Getting soft on crime
How the justice system in Canada punishes less than it used to
We started this series with a paradox of our criminal justice system. Canadians are (rightly) concerned about “revolving door” justice – that is, a perception that people doing foul deeds are arrested, charged, and quickly released. Many – most notably provincial governments – have attributed this to bail laws that favor pre-conviction release over remand custody.
We come to a paradox. As concern over “revolving door” justice grows, so has the number of people held in remand custody. These people – legally deemed innocent – are locked up pending their day in court. There are more legally innocent people in provincial jails than there are those convicted. It’s been that way for a decade and a half – and the numbers are not close.
That’s the paradox. A perception of “soft” bail laws combined with the “tough” outcome of lots and lots of folk in remand custody.
I’ve been making the argument that the focus on bail law is misplaced. Politically, focus on the bail provisions of the criminal code and individual bail/release decisions by judges have – for the provincial governments – the convenient effect of letting the provinces off the hook for problems in the criminal justice system.
And make no mistake about it. Provincial governments have more impact on the criminal justice system than does the federal government.
We looked at how the increasing slowness of our criminal justice system has caused the number of people in remand to increase. The variation by province shows the importance of provincial governments. It takes Manitoba twice as long to process criminal cases as does Saskatchewan. As a result, Manitoba has more people in remand custody.
We also looked at how the Criminal Code’s curious absence of the prior record in factors to be considered by judges in sentencing contributes to an inability of judges to deal with chronic and repeat offenders.
Today, let’s look at the deterrence effect of our criminal justice system.
Punishment and deterrence for crime
Classical deterrence theory was developed by a guy named Jeremy Bentham in the 1820s. Bentham argued that if people were punished for crimes, they would be less likely to commit crimes. For deterrence to work, Bentham argued that the three key factors were:
Certainty of punishment.
Speed of punishment.
Severity of punishment.
Bentham argued that severity was actually the least important aspect. At the time he was writing, Britian’s criminal law was known as the “bloody code.” Over 200 offences resulted in hanging. At the same time, crime rates were extraordinarily high. Why? Partly because Britain had no police forces. People were not deterred by severe punishments if they didn’t think they would get caught.
Bentham’s insights have stood the test of time for over two centuries. Not for every type of crime, but for lots of them.
So let’s look at how the Canadian justice system does with deterrence.
We’ve already looked at speed of punishment. In Canada, three-toed sloths move faster than our justice system and it’s getting slower.
That leaves us certainty and severity.
Measuring the level of deterrence
I’m going to construct some measures here. You won’t find these in criminology textbooks, but I think looking things this way can be useful.
Here’s the methodology discussion. You can skip it if you want to get to the bottom line.
Statistics Canada reports the number of crimes reported to the police. Reporting is done on a calendar year basis and is broken out to the criminal code classifications. Outcomes of the criminal justice system are recorded with something called the Integrated Criminal Court Survey. Clerks in courthouses across the land are required to fill in reports about each and every criminal court docket. Reporting is on a government fiscal year basis (April 1 to March 31). Coverage is incomplete. Quebec stopped reporting this data in 2021/22 fiscal year. Manitoba seems to have screwed up the reporting on sentence lengths to the extent Statistics Canada has “revised” the province’s reports of average sentence lengths by withdrawing the data completely.
But the data is still useful. We can look at the outcomes for different types of crime to compare over time and between provinces. There are limitations. In addition to the combination of intransigence and incompetence by Manitoba and Quebec, we have the issue that one person can be charged with several offences, or several people can be jointly charged as a result of one crime.
Timing is a problem. Given the slowness of the justice system, we don’t know when the crimes being resolved by the courts actually occurred. To address this, I’ve made the arbitrary decision to assign the number of reported offences in a specific calendar year to the court decisions of the overlapping government fiscal year. The Integrated Criminal Court Survey data run starts in 2005/06, so I begin by comparing the number of crimes committed in calendar year 2005 with the court outcomes in the fiscal year of April 1 2005 to March 31 2006. There will be leakage both ways, but this should more or less average out. Anyways, it’s the best I can do given the available data sources. If anyone has a better idea, let me know.
To conclude: what follows ain’t perfect, but it does provide a consistent basis of comparison over time and between provinces. With the exception the impact of Quebec’s separation from this aspect of confederation in 2021 on the national data, we are comparing apples to apples and oranges to oranges.
Let’s look at major assaults. These are Level 2 and Level 3 assaults.
Level 2 assault is defined in the Criminal Code in which the assailant “carries, uses or threatens to use a weapon or an imitation thereof” or “causes bodily harm to the complainant” or “chokes, suffocates or strangles the complainant.”
Level 3 assault is defined as one that “wounds, maims, disfigures or endangers the life of the complainant.”
We’re talking serious stuff here. These assaults are a lot more severe than a slap on the face. Each one carries the risk of turning fatal.
Here are the numbers.
