Committed in 2012. Solved in 2019.
A seven-year-old Grand Traverse County robbery case solved this year demonstrates the promise — and the limitations — of forensic science.
By Patrick Sullivan | Dec. 14, 2019
When a man passed a note through a Walgreens drive-thru saying that he’d rigged the store with a bomb and needed all the morphine and Ritalin the Grand Traverse store had, an alert pharmacy clerk and quick action by detectives soon led to all the evidence needed to identify a suspect and connect him to the crime.
The clerk handed over one bottle of each of the drugs, saying that was all they had, and she jotted down the license plate as the Pontiac Grand Prix sped away. Her effort helped police quickly discover that the vehicle had been stolen. Soon after, they found it abandoned on a nearby side street.
Fingerprints lifted from inside the car would later connect the crime to a suspect with a criminal history — Jason David Cook, now 42, a man was on probation in Grand Traverse County for felony theft charges when he held up the Walgreens.
Open-and-shut case, right?
The evidence was gathered immediately after the May 2012 robbery, but Cook was not tied to the crime until seven years later, when a scientist at the Michigan State Police Grayling Crime Lab matched the latent crime scene fingerprints to Cook.
‘GETTING HITS LEFTS AND RIGHT’
How did it take so long to match the Grand Prix fingerprints to Cook’s fingerprints, which, in 2012, would have been already included in a database of anyone who had been arrested.
The reason for the delay was not because of a backlog at state police forensic science crime labs, said Connie Swander, director of the Grayling crime lab, which is one of only eight crime labs in the state (and which serves state and county police departments in 26 northern Michigan counties).
The seven-year pause was due to a far more frustrating reason, one anyone who relies on computers to do their job can understand: The software was faulty. It often missed matches, Swander said.
State police fingerprint analysts across the state, all of whom used the same program, recognized the issue, and the state switched to another software vendor in late 2018. After the switch, the analysts began to re-run latent prints from cold cases. They got lots of hits.
“Over the years they have found out that, like any technology, if somebody doesn’t keep it up, for whatever reason, it starts not being as good. It’s missing a lot of things,” she said. “Now that we’ve got the new system, and we fed all of our old samples that have been in there for years over to the new system, and we’re getting hits right and left.”
That doesn’t mean that old cases are being solved left and right. In many instances, investigators had solved the cases through other means. But the new fingerprint analysis system is solving some cold cases, like the Walgreens holdup and, possibly, a cold-case homicide that Swander said she could not discuss because the case is still open.
A RAPIDLY DEVELOPING FIELD
Fingerprints have been used in criminal investigations since the 1890s. Forensic DNA dates back only to the 1990s, but it’s nonetheless overtaken fingerprints by a wide margin: Today, cells can be sequenced in a way that a DNA result can become something like a barcode, an exact mathematical expression of a person’s identity. Meanwhile, fingerprints remain a swirling maze of lines that are much more complicated to express scientifically.
The Walgreens case shows just how far forensic science has come in a short span of time. It is an ever-evolving science, prone to unpredictable developments and unfortunate setbacks.
Swander worked as a medical technologist in a hospital before she joined the state police crime lab in 1985. She started in a department that was then, in the pre-DNA days, called the serology department (serology is the diagnostic examination of blood serum), and is today called the biology/DNA department. She was promoted to lab director in 2003.
Over those years, Swander has watched the capabilities of forensic science expand exponentially.
In the early days, latent fingerprints couldn’t be run through a database of people who had been arrested for crimes because there was no database; instead, there were binders and stacks of fingerprint cards.
“When I first started here, we didn’t even have a computer,” Swander said. “Now we’re completely electronic. All our reports are done electronically. You know, your computer is like, attached to your side. I’ve seen a ton of changes.”
Early in her career, the crime lab tested blood to determine whether or not it was human and then ran an ABO blood-type analysis so that suspects could be either ruled out or ruled in — but not positively identified.
Similarly, fingerprint analysis was more art than science. Fingerprints had to be matched by a trained eye concentrating on two patterns under a magnifying glass.
That final aspect of fingerprint matching remains true today; fingerprint matches are still made by a trained professional who optically compares two fingerprints. Nowadays, however, a computer program tips off the analysts about which fingerprints in the database should be compared to which, making more matches possible than would have been imaginable when Swander started in 1985.
“Back when I came in, you had your question print, and you actually have to have a print card from everybody that you wanted to compare,” Swander said. “Now, they can have a question print and can search it against hundreds of thousands of knowns that are in there.”
Developments in biology have been even more mindboggling during Swander’s career. She was there for the introduction of DNA into forensic science, and she has watched it become more precise with each passing year.
When DNA was first used in criminal prosecutions in northern Michigan in the late 1990s, there were a lot of limits on what the evidence could do.
