Skip to content

StatsCan under fire from Yukon’s chief statistician

Following another unflattering study of the North by Statistics Canada, the agency will hear criticism from Gerry Ewert when Canada’s top…

Following another unflattering study of the North by Statistics Canada, the agency will hear criticism from Gerry Ewert when Canada’s top statisticians gather later this month.

“It’s the lumping together of the North as if we’re all the same,” said Ewert, director of the Yukon’s bureau of statistics, noting his ongoing objection with StatsCan research methods in the territory.

On Monday, Canadian newspapers ran headlines proclaiming people in the North to be three times more likely to be victims of violent crimes, like sexual assault and robbery, than southerners.

The stories originated from a new StatsCan study, Victimization and Offending in Canada’s Territories.

Drawing from police data, it grouped the Yukon with Nunavut and the NWT.

When generalized statements are drawn using numbers compiled from the three territories, the Yukon appears to have far different social issues than it does, said Ewert.

That leads to false perceptions of the territory across Canada, and poor research foundations for public policy to be built upon, he said.

“By getting lumped together we, in many ways, get the short end of the stick,” said Ewert.

While the study reports northerners are three times more likely than southerners to be victims of violence, a closer look suggests the Yukon is very different than its neighbours.

Totals in the study for criminal code incidents, like homicide, assault, sexual assault and robbery in 2005, were 957 in the Yukon, compared with 2,843 in the NWT, and 2,112 in Nunavut.

The Yukon stats bureau has been tracking RCMP data since 1995 and has noticed a decrease in the number of men charged with spousal assaults, said Ewert.

But you wouldn’t get that impression from StatsCan’s research, he said.

“It’s less of a problem here than it is in the other two territories, but we get lumped together with the others — so when they report for the North, they’re reporting a rate that’s substantially higher than it is for the Yukon.”

If a politician or bureaucrat were to design a family violence program in the Yukon, the skewed statistics may lead them to believe there is a bigger problem than there really is, leading to flawed policy, he explained.

During the upcoming roundtable meetings in Ottawa, Ewert will argue StatsCan should stop grouping the territories together and use different research methods in the North.

“The three territories are unique, discrete, different and deserve to have the same level of information as any other jurisdiction in Canada,” said Ewert.

Canada’s stats agency used to hear similar complaints from Manitoba, Saskatchewan and Alberta.

The three were once grouped into a region called “the Prairies” in StatsCan research, despite their differences, said Ewert.

“It just drove them crazy because they’re quite different, in terms of both their population and demographics, and also the structures of their economies.”

The same holds true North of 60, he said.

But there have been small statistical victories for Ewert.

StatsCan once reported Yukon alcohol sales numbers as if they reflected overall alcohol consumption rates in the Yukon.

The small population mixed with masses of tourists led to hugely skewed numbers, he said.

“Yukon has a population of under 31,000 people, and we get close to 350,000 people going through here in a year.

“Now, in order for them to take retail sales of alcohol and say it’s ‘consumption’ and therefore, that Yukoners are the biggest boozers in the world, they have to assume that not one of those 350,000 had a beer.

“We finally managed to get them to stop talking about it as consumption,” he said.

“They did that for years and years, and I screamed and yelled, and every time a newspaper in the South wrote it up, I would write a letter to the editor.”

Though the grouping method is not ideal, small sample sizes in the North often require the three territories to be lumped together, said Jodi-Anne Brzozowski, the author of the recent study on victimization in the North.

“We do what we can to release the information at the level we can, but we are restricted, because we have strict guidelines,” she said.

“So, when we get into small sample sizes, what we have to do as an alternative is to group together categories.”

More than 1,300 people in the three territories were polled through telephone surveys for the study.

The tradeoff is always between accuracy and disseminating knowledge for sparsely populated areas, explained Brzozowski.

“We have the data. But if we split it by territory then we can’t release it, because the sample size becomes too small.

“This is the first time that we’ve been able to disseminate information about self-reported victimization in the territories.

“Yes, there are variations (between the territories) but that happens in the provinces as well when we’re reporting overall,” she said.

 “We’re not just painting a blanket picture of the territories; we have the detailed information where it warranted and where it is releasable.”

In many cases, however, the Yukon’s crime statistics are “much lower than the other territories,” she said.

Ewert will meet with StatsCan officials in Ottawa between November 14 and 17.

He’s not holding out for much change.

“Statistics Canada prefers to use a consistent methodology for all jurisdiction, whether or not it is appropriate for the size of the jurisdiction,” he said.