DNA is Not Infallible

DNA is portrayed in movies and real life as the gold standard of evidence. DNA found at a crime scene is incontrovertible proof that the persons whose DNA matches the sample was there. But, what if the odds for certain DNA matches were really only 1 in 3?

This was the case in a recent California rape conviction of a 70-year old man for a crime committed over 30 years ago. Without an eyewitness, without a medical examiners testimony, without the other likely suspect having his DNA tested, and without the victim herself identifying John Puckett as her rapist, he was convicted based solely on flimsy evidence acquired through DNA data mining.

How could that happen? Simply, the prosecution suppressed the real odds and the jury was told the odds of a false match were considerably smaller than 1 in 3. They were told it was 1 in 1.1 million.

The LA Times summaries the semantic tap dancing the prosecution had to go through to secure a conviction:

The chance that two unrelated people will share the same 13 markers can be as remote as 1 in a quadrillion -- a number with 15 zeros. Because the match in Puckett's case involved only 5 1/2 genetic locations, the chance it was coincidental was higher but still remote: 1 in 1.1 million.

But Barlow thought this figure vastly exaggerated the strength of the evidence. It did not take into account how Puckett had been identified: through a search of a large database.

The general-population figures used by prosecutors portray the odds of matching crime-scene DNA to a single, randomly selected person.

But because database searches involve hundreds of thousands or millions of comparisons, experts say using the general-population statistic can be misleading.

Think of a lottery. If you buy a single ticket, your chances of hitting the jackpot are remote. If you buy many tickets, your odds improve with each purchase. In Barlow's view, the prosecution in effect bought hundreds of thousands of lottery tickets to find the match with Puckett. She contended that this greatly increased the odds of a match to an innocent person.

Barlow argued during pretrial hearings that the jury should be told about the recommendation of two leading panels of scientific experts, one convened by the National Research Council and the other by the FBI. Both committees settled upon a statistical remedy to adjust for the many individual comparisons made during a database search. It has been widely but not universally embraced by scientists.

In every cold hit case, the panels advised, police and prosecutors should multiply the Random Match Probability (1 in 1.1 million in Puckett's case) by the number of profiles in the database (338,000). That's the same as dividing 1.1 million by 338,000.

…

Through dueling experts, the prosecution and defense offered jurors a dizzying array of numbers to consider in weighing the DNA match.

A "likelihood ratio" presented by a prosecution expert placed the chance of a coincidental match at 1 in 1.7 million. A "combined probability of inclusion" put it at 1 in 152 billion.

The numbers all pointed to the virtual certainty that the DNA at the crime scene was Puckett's.

In an interview after the trial, Ranajit Chakraborty, the prosecution's DNA expert, told The Times that he generally favors giving jurors the database adjustment. He did not present an adjustment in this case because the judge, like most others, would not allow it.

The full article is well worth reading since it shows the danger of making convictions solely on DNA.

High-tech methods of crime scene analysis will make it important to be increasingly skeptical of the proof presented. It’s only a matter of time before DNA will be taken from every crime scene and analyzed. Around that time it will be the MO of the smart criminal classes to leave someone else’s DNA at the scene to throw the investigators off. The odds may be several million to one that DNA matches, but that shouldn’t mean guilt is certain.

Sleaze in LA Federal Prosecutors Office

Since bureaucracies do not operate with free market incentives, there is no objective way to measure their success or failure. Instead of being measured by profits, bureaucracies measure success by their size and reach. Unlike a profit maximizing business, a bureaucracy wants bloat. Efficiency is a dirty word to a bureaucracy. If they get too efficient they might cease to matter, or, even worse, they might not get to suck up as many tax dollars as before.

So a bureaucracy – any bureaucracy - will rely on proxies for efficiency and end up with workers gaming the measured metrics.

In this case, the prosecutors office concentrated on easy cases to hit their numbers. It remains unclear how many weak cases were pushed through for the sake of quotas:

U.S. Atty. Thomas P. O'Brien is facing sharp criticism from prosecutors within his office who say he is pressuring them to file relatively insignificant criminal cases to drive up statistics that make the office eligible for increased federal funding.

The prosecutors said O'Brien's effort to increase filings amounts to a quota system in which lawyers face possible discipline and other career consequences if they fail to achieve their numbers.
...

The disgruntled prosecutors in Los Angeles say they are now spending an exorbitant amount of time working on less significant cases -- mail theft, smaller drug offenses and illegal immigration -- to reach quotas. They cited the recent disbanding of the office's public integrity and environmental crimes section, a unit with a history of working on complex police corruption and political corruption cases, as evidence of a shift toward high-volume, low-quality prosecutions.

"It's all about the numbers," one prosecutor said.

One former supervisor put it this way: "I can't remember how they sugarcoated it, but the feeling around the office was, if you got your quota, then you could work on your real cases without being hassled."

Richard LaPointe - An Innocent Man Serving Life

The details of the case are shocking. Two years after the murder of Bernice Martin, Richard Lapointe, a mentally handicapped man, was snatched from his family in the middle of a picnic. For the next ten hours an unrecorded interrogation took place and three contradictory "confessions" were extracted. Mr. Lapointe was then jailed for over two years before the trial since he was unable to afford bail. He was convicted of murder.

It's seems clear why he was targeted. His handicaps made him an easy target for trained interrogators who had no physical evidence to rely on.

His supports have their own take:

They say the case is preposterous, that Mr. Lapointe, a victim of numerous mental and physical infirmities, was not capable of such conduct. They say to commit a crime of astounding brutality, Mr. Lapointe, who had no history of violence of any kind, would have decided to sneak out in a 45-minute interval while his wife and son were upstairs. He would have taken a short walk to Mrs. Martin’s apartment, raped her, stabbed her, strangled her with a ligature that required far more dexterity than he had ever shown and set the place on fire.

And then, with no one seeing him come or go, he would have returned home, with no wounds, no blood, no smell of smoke, in time to calmly watch a National Geographic special with his family.

Video taping interrogations was cost prohibitive in 1989. That is not the case today. With storage and equipment costs so low the police have other reasons to avoid taping interrogations today. If the jury saw the techniques that were used to get confessions, a written confession would lack its current stand alone veracity. You can bet that if the jury had to watch 10 hours of intimidation and bullying of a handicapped man, the verdict would have gone a different way.