In 2005, there were 53,451 of these major assaults recorded by police. This resulted in 31,223 charges being laid by police. This resulted in:
20,661 criminal cases before the courts. Some reported assaults resulted in more than one person being charged (and hence more than one case). Some people were charged with more than one assault within a single case. In some cases, prosecutors decided not to proceed with charges – either because the prosecutor thought there was not a provable case, because the complainant did not wish to proceed with charges, or because the charge was diverted into some sort of non-court or mediation process.
11,170 cases resulted in a guilty finding. There were 1,641 acquittals and 7,475 charges were stayed/withdrawn. (The numbers don’t completely add up because a few people died while waiting for their day in court and so on).
Of those convicted, 4,653 were sentenced to some jail or prison time.
Of those sent to the slammer, the average sentence length was 166 days.
Now fast forward to 2023. In that year, 85,973 cases of major assault were reported to police. This resulted in 52,285 charges laid by police. This, in turn, resulted in:
21,122 cases before the courts.
8,624 convictions, 376 acquittals, and 12,830 cases being stayed/withdrawn.
Of those convicted, 3,725 were sentenced to some jail or prison time.
Of those sent to the slammer, the average sentence length was 188 days. This apparent “toughening” of sentences is likely a statistical artifact. As more cases are stayed/withdrawn and fewer people are sentenced to jail, those sentenced to incarceration will have committed assaults that are, on average, worse than they were in 2005.
Let’s now apply the new Stobbe statistical analysis for major assaults.
In 2005, 20.9 percent of reported major assaults resulted in a conviction. By 2023, this dropped to 9.6 percent.
In 2005, 8.7 percent of reported major assaults resulted in someone going to jail. By 2023, this dropped to 4.1 percent.
In 2005, there were 14.5 days of incarceration per recorded major assault. By 2023, this dropped to 9.7 days.
The bottom line is that in less than two decades, our criminal justice system has cut its punitive effect roughly in half – no matter how you measure it.
And remember.
Ottawa had basically nothing to do with this. The law here has not changed.
I’ve done this exercise for a number of crimes. I won’t go through the full process for each but will just report on bottom lines.
Robbery
Robbery combines theft with violence or the threat of violence. It’s a serious crime. Every robbery has the potential to turn into a homicide since it based on the statement of “your money or your life.” Sometimes the robber kills even after getting the money.
In 2005, 9.2 percent of reported robberies resulted in a conviction. By 2023, this dropped to 5 percent.
In 2005, 6.8 percent of reported robberies resulted in someone going to jail. By 2023, this dropped to 3.5 percent.
In 2005, there were 46.3 days of incarceration per recorded robbery. By 2023, this dropped to 17.5 days.
Common Assault (Level 1 Assault)
This is the least serious of violent crimes such as a fairly ineffective punch on the nose or a threat to do so.
In 2005, 10.6 percent of reported common assaults resulted in a conviction. By 2023, this dropped to 4.9 percent.
In 2005, 1.5 percent of reported common assaults resulted in someone going to jail. By 2023, this dropped to 0.8 percent.
In 2005, there were 0.7 days of incarceration per recorded common assault. By 2023, this dropped to 0.3 days.
Sexual Assault (All three levels)
When I was teaching criminology, my students had a great deal of difficulty in coming to terms with the fact that there is no such thing as the crime of “rape” in Canada. Instead, there are different levels (three, to be precise) of sexual assaults. These range from slapping the butt of a server in the bar to the most hideous of – well – rapes. For court data, Statistics Canada lumps all three levels together.
In 2005, 8.1 percent of reported sexual assaults resulted in a conviction. By 2023, this dropped to 3.1 percent.
In 2005, 4.3 percent of reported sexual assaults resulted in someone going to jail. By 2023, this dropped to 1.3 percent.
In 2005, there were 22.8 days of incarceration per reported sexual assault. By 2023, this dropped to 9.1 days.
Theft
Theft is, well, theft. Stealing by stealth rather than force or fraud. Statistics Canada’s court data presents data for total thefts rather than dividing it by the Criminal Code distinction of over and under $5,000.
In 2005, 7.8 percent of reported thefts resulted in a conviction. By 2023, this dropped to 2.2 percent.
In 2005, 3 percent of reported thefts resulted in someone going to jail. By 2023, this dropped to 1 percent.
In 2005, there were 1.7 days of incarceration per reported theft. By 2023, this dropped to 0.3 days.
Breaking and Entering
Breaking and entering is a crime that speaks for itself. Sometimes it is done for the purposes of theft. Sometimes there is intent to commit assault or sexual assault. Sometimes people break and enter for mischief or curiosity.
Breaking and entering for theft is more serious than straight-forward theft. People feel the harmful effects of intrusiveness if some steals something from their bedroom than their front steps. Breaking and entering also carries with it escalated risks of violent outcomes. In most American states, breaking and entering of a residence (especially at night) is categorized as a serious violent offence because of potential deadly outcomes. We don’t have this formal distinction in Canada, but breaking and entering is considered a serious property crime.
