Thursday, June 12, 2008

Recording full interrogations would have saved a lot of grief in Austin's Yogurt Shop case

It's been 17 years since the infamous Yogurt Shop murders rocked the Texas capital, and as time passes it's becoming more difficult to determine where investigators went wrong in building their case against Robert Springsteen and Michael Scott.

But go wrong they did.

A combination of coerced confessions and new DNA evidence appear to be causing their case to crumble. The Austin Statesman's Steven Kreytak reports ("Yogurt defense lawyer lashes out at prosecutors," June 12):

A lawyer for Robert Springsteen, facing retrial in the 1991 Austin yogurt shop slayings of four teenage girls, chided prosecutors Wednesday for failing to disprove his contention that DNA evidence disclosed this year exonerates his client.

Joe James Sawyer spoke outside court after an otherwise uneventful pretrial hearing in the case of Springsteen's co-defendant, Michael Scott, whose lawyers say also is cleared by the DNA evidence.

Both men disavow confessions they gave Austin police in 1999, saying they were coerced. An appeals court threw out their previous convictions, saying Scott's confession was improperly allowed at Springsteen's trial and vice versa.

They stand accused of capital murder in the deaths of Amy Ayers, 13; Eliza Thomas, 17; and sisters Sarah and Jennifer Harbison, 15 and 17. The girls died during a robbery at the I Can't Believe It's Yogurt store near Northcross Mall in North Austin.

In April, prosecutors said that in preparation for the retrials, they ordered new DNA testing of swabs taken from Ayers' body. Prosecutors said previously undiscovered DNA was found and it did not come from Scott, Springsteen or two men initially charged as co-defendants: Forrest Welborn and Maurice Pierce.

Prosecutor Gail Van Winkle said at the time that the discovery does not prove that Scott and Springsteen are innocent and said prosecutors were having testing done to determine the source of the newly discovered DNA.

On Wednesday, she said that testing had not been completed.

Sawyer said they are moving too slow.

"The law has passed them by," Sawyer said, referring to the appeals court rulings.

"The science has surpassed their arguments," he said of the DNA evidence.

False confessions are much more common than most people think, and in this high-profile case they were a particular problem. More than fifty different people confessed to the Yogurt Shop murders! By definition most people who confessed weren't involved. In that context, virtually any confession becomes suspect.

Now that DNA evidence found at the scene failed to match any of the suspects in the state's theory of the crime, the case against Springsteen and Scott appears to be wearing awfully thin, especially since the appellate courts have said they can't use one another's confessions against each other.

The critical piece of evidence in the case - Michael Scott's confession - was videotaped, but only the end result, not the full interrogation which lasted more than 22 hours. The Court of Criminal Appeals later agreed it should be excluded. If the entire interrogation were recorded, we'd either see a coercive situation were the defendant was pressured to confess, or clear evidence the police did the right thing. We wouldn't be wondering after the fact whether coercion took place: We'd know.

I'm learning in a book I'm currently reading, Police Interrogations and American Justice by Richard Leo, that part of the reason false confessions happen is that US police are trained to use an inquisitory instead of an information gathering approach to interrogations like police in other countries. Under the American model, the investigative target is accused of the suspected offense as a routine part of the investigation. Denials are cut off by interrogators, sometimes angrily. Crime details or even false information may be fed to suspects in order to trap them into a lie.

According to Leo, about half of all criminal convictions (45-55%) result from confessions to police. "American detectives assume that the suspects they interrogate will initially lie to them and deny their guilt," he writes. "Detectives assume suspects will lie because it is not in their rational self-interest to provide the state with testimonial evidence that almost certainly will lead to his arrest, prosecution, conviction and incarceration. The modern methods of psychological interrogation have been designed to break down a suspect's denials of guilt by persuading him that he has no meaningful choice under the circumstances but to comply with detectives, and that contrary to all appearances, logic, and common sense he is better off by confessing."

So applying that observation to the current instance, Austin detectives used these inquisitorial techniques to aggressively accuse hundreds of possible suspects during their investigation, in many cases refusing to let them out of the interrogation room for many hours. The questioning was so aggressive that many innocent people threw up their hands and agreed, "I did it." That's how you get dozens more people confessing to a crime than could actually have been involved, as in the Yogurt Shop case.

Remember, if Springsteen and Scott didn't do it, that means the real killers may still be out there.

Recording interrogations helps prevent wrongful convictions and protect officers from false allegations. In most cases it actually helps the prosecution more than the defense, at least when they've accused the right person. Given how cheap recording technology has become, I can see little justification for not routinely recording what happens in police interrogation rooms. Not only does the practice prevent false accusations of coercion or misconduct during questioning, it captures tidbits of information that might be missed if an interrogating officer must write up the event from memory or notes afterward.

It's just impossible to know what went on in the hours before Michael Scott's confession, and that's the main reason these cases appear headed for the rubbish pile.

If the pair is actually innocent, it's taken far too long to figure it out and a tremendous injustice has been done to them. If they're guilty, the failure to record what happened in that interrogation room back in 1999 will most likely be the reason the state can't convict them.


Michael said...

