Backing Up

Date Published 06.03.15

I was once forwarded an email written by a retired California police lieutenant. Details in the email were supposed to help clarify the chronology of a cold case investigation. The timeline information was helpful, but what stuck with me was an aside the lieutenant made, a lament about disorganization and missed opportunities, about the tendency of cops to go it alone and not share what they know.


“Controlling cops is a lot like herding cats,” the lieutenant wrote. “Like blind hogs they sometimes find an acorn, but mostly they just tear up the ground and muddy the waters.”


Muddy waters are, of course, the plague of unsolved cases, and the murk thickens from any array of sources, not just disorganized and uncommunicative cops. Well-meaning tipsters can flood the phones with misleading information. A single wrong detail, that a suspect was in custody when he was actually out on furlough, for example, can derail a case for years.


Robert Keppel is a former homicide detective who helped apprehend Ted Bundy. In The Riverman, Keppel’s book about Bundy and Gary Ridgway, the Green River Killer, he says they had Bundy’s name as a possible suspect within days of the abductions of two women from Lake Sammamish on July 14, 1974, but Bundy wasn’t arrested until more than a year, and a dozen victims, later. Keppel points out it’s not uncommon for a killer’s name to be buried in the files somewhere within the first two weeks of a homicide investigation.


I’ve been thinking about Keppel’s observation in light of three high-profile cold cases that have recently experienced apparent breakthroughs. Two of the cases are not only decades old but are also child abductions in which the bodies were never recovered, making them particularly challenging to solve. In these two cases it’s not clear the police and prosecutors’ theories are correct, but in both of them the name of the suspect was in the police files --- reports filed forty years ago and thirty-six years ago, respectively --- within days of the crime.


The third case isn’t as old, the bodies were recovered, and strong forensic evidence ties the suspect, who’s in jail awaiting trial, to the crimes. But after reading up on the case I wondered why the man’s name wasn’t flagged earlier, after the first murder. He was in a relatively small pool of people that police would have had an obvious interest in tracking down and interviewing.


If they had talked to him, and taken the extra step of looking into his past, a young co-ed might still be alive.


First case: The Lyon Sisters, 1975


Sheila Lyon, 12, and her sister Katherine, 10, vanished on Tuesday, March 25, 1975 during a trip to a local shopping mall near their home in suburban Washington D.C. It was the girls’ spring break and their plans were low-key and ordinary: see the Easter decorations, eat some pizza, and maybe shop a little. They left for the half-mile walk to the mall in late morning with strict instructions from their mother to return home by 4 p.m. The girls never returned. They were never seen again.


The disappearance of the Lyon sisters has been one of the most haunting and well-publicized unsolved cases in the D.C. area. In retrospect the case may have been hindered from the start by one of the most common problems that misdirect investigations, particularly high-profile ones: the vivid red herring.


Almost immediately witnesses stepped forward and told police they saw the two girls speaking outside the mall to an unidentified older man in a brown suit. He was estimated to be between 50 and 60 years old and was carrying a briefcase with a tape recording device in it. The Lyon sisters, along with other children, appeared to be speaking into a microphone at the man’s request.


Tape Recorder Man quickly became the police’s prime suspect, especially after accounts came in of what sounded like the same man at other local malls, approaching children and asking them to record an answering machine message.


A massive investigation was in full chaotic force a week after the sisters’ disappearance when Lloyd Welch, 18, came forward and told the mall’s security guard that he’d been there that day and witnessed the girls get into a car with a man and leave. Detectives questioned Welch; something about his account stirred their interest enough to give him a polygraph. He failed. At the time Welch was a ne’er do well carnival worker. He was frequently homeless. It’s easy to imagine the impression Welch made on detectives. How would this disheveled, hitchhiking punk successfully pull off abducting two schoolgirls in broad daylight? Tape Recorder Man was older, better dressed, and looked like he had a well-honed ruse. He was certainly their man. The page on Welch was flipped. The investigation moved on.


Age may have been a factor that contributed to police overlooking Welch and concentrating on Tape Recorder Man. The “dirty old man” stereotype of child molesters persisted for a long time. Research into the subject, and subsequent training of law enforcement, has led to a better understanding of these kinds of offenders. The truth is that 30 to 50 percent of all child molesters are adolescent males.


