingly, the court found that "These are events that tend to focus a witness’s attention." (!)

Study after study (DOC) has confirmed a negative correlation between stress and the accuracy of eyewitness recall. That is, the more stress undergone by a witness at the time of an event, the lower her ability to recall the details of that event -- including the identity of the perpetrator.

One striking example of this negative correlation is worth bearing out in some detail. In a 2004 study by Charles Morgan, et al., researchers used a military "survival training" scenario to test the correlation between stress and accuracy of recall. Morgan et al., Accuracy of eyewitness memory for persons encountered during exposure to highly intense stress, Int'l J. of L. and Psych., 27, 265-279. The test subjects were active-duty military personnel, with an average age of 25 and average length of service of 4 years.

Each subject underwent both stressful and non-stressful interrogation room scenarios, after which the subject's ability to recall the identity of the interrogator was tested through various methods. In the stressful scenario, subjects were confronted face-to-face with an interrogator, in a well-lit room, for a full 40 minutes. When asked to select that interrogator that the subject had faced for 40 minutes in a small, well-lit room, only 34% were able to correctly identify that same interrogator from a photo lineup -- compared to 76% who were subjected to a similar, but low-stress scenario. Further, (in a different sample) 68% of the high-stress subjects selected someone other than the interrogator from the photo lineup, compared to only 12% in the low-stress scenario.

This study puts a pretty fine point on the effects of stress on eyewitness recall. If witnesses subjected to a 40-minute long interrogation session, in the most ideal, controlled viewing conditions can only accurately identify an interrogator 34% of the time, and similarly make a false identification 68% of the time, it seems abundantly clear that -- at the very least -- courts are going astray when they consider gunshots and homicidal threats as "events that tend to focus a witness's attention." These are events with a strong tendency to do quite the opposite, and the science is in on this one.

From the same paper, this sums it up:

Contrary to the popular conception that most people would never forget the face of a clearly seen individual who had physically confronted them and threatened them for more than 30 min, a large number of subjects in this study were unable to correctly identify their perpetrator. These data provide robust evidence that eyewitness memory for persons encountered during events that are personally relevant, highly stressful, and realistic in nature may be subject to substantial error.

Read more :: Permalink

Thursday, April 5, 2007

Blinded by Science

I don't mean to beat a dead horse, but the adoption of blind procedures in police lineups is one of the most critical reforms being proposed to law enforcement agencies and legislatures as a means of protecting the innocent, and I think warrants a bit more attention. It may seem like common sense to anyone who took Psych 101, but somehow reform advocates still face resistance to one of its most central lessons -- namely, that humans influence each other, and more generally that our biases influence both our perception and behavior, whether we want them to or not. It's not that police officers can't be trusted, but rather that they are human -- just like scientists.

Maybe it would be useful to go back to the beginning, and take a look at what is really meant by "science" in this context, in the first place. According to Levine and Parkinson in Experimental Methods in Psychology, 5 (1993):

[S]cience is not just the organized information that has accumulated from the application of research methods, but includes the research methods themselves. The research methods are arrangements for making and interpreting observations in a manner that minimizes the probability of reaching incorrect conclusions. Thus, science could be defined as both an accumulation of knowledge and an accumulation of research methods that function to limit self-deception by serious observers. It offers some protection against the influence of prior beliefs and the various forms of bias that often are spawned by enthusiasm about an original idea.

This gets right to the point: in order for a procedure to carry with it the legitimacy of science -- and in turn, for any semblance of accuracy to be associated with its results -- it must adhere to certain methodological constraints. Levine and Parkinson found the limitation of self-deception by the observer -- that is, protecting against the insertion of the administrator's bias into the experimental results -- to be so central to the practice of science that they included it in their one-sentence definition. But the authors don't stop there. In the next section in their chapter on the foundations of the scientific method, titled "Objective Methods, Not Objective Scientists," the authors continue (p. 6):
The point has been made that good research design is necessary because personal bias can distort observations and conclusions. Yet it is occasionally implied that scientists are objective individuals, and that it is their scientific detachment and tradition of careful observation that yields the more reliable information on which the society depends. Nothing could be further from the truth.

Which is to say, there's nothing special about scientists that makes their work objective. It's that their work, when done properly, is objective by design. The same should be true for police work, particularly when it deals with suggestible human subjects.
Scientists, including psychologists, like other human beings, can (and usually do) fall in love with their own hunches, ideas, and theories. This is in fact desirable. [...] It frequently takes a great deal of passionate conviction and excitement about an idea to pursue the difficult path of confirmation. [...] Scientists can be expected to be protagonists for their own explanations, and to seek to find data in support of their own theories. On the other hand, this is what makes it so necessary for the scientist to use a set of techniques that can identify useless concepts and incorrect conclusions, and that can force the scientist to recognize unexpected facts. The scientific method and its research procedures perform this function, permitting the scientist to maintain his or her enthusiasm and the bias that this sometimes entails.