Grits: can you refresh my memory of the wrongful convictions based on coerced confessions in the Pizza Hut case out of Austin?

Gritsforbreakfast said...

Not offhand.

Anonymous said...

Michael -

Richard Danziger and Christopher Ochoa were convicted of the murder of Nancy Depreist after being interrogated by Hector Polanco (believed to be the first officer that ever spoke with Maurice Pierce of the Yogurt Shop case). The convictions were later overturned when the actual murderer "doscovered God" while in prison on an unrelated case and confessed to the DePriest murder.

Gritsforbreakfast said...

You're right, 12:51, I didn't remember the location. But that story, particularly what happened to poor Richard Danziger, is truly the stuff of nightmares. Unimaginable.

Michael said...

Thanks, 12:51. That's the "Shut your face" cite for the know-it-alls who proclaim the infallibility of confessions.

SB said...

Police depts have the video and audio equipment and it gathers dust. Such recordings would stop the " your buddy told us what you did' and " we have proof" and "Tell us what you did and the judge may go easier on you". It is the prosecutors who offer a deal that one cannot refuse " Take it to court and you will be an example to show others why this is a bad idea." I have seen all of this and more.
A 14 year old accused her father of sexual assault. I didn't take long for the police dept to get him to confess. Police said they had evidence and he did not think he daughter would lie. Upon leaving the jail he went directly to a mental hospital where he confessed repeatedly and wanted the harshest term possible. He told his court appointed attorney that he was guilty and was directed to start sex offender classes which are only available with another confession of guilt.
As time passed he began to question his guilt and then the girl's diary was found. It told of a long term that was being used to get CPS involved and get her out of the house. This diary brought into question whether the incident ever happened.
The diary was rushed to the court appointed attorney who turned it over to the DA's office who mailed it to the accuser. There are records of this transaction.
Guilty or not guilty? We aren't 100% certain and nobody else cares. Meanwhile, he is serving 3 20 year sentences. Nobody bargained for a lower sentence and those confessions were not going to be allowed to go to waste. DNA does not help in these type cases.

Michael said...

I think some of the interrogation techniques you mentioner are legitimate, by the way. While I oppose coercion, I don't have objections to the "your buddy confessed already" technique.

Gritsforbreakfast said...

I'm coming to believe, Michael, that the problem with that tactic is it's part of a broader accusatory approach where police are trying to "trick" someone into confessing instead of gathering evidence in a more neutral fashion.

In particular, the practice of cutting off a suspect's denials with lies such as "Don't tell me you didn't do it, your buddy already confessed," opens the door wide to invite false confessions. I see the tactic as problematic not from an ethical standpoint because of lying per se, but because it encourages the Chris Ochoas of the world to falsely accuse the Richard Danzigers, etc..

Finally, discussing Leo's book, a detective recently told me that he didn't like to lie to suspects because if they catch an interrogator in a lie the cop loses credibility and the interrogation (and possibly the case) is over - "your buddy confessed" might work in the short term, but sometimes lying can backfire.

The flip side is, if investigators don't cut off denials and let suspects talk, recording information instead of demanding a confession, they gather data for further investigation that can either be confirmed or refuted and build an actual case based on evidence, not just a coercive interrogation.

Anonymous said...

As someone who has done multiple interrogations, I think it is very naive to think that simply asking neutral questions will get confessions. Almost every interrogation I have been in has started out with denials. Many are done despite real clearcut evidence being shown to the suspect.

Yeah we try to trick the bad guys into admitting their guilt by making them think that they are already caught. I've tried the "I have you on tape" approach with innocent people and their reactions are totally different from that of guilty people.

In addition, if you read up on the Reid technique of interrogation, it doesn't initially call for cutting of denials. You question the suspect first with a variety of questions. Once you get certain answers that are guilty indicators, then you start the actual interrogation.

You talk about using actual evidence instead of confessions and I agree that is the bet route. However, in many cases there is not enough evidence. Sometime the police get lucky with prints, murder weapons or whatever but this is not CSI.

What about a rape case in which the suspect didn't ejaculate or leave any DNA? There is no tape but the victim provided a good description and the suspect was located nearby. If the victim identified the guy, what approach do you recommend the police use?

I do agree that all interrogations should be recorded. This will keep the suspect from trying to recant any statements made (and this always happens).

Gritsforbreakfast said...

FWIW, Todd, I didn't say asking neutral questions would get confessons. I said it would gather information. Personally I think it's naive for you to believe the way you work now is the only way things can happen. Many police in other countries don't use inquisitorial interrogation techniques and still solve crimes.

Re: "Guilty indicators," the interrogator is neither judge nor jury. I'm sure Chris Ochoa gave off some "guilty indicator," and look where it got him and Richard Danziger. The Reid technique is a significant cause of false confessions, IMO, and it's not the only possible interrogation approach.

Finally, your example of a rape case with no evidence is the perfect point. Accusing people can cause innocent ones to confess (dozens in the yogurt shop case). Most of the false conviction cases getting out on DNA were rape cases where some cop was sure he had the right guy, pressed hard, and either got a false confession or pressured a witness into falsely accusing. When you have no evidence, the risk of false confession is that much greater and makes these tactics more dangerous for the innocent.