Brought to you by the same party that did away with the ability of Statistics Canada to rely on the long form consensus as facts have a nasty habit of getting in the way of the Harper Con ideology du jour.
Maybe the Conservatives should be asking us whether we are more afraid of a party leader who happened to work in the US or bringing US Style Crime and Punishment to Canada as their party is so intent on doing.
The sad thing is that we may very well be fighting an election where the two major political parties beat the drums about how Canada approaches controlling crime and we’ll be seeing an awful lot of bullshit studies being prepared and thrown into the fray.
[quote] "Study claiming crime on the rise gets hung out to dry by experts in the field
…OTTAWA - Criminologists say a study that attacks the long-standing measurement of Canada’s crime rate is “highly politicized” and without statistical merit.
Crime and punishment appears to be shaping up as a defining ballot question whenever Canadians next go to the polls, so the statistically unorthodox claim that violent crime is on the rise is noteworthy.
Scott Newark, a former Harper government adviser who counts himself a contributor to the Conservatives’ 2006 justice platform, released the 29-page study last week questioning Statistics Canada’s methodology on compiling crime stats.
Newark asserts that “many of the most common conclusions that are drawn about crime in Canada are in fact incorrect or badly distorted.”
“Serious violent crime is increasing,” the former executive officer of the Canadian Police Association flatly asserts.
While Newark’s report for the Macdonald-Laurier Institute was given prominent coverage by both the Globe and Mail and National Post newspapers, the wider academic community that relies on the data was not consulted.
Their reviews are scathing.
“It’s really badly done. It’s embarrassing, actually,” said Neil Boyd, a criminologist at Simon Fraser University in Vancouver.
“For anybody who follows what Statistics Canada puts out, it’s a grossly unfair characterization. It mixes up key concepts…”
“That’s what struck me. This is a highly politicized document that isn’t paying attention to relevant data.”
James Hackler, a criminology professor at the University of Victoria, wrote a letter to the Globe on Friday saying the study illustrates “ideological bias.”
In an interview, Hackler said studies across the Western world clearly show crime rates in decline, although criminologists differ on the reasons.
Newark’s study, he said, “is like saying the world is flat. This is not a credible report.”
And Rosemary Gartner, a criminologist whose course load at the University of Toronto includes methodology, appeared dumbfounded by Newark’s repeated insinuations that Statistics Canada is manipulating the data.
“He makes a reference how indeed it serves their institutional purposes to show that crime is going down,” Gartner said in an interview.
“What the hell does that mean? Did it serve StatsCan’s institutional needs when crime was skyrocketing in the early 1990s?”
Newark, whose political advocacy for “tough on crime” policy goes back to the Mike Harris Conservative government in Ontario, bases his critique on some key assertions:
— Under-reporting of crimes because only the most serious charge from any one incident is counted.
— Data that Statistics Canada once included in its annual Juristat report is no longer published.
— The Crime Severity Index, based on sentence length for particular crimes, is a subjective measurement influenced by lenient judges.
— StatsCan’s after-the-fact upward revisions of crime data tend to exaggerate falling crime rates from year to year.
Julie McAuley, director of the Canadian Centre for Justice Statistics at Statistics Canada, and John Turner, chief of the agency’s policing services program, addressed Newark’s claims one by one:
— Reporting only the most serious charge from a multiple-charge incident is somewhat problematic, McAuley said, but it’s been handled this same way since 1962 and conforms with other jurisdictions such as the U.S. The recording method reduces variations in charging practices between police services.
— All StatsCan data from the past 30 years is available upon request. It just can’t all go in a single publication.
— The Crime Severity Index uses national sentencing averages to avoid local judicial variations. The index was adopted in consultation with the Canadian Association of Chiefs of Police and all the provinces. “The principle founding mandate behind the whole crime index was that it had to be objective,” said Turner.
— Over the last decade, annual data revisions have increased crime stats six times and decreased them four times, so the revisions cut both ways.
Gartner, who teaches methodology, also took Newark to task for continually switching his comparison years, and for selectively comparing incident numbers without accounting for population growth.
She also notes that Newark uses StatsCan’s General Social Survey findings to indicate unreported crime is a problem. But, she said, he “completely ignores the fact that the (same) GSS victimization survey shows violent crime and other crimes have been going down.”
Boyd, moreover, said the survey’s findings on unreported crime have barely budged over the years.
Gartner and Boyd both agree with Newark that it would be great to have far more detailed crime data, particularly profiling who is committing crime.
“The revisions he’s calling for would require a huge infusion of cash to every single police force in Canada, as well as to StatsCan,” said Garner. “I’d love it. I’m a criminologist. The more data the better.”
"But the (report’s) implication is: ‘They can do this, they should have been doing this, it’s not clear why they haven’t been doing this.’
“He knows better.”""[/quote]