Thank you very much.
To understand whether increased penalties affect crime, I would suggest that you have to look at the overall weight of evidence. The conclusion that Professor Webster and I came to, based on a thorough survey of the evidence, especially that which was carried out in the last fifteen years, was that variation in the severity of sentences does not affect crime rates.
The reviews that come to different conclusions have generally looked selectively only at the occasional paper that finds some evidence that harsh sentences deter. In other words, for more than 25 years, the overwhelming weight of evidence has been consistent with the conclusions that harsh sentences, in legislation or in practice, will not have any consistent or appreciable impact on levels of crime in the community.
I would now like to turn to two sets of data that have received a lot of prominence here in Ottawa. Each of these has been used to demonstrate that harsh sentences deter. The first example comes from an analysis of a paper that has been referred to in the context of the current bill. It seems that when the deterrent impact of harsh sentences is raised, the name Steven Levitt, one of the authors of the best-selling book, Freakonomics, is mentioned, and his paper with Daniel Kessler, published in 1999, is cited.
On the basis of their evidence of the effects of the June 1982 California initiative, Proposition 8, these two economists concluded that the increase in sentence severity that came into effect in June 1982 was responsible for the reduction in crime in California. As a result of this 1982 change in sentencing laws in California, sentences for certain crimes committed by repeat offenders were made considerably longer. From a deterrence perspective, the change in sentencing laws was seen as a good opportunity to test the deterrence theory, since the change in the sentencing was dramatic, sudden, and well publicized. The typical finding is what's shown here, and it is also shown in the translated document.
These are data from Kessler and Levitt's original paper, and what you see is the timing of the law changes marked by the vertical line. From this graph one could easily conclude that crime was going up until the time the sentencing law in California became dramatically harsher. Crime then dropped dramatically, immediately after the law, one could conclude, and these would obviously be quite impressive results.
The findings are similar for four other crimes that were covered by the change in law. Crime went up before the change in the law, Kessler and Levitt's data would show, and then dropped dramatically afterwards. It's no wonder that the supporters of the current bill have repeatedly cited this single study by a quite famous economist, but I'd like you to look again at these data.
Look at this curve carefully and what you'll see is something that's quite peculiar. Levitt, in the published paper, presented data only for the odd-numbered years. That's what's in the figure; that's what was in his paper.
Let's look at the full set of data that were never publicized and never presented in this very highly cited paper on deterrence. This slide simply adds the even-numbered years. The data for the odd-numbered years is identical to what you saw before. Again, the vertical line is when the law change occurred. Unlike the partial set of data, which Kessler and Levitt relied on, what you see is that crime was going down, and started going down before the law changed. The other offences examined by Kessler and Levitt, which were subject to these especially new harsh sentences, show the same kind of pattern.
On the left of these panels, I've presented the data as published by Kessler and Levitt. On the right panel, all I've done--all, in fact, Cheryl Webster, from the University of Ottawa, and Frank Zimring, from California, and I did--was to add the data for the even-numbered years. By choosing, as Kessler and Levitt did, to present the data only for the odd-numbered years, they gave you a picture of the trends that is completely different from the picture you see when all of the crime rates for all of the years are included.
If you wish, you can look at the monthly data to get a more exact estimate of when the crime drop occurred. We did this as well. Here's one example of it. Again, we marked the time when the law changed by the vertical line, and what you see is that the crime drop started before the law changed, not after, as you would expect if it were the law that was responsible for the change in crime rates.
These graphs are part of a paper that I co-authored with Professor Webster, and Professor Franklin Zimring, from the law school at the University of California, Berkeley. Professor Zimring is one of the world's experts on deterrence, having written extensively on deterrence, beginning with his classic book on the topic in 1972.
As you can imagine, Professor Levitt is not very pleased with our analysis. The best one can conclude I think from Levitt's very interesting, very selective use of data is that it would be risky to base any policy on a study such as this.
About a year ago, in the last session of Parliament, when you were examining Bill C-215, you had a witness before you who indicated that sentence enhancements had helped to drive down the rate of violent crime in Florida. His evidence, like that of your local witness last week, concluded that Florida's 10-20-life law may have sounded convincing. The implication of their statements is clear: tough sentencing regimes drive down crime.
I'm old-fashioned. I think you should look carefully at the data. From the data I presented to you at the outset, you should already be skeptical about such assertions. Crime was already on the downward trend in the United States. Violent crime peaked in the United States and Canada in the early 1990s and then drifted downward.
So let's look at this trend in California. The next two figures show total index crime, which is a measure of the total more serious kinds of crime, and index violent crime for Florida in the 1990s. The timing of the implementation of the so-called 10-20-life law in Florida is marked again by the vertical line.
If you look at this figure, or the next one, which deals with violent crime, the problem with the inference that the law created a change is immediately evident. Crime was going down anyway. If these two figures didn't have a vertical line in them showing where the law change took place, you wouldn't have any idea that anything special was happening. Crime was going down in Florida, just the way it was going down in other parts of the United States and in Canada. There is no evidence that the change in law changed anything.
There are obviously many more studies on this topic. The best research examines more than one jurisdiction and attempts to control for other factors known to correlate with crime rates. Considerably more sophisticated studies have been carried out.
In the United States in the 1990s, largely as a result of the popularity of the so-called three strikes laws, many U.S. jurisdictions brought in very harsh sentencing regimes for at least some offences. Some studies have looked, overall, at the impact of these sentencing changes. One set of investigators, for example, examined the impact of the sentencing changes on seven different crimes in 21 states, using the data from states where no changes were made as a form of a control.
This slide shows you a summary of their findings. The authors report that there were as many increases in crime as there were decreases that followed the imposition of three-strikes sentencing laws. Clearly, it is just as inappropriate to focus on only those changes in the law where crime decreased as it would be to focus on those instances where crime increased after sentencing got tougher. But these findings do show you the dangers of taking isolated findings out of context.
There are two other sets of problems with mandatory minimum sentences that I would like to mention. It is almost inevitable that mandatory minimum sentences will result in disproportionate sentences for at least some offenders. We already have a requirement in the Criminal Code that sentences must be proportionate to the severity of the offence and the offender's responsibility for that offence. It is my impression that most Canadians endorse proportionality in sentencing.
Clearly, Parliament, in attempting to constrain judges with mandatory minimum sentences, is purposefully sending the message that it does not trust judges to judge the severity of offences. But in addition, mandatory minimum sentences almost certainly force judges to hand down sentences that violate section 718.1, the proportionality principle in sentencing.
If the proportionality principle needs strengthening in the Criminal Code--and I, for one, believe it does--then there are ways in which this can be done. But mandatory minimum sentences have been shown repeatedly not to be an appropriate tool to accomplish this goal. There are other harms that can come from proposals such as this one. If the Parliament of Canada were to approve Bill C-10, it would be telling Canadians that Parliament can make our communities safer by increasing mandatory minimum sentences. This is, quite simply, a false promise. If you were to vote in favour of this bill, therefore, you would be, in my opinion, making a promise to Canadians that is known to be false. But it is worse than that. Focusing on such matters as mandatory minimum sentences also distracts you, the Parliament of Canada, from considering approaches to crime prevention that might actually make our communities safer. In other words, by convincing yourselves and others that the proposals such as this one will improve our communities, you necessarily do not adequately consider approaches to crime prevention that would improve our communities.
Thank you very much.