Dan Gardner is the New York Times best-selling author of Risk, Future Babble, Superforecasting (co-authored with Philip E. Tetlock), and How Big Things Get Done (co-authored with Bent Flyvbjerg). His books have been published in 26 countries and 20 languages. Prior to becoming an author, Gardner was an award-winning investigative journalist. More >

Making Sense of Sentencing

I understand why people are furious at the two-year sentence imposed on Graham James but all the angry shouting and demands for tougher punishments won't do anything to correct the very real problems with criminal sentencing in Canada. It could even make those problems a little worse.

Start with that sentence. Most people think it's unjust. I suspect they're right, although I'd hesitate to form an opinion without knowing a lot more about the case than I do. But people also think the James sentence demonstrates that the justice system is far too soft. And that is wrong.

Why? Because for the James sentence to say something about criminal justice in general it can't be an aberration. It has to be typical. And we don't know if it's typical.

It may be. Or it may not. The simple truth is: We don't know.

Yes, pundits cite other cases and sentences. The judge undoubtedly considered some, too. But that is analysis by anecdote and impression, which is a very bad idea. There's only one way to really know what a typical sentence is in a case like this: Consult a comprehensive, publicly available, database of sentences.

There is no such database. "An interested, intelligent member of the public or an interested, intelligent judge cannot, apparently, get systematic information about what is happening in Canada's courts," University of Toronto Criminologist Tony Doob wrote last year in the Canadian Journal of Criminology and Criminal Justice.

In the 1980s, Doob sat on the Canadian Sentencing Commission, which reviewed every aspect of criminal sentencing. The commission's 1987 report was scathing. Doob is just as scathing today. Because nothing has changed.

A basic principle of justice is that similar crimes should be punished similarly, the commission noted, but that requires a proper database. And there wasn't a proper database. A quarter-century later there still isn't.

Of course similar cases could still be punished similarly if the Criminal Code gave judges clear guidance on sentencing. But it didn't. A quarter-century later, it still doesn't.

For most offences, the range of possible sentences is wide open. There is no minimum. And the maximums - typically two years, seven, 10, 14, or life - are often set absurdly high. A judge sentencing someone for breaking into a house can give him essentially nothing, or life in prison, or anything in between.

The commission also condemned the Criminal Code's incoherence. Both manslaughter and breaking into a house, for example, are punishable by up to life in prison. Both sexual assault with a weapon and possession of counterfeit money are punishable by up to 14 years in prison. I doubt there is anyone in the country who thinks these are comparable crimes that should be punished comparably. And yet that's what the law says. And there are countless more examples like them.

So how did that happen? Did Parliament really decide that manslaughter and break-and-enter are crimes of equal gravity? That possession of counterfeit money is as serious as sexual assault with a weapon? Of course not. Parliament doesn't review the Criminal Code to ensure its internal coherence. It just cranks out news laws - often in response to the latest headline - with little or no consideration of what's already there.

This isn't a new problem, as the Sentencing Commission noted in 1987. Parliament has been operating this way since the Criminal Code was created in 1892, which is why the Code was a big mess in 1987, and an even bigger mess today.

"The search for just sanctions (is) at best a lottery," the Alberta Court of Appeal declared in a powerful 2010 decision, "and at worst a myth."

The Sentencing Commission recommended sweeping change.

First, wipe out all existing sentencing provisions. Next, state whether there will be a presumption of incarceration or not for each offence.

Finally, for each offence where there is a presumption of incarceration, create a relatively narrow "presumptive range" for sentence lengths.

These changes would ensure that similar cases would be treated similarly. But flexibility is also critical to ensure that unusual cases are not punished inappropriately. (This is the main reason why mandatory minimum sentences are a terrible idea.) So the commission recommended that judges be allowed to deviate from the presumptive range provided they explained why they were doing so. (The deviation and explanation could be appealed.)

The Sentencing Commission also called for the creation of a permanent commission whose job would be to help create the new sentencing guidelines, to collect sentencing data, and to work with the House of Commons to ensure sentencing practices remained rational and just.

The result would be a justice system that is transparent, coherent, predictable, and accountable - four qualities sorely lacking in the status quo.

This wasn't a utopian dream. Minnesota introduced a system somewhat like this in 1980. It works well. England made a similar change in 2009.

But Canadian governments weren't interested. Not the Mulroney Conservatives. Not the Chrétien Liberals.

The Liberals did add a list of sentencing principles to the Criminal Code in the mid-1990s, but these were too vague to make any real difference. Worse, the Liberals reacted to the headlines of the day by passing a barrage of mandatory minimum sentences, which helped to turn this rarely used form of sentencing into a standard political talking point. Entirely predictably, and against all expert advice, the Conservatives have salted the Criminal Code with mandatory minimums - steadily adding to the Code's complexity, arbitrariness, and incoherence.

Now the headlines are all about the Graham James sentence. Guess what comes next? Right. Another mandatory minimum. Bet on it.

What is most unlikely is real reform. "Sentencing has been neglected as a serious policy area," Tony Doob wrote. "Sentencing has, instead, been used as a political resource with little, if any, concern about its overall coherence."

There's not the slightest sign that will change. And so the sentencing system will remain, in Doob's words, "an unprincipled and unpredictable mess."