Law & Disorder Read online

  The first report of the fire came from eleven-year-old Buffie Barbee. About ten minutes after ten on the morning of December 23, 1991, two days before Christmas, she was playing in her backyard with her sister and a friend when she smelled and saw smoke. She ran inside to tell her mother, Diane, that the house next door was on fire. Diane raced out the front door and realized it was not the neighboring house, but the Willinghams’, two doors down.

  Diane saw Todd outside the door on the front porch, wearing only a pair of jeans. He was yelling, “My babies are burning up!” and asked her to call the fire department. While Diane ran for a phone, Todd grabbed a stick and jabbed holes in the bedroom window, from which flames shot out. Moments later, all five of the bedroom windows exploded outward and the house was engulfed in flames. Some witnesses said Todd seemed to retreat into a trance at that point. But by the time firefighters arrived a few minutes later, others recalled that he had to be restrained from trying to get back inside.

  The firemen immediately laid out their hoses and equipment and attacked the fire. Burvin Terry Smith asked him if there were any other people in the house. Paramedic Ron Franks and firemen Charles Dennison and Steven Vandiver, equipped with oxygen masks and tanks, entered the house through a front window and fought their way through the blaze, trying to “knock it down,” in firefighter parlance. Vandiver emerged with Amber’s limp body in his arms. She died at the hospital without regaining consciousness. When Assistant Chief Douglas Fogg reached the children’s bedroom, he found the one-year-old twins dead. A medical report would find that all three children died of acute carbon monoxide poisoning as a result of smoke inhalation.

  Todd told firemen that Stacy had gone out early to shop for Christmas presents for the children and that he was asleep when Amber rushed into his bedroom and woke him up, shouting how the house was on fire. When he opened his eyes, the room was full of smoke and he shouted to Amber to get out of the house.

  Almost as soon as the blaze was extinguished, the fire department began investigating. Assistant Chief Fogg went through the house with a video camera, narrating on the tape as he focused on each surface. On December 27, he went back with Manuel Vasquez, a state deputy fire marshal who was one of the most experienced fire investigators in Texas. Both men had impressive backgrounds. Fogg, a local boy, had been a navy medical corpsman in Vietnam, with four Purple Hearts attesting to his dedication and courage. Vasquez was a veteran of army intelligence who had more than twelve hundred fire investigations to his credit.

  At the very front of the house, the damage seemed mainly smoke and heat-related, including the master bedroom about halfway back, where Todd had been sleeping and Amber had been found unconscious. The farther down the main hallway they went, the more suspicious they became. There was charring along the base of the walls and the floors, even though fire usually burns upward. The “burn trailer” seemed to go from the children’s bedroom in the left front, into this hall and then out the front hall, suggesting the fire started there. Moreover, the intensity of the burn marks on and under the floor of that bedroom suggested that accelerant could have been poured in the middle. The weblike patterns on the broken glass were also evidence of an intensely hot fire in this particular location, further bolstering the accelerant scenario.

  As they followed the burn trailer back out to the front, they noted brown staining on the concrete floor of the porch, also consistent with the use of an accelerant. V-shaped soot patterns in the children’s bedroom, the hallway and the front porch further documented the three places where the fire had started. If accelerant had been poured at each of those locations and then lit, it would be nearly impossible for anyone inside to escape the burning building alive.

  Before they left, Fogg and Vasquez reached an inescapable conclusion: This fire had been intentionally set.

  Of all the major crime categories, arson stands alone in one key respect. Except for a rare bombing in which a gas leak might initially be considered a possible cause, arson is the only violent act in which the first stage of the investigation is to determine whether a crime has actually been committed. With a murder or robbery, we know that a crime has taken place and our first priority is to figure out who did it. Even with poisoning, either direct or a product tampering, like the Tylenol case of 1982, experienced investigators can tell very quickly that the sickness or death was intentionally caused.

  I have worked many arson cases during my career, usually teamed with an experienced specialist. Dave Icove, a special agent with a Ph.D. in engineering science and mechanics, headed up the Arson and Bombing Investigative Services (ABIS) subunit in our underground lair of offices at the FBI Academy. Icove was an expert I knew I could always count on. His résumé of publications, patents and funded research is many pages long. Later Gus Gary, a special agent for the Bureau of Alcohol, Tobacco and Firearms, came over to Quantico and worked closely with my Investigative Support Unit. As difficult as arson cases can be, Dave, Gus and I had an advantage that most investigators don’t have; that is, we already knew it was arson.

  By the time we would be called in to profile the UNSUB, there was usually a pattern of fires in a given area, meaning that we knew they were intentionally set. For example, when we were presented with a string of fires in synagogues and rabbis’ homes in Hartford, Connecticut, in 1983, we knew these blazes didn’t just all happen accidentally.

  Our job in arson cases was to use locations, types of fires, eyewitness accounts and all of the other tools and tricks in our arsenal to figure out what type of individual was responsible and how to flush him out. In an early form of what is now known as “geographical programming,” Dave used a map, ruler and compass; through marking the location and order of each fire and drawing lines between them, he was able to predict with a high degree of specificity where the offender lived.