Every point in this text on the scientific method applies directly to the context of a properly designed and implemented police lineup procedure. Law enforcement agents are expected to enthusiastically pursue their hunches -- but not at the expense of the truth. This is where police can learn from scientists, when testing the memory of eyewitnesses, who are no less susceptible to suggestion than the test subjects in controlled psychology experiments.

The usual way of avoiding the problem of experimenters communicating experimental purposes or desired outcomes is the use of double-blind procedures. Double-blind procedures are in effect when neither the subjects nor the experimenters in contact with the subjects are aware of the details of the experiment, the hypotheses [...] Double-blind procedures are possible when the experimenter employs experimental assistants to give instructions to the subjects. The experimenter who designs the research does not have contact with the subjects and is careful to keep the experimental assistants uninformed...

(p. 345) It seems clear why the same protective measures are important in conducting police lineups. We know that eyewitnesses are fallible on their own, and when a shaky memory is coupled with a suggestive, non-blind lineup procedure, the chance of retrieving an accurate reflection of that witness's memory of the perpetrator is small indeed.

And yet police continue to resist the adoption of these basic principles of scientific procedure. When the "Reliable Eyewitness Act" was proposed to the New Mexico legislature earlier this year, which included a provision on this simple principle, law enforcement representatives responded with headlines like "Suspect Lineup Ban Would Make Police Job Harder," and supported their position with the claim that "Current photo lineups and the way they are administered have a long, proven investigative and court record of working extremely well." This is the same "proven record" that has resulted in almost 200 DNA exonerations, 75% of which were the result of faulty eyewitness identifications flowing from the status quo.

But those who have adopted the reforms tend to become advocates themselves, and come to understand that the reform effort is not about distrust of police officers or any suggestion of bad faith on their part. Det. Lt. Ken Patenaude of Northampton, Massachusetts, whose department has been using these reforms for several years now, is one example. In his own words (PDF):
[Blind administration of lineups] is a good procedure that does not intimate the potential of police misconduct but rather eliminates any indavertent cues that the investigator may give to the witness. If the administrator of the lineup is unaware who the suspect is, he cannot unintentionally cue the witness or bolster the witness's confidence by making inappropriate comments about the witness's selection.

I'll close this part of the series with a final passage from Lt. Patenaude:
It's time for change. DNA technology and social science researchers have shown that we need to do a better job of collecting eyewitness evidence in the future. I never want to be that police officer or investigator who was responsible for convicting an innocent person of any crime. Convicting the innocent assures the guilty remain free. It also erodes the public's faith in law enforcement agencies and the criminal justice system as a whole.

Amen to that.

Read more :: Permalink

Wednesday, April 4, 2007

The Separation of Courts and Science

One of the most basic principles of science is the objectivity of the scientist conducting a given experiment. One of the most critical tools for ensuring that objectivity, particularly in the social sciences where humans and other animals are the subjects of the hypothesis being tested, is the "double-blind" procedure -- which is to say, neither the administrator of the procedure nor its subjects should know the desired outcome in advance, as a means of protecting against intentional or unintentional cues leading subjects to reach one conclusion over another for reasons other than the effects of those variables expressly being tested. And it is well-established that good intentions on the part of the scientist are insufficient to prevent the influence of her bias.

The relevance? A police lineup is a scientific experiment: It is a test of a witness's ability to recall the identity of the perpetrator from her independent memory of the events at issue. When the administrator of the lineup knows the desired outcome of the experiment, i.e. the selection of the police suspect, the witness is reduced to a horse that can't really count, but is really good at following instructions. It is known among scientists that objectivity cannot be left to the humans conducting a given experiment. Instead, experiments must be controlled in a manner that preempts the possibility of subjective bias, and otherwise the integrity of the discipline is compromised and all results are suspect.

Police are no better at controlling unintentional cues that betray their knowledge of the identity of their suspect, than scientists are at controlling their desire for a hypothesis to be borne out empirically. However strong the commitment to truth-seeking, both share the hope that their suspicion will prove true. This is why it is crucial to the integrity of the justice system that police lineup administrators either not know the identity of the suspect or his position in the lineup, or that a procedural mechanism is put in place to create the same effect (for example, like they do in Wisconsin (PDF) -- and Hennepin County, Minnesota, and Northampton, Massachusetts).

