Mindhunter

by John Douglas, Mark Olshaker

  From a car found near the shore, the three bodies were identified as Joan Rogers, thirty-eight, and her two daughters, seventeen-year-old Michelle and fifteen-year-old Christie. They lived on a farm in Ohio, and this was their first real vacation. They had already been to Disney World and were now staying at the Day’s Inn in St. Petersburg before returning home. Mr. Rogers didn’t feel he could spare the time away from the farm and hadn’t accompanied his wife and daughters.

  Examination of the dead women’s stomach contents, correlated with interviews from restaurant workers at the Day’s Inn, fixed the time of death to have been about forty-eight hours previously. The only tangible piece of forensic evidence was a scribbled note found in the car giving directions from the Day’s Inn to the spot where the car was found. On the other side were directions and a drawn map from Dale Mabry, a busy commercial street in St. Petersburg, to the hotel.

  The case instantly became a major news event, involving the police departments of St. Petersburg and Tampa and the Hillsborough County Sheriff’s Department. Fear among the public was high. If these three innocent tourists from Ohio can be killed like this, everyone reasoned, then anyone can be a victim.

  Police tried to follow up on the note, matching the handwriting against that of hotel employees and people in shops and offices around the area on Dale Mabry where the directions began. But they came up with nothing. The brutal, sexual nature of the killings, however, was alarming and indicative. The Hillsborough Sheriff’s Office contacted the FBI’s Tampa Field Office, saying, "We may have a serial case." Still, the combined work of the three police jurisdictions and the FBI produced no significant progress.

  Jana Monroe was an agent in the Tampa Field Office. Before coming to the Bureau, she’d been a police officer and then a homicide detective in California. In September 1990, after Jim Wright and I interviewed her for an opening in the unit, we requested her reassignment to Quantico. Jana had been a profile coordinator in the field office, and once she joined the unit, Rogers became one of the first cases she did for us.

  Representatives of the St. Pete police flew up to Quantico and presented the case to Jana, Larry Ankrom, Steve Etter, Bill Hagmaier, and Steve Mardigian. They then developed a profile, which described a white man in his mid-thirties to mid-forties; in a blue-collar, home maintenance-type occupation; poorly educated; with a history of sexual and physical assault and precipitating stressors immediately prior to the murder. As soon as the heat was off the investigation, he would have left the area, but like John Prante in the Karla Brown case, he might later have returned.

  The agents were confident of the profile, but it didn’t lead to an arrest. Little progress was being made. They needed a more proactive approach, so Jana went on Unsolved Mysteries, one of the nationally syndicated television programs that often have good results in locating and identifying UNSUBs. Thousands of leads were generated after Jana’s appearance and description of the crime, but still, none of them panned out.

  If one thing doesn’t work, I always tell my people, you try something else, even if it’s never been tried before. And that’s what Jana did. The note of scribbled directions seemed to be the one item linking the victims to the killer, but so far it hadn’t been very useful. Since the case was well known in the Tampa-St. Pete community, she came up with the idea of blowing it up on billboards to see if anyone recognized the handwriting. It’s accepted in law enforcement circles that most people will not recognize handwriting outside their immediate family and close friends, but Jana figured someone might well come forward, particularly if the subject had been abusive and a spouse or partner was looking for a reason to turn him in.

  Several local businessmen donated billboard space, and the note was reproduced for all to see. Within a couple of days, three separate individuals who had never met each other called the police and identified the handwriting as belonging to Oba Chandler, a white male in his mid-forties. An unlicensed aluminum-siding installer, he had been sued by each of these three people when their newly installed siding had come loose after the first heavy rain. They were so sure of the ID because each had a handwritten copy of his legal response to their charges.

  In addition to the age and profession, he fit the profile in other key areas. He had a previous record of property crimes, assault and battery, and sexual assault. He had moved out of the immediate area after the heat was off, though he hadn’t felt a need to leave the region. The precipitating stressor was that his current wife had just delivered a baby he didn’t want.

  And, as often happens once you can do something to break a case open, another victim came forward after hearing the details of the murder. A woman and her girlfriend had met a man matching Chandler’s description who wanted them to come out with him on his boat in Tampa Bay. The girlfriend had a bad feeling about the whole thing and had refused, so this woman went alone.

  When they were out in the middle of the bay, he tried to rape her. When she tried to resist, he’d warned her, "Don’t scream or I’m going to put duct tape on your mouth, tie you to a cinder block, and drown you!"

  Oba Chandler was arrested, tried, and found guilty of the first-degree murder of Joan, Michelle, and Christie Rogers. He was sentenced to death.

  His victims were ordinary, trusting people whose selection was almost random. Sometimes the selection is completely random, proving the frightening assertion that anyone can be a victim. And in situations like these, as in the Rogers case, proactive techniques become all-important.

  In late 1982, people were dying suddenly and mysteriously in the Chicago area. Before long, Chicago police came up with a connection between the deaths and isolated the cause: the victims had all taken Tylenol capsules laced with cyanide. Once the capsule broke down in the stomach, death followed quickly.