  What we didn’t have to figure out, in most instances, was whether these fires were arson or mere accidents. In other words, we didn’t have to determine whether or not a crime had even been committed.

  Local investigators of individual fires don’t have that foreknowledge, so their responsibility is grave and their judgment is critical. Fogg and Vasquez had no doubts about their judgment on West Eleventh Street, though. They compiled and documented a list of twenty separate indicators of arson.

  Meanwhile, the behavioral evidence was piling up. Even though it is a town of about twenty thousand, representing more than half the population of Navarro County, Corsicana is a small and close-knit community, where residents pride themselves on knowing most of the other folks in their neighborhood. Knowing that Todd and Stacy had very little money, Vicki Prater, the proprietor of a local bar and tavern, got together with some of her neighbors to hold a benefit to raise money for the children’s funeral and burial.

  It was an extremely kind and generous act. However, the way she saw Todd acting at the event left her deeply troubled. He seemed heavily into playing darts, one of his normal pastimes, and generally “too involved in the fun” for someone who had just lost twin babies and a toddler under horrific circumstances.

  Fire department lieutenant and paramedic Ron Franks thought it was odd that after the fire was put out, Todd seemed unusually upset about a dartboard and set of darts and asked if he could go back in to look for them.

  The Corsicana police had some questions of their own. According to Todd’s account, when Amber woke him up, he took just long enough to pull on a pair of jeans. When he was seen outside, that, in fact, was all he was wearing.

  So why weren’t his feet burned?

  As hot as the fire investigators said that hallway floor must have been to leave the burn patterns it did, how could Todd have gotten out of the house without blistering the soles of his bare feet? Other than some minor burns and singeing of his hair, he had escaped virtually unscathed.

  When police brought him to the station for questioning on December 31, he replied simply that when he got out, the fire was mostly higher up; and while the
floor was hot, it wasn’t burning hot at that point. Then it must have gotten burning hot awfully fast after that, the officers concluded skeptically. Fogg and Vasquez were both there for the questioning, along with police officer Jimmie Hensley. As far as the cause of the fire, Todd mentioned the three space heaters he and Stacy used to keep the house warm, one of them in the children’s bedroom; maybe one of them had shorted out, or the kids had upset it or something. But during his careful examination of the house, Manuel Vasquez had noted the dial on the heater in the children’s bedroom was in the “Off ” position. Maybe it was something electrical in the house, Todd suggested. He had heard something like a popping or cracking sound.

  Todd acknowledged his bad treatment of Stacy. He said that they had split up several times over it, but what kept bringing them back together was the kids. When he was asked why anyone would want to hurt his family, he could not come up with a single thought.

  The investigators’ impressions of his behavior were similar to Vicki Prater’s. Despite his insistence that the children were what kept him with Stacy and led them to marriage, instead of displaying grief or even remorse, he seemed almost cocky. It was as if he were bragging about how he’d rushed back into the house to try to save his children—an action no one had witnessed. The whole thing seemed too rehearsed and calculated. The only time he broke down was when they showed him photographs of his dead children.

  This is not an unusual reaction. I have used and suggested this technique and its variations many times when interrogating suspects. I have also coached detectives in its use. On one level, you’re trying to get through to something human within the suspect, appealing to his most primal sensibilities. Maybe the comprehension of the moral horror of his act will prompt some feelings of guilt. Maybe there are feelings of guilt because he felt he did not do enough.

  On another level, you’re evaluating his reaction. Who are the tears for: the victim or himself? Is he sobbing because he is sorry for the victim, or is he sorry for himself that he got caught? I have seen it both ways. But with the predators I hunted throughout my career, the second reason is far more common.

  If a case like this had been brought to my unit at Quantico, I think the first thing I would have asked would be “Why are you coming to us? If you know it was a set fire and only one individual was in a position to set it, then you’ve got a slam dunk here.”

  Nor would there have been much doubt as to what type of arson it was. We break arson down into several general categories: insurance fraud, pyromania and thrill seeking, vandalism, concealing other crimes, revenge, civil disorder and murder. If this one was arson, then it was arson as murder.

  When the physical evidence and the behavioral evidence match up, and there aren’t any other likely suspects given the time and location of the occurrence, that’s a pretty good package to hand to a district attorney (DA).

  But if we had been called in, here’s where I would have focused: While I would have been skeptical of Todd Willingham’s tears at the sight of his kids’ photos, I would have also warned about placing too much meaning in his emotional reactions during and immediately after the fire—whether they were “appropriate” or not. Experience and the study of many cases have shown us that reaction to the death of a loved one is a very individual, very personal thing.

  When his infant son was kidnapped, Charles Lindbergh was criticized by many and actually suspected by some for seeming so cold and distant. In reality, this was merely his way of coping. It isn’t surprising when you consider that this was a man who had been in many harrowing aviation situations and, against all odds, had flown solo across the Atlantic for more than thirty-three sleepless hours in a rickety plane without even a front window. If he weren’t able to control his emotions in moments of crisis, he wouldn’t have been the man he was. Everyone reacts differently, and observation during this kind of crisis is one of the least reliable ways of determining personality and motive.