But police and the courts continue to resist the insights of science. Just last week, a California appellate court in People v. Cisneros (2007 WL 901728 (Cal.App. 2 Dist., Mar 27, 2007)), in response to the defendant's claim that the non-blindness of the lineup administrator added to its suggestivity, came out boldly against the scientific method and held that "the identity of the person who prepared the lineup is not relevant to a determination whether the lineup was unduly suggestive."

One hundred years after Hugo Munsterberg made his pitch that the court open its doors to the sciences, statements like the above make it clear that the doors remain largely shut.

Read more :: Permalink

Monday, April 2, 2007

A "Policy Review" of Eyewitness ID Procedures

The Justice Project just released a Policy Review, in which it advocates for a collection of reform measures relating to the manner in which police conduct lineup procedures.

Some of the well-settled best practices recommended in the Policy Review include:

The careful documentation of the identification, which decreases possible manipulation of witness certainty. Often, a witness’s initial confidence in an identification is rather low and provisional, but reinforcing statements and behaviors from authorities can exaggerate that certainty -- sometimes greatly -- by the time the witness testifies at trial. Documenting a witness’s "certainty statement" prior to any such feedback helps the jury to assess the eyewitness evidence appropriately.

Here's a question: if police fail to document an eyewitness identification procedure, why should the results of the mystery procedure be admissible in court? Courts don't let in DNA test results without procedural records, and the same standard should apply to eyewitness evidence collected by way of police procedures. New Jersey's high court agrees -- in State v. Delgado, it exercised its supervisory authority under the NJ constitution to require detailed records of an ID procedure as a prerequisite to admissibility, "to enhance the reliability of the factfinding process in [its] courts." Delgado, 902 A.2d 888 (N.J. 2006).
The use of cautionary instructions. Prior to presenting the lineup members, the eyewitness should be instructed that the perpetrator may or may not be included in the lineup. Cautionary instructions remove some of the pressure on the eyewitness to choose a suspect when the culprit may not be in the lineup.

According to a meta-analysis of available psychological research, cautioning witnesses with this simple instruction prior to an ID procedure reduces misidentifications by 41.6% in lineups in which the real culprit isn't present. Connecticut's high court used its supervisory authority to require this procedural safeguard, in State v. Ledbetter (PDF). New Jersey followed that decision in State v. King earlier this year, as reported here. Other courts would be well-advised to follow suit.
The effective use of fillers. Fillers, or non-suspect individuals presented to an eyewitness as part of a lineup, if chosen correctly, allow authorities to judge the reliability of an eyewitness. The effective use of fillers is critical to ensuring that an innocent individual is not identified simply because of the composition of the lineup.

Using fillers in a manner that doesn't make the suspect stand out seems like a no-brainer, yet we still have cases like Antonio Beaver, exonerated just last week after over 10 years in prison, who was wrongfully convicted based on a faulty eyewitness identification procedure in which he was the only member of the lineup with abnormal teeth, where the witness described the culprit as having a "David Letterman-like gap." Mr. Beaver didn't commit the crime, but he stood out in the lineup as the only one with abnormal teeth, a trait he just happened to share with the real culprit. A poorly designed lineup and the force of one eyewitness's testimony landed him a prison sentence of 18 years, over 10 of which he served before DNA testing exonerated him. A simple procedural safeguard could have prevented this inexcusable injustice, and unfortunately Mr. Beaver's case is not unique.
The use of an administrator who does not know which person is the suspect. Also called "double blind," this procedure prevents well-intentioned officials from giving inadvertent clues to the witness as to which person in the lineup is the police suspect.

Every scientific experiment involving human subjects is conducted in double-blind format for a reason: when humans know the desired outcome, they tend to betray that knowledge, just like Oskar Pfungst and his dutiful horse, Clever Hans, the "counting" horse. Eyewitnesses respond to human cues just like horses do, and just as Clever Hans lost his ability to count when Dr. Pfungst was concealed behind a wall, so will the likelihood that eyewitnesses will incorrectly identity a police suspect when the lineup administrator doesn't know the identity of that suspect.

And finally:
The presentation of the lineup members "sequentially" (one at a time) rather than all at once. This procedure enhances accuracy by reducing the tendency for "comparison shopping" in favor of a more direct assessment of whether the lineup members match the witness’s memory of the perpetrator. It also takes pressure off the witness to choose a suspect even though the attacker might not be in the lineup.

