Statistics Canada reported on October 19 that gasoline was 27.8% more expensive in September than a year earlier. Excluding gasoline, consumer prices rose 2.2%. The two core measures the Bank of Canada watches most closely eased, with CPI-trim at 1.8% and CPI-median at 1.9%.
Households were answering a different question. In the Bank of Canada survey of consumer expectations, published the same day, five-year inflation expectations jumped more than half a point to 3.91% from 3.39% in the second quarter, rising to match elevated short-term expectations. Most interviews were completed before trade talks with the United States broke down, so the tariff fight cannot explain most of the jump.
A gap of just over two percentage points between the 1.9% core reading and the 3.91% households expect over five years is not evidence of irrationality. It is the predictable output of how people build an inflation estimate, and behavioural finance has been documenting that process for more than fifty years.
The Availability Heuristic Is Doing the Arithmetic
In 1973, Amos Tversky and Daniel Kahneman showed that people judge how common an event is by how easily examples come to mind, a shortcut they called the availability heuristic. It works when memory mirrors reality. It misleads when one kind of evidence is far more vivid than the rest.
Gasoline is the most vivid price in the economy. It is bought every week or two, it is posted in large digits on street corners, and it is paid at the moment of purchase. Gasoline rose 27.8%, and a driver is reminded of it each time the tank empties.
The basket tells a quieter story. Shelter costs rose 1.8% and food bought in stores rose 1.7%, the smallest annual gain since May 2024. Those prices arrive as a mortgage payment or a grocery receipt, with no number on a pole to announce the trend.
The grocery aisle carries its own distortion. In a 2021 paper in the Journal of Political Economy, Francesco D’Acunto, Ulrike Malmendier, Juan Ospina and Michael Weber found that the weight a consumer gives a price change depends on how often the good is bought, not on its share of spending, and that price increases loom larger than decreases. Groceries are still 33.0% more expensive than in September 2020, so a shopper can watch the rate of increase slow and still feel the level.
Experienced Inflation Is Local, and So Is the Memory
Ulrike Malmendier and Stefan Nagel reported in the Quarterly Journal of Economics in 2016 that individuals overweight the inflation they have personally lived through. Using 57 years of the Reuters and University of Michigan consumer survey, they found that differences in lived experience strongly predict differences in expected inflation. What a household remembers shapes what it expects.
Twelve of the 19 cities Statistics Canada reports ran above the national 3.1% in September, with readings from 2.1% in Toronto to 6.1% in Whitehorse that span twice the width of the Bank of Canada one-to-three-per-cent control range.
Twelve of the 19 cities reported ran above the national rate, and Statistics Canada said the fuel oil increase weighed more heavily in Atlantic Canada, where that fuel is more widely used. Statistics Canada cautioned that city figures rest on small samples and can fluctuate widely.
Fuel oil explains part of the pattern. Excluding energy, inflation was highest in Nova Scotia at 3.3%, against a national 2.1%, and Statistics Canada said the fuel oil increase had a greater impact in Atlantic Canada. Halifax at 5.4% and Charlottetown and Summerside at 5.6% are living through a different inflation than Toronto at 2.1%.
City figures rest on small samples, so any single reading deserves caution. The direction is still the point: a household in Halifax and a household in Toronto are not reading the same inflation, yet both are asked about the same national target.
Why the Expectation Matters More Than the Print
A single inflation print is history. An expectation is a forecast that people act on, feeding into wage demands and pricing decisions, which is why the Bank of Canada tracks it closely. Roughly half of the firms in its Business Outlook Survey plan to keep passing some or all cost increases on to customers.
Markets moved on this logic within days. On October 19, overnight swaps priced an 87% chance of a quarter-point Bank of Canada hike on October 28, according to Bloomberg data cited by the Financial Post, up from about 30% a week earlier. Governor Tiff Macklem had told reporters on October 17 that he expects the Bank to raise its inflation forecast, with oil far above the level assumed in July.
The labour market complicates the picture. Canada shed 68,000 jobs in September, the second straight monthly decline, and unemployment rose to 6.5%. Households facing higher fuel costs and a less secure job market are not wrong to feel squeezed. The behavioural question is which signal they weight most when deciding what to do with their savings.
Salience Fades Slowly, and October 28 Is the Next Test
If lived experience is overweighted, as Malmendier and Nagel found, expectations formed during a year of 27.8% gasoline inflation are unlikely to reset the moment the pump price eases. The gap closes one of two ways: household expectations drift down toward the 1.8% to 1.9% core readings, or core inflation drifts up toward what households already expect.
The Bank of Canada announces its rate decision on October 28 alongside a full Monetary Policy Report. Its updated inflation forecast will show how the central bank reads the gap, with the policy rate held at 2.25% since October 2025.