Why Markets Got Canadian Jobs Wrong

• Canada lost another 40k jobs…

• ...but the concentration in the education sector should be faded…

• ...while the BoC will be more concerned about further evidence of a wage-price spiral

• CDN jobs / UR, m/m 000s // %, SA, August:
• Actual: -39.7 / 5.4
• Scotia: 0 / 5.0
• Consensus: 15/ 5.0
• Prior: -30.6 / 4.9

I’m inclined to fade most of the loss of employment last month and place greater emphasis upon persistently strong wage gains that are exceeding the rate of inflation. In my view that’s how the Bank of Canada is likely to treat the report. For this reason, I don’t agree with the market reaction that drove short-term yields lower including a 9bps decline in the yield on 2-year GoC bonds and the Canadian dollar about half a penny weaker and would fade the reactions. See chart 1 for some initial highlights. Also remember that the high noise factor to this report indicates that jobs might have declined by anywhere from about –100k to a mild gain of under +20k based upon confidence intervals around the estimates.

The quality of the data behind the job loss should be treated with care. That’s because the 40k drop in total employment was heavily driven by a 50k drop in education sector jobs. Where did all the teachers go?? Possibly nowhere other than back to school as it may just be a data distortion. That likely has something to do with messy contract negotiations and the staggered expiration of teacher contracts across the country. For instance, teacher contracts expired on June 30^th in BC and Manitoba and August 31^st in Ontario and Alberta as examples. Each of the biggest provinces saw declines (chart 2). Since it’s a household survey, we have to be careful toward the reliability of responses when teachers are asked whether they are presently employed as some might have said they did not have a contract so were not technically employed and, some might have had a soon-to-expire contract but responded they were essentially without one during the August reference week. It’s also possible that seasonal adjustments don’t adequately adjust for these issues over the years. Tisk tisk, teachers not necessarily knowing how to answer surveys…

The education sector drove the decline in overall public sector payroll employment that fell by about 28k. Private payroll employment fell by 4.4k which is not statistically meaningful. Self-employed jobs were down by about 8k. Thus, even after taking education out of the picture this would still not have been a great report, but better than the headline loss indicates.

The strong wage figures, however, will fan concern at the Bank of Canada that wage-price spiral dynamics are being driven. Average hourly wages for permanent employees were up by 9.6% m/m at a seasonally adjusted and annualized rate. 9.6%! This measure has exploded over the past four months with gains of 10½% in May, nearly 12% in June, 8½% in July and 9½% in August. See chart 3. The trend has generally been strong for over a past year with a brief interruption during the period of initial omicron restrictions and reopening effects. The longer term trend in wage growth confirms the impression that the job market remains very tight. Add in poor productivity growth and the combination of the two is disconcerting if the aim is to get the cost of living under control.

That rate of growth in nominal wages is now exceeding inflation over this period such that real (inflation-adjusted) wages are rising. By this same measure, headline CPI was up by 3.2% m/m SAAR in July (August pending), 6.5% in June, and by 11–13% in each of March, April and May while core CPI (ex-food-and-energy) has been up by 6.1% in July, 5.2% in June, and 8–9% in March through May. This will complicate the Bank of Canada’s fight against inflation.

Hope lies in the fact that the labour force expanded by about 66k and when combined with the job loss drove the unemployment rate up a half point to 5.4%. This could be a positive if wage gains are pulling people off the sidelines back into the job market which could raise contestability and dampen wage gains. That’s a highly premature argument to make just yet since last month’s gain in the workforce followed an almost 130,000 decline in the size of the labour force over the prior two months. The participation rate reflects this (chart 4). 

Hours worked are a further warning sign on GDP growth. They were flat in August which matters because GDP is an identity expressed as hours worked times labour productivity which itself is hardly Canada’s strong suit! On a quarterly annualized basis, growth in hours worked has been steadily slowing and that signals waning economic growth (chart 5).

Chart 6 shows a breakdown of the change in employment during August by sector. Chart 7 shows that even with recent weakness, overall employment remains 383k higher than in February 2020 just before the pandemic’s effects were felt. The job market remains tight. If the teacher distortion shakes out in the next report or two—which depends upon the state of labour negotiations—then we could see this sector drive a rebound in employment. We should all hope that the two sides will come together and avoid further disruptions to in-person learning.