Canada employment data: impossible to analyze

I should probably stop even looking at the Canadian employment data when they come out each month, because the numbers are so astoundingly erratic that they make my brain hurt.  Let's take a look at the September numbers, which were released this morning.

First the headlines: total employment in the economy rose by 67,000 in the month, way above the market's expectations.  Using the usual 10-to-1 US-Canada comparison, this would be equivalent to a 670,000 rise in US non-farm payrolls, whereas we know that the actual gain in US employment for the month was less than a quarter of that.  Canada's unemployment rate, remarkably enough, was unchanged at 7.0 percent, because according to StatsCan's survey, labour force participation rose sharply.  That's usually considered to be a positive sign -- people don't start looking for work if they know there's no work to be had -- but once you look behind the headline numbers, the picture looks a lot less rosy.

For one thing, most of the employment gain was in part-time rather than full-time work.  That has, in fact, been the case throughout the past year.  Although the number of persons employed has risen by 0.8 percent in the past twelve months, aggregate hours worked are only 0.2 percent higher.  That's consistent with the generally sluggish GDP growth that has been evident throughout the period.

The service sector accounted for most of the job gains in September (about 55,000).  The modest rise in employment in the goods producing sector is further evidence that the hoped-for rebalancing of the economy, in response to the weak exchange rate, is proceeding at a glacial pace.

And one more thing: if StatsCan is to be believed, fully 50,000 of the jobs created last month fell into the category of self-employment.  Given that the number of self-employed in Canada is around 2.8 million, that's a remarkable gain for a single month.  If experience tells us anything here, it's that a large percentage of the newly self-employed will wind up back in regular paid employment within a month or two.  This self-employment series is perhaps the most volatile number in a notoriously erratic data set.

The best takeaway from today's numbers is almost certainly not the headline job gain, which will surely not be sustained.  Rather, it's the year-on-year change in employment and in hours worked.  Those numbers show that, with Federal spending initiatives just starting to kick in, the economy is growing, but very slowly.