YOU CAN SEND CRIMINALS A MESSAGE, BUT THEY WON'T HEAR IT

Even among his detractors, Stephen Harper is said to be an ideas man.
A policy wonk. The sort of guy who reads spreadsheets on the beach.
For Mr. Harper, politics is the distasteful process one must engage
in if one wishes to do the really juicy work of crafting and
implementing public policy.

Everyone says this so it must be true. And that's why, today, I'd
like to bring to Mr. Harper's attention a fascinating paper published
last year in an academic journal.

It's filled with numbers, tables and those funny-looking algorithms
that make scientists such babe magnets. There's even a discussion of
"hierarchical linear modelling techniques." Cool.

Oh, it's also interesting because it shows that Mr. Harper's plan to
revolutionize the criminal justice system won't work. As a wonk who
values public policy above all else -- especially cheap, low, tawdry
politics -- I'm sure Mr. Harper will be as shocked as I am by this
and re-think the whole thing.

The paper appeared last August in a journal called Criminology. The
title is "The Missing Link in General Deterrence Research" -- which
is not the most exciting title around, but a true wonk like Mr.
Harper knows that the duller the title, the sexier the stats.

The author, Gary Kleck, is a criminologist at Florida State
University. Like Mr. Harper, he is interested in what experts call
"deterrence." It's a simple idea that seems to make a lot of sense:
If you impose tougher sentences on convicted criminals -- send more
of them to prison and keep them there longer -- many would-be
criminals will decide that crime really doesn't pay. Crime will drop
and we'll all be safer.

Mr. Harper's big plan for crime -- one of his famous five priorities
-- calls for a long list of tough new mandatory minimum sentences.
The main rationale is deterrence. Send a message that serious crime
means serious time, put the fear of God and the prime minister into
the bad guys, and crime will drop.

As I said, this stuff seems to make sense. But things aren't always
what they seem. They have to be tested.

There are many ways to test whether deterrence works. For his paper,
Mr. Kleck focused on one simple fact: For people to be influence by
the law, they have to know about it. What matters isn't how tough the
system is; what matters is how tough people perceive the system to be.

Mr. Kleck and his colleagues surveyed more than 1,500 people in 54
major American cities. They asked, how likely is it that someone in
your city who commits a crime known to police will be arrested? How
likely is it they will be convicted? How likely that they will be
sent to prison? And how long is the sentence likely to be?

The researchers asked these questions for five major crime types.
Then they compared people's answers to the official statistics.

"We wanted to know how close the relationship was between perception
and reality," Mr. Kleck says on the phone from Florida. "And the
answer was, there was no relationship. It was almost exactly a
correlation of zero. It could scarcely have been a weaker
relationship. Put another way, people's perceptions might as well
have been completely random guesses."

This is pretty amazing stuff. Over the past 15 years or so, American
media became obsessed with crime and punishment, and American
politicians turned boasting about tough new laws into a campaign
cliche on par with kissing babies. But despite all this talk and
attention, the American public still has no clue about the basic
realities in courtrooms.

Whatever, you might say. It's not the ordinary person we want to
deter. It's criminals. And criminals know the system inside and out.
Pass tough new laws and you can bet the word will get around.

Mr. Kleck thought of that, too, so he asked people if they had ever
been arrested for a non-traffic offence. Then he compared the answers
of people who had been arrested with those who hadn't. Result: Those
who had been arrested were, if anything, even more ignorant than the rest.

That's less surprising since there's lots of research that shows
criminals aren't the jailhouse lawyers we imagine. In fact, they tend
to be young, semi-literate and dumb. They don't subscribe to
newspapers. They don't watch Question Period. They don't read
criminology journals or the latest amendments to the Criminal Code.
What they know about the system tends to come from equally clueless
buddies "boasting about what they did and got away with," says Mr.
Kleck. "And that's probably not too sound a foundation."

Without a connection between people's perceptions of how tough the
system is and how tough it really is, politicians don't have the
power they think they have. They can make the law tougher. But they
can't make people think the law is tougher. And that means their
tough new laws cannot deter anyone.

To be fair to Mr. Harper, there is another way that tougher sentences
can reduce crime. Experts call it "incapacitation": If a thug is
locked up, he can't commit new crimes. It seems obvious that longer
sentences mean more incapacitation and less crime. Simple, right?

Wrong, unfortunately. In reality, incapacitation is a big,
complicated issue and longer sentences deliver diminishing returns.
More on this in a future column.

But for Canadians who aren't interested in the hows and whys of
crime, there is a clear bottom line here: Are Mr. Harper's tough
mandatory minimums worth the cost? Will they make people safer?

Vic Toews, Mr. Harper's justice minister, insists they will. They
proved themselves in the United States, Mr. Toews told reporters a
few weeks ago. It was tough mandatory minimums that drove down crime
in the 1990s.

But Mr. Kleck says that's not true: "The consensus of American
experts who have looked at that is that the mandatory minimums didn't
help and may well have hurt."

Ouch. A wonk consensus.

I feel for Mr. Harper, I really do. It can't be easy to admit you got
a major issue all wrong. But he's no mere politician. He's an ideas man.

I'm sure he'll do the right thing.