“You needed a blood stain about the size of a nickel to be able to get a sufficient amount of nuclear DNA to get a result on your sample. Nowadays we can take a spot the size of a pinhead, and we are able to get a DNA pattern,” Swander said. “At first, DNA was maybe a step up from [blood-typing]. It could give you some statistics. But as that technology has evolved and gotten more and more finetuned, we’ve gotten down to where you can get numbers as high as one in six quadrillion, the chance of someone else having that same DNA pattern, which is greater than the population of the world, which is therefore a pretty good indicator that you’ve got the right suspect.”
BACKLOGS AND NAGGING PROSECUTORS
Staff at the Grayling lab attempt to get results turned around as quickly as possible, but there are a lot of forces that work against them. For instance, when a crime scene needs to be studied, the techs have to leave their workplace and travel somewhere, often over great distances, to collect evidence.
For most of the forensic disciplines, turnaround time runs up to 30 days. Forty days is considered late.
But DNA is different. Advances have made it so much more useful and commonplace that there is much greater demand across the state for DNA test results, and those results can take much longer, partly due to the fact that the scientific tests themselves can take weeks.
It is delayed in another way because DNA testing is only conducted at three locations in the state — at DNA labs in Grand Rapids, Northville, and Lansing. There are eight other crime labs across the state similar to the Grayling lab, which each include four departments: a drug identification unit, latent prints, firearms ballistics, and biology.
What sets Grayling apart from the other labs, perhaps, is that it serves 26 counties in northern lower Michigan and two counties in the eastern U.P. There are so many police agencies in that vast area that there is always pressure to process evidence quickly.
“We constantly have this pull,” Swander said. “Prosecutors are just like doctors. I worked in a hospital for 18 years, and every doctor wants their lab work done first, before all the other ones. And I have found prosecutors are the same way. They want their cases done first before any other prosecutors in the state. So, we have that to deal with, too.”
What’s more, now that DNA can be extracted from invisible cells left behind on anything a person touches, demand for DNA tests from police departments has gone up, and backlogs for DNA tests have increased.
While the outlying crime labs don’t conduct the actual DNA tests, they are critical to making sure that DNA can be used as evidence because they are trained in collecting the samples needed from crime scenes and maintaining “chain of custody” that assures the integrity of the evidence doesn’t come into question.
“That’s the most important part of it,” Swander said. “The case is only as good as the evidence that you get on it, and the evidence is only as good as the chain that you can maintain on it.”
Across all of the disciplines, submitted evidence at forensic labs is first-come, first-serve, unless there is good reason to make something a priority, like a test needed to identify a murder suspect or if something is urgently needed for an upcoming court date.
Swander said crime labs have had to become more discriminating about what DNA evidence they will accept. That means violent crime always takes priority over property crime.
Swander said the rate at which forensic science has developed over her career is astonishing, and she can’t imagine what the future holds.
“I’ve been in 34 years. I can’t even imagine — I won’t be around, but I can’t imagine — what it’s going to look like in another 34 years,” she said.
DETAILS FADE AWAY
The technological glitch that delayed justice in the Walgreens robbery case for seven years also presented some obstacles that had to be overcome by the team responsible for prosecuting Cook, the suspect.
Memories had grown distant in the seven years since the note had been handed over at the drive through.
“Obviously, when we got the latent fingerprint hit fairly recently, we were wondering as to why it didn’t hit years prior, because it’s not like Mr. Cook has a clean record, by any means,” said Grand Traverse County Sheriff’s Capt. Randy Fewless, who oversaw the investigation. “I guess better late than never.”
Fewless said the biggest challenge posed by such a delay is the fragility of witness’ memories.
Ultimately, Cook pled guilty to larceny in a building, and he was sentenced to nine months in jail.
Kyle Attwood, the assistant prosecutor who handled the case, said he couldn’t say if the case would have turned out differently had it been prosecuted in 2012, but he said the lag definitely presented challenges.
“Witnesses’ memories fade, and the ability to conduct any necessary follow-up is impaired,” he said in an email. “I can’t say for certain what would have happened if we’d been able to charge the case immediately; however, I'm confident we would have had an easier time meeting our burden.”
When Thomas Holcomb, the lead fingerprint examiner at the Grayling lab, testified at the preliminary hearing, he had to overcome questioning from Cook’s defense attorney, Janet Mistele, which suggested that, when he confirmed the fingerprint match in 2019, he was coming to a conclusion that differed from the one reached by an examiner in 2012. In fact, he testified, the latent prints were never compared to the known Cook prints until 2019.
At the hearing, Holcomb explained how, in recent years, fingerprint examiners had come to realize that the computer matching system was no longer working properly, prompting the switch to the new system in November 2018.
“We knew there were problems, and that’s why we ended up going to a new vendor,” he said.