  I would have counseled investigators to concentrate on what we generally refer to as postoffense behavior. In the days and weeks following the event, how has the subject’s behavior changed? Has he either avoided investigators or inserted himself into the investigation? Has he begun drinking more or abusing drugs? Has his weight changed? What is he doing with himself during the day and at night?

  Then I would have asked if the police were sure of the arson investigators’ report. Personally, I have no expertise in structural or chemical engineering, biochemistry or metallurgy, just as I have no expertise in fingerprint analysis or ballistics. So when those disciplines come into a case investigation, I rely on the best experts I can find. In the Bureau, we are fortunate to have some of the best in-house, and we have called on many others. For example, I had a close working relationship with Dr. James Luke, the former ME for Washington, DC. We so completely trusted each other’s methods and analysis that I often called him in on cases in which forensic pathology might be a determining factor.

  If the locals expressed any doubt about the quality of their own forensic science, I would have asked Dave Icove or Gus Gary to take a look. If either man saw anything that raised red flags in the reports, photos and analysis, he could go down to Texas himself and take a look.

  The only other service I would have offered would be prosecution strategy; that is, help figure out how to explain motive to a jury.

  Technically, the prosecution doesn’t have to establish motive in a criminal trial. If the evidence indicates beyond a reasonable doubt that the defendant did the deed in question, it doesn’t matter why he did it. But in the real world of criminal trials, most jury members, consciously or subconsciously, consider motive part of the proof. If they can’t figure out why the defendant would have committed such a crime, they have trouble convicting. That’s why my unit was often called in by prosecutors, even after the police had already delivered a suspect.

  Had this case been brought to us, since the science had already established the fact of arson and there were no suspects other than Todd Willingham, we would have deemphasized his immediate behavior at the fire scene. Instead, we would have pointed to his conduct at the benefit at Vicki Prater’s bar as evidence that he was relieved rather than grief-stricken at the loss of his three girls. He was no longer stuck at home, changing three sets of diapers; he could be out doing what men his age normally want to do. Many people saw a similar motive in the behavior of Casey Anthony when she was tried for the alleged murder of her young daughter, Caylee, in Orlando, Florida. While an apparently agonized jury found her not guilty of the charge, it was extremely difficult for many of us to square the image of a young woman partying, drinking and participating in wet T-shirt contests while knowing that Caylee was (depending on which version you believe) either missing or dead.

  The prediction of future violence is far from an exact science, but we live by the concept that the best predictor of future violence is past violence. And here, Willingham easily qualifies. Any man who regularly beat and mistreated his young wife is certainly capable of other types of violence. That doesn’t mean he will do it, but he is certainly capable. And once you’ve harmed your wife, harming your children is merely an escalation, not a diversion from form.

  There is one detail from the reports that I think would have given me pause, and that is the fact that two-year-old Amber was found in her parents’ bedroom, while the baby twins were found in the children’s bedroom, with more severely burned bodies. This fact would not only tend to confirm Todd’s description of how he learned of the fire, but also call into question his methodology if the fire was set. That is, I would have expected to find all three children together if he was trying to kill them. A two-year-old is far more mobile than one-year-olds, but I wouldn’t expect her to be able to escape the primary fire site if it had been started by accelerants and therefore burst into flame all of a sudden.

  When Amber was found in the master bedroom, she was clad only in underpants and one sock. Had the other sock been found i
n the debris, it might have indicated her path. But the fact that she was so lightly clad also points to the space heater in the children’s bedroom having been on, regardless of what the switch indicated. Malfunction or Amber doing something to it could have been a cause of the fire. Stacy told investigators that several times Amber had played with the space heater and been punished for it. The house was not very well insulated, so it’s doubtful Stacy and Todd would have put the kids to bed in so little if a heat source was not nearby.

  On the other hand, I would have questioned Todd’s just putting on his jeans, but not also picking Amber up and carrying her out under his arm. According to his account, he told her to run right out, and the smoke might have been so thick that he thought she did, but I would certainly want to know more about that moment and press the investigators for the complete evidence.

  On January 8, 1992, police arrested Todd Willingham and charged him with three counts of murder in the first degree. In Texas, the multiple counts made him eligible for the death penalty. Since he couldn’t afford an attorney, the court appointed two: David Martin and Robert Dunn.

  Prosecutor John Jackson thought he had a pretty solid case, and I would have agreed with him. But before the trial, he approached Martin and Dunn with an offer to take execution off the table in exchange for a guilty plea and a life sentence. Jackson was personally against capital punishment, believing it was an ineffective deterrent and a waste of court time and taxpayer money. It also haunted him that there was always a possibility that an innocent man or woman could be executed. If he could get a conviction without a death penalty, that was fine with him.

  The defense attorneys, who understandably thought their client was guilty, urged him to take the plea, save his own life and spare the families and the community the anguish of a trial. Willingham steadfastly refused.