In 2005, 3.4 percent of reported break and enters resulted in a conviction. By 2023, this increased to 3.8 percent.
In 2005, 2.1 percent of reported break and enters resulted in someone going to jail. By 2023, this increased to 2.3 percent.
In 2005, there were 5.2 days of incarceration per reported break and enter. By 2023, this dropped to 3.4 days. While conviction and incarceration rates increased a bit, average sentence lengths dropped from 250 days to 149 days.
The bottom line
When dealing with pretty much every type of crime, the Canadian justice system delivers a less punishment than it used to.
Why?
It’s been conscious policy decision by provincial governments to put fewer people in jail. Police are solving pretty much the same proportion of crimes as they used to. However, after arrest and charging, prosecutors are converting fewer of these arrests into convictions and incarceration.
There has been a policy decision by provincial governments to “divert” cases from the courts into some other kind of process. Even after charges have been laid, a much higher proportion of cases results in charges being stayed/withdrawn. This is something completely under the control of the prosecutors – who are employees of provincial governments.
Finally, for most offences, even when someone is sentenced to jail, the sentences are lower (this is disguised a bit since only the more heinous cases make it this far). This results from judicial sentences – but judges generally feel constrained by the prosecutors’ recommendations. They rarely give harsher sentences than the prosecutor recommends, and when they do, it has to be explained and justified.
Opinions differ on whether the decline in punitive/deterrent effect in our criminal justice system is a good or bad thing. Sometimes there is a good argument against sending someone to jail. But sometimes jail is called for.
Has the decline in punishment gone “too far”?
I spend a lot of time looking at criminal cases. I often describe a case to my wife and ask what she thinks the appropriate punishment should be. Please understand that my wife is a kind and gentle person. She identifies with the plight of the poor and oppressed. She is the polar opposite of a vindictive or retributive person. If you called her a “bleeding heart”, she would take it as a compliment.
Here’s the thing.
Her assessment of what penalty would be appropriate in a case is almost always “tougher” than the one imposed by our criminal justice system.
When our justice system is “softer” and more lenient than my wife, perhaps we’ve swung too far towards leniency. When we incarcerate for only 1.3 percent of sexual assaults, perhaps we are looking at a door that is locked to keep people out rather than a “revolving door.
I’m not going to dive too deep into this, but as we’ve cut back on the use of jail as a punishment, we’ve almost completely eliminated the use of fines for some criminal offences. In 2005/06, fines were imposed in 24.4 percent of all criminal cases where the person was found guilty. Almost two decades later, this is down to 15.3 percent. We still give fines for things like impaired driving (75.4 percent of people found guilty) but we’ve pretty much abandoned the tool to punish things like common assault or the theft. When I was a kid, I’d follow the court news in the paper. It was clear that a common assault generally resulted in a fine of $50 to $200. Last year, fines were imposed on only 0.5 percent of people found guilty of common assault and 4.7 percent of people found guilty of theft.
The role of provinces
I’ve repeatedly made the case that the decline in punishment (for better or worse) has been the result of decisions made by provincial governments rather than by Ottawa.
Let’s look at some empirical evidence for this claim.
Police in different provinces are basically equal in their ability to “solve” crimes.
Canada has one criminal code. They law is the same across the country. If Ottawa was solely responsible for outcomes, they would be consistent from sea to sea to sea.
But that’s not the case.
Let’s look at just one crime. Major assault (defined as Level 2 and Level 3 Assault). This is a serious crime. It results in physical damage to people. As we saw earlier, for every Level 2 and Level 3 Assault in 2023, there was a total of 9.7 days of sentenced jail time. Now let’s look by province:
Newfoundland and Labrador: 8.4 days
Prince Edward Island: 13.2 days
Nova Scotia: 14.3 days
New Brunswick: 21.8 days
Quebec: Does not report – was 15.8 days in 2020/21
Ontario: 8.8 days
Manitoba: Does not report on sentence length
Saskatchewan: 13.2 days
Alberta: 7.6 days
British Columbia: 4.3 days
With the same law in place, British Columbia produces one fifth as much jail per assault as does New Brunswick. We have a province more than twice as “tough” as the national average – and one less than half as “tough”. I won’t start using fancy-pants descriptors for this such as looking at standard deviations. I’ll just say that’s a lot of difference in outcomes – far too much to be attributed to chance.
When you see provincial premiers and justice ministers calling on Ottawa to “fix” the criminal justice system with something like “tougher bail laws”, you are seeing a politician deflecting attention and blame from the actual administration of justice – which is something within provincial jurisdiction and control.
This is the sixth post in a series looking at Canada’s justice system. Previous posts can be found here, here, here, here, and here.
If you want an in-depth discussion of remand custody, check out my book Lessons from Remand. I’ve also written books on the 'Mr. Big' Investigative technique and how convictions for murder can be obtained in cases where the killer disposed of the victim's body.