Canada may be fighting trade wars on two fronts, but its job market didn’t get the memo. Statistics Canada (StatCan) data show the country added tens of thousands of jobs in September. Heck, the data even shows a 6-figure addition of full-time jobs over the 30-day period. Suspicious? You’re onto something. Let’s talk about the overadjustments that likely just occurred.
Canada Gained 60k Jobs In September, Driven By Manufacturing
Canada’s job market was booming last month, despite the flood of negative news stories. Seasonally adjusted employment grew 0.3% (+60.0k) to 21.02 million in September, which is a remarkable month. Not enough to roll back losses in the past two months, but this would have been one of the biggest months on record pre-2020.
The deeper the data, the better the news. Seasonally adjusted employment’s biggest driver was manufacturing—which grew 1.5% (+28k jobs) in a single month. It was followed by healthcare (+0.5%; +14k jobs), and agriculture (+6.1%; +13k jobs). Getting skeptical considering manufacturing and agriculture were both simultaneously hit by shocks? We trained you well. Let’s talk about how the sausage was made.
Canada May Be Overseasoning Those Adjustments
Seasonal adjustments are designed to eliminate the volatility of predictable seasonal patterns. People buy homes in the Spring and do Christmas shopping in December. Schools are back in session in September, which means summer employment will decline. These are predictable trends that can arguably be forecast, so they can be accommodated by adjustments. In Canada, this means generally suppressing growth in the summer and boosting the winter contraction. The end result is a trend smoother than a policymaker’s brain.
What’s not recurring? Canada engaged in simultaneous trade wars with its two largest trade partners, who also happen to be the two largest economies in the world. That’s great news for our sanity that this isn’t a regular occurrence, but the irregularity means StatCan most likely overcorrected the data.
Canadian Employment Data Overadjusted, Overstating Growth
Let’s take a look at the unadjusted monthly growth over the same period. Unadjusted employment saw a monthly drop of 0.22% (-46.9k jobs) in September 2025, and 0.21% (-43.4k jobs) over the same period in 2024. A 1 basis point (bp) variance is roughly the same seasonal drop, with almost no difference in the raw behaviour.
Seasonally adjusted monthly growth came in at 0.29% (+60.4k jobs) in September 2025, and 0.18% (+36.7k jobs) over the same period. The unadjusted 1 bp variance translates into an 11 bp increase, which is substantial. Roughly the same move in the unadjusted data translated into 61% more growth this year. The odds of this being an overcorrection are looking pretty good.
Let’s review that banger for full-time jobs. Seasonally adjusted full-time workers saw monthly growth of 0.6% (+106.1k jobs) in September, and 0.5% (+92k jobs) over the same period in 2024. Last year was impressive, but this year is a mindblowing 6-figure addition of full-time employment… in theory.
Unadjusted monthly full-time employment contracted by 2.1% (-361.6k jobs) in September 2025 compared to a contraction of 2.0% (350.1k jobs) in 2024. The unadjusted rate fell an additional basis point this year, but resulted in seasonal adjustments adding 1 bp to growth. A larger unadjusted loss rate resulted in bigger seasonally adjusted gains. This appears to be another heavy-handed overadjustment. Tragically, workers who only exist in seasonal adjustments are kind of like billionaires—they don’t pay taxes.
None of this is to suggest anything malicious, but to clarify that data is data—until it isn’t. This is a well-understood phenomenon we’ve discussed before in the context of recessions, and it’s no surprise to see it occurring here. It’s important to understand that when data appears too strong or too weak, it’s worth double-checking the unadjusted data. It’s also important to understand that StatCan’s seasonal adjustment model is widely used in economics, so an overadjustment here will likely be seen in other areas—from GDP to home sales, and even home prices. Good luck filtering the noise.