Tape Recorder Man was never identified. The Lyon case went cold. Wheaton Plaza was eventually renamed Westfield Wheaton. Forty years later a new detective assigned to the Lyon cold case sat down and began reading from the beginning. He came across Lloyd Welch’s statement and, his curiosity piqued, looked into him. Welch had lived a largely itinerant lifestyle as a ride operator for a traveling carnival. He also had a long rap sheet. Not just long, but interesting: a history of convictions involving sex crimes against young girls.


Old mug shots are sometimes hard to acquire even for active detectives, so it’s unclear when the current team of detectives that began looking into Welch first saw his mug shot for a residential burglary in 1977. I’m sure they experienced a spike of adrenaline when they saw it. Here’s Welch’s mug shot next to a sketch based on a witness’s description of a man who was staring at the girls and following them at the mall the day they disappeared.





The more detectives learned about Welch the better he looked as a suspect. Luckily they knew where to find him. He’d been incarcerated since 1997 in Delaware for sexually assaulting a 10-year-old girl. Detectives went and interviewed him there several times.


According to police affidavits, Welch admitted to detectives that he left the mall with the Lyon sisters in a vehicle that day. But he pointed the finger at his uncle, Richard Welch, characterizing him as the main culprit in the kidnappings. Lloyd Welch insisted he was dropped off at home and his uncle and a younger family member drove away with the girls. Welch told detectives he saw his uncle sexually abusing one of the sisters the next day at his uncle’s home. After that, he said, he never saw the Lyon sisters again.


Though no charges have been filed, a flurry of law enforcement activity began last autumn after Lloyd and Richard Welch were publically named persons of interest. Police began searching land owned by the Welch family in rural Bedford County, Virginia. A grand jury was convened to investigate, and Richard Welch’s wife, Patricia, was charged with perjury related to her testimony.


The strength of the evidence against Lloyd and Richard Welch is unclear, and the case appears to be in limbo for the time being. But the question remains, why was a young man that admitted to being at the mall, closely resembled one of the suspect composites, and failed a polygraph dismissed at the beginning of the investigation? Why did it take forty years to become interested in him again? My best guess is that someone with influence in the department decided back in 1975 that Welch couldn’t have pulled off this kind of crime. A dutiful detective back in the day made some kind of note next to Welch’s name in the report, maybe something as simple as “no” or “eliminated.” Years went by and the original team of detectives retired, or died, or forgot altogether why it was the unkempt kid didn’t raise more flags. Especially in the early days of the highly publicized case, when phones were ringing and tips flying, the Welch lead got buried and overlooked.


I understand how that can happen, but it seems to me there should be a system in place for cold case investigations. Suspects, dubious witnesses --- anyone that sets off even the tiniest alarm --- should be prioritized and kept on a special list that's set aside. The list should be reviewed, even if it’s just once a year, and the names researched and run through the criminal databases. Had that system been in place on the Lyon case, Lloyd Welch likely would have been a suspect much earlier, possibly within two years. He was arrested for burglary in 1977; that type of crime doesn’t necessarily suggest an inclination for child abduction, but it does reveal criminality. A follow-up interview with Welch, who could have been looking to negotiate a deal in his burglary arrest, may have yielded answers.


The difference between two and forty years in a criminal investigation is immense. In March the media reported that investigators returned to dig again on a mountain in rural Virginia where they believe the girls may be buried. None of it --- the mountain, failing memories, destroyed or lost records --- will be easy to dig through now.


Coming Up: Cases Two and Three

Also: I was recently a guest again on the excellent Jackie Kashian's podcast The Dork Forest.




The Feed

RT @emilynussbaum: The artful @hodgman's straightforward case for Hillary:
@Twaikuer @pattonoswalt @daveanthony Know what he does believe in? PAC $. Took 10K from HRC pac 2006. That means he's in her pocket.#BSLogic
@Twaikuer @pattonoswalt @daveanthony Good one. Unfortunately Bernie on record as not believing in charity.
@johnlevenstein Thanks for asking, btw. That's the kind of elevated discourse missing lately. A lot of mud slinging. #I'mNotAboveItEither
@johnlevenstein Can't convey it all thru Twitter but yes, she has flaws. Too poll-driven, burned needed bridges, trouble owning mistakes.