Lululemon murder: A plot twist reminds us that most crime isn't random.

Murder, theft, and other wickedness.
March 22 2011 2:00 PM

Stab in the Dark

A twist in the Lululemon murder story reminds us that most crime isn't random.

Brittany Norwood.
Brittany Norwood

The killing of Jayna Murray at the Lululemon Athletica store in Bethesda, Md., on March 11 at first seemed random: Masked men broke in, beat, and sexually assaulted Murray and her colleague, Brittany Norwood, and left Murray dead. As it turned out, the crime wasn't random at all: Police are now fingering Norwood as the killer. The development was billed as a "bizarre twist" in the case. But given the extreme rarity of random murder, it shouldn't be surprising that Norwood is now the prime suspect.

The story began when a Lululemon employee arrived at the store last Saturday morning to find Murray dead and Norwood tied up. Norwood told police she and Murray had returned to the store Friday night after it had closed, to retrieve something Norwood had forgotten. When they did, she said, two men had followed them into the store, assaulted them both, and killed Murray. A manhunt ensued. Police looked into local data on ski mask sales. They offered a $150,000 reward. They even tailed a man who fit the description of one of the killers.

By Friday, March 18, police had arrested Norwood. Her story didn't add up. The only bloody footprints police found were Norwood's and those made by shoes that were found at the crime scene. Her bruises appeared self-inflicted, and she'd been tied up in such a way, police said, that she could have done it herself. Neither woman, it turned out, was sexually assaulted.

Bethesda, a wealthy suburb just north of Washington, breathed a sigh of relief: Shoppers were at risk of assault not by masked killers on the loose, but merely by their coworkers.

Which shouldn't come as a surprise. Random homicide is extremely rare. So rare that police probably suspected Norwood's story from the start—although, as the police chief said, they had to assume at first she was telling the truth. Only 15 percent of homicides reported every year are committed by someone who doesn't know the victim, according to the Bureau of Justice statistics. And even then, the two people usually have mutual friends and acquaintances, says Richard Rosenfeld, a criminology professor at the University of Missouri-St. Louis: "That explains why they're in the same place at the same time." And yet, we often assume randomness, and treat the discovery that a murderisn't random as news.

Why do we assume that so many homicides are random? The FBI is partly to blame. In the early 1990s, the bureau released a report claiming that half of all homicides were committed by strangers. But that report was flawed, says Rosenfeld: It conflated police reports in which it was established that victim and offender didn't know each other, and unsolved reports that didn't establish their relationship one way or another. The more random the crime, of course, the more necessary the FBI. "Randomness is scary," says Joel Best, professor of sociology and criminal justice at the University of Delaware. "It implies that you can't protect yourself."

Politicians like to play up the randomness of violence, too. "It gives everyone a stake in the problem," says Best. Even liberals and conservatives can agree that randomness is good: "If you're conservative, talking about random violence allows you to gloss over the obvious relationship between victimization and class," says Best. "If you're a liberal, you can gloss over the relationship between [committing] violence and race." Everybody wins. Except, you know, the victims.

The media does its part, of course. Murders don't typically make headline news, unless there's something unusual about them—for example, that they occur in an upper-middle-class suburb. (Slate's Timothy Noah calls this genre of news coverage "When Bad Things Happen to White People.") By covering random crime, news organizations help to create the impression that most crime is random.

Thank goodness they're wrong. Police rely on crime's non-randomness in order to stop it. CompStat and other data analysis tools examine patterns in crime and tell police which hotspots to patrol. Crime is so predictable, some departments have introduced "predictive policing" programs that would provide crime forecasts, like the weather. Statistically, says Rosenfeld, "the default assumption should be that a homicide involves two young men who know one another, both of whom have extensive criminal records." Even the occasional victim of a "random" homicide—the innocent bystander killed in a gang shootout, for example—is usually less random than people realize: He may not be connected to his killer, but he is from the neighborhood or one close by.

That doesn't mean masked men never assault people they don't know in chichi neighborhoods. In 2006, a man had his throat slit in a random attack in Georgetown. The incident led to fears of a spike in crime. But the numbers told a different story: The murder rate that year in the District of Columbia was the lowest in 20 years. And the vast majority of murders that did occur in Washington, needless to say, weren't in Georgetown.

Christopher Beam is a writer living in Beijing.