What’s an easy way to trip up an economist? Ask one to forecast Canadian job growth.
Take the most recent edition of the Labour Force Survey. It showed employment surged by a net 66,800 positions in January, paced by the largest one-month gain in private-sector hiring on record. What did economists call for? Their consensus estimate was a scant 5,000 jobs added, or 7.5 per cent of the reported figure.
Granted, a one-month gaffe is not particularly noteworthy, given how difficult job creation is to predict. But for some time, economists have been seriously low-balling Canadian job growth – and it’s only gotten worse. That’s worth keeping in mind when the next LFS comes out Friday, in which economists expect a meagre 600 positions added.
Derek Holt, head of capital markets economics at Bank of Nova Scotia, pointed out the chronic shortfall in estimates after January’s report. Since the end of 2015, the economy had added nearly 900,000 jobs, as measured by the net change in employed persons. Economists called for 325,000 positions, or less than 40 per cent of the total, based on the monthly consensus forecast of those surveyed by Bloomberg.
“No wonder why so many are perennially bearish toward the economy and relatively less bearish or more bullish on [interest] rates: consensus simply sucks at forecasting jobs and has been ignoring this fact for years!” Mr. Holt wrote in a note to clients. “That may well be among the best examples of confirmation bias that one will find anywhere these days.”
Economists have long struggled to pin down job gains. To some degree, that’s defensible. The LFS is often derided as a “random number generator,” largely because the monthly change in employment comes with a hefty margin of error, typically around 29,500. For some perspective, since 2000 the Canadian economy has added an average of 18,600 jobs per month, or substantially lower than the margin of error.
Put another way, we typically don’t know whether Canada truly added or lost jobs from one month to the next. December provides a good example. The LFS said a net 8,700 jobs were created that month. But here’s another way of describing what happened: the true figure may have fallen between a gain of 37,200 jobs and a loss of 21,600 positions, after taking into account the margin of error. Even then, there’s a chance the true figure fell outside that range.
For economists, their forecasts have hit a particular rough patch. Consider that from 2000 through 2015, the consensus amounted to 58 per cent of the total number of jobs created during that period, versus 37 per cent since then.
So, what’s going on? Douglas Porter, chief economist at BMO Nesbitt Burns, points to population gains. In a February research note, he noted that Canada’s population aged 15-plus increased by 432,100 in January from a year earlier, the largest 12-month increase in 43 years of records.
Indeed, the federal Liberal government has ramped up immigration levels to help stoke economic growth, and in turn that’s driven much of the population increase. Canada’s immigration system also tends to favour higher-skilled newcomers than other comparable countries. As such, the employment rate for recent immigrants (those who landed five or fewer years back) is higher than for those born in Canada, though the newcomers’ rate still lags in the prime working ages of 25 to 54.
“In other words, Canada’s provinces, particularly its largest cities, have done a fine job plugging landed immigrants into the labour market,” National Bank Financial economists Taylor Schleich and Warren Lovely wrote in a client note last month. “That’s no doubt partly due to the sheer quality of the immigrants we’re taking in.”
There’s also another factor at play: job growth, per the LFS, might simply be playing catch-up. Statscan has multiple measures of employment, including the payroll survey of employers, which many economists consider a more reliable gauge of job growth. Since the end of 2015, cumulative job growth was fairly similar between the surveys, before the LFS started to lag in 2018. January’s outsize gain went a long way to narrowing the gap.
“The [payroll] data showed a dip in December that may be capturing a softer near-term trend for hiring now that growth is slowing,” said Avery Shenfeld, chief economist at CIBC Capital Markets, in a research note. However, on Friday “we could still get a climb in the LFS jobs picture that would be catching up to earlier [payroll] data.”
Then again, Canada’s “random number generator” might have something else in mind.
Matt Lundy