• Canada gained 60k jobs in September with some decent details
  • Seasonality quirks should dampen confidence in the data
  • Why October’s BoC decision may still be live
 
  • Canada employment m/m 000s / UR %, SA, September:
  • Actual: +60.4 / 7.1
  • Scotia: -50 / 7.2
  • Consensus: +5.0 / 7.1
  • Prior: -65.5 / 7.1

There are few—but meaningful—holes to poke in a very strong Canadian employment report that adds to evidence in favour of a hold by the Bank of Canada on October 29th.

That’s more probable now, but not assured. We still need to see CPI on October 21st and the BoC’s quarterly surveys the day before, but at this point the odds of skipping the meeting have gone up. The economy is still weak and building spare capacity which could motivate the BoC to opt in favour of additional insurance. The BoC looks at job market trends that are still weak with employment down 46k in Q3 and there is a strong reason to be careful interpreting this round of jobs numbers.

DETAILS

Chart 1 provides some summary metrics. Total employment was up by +60.4k. Almost all of that was in payrolls (+53k) with public sector payrolls up by 31k and private sector payrolls up by 22k but self-employed was also a little higher at +8k.

Chart 1: Canadian Jobs Break Down

Chart 2 shows the breakdown of job growth by sector. Somehow manufacturers added 28k jobs. The rest was mostly in health care and social assistance plus 'other' services ex-public admin. There was decent breadth to small gains in some sectors. The only notable downside was the retail/wholesale sector.

Chart 2: September Changes in Canadian Employment Levels by Sector

Wages grew by 2.5% m/m SAAR (chart 3). That’s the second consecutive slow down, but with the same cautions that collective bargaining exercises remain skewed toward above-inflation wage gains for years to come.

Chart 3: Canadian Hourly Wages

All of the gain was in full-time jobs at +106k with part-time losses at -46k.

Go Alberta! That one province’s +43k job gain was the dominant driver. BC was up 8k. Other provinces were a combination of small gains and small losses. Chart 4. The manufacturing gain was in Ontario (+12k) and Alberta (+7.9k). Alberta's 43k job gain was 26k in goods sectors led by construction (+8.4k) and manufacturing (+7.9k) plus services (+16.7k) led by prof/scientific/technical (+6.5k), education (+7.4k), public admin (6.6k), 'other' services (6.6k) and health care (+4.2k).

Chart 4: Job Growth in Canadian Provinces

Where I’m skeptical is toward the seasonality aspects of the report. For one, there was an unusually light m/m NSA drop in a month that's usually much bigger (chart 5). That’s the second year in a row when that’s happened. In fact, the month of September has seen the strongest seasonally unadjusted readings— either the biggest gains or the smallest dips—for overall employment on record in each of the pandemic and post-pandemic years from 2020 through to 2025 as shown in the chart. Is this because of changes in seasonal hiring patterns that are genuine in nature only since the pandemic, or that are not being properly controlled? It’s hard to tell and Statcan offers nothing to back up whether or not this is credible.

Chart 5: Comparing Canada Jobs for All Months of September

Further, the SA factor moved HIGHER which negates the whole argument about shedding the pandemic-era weights in a changing recency bias that was previously showing signs of moderating more toward averages (chart 6). This September’s SA factor was among the highest recorded for like months of September. Anything lower would have strongly dented the headline jobs number. 

Chart 6: Comparing CA LFS SA Factor for All Months of September

In fact, had Statcan stuck with the September 2024 SA factor, then jobs would have risen by about 20k, or one-third of the reported pace. Again, the question is whether seasonal adjustments are properly controlling for changes in hiring patterns. Statcan will simply say it’s the SA factor that is spat out of an X12A seasonal adjustment calculation but missing is any intuitive explanation for why this September pivoted back to some of the highest seasonal adjustment factors since the pandemic and its immediate aftermath. Take the numbers with a lot of salt in my view.

Hours worked are tracking softly for purposes of estimating September monthly GDP and Q3 GDP. Hours were down by 0.2% m/m SA. They were only up by 0.4% q/q SAAR rise in Q3 overall (chart 7). Since GDP is an identity defined as hours times labour productivity, such meek hours worked figures indicate a soft economy especially since productivity is hardly Canada’s strong suit.

Chart 7: Canada Total Hours Worked

By cohort, the gain was led by women 25+ (+40k) followed by men 25+ (25k) as youths 15–24 lost a small amount (-4k).

The unemployment rate held stable at 7.1% this month compared to the upward trend since 2022–23 that was primarily driven by an excessive number of temporary residents. Chart 8.

Chart 8: Canada's Unemployment Rate

Shortly after the numbers, Ottawa announced a limited set of three policy measures that carry micro, not macro implications. It will make automatic federal benefits entitlement to low-income workers with automated tax filings, make a national school food program permanent and renewed the Canada Strong travel pass. These are not judged to be significant to the macro landscape.