It is worth noting that the Innocence Project and other advocates endorse the sequential lineup methodology only when it is coupled with a blind administrator. Otherwise, the Clever Hans risk becomes even more pronounced, when cues from the lineup administrator can only be interpreted as being directed at the single person being observed by the witness at a given time.

Lots of recurring themes in the Justice Project's review, but maybe if they're repeatedly frequently enough, more law enforcement agencies, courts, and policymakers will begin to take heed. The full report is here (PDF).

Read more :: Permalink

Another Bad ID, an Unlikely Exoneree, and the Problem with DNA

Last week another faulty eyewitness identification was exposed, and another wrongfully convicted man was set free, thanks to the continued good work of the Innocence Project. Antonio Beaver spent more than ten years in Missouri prisons, after being convicted for a carjacking he didn't commit.

In some ways Mr. Beaver's case is much like other wrongful-conviction cases in which the (eventual) exoneree was convicted based on a bad eyewitness identification procedure. The police placed him in a live lineup in which he was one of two people wearing a baseball cap, and the only person with noticeable abnormalities to his teeth. The witness had described the perpetrator as an African-American man wearing a baseball cap with a gap in his teeth. Not surprisingly, he was selected from the lineup, and was convicted by a jury with no other evidence linking him to the crime.

What's different about Mr. Beaver's case is the nature of the crime for which he was convicted. In cases of carjacking and other similar offenses, it is very rare that testable DNA evidence is recovered from the crime scene, and thus it is similarly unlikely that those wrongfully convicted of similar crimes will have any hope of exoneration. However tenuous the foundation for the eyewitness evidence in the case, once a jury is convinced that the witness "will never forget that face," the defendant's fate is typically sealed. Mr. Beaver's sole good fortune in his unfortunate case was that the real perpetrator was observed to have bled on the door of the car, and a sample of that blood was preserved and eventually tested, clearing Mr. Beaver from any involvement.

But Mr. Beaver's case brings into focus an important fact that is often overlooked in discussions of DNA exonerations and the causes underlying wrongful convictions. While robberies rarely result in testable DNA evidence, they account for almost four times the number of arrests every year in the United States, as compared to rapes -- where DNA is often available for testing. More to the point, in a study (PDF, see p. 530) that examined all known cases of misidentification prior to the advent of DNA testing, misidentifications in robbery cases outnumbered those in rapes by more than two to one. Since the advent of DNA testing, however, exonerees falsely accused of rape account for more than 20 times the number of those falsely accused of robbery.

The difference? Robbers don't typically leave enough DNA at the crime scene for testing. But there's no reason to think any fewer people are wrongfully convicted of robberies now than before the advent of DNA testing. In fact, if we assume the ratio of misidentifications in robbery cases as compared to rape cases hasn't changed, and we make the (substantial) assumption that every person wrongfully convicted of rape between 1983 and 2003 was eventually exonerated, we are left to conclude that over 17 people per year are wrongfully convicted of robbery, or over 300 since 1989. In fact, only 6 of the exonerees between 1989 and 2003 were convicted of robbery. (See this Gross et al. article chronicling exonerations in the U.S., both pre-and post-DNA testing.) In the absence of DNA evidence to clear them and against the well-known force of a confident-but-wrong eyewitness at trial, their chances of walking free before their sentences are complete are small indeed.

All of this goes to illustrate that DNA is not enough to protect the innocent, because it only helps the "lucky" few. Which means that we need to focus on fixing the policies, procedures (PDF), and laws governing the collection and use of eyewitness evidence in criminal trials, and stop wrongful convictions before they happen in the first place.

Read more :: Permalink

One, Two, Three: DNA Exoneration Of Anthony J. Capozzi

The exoneration of Anthony Capozzi for a crime in Buffalo is just the latest in a long string of DNA exonerations for convictions based primarily on eyewitness testimony. In Mr. Capozzi's case, three eyewitnesses confidently identified him at trial as the rapist.

The quote highlighted by Talk Left is the most telling part of the story:

“Maybe one’s wrong, maybe two are wrong, but jurors sit there and say how can three people possibly be wrong?” he said. “They were all so positive it was

But there is another quote that struck me too. At the time of his trial, Mr. Capozzi sent a warning out to the people in Buffalo:
I said women in Buffalo shouldn't take comfort in the fact that Anthony Capozzi was convicted of these crimes," he said. "He didn’t do them. I suggest that you not run in Delaware Park alone. Because the person who did these is still out there.

Twenty-two years in prison for a crime he did not commit. Plus no increase in public safety as the real culprit remains free. The cost of mistaken eyewitness testimony.

Read more :: Permalink