  Ed Hagarty, the Chicago SAC, asked me to come into the investigation. I’d never worked a product-tampering case, but as I thought about it, I figured that much of what I’d learned from the prison interviews and experience with a variety of other types of offenders should apply here, too. In FBI code, the case became known as "Tymurs."

  The primary problem facing the investigators was the random nature of the poisonings. Since the offender neither targeted a specific victim nor was present at the crime scene, the type of analysis we normally did wouldn’t reveal anything directly.

  The homicides were apparently motiveless—that is, they weren’t motivated by any of the traditional, recognizable motives such as love, jealousy, greed, or revenge. The poisoner could be targeting the manufacturer, Johnson & Johnson, any of the stores selling the product, one or more of the victims, or society in general.

  I saw these poisonings as the same type of act as a random bombing or throwing rocks down from an overpass onto cars below. In all of these crimes, the offender never sees the face of his victim. I pictured this offender—much like David Berkowitz shooting into darkened cars—as more concerned with acting out his anger than with targeting a particular type of victim. If this type of subject were ever made to see the faces of his victims, he might have second thoughts or show some remorse.

  Given the ready comparison with other random, cowardly crimes, I felt I had an understanding of what the UNSUB would be like. Even though we were dealing with a different type of crime, in many ways the profile was a familiar one. Our research had shown us that subjects who kill indiscriminately without seeking publicity tend to be motivated primarily by anger. I believed this guy would have periods of severe depression and would be an inadequate, hopeless type who would have experienced failure throughout his life in school, jobs, relationships.

  Statistically, the subject would probably fit the assassin mold—a white male in his late twenties to early thirties, a nocturnal loner. He would have gone to victims’ homes or visited grave sites, possibly leaving something significant there. I expected him to be employed in some position as close to power and authority as he could come, such as ambulance driver, security guard, store detective, or
auxiliary policeman. And he would probably have some military experience, either Army or Marines.

  I thought he’d have had psychiatric treatment in the past and have been on prescription drugs to control his problem. His car would be at least five years old, not well maintained but representing strength and power, such as the Ford model favored by police departments. Near the time of the first poisoning—around September 28 or 29—he would have experienced a precipitating stressor for which he may have blamed society in general, fueling his anger. And once the case became public, he would discuss it with whoever would listen to him in bars, drugstores, and with police. The power these crimes represented was a major boost for his ego, which indicated he might keep a diary or scrapbook of media coverage.

  I told the police it was also likely he’d written to people in positions of power—the president, the director of the FBI, the governor, the mayor—to complain about perceived wrongs against him. In early letters, he would have signed his name. As time passed without what he considered an appropriate response from anyone, he grew angry over being ignored. These random killings could be his way of getting back at all those who didn’t take him seriously.

  Finally, I warned against reading too much into the selection of Tylenol as the means of poisoning. This was a crude, sloppy operation. Tylenol was a common drug and the capsules were easy to open. It was at least as likely that he liked the packaging as that he had any particular grudge against Johnson & Johnson.

  As with serial bombers, arsonists, and other such cases, in a large city like Chicago many people would fit the general profile. Therefore, like the Rogers case, it was more important to focus on proactive techniques. The police had to keep pressure on the subject and not let him cope. One of the ways they could do this was by issuing only positive statements. At the same time, I warned them not to provoke him by calling him a madman, which, unfortunately, was already happening.

  More important than that, though, would be to encourage the press to print articles humanizing the victims, since the very nature of the crime tended to dehumanize them in the UNSUB’s mind. In particular, I thought he might begin to feel some guilt if forced to confront the human face of a twelve-year-old girl who had died, and we might be able to get to him through that.

  As a variation on what we’d tried in Atlanta and in the Shari Smith case, I suggested holding a nighttime vigil at the grave sites of some of the victims, which I thought the UNSUB might attend. Recognizing that the subject probably didn’t feel good about himself, I also advised giving heavy press to anniversaries associated with the crimes.

  I thought we could encourage him to visit specific stores in the way we’d been able to "direct" bank robbers in Milwaukee and Detroit to hold up specific bank branches where we were waiting for them. For example, the police could leak information about steps being taken to protect customers at one particular store. I thought the guy might feel compelled to visit that store to see firsthand the effects of his actions. A variation on that would be to publish an article about an arrogant store manager who would publicly state how confident he was in his establishment’s security and that it would be impossible for the Tylenol poisoner to tamper with any product on his shelves. Another version of this ploy would be to have police and FBI agents respond to a "hot tip" at a particular store, with attendant publicity. This would turn out to be a false alarm. But the police official would then state for the cameras that his department’s intelligence capability is so efficient that the unknown subject decided against planting the poisoned Tylenol. This should provide him with an indirect challenge he might find difficult to pass up.

  We could put forth a bleeding-heart psychiatrist who would give an interview professing great support for the subject, categorizing him as a victim of society and thereby providing him with a face-saving scenario. The subject would be expected to call or drive by the doctor’s office, where we’d be ready to trap and trace.

