Yukon and N.W.T. both claim lowest unemployment in Canada. What gives?

Last week, the N.W.T. Bureau of Statistics put out a release celebrating some good news for a change: the territory had the lowest unemployment rate in the country.

But there was one, tiny problem — the Yukon claimed the same thing.

According to their numbers, Yukon's unemployment rate sat at 5.6 per cent, to the N.W.T.'s 5.8 per cent.

Over in Yellowknife, statisticians said different — that the N.W.T. could claim unemployment of 5.9 per cent, to Yukon's 6.1 per cent.

So what gives? It all comes down to something called seasonal adjustment.

How unemployment figures are made

Since 1945, Statistics Canada has surveyed Canadians to produce a general estimate of how many people are employed and how many are looking for work.

Every month, they release a new assessment, and statistics departments in each province and territory line up to release their own interpretation of the data.

There's often something interesting for reporters to root out in the numbers — for example, this month, there was the news that despite plummeting private sector employment, hundreds of new government jobs kept unemployment numbers low in the N.W.T. Interesting, right?

But in the words of Mark Twain, "facts are stubborn things, but statistics are pliable" — in other words, when it comes to describing the world in data, there's more than one way to skin a cat.

Lies, damn lies, and statistics

So how does seasonal adjustment factor in here?

Well, employment numbers are seasonally adjusted when they take into account jobs that are lost at roughly the same scale at roughly the same time each year.

For example, there aren't a lot of people farming in January. So to count the loss of those agricultural jobs against Ontario's overall employment, for example, would be a little harsh — and maybe even create the false impression that the province's economy was faltering.

"In the Yukon, we have a lot of seasonality," explained Gary Brown, the chief information officer for the Yukon Bureau of Statistics. "There's a lot of swing from the winter to the summer."

Powering that swing are big job gains in tourism, mining exploration, and road construction in the summer months.

"Using seasonally adjusted numbers you're able to compare, say February to July," he said.

Brown said "both stats have merit," but when it comes to unemployment, more places use seasonally adjusted numbers than the alternative — including Statistics Canada itself.

But that's not true for sector-by-sector statistics, which are usually given in raw, unadjusted format.

An 'insignificant difference'

The N.W.T. and Nunavut, where seasonally unadjusted numbers are used for all data, are the outliers here. And that makes a certain amount of sense.

Seasonal work in those territories accounts for a smaller percentage of the labour force. And in a global pandemic, when not many people are moving around for any reason, there's a case to be made that unadjusted numbers give a more accurate picture of the way things are.

But even if they can't agree on the numbers, don't expect the territories to go to war over these statistics.

"When we get into who's lower … the numbers are so small, in both territories it's just a handful of people that would make a difference either way," said Brown. "It's basically an insignificant difference between the two."

Still, he can't resist a humble brag about the Yukon's stellar employment.

"I don't want to brag, but … in the last five years … we've been lowest in the country, probably all but three or four of those months," he said.

Okay, Yukon, we get it — your job numbers are great. Just spare a little celebration for the N.W.T., where unemployment, unadjusted, is among the lowest it's ever been.