--------------------------------------------------------------------------------

Scary anecdotes don't tell the real story on gun violence

--------------------------------------------------------------------------------

Two weeks ago, Jordan Manners, a 15-year-old-boy, was shot dead in the halls of C.W. Jefferys Collegiate, a Toronto high school. Newspapers were filled with shocked reactions. "I grew up in a time where the notion of a school lockdown was a foreign one," said Ontario PC leader John Tory. "Now it seems they're almost commonplace."

The Ontario premier and Toronto mayor railed against handguns. A few commentators pointed to absentee fathers. Underneath all the talk lay the sense that schools are sliding into chaos. "Canadians are smugly appalled at the grim conditions of inner-city schools in the United States, where cops roam the halls and surveillance cameras and metal detectors are standard equipment," wrote Margaret Wente in the Globe and Mail. "So much for our illusions. It turns out we have the worst of both worlds -- guns in schools, but no metal detectors."

Then came the inevitable poll. Half of Canadians think schools are more violent now than five years ago. The only surprise is that half didn't. Open a newspaper and what do you see? A shooting, a lockdown, a columnist saying schools have been taken over by "thugs." It's obvious things are getting out of control. Isn't it?

Maybe. But let's get a little perspective on this.

There were 22 students murdered in American schools in the school year 2004-2005. That's a lot of tragedy. But is it also a lot of danger? Not really. There are more than 50 million children in American schools. So the chance of a student being murdered at school that year was about 0.00004 per cent. The average American child has a greater risk of drowning in the bathtub.

There are an estimated 680,954 students in Ontario's high schools. In Toronto alone, there are 114,392 secondary students. So what does the murder of one student -- the first of its kind in the history of Toronto -- tell us about the safety of hundreds of thousands of others? Without further information, nothing. Nothing at all.

I realize it sounds cold to talk about numbers when a child's life is lost. But if we are to understand tragedies and respond intelligently to them, we need to see reality as it is. And numbers are essential for that.

Unfortunately, we don't have good numbers. To really get the picture on school violence, we need data broken down by types of violence and we need them over a period of time so we can tell what the trends are. As far as I can see, those numbers don't exist.

Even more unfortunately, politicians and journalists have substituted anecdotes for the missing data. Margaret Wente says there was a knifing at her old high school last year. A teacher at C.W. Jeffreys wrote an open letter saying administrators ignored assaults in the past. Some teachers at other schools agree. Along with the shooting of Jordan Manners, this handful of anecdotes proves violence is rising. Or so many people believe.

Is violence rising? Without reliable research, I can't really say. But I do know that screaming headlines, sad stories and anecdotes can be very misleading.

In 1998, two students in Jonesboro, Arkansas, shot to death four classmates and a teacher. The reaction in the U.S. was almost identical to what we have seen here over the last two weeks, only an order of magnitude bigger. Children were in danger. This was a crisis.

And the numbers? As in Ontario today, there were no numbers, at least not that anyone had collected in one place. So Congress ordered researchers to put together an annual report on school violence.

Before the first edition of that report was complete, the Columbine massacre happened in suburban Denver -- and the storm grew into a hurricane. Across the U.S., kids practised lockdowns and school boards blew budgets on guards and metal detectors. That there was a crisis was beyond dispute. The only debate was over the source of the bloodshed. Was it guns? Video games? Broken homes?

Meanwhile, journalists hunted for anything that even hinted at school violence. Anonymous threats, kids caught with guns, students sharing violent fantasies -- stories that would have been trivial a year before got front-page treatment. And Americans worried. A Gallup poll taken one month after the massacre found 52 per cent of parents feared for their children's safety at school.

Finally, the first school violence report was finished. And it was startling.

The risk of a student being murdered was microscopic. And violence wasn't climbing in American schools. It was dropping. Fast. The rate of what the report called "serious violent crime" was 13 incidents per 1,000 students in 1994. By the time the 2004 edition of the violence report was released, that number had dropped all the way to four per 1,000 students. In 1993, 12 per cent of kids told researchers they had carried a weapon of some kind onto school property within the last 30 days; a decade later, that had fallen to six per cent.

How is it possible for perception and reality to diverge so radically? I've written a book to answer that question, so it's not something I can put in a couple of paragraphs. But a central part of the answer lies in the fact that journalists are human and humans are moved by stories, not numbers.

Remember the 2005 "summer of the gun" in Toronto? A wave of gun-related murders culminated in the Boxing Day shooting that took the life of 15-year-old Jane Creba. Such terrible, tragic stories. The media coverage was massive and worries about gun crime rose across the country.

But then, in 2006, gun killings in Toronto plummeted 46 per cent. Didn't know that? Of course not. That number was scarcely reported. It's only a number, after all. It's not an emotional story about people and it's pretty hard to turn into one. How do you tell the gripping story of a man not shot? An innocent bystander not gunned down? A family not in mourning?

Scientists like to say "anecdotes aren't data," meaning numbers can prove things that stories can't. But human nature sees things the other way round. Stories matter. Numbers don't. And the inevitable result is a profound bias in favour of negative information.

With the shooting of Jordan Manners and the furor about violence in schools, I suspect we are witnessing another demonstration of this human flaw.