  And I thought that if officials set up a volunteer civilian task force to help the police with all the phoned-in tips, the subject would likely volunteer to help man it. Had we been able to set up something like that in Atlanta, I think we would have seen Wayne Williams. Ted Bundy, in his time, had volunteered at a Seattle rape crisis center.

  There is always some squeamishness on the part of law enforcement about cooperating too closely with—or using—the media. This has come up a number of times in my career. Back in the early 1980s, when the profiling program was relatively new, I was called up to headquarters to meet with the Criminal Investigation Division and Bureau legal counsel to explain some of my proactive techniques.

  "John, you don’t lie to the press, do you?"

  I gave them a recent example of how a successful proactive approach to the media had worked. In San Diego, a young woman’s body was found in the hills, strangled and raped, with a dog collar and leash around her neck. Her car was found along one of the highways. Apparently, she had run out of gas and her killer had picked her up—either as a Good Samaritan or forcibly—and had driven her up to where she was found.

  I suggested to the police that they release information to the press in a particular order. First, they should describe the crime and our crime analysis. Second, they should emphasize the full thrust of FBI involvement with state and local authorities and that "if it takes us twenty years, we’re going to get this guy!" And third, on a busy road like that where a young woman was broken down, someone had to have seen something. I wanted the third story to say that there had been reports of someone or something suspicious around the time of her abduction and that the police were asking the public to come forward with information.

  My reasoning here was that if the killer thought someone might have seen him at some point (which they probably did), then he would think he had to neutralize that with the police, to explain and legitimize his presence on the scene. He would come forward and say something to the effect of, "I drove by and saw she was stuck. I pulled over and asked if I could help, but she said she was okay, so I drove off."

  Now, police do seek help from the public all the time through the media. But too often they don’t consider it a proactive technique. I wonder how many times offenders have come forward who slipped through their fingers because they didn’t know what to look for. By the way, this is not to imply that genuine witnesses need have any fear of coming forward with their stories. You will not become a suspect, but you may very well help lead to the arrest of one.

  In the San Diego case, the technique worked just as I had outlined it. The UNSUB injected himself into the investigation and was caught.

  "Okay, Douglas, we see your point," the FBI headquarters staff responded begrudgingly. "Just keep us informed whenever you think you’re going to use this approach." Anything new or innovative can be scary to a bureaucracy.

  I hoped that in one way or another, the press could help bring forth the Tylenol poisoner. Bob Greene, the popular syndicated columnist of the Chicago Tribune, met with the police and FBI. He then wrote a moving article about twelve-year-old Mary Kellerman, the poisoner’s youngest victim and the only child of a couple unable to have more children. As the story appeared, police and FBI agents were ready with surveillance on her home and the grave. I think most of the people involved thought this was bullshit, that guilt-ridden and/or happily reminiscing killers don’t actually return to grave sites. But I urged them to give it a week.

  I was still in Chicago when the police staked out the cemetery, and I knew I’d face their ire if they didn’t come up with anything. Stakeouts are boring, uncomfortable work under the best of circumstances. They’re even worse in a graveyard at night.

  The first night, nothing happens. It’s peaceful and quiet. But sometime during the second night, the surveillance team thinks they hear something. They approach the grave, being careful to stay out of sight. They hear the voice of a man just about the age the profile predicted.

  The man is tearful, apparently on the verge of sobbing.
"I’m sorry," he pleads. "I didn’t mean it. It was an accident!" He begs the dead girl to forgive him.

  Holy shit, they’re thinking, Douglas must be right. They pounce on him.

  But wait a minute! The name he uses isn’t Mary.

  This guy is scared out of his wits. And when the police finally get a close look, they see he’s standing in front of the grave next to Mary’s!

  It turns out that buried next to Mary Kellerman is the victim of an unsolved automobile hit-and-run, and her unwitting killer has come back to confess his crime.

  Four or five years later, Chicago PD used the same ploy with an unsolved murder. Spearheaded by FBI training coordinator Bob Sagowski, they began giving information to newspapers around the time of the anniversary of the murder. When police apprehended the murderer at the grave, he commented simply, "I wondered what took you so long."

  We didn’t catch the Tylenol poisoner this way. We didn’t catch a murderer at all. A suspect was apprehended and convicted on extortion charges linked to the murders, though there wasn’t sufficient evidence to try him for the murders themselves. He fit the profile, but had been out of the Chicago area when police conducted the cemetery stakeout. After his incarceration, however, no more poisonings were reported.

  Of course, since there was no trial, we can’t say with any legal certainty that this was our man. But it is clear that a certain percentage of the perpetrators of unsolved serial murders are actually caught, unbeknownst to the officers and detectives investigating the cases. When an active killer suddenly stops, there are three strong explanations aside from his simple decision to retire. The first is that he’s committed suicide, which can be true for certain personality types. The second is that he’s left the area and is actually plying his trade somewhere else. With the FBI’s VICAP (Violent Criminal Apprehension Program) computer base, we’re working to prevent that from happening by giving the thousands of police jurisdictions around the country the ability to share information easily with one another. The third explanation is that the killer has been picked up for some other offense—generally burglary or robbery or assault—and is serving time on the lesser charge without authorities having connected him to his most grievous offenses.