Gas prices cooled in August. Inflation held at 3% anyway.

Canada's August inflation number landed on September 14 and most of the coverage said the same thing: 3.0%, unchanged from July, nothing to see.

The headline didn't move. What's underneath it moved a lot, and it moved in the exact direction the Bank of Canada had warned about twelve days earlier.

Gasoline cooled. The number didn't.

Three figures from one Statistics Canada release.

Gasoline was up 22.8% year over year in August, down from 25.7% in July. Cooling.

Inflation excluding gasoline was 2.4% in August, up from 2.2% in July. Warming.

All-items inflation was 3.0% in both months.

So the wedge between the headline and the ex-gasoline number shrank from 0.8 points to 0.6 points. You want that wedge to shrink, and you want it to shrink because gasoline came down to meet the rest of the basket. What happened is the rest of the basket came up to meet gasoline.

Same 3.0%, different reason for it. Nobody said that on the day.

What gasoline is actually worth

You can back that weight out of the published numbers.

If the headline is a blend of gasoline and everything else, then 3.0 = w x 22.8 + (1 - w) x 2.4, where w is gasoline's share of what a household spends. Solve it and you get 0.6 = w x 20.4, so w is about 2.9%.

Run the same thing on July: 0.8 = w x 23.5, so w is about 3.4%.

The two don't land on the same number, because Statistics Canada reweights the basket and the blend isn't perfectly linear. But they bracket it. Call it roughly three cents out of every dollar a household spends.

Three cents out of every dollar is producing 0.6 of the 3.0 points. A fifth of the national inflation rate, out of 3% of the basket.

The Bank told us what to watch for, and then this happened

On September 2 the Bank of Canada held the policy rate at 2.25% and noted that excluding gasoline, inflation had been 2.2% in July. It then flagged the thing it was actually worried about: the longer high oil prices and fat refinery margins stick around, the more likely they spill into the price of everything else.

August is the first reading since that sentence was written. Ex-gasoline inflation went up 0.2 points.

Before anyone panics, a 0.2 point move in one month is not a spiral, and some of it has an innocent explanation. Travel tour prices jumped from 15.2% to 26.1%, which Statistics Canada puts mostly down to a base-year effect: Canadians stopped travelling to the United States in 2025 and that collapse has now dropped out of the twelve-month window. Calendar arithmetic. Core is still behaving too, at 2.1% excluding food and energy and 2.3% excluding energy.

I made the case three weeks ago that the 3% headline is a gasoline number and that Ontario housing is being priced by an oil shock it has nothing to do with. Still true. Just 0.2 points less true than it was in July, and the direction of travel matters more than the size of any one month.

Shelter is the thing being disinflated, not the thing causing it

Shelter inflation was 1.5% in August. Below the headline, below ex-gasoline, below every core measure Statistics Canada publishes.

Here's the version that tells you something. Divide 1.015 by 1.030 and you get 0.985. Relative to everything else Canadians buy, shelter got about 1.5% cheaper over the past year.

Check it on the index levels if you don't trust the shortcut. Shelter sat at 188.0 last August against an all-items index of 164.8, a ratio of 1.141. This August it's 190.9 against 169.8, a ratio of 1.124. Down 1.4%. Same answer through a different door.

Now transportation, which came in at 7.5%. Divide 1.075 by 1.030 and it got 4.4% more expensive in real terms. Index check: 1.048 last year, 1.094 now, up 4.4%.

Shelter down 1.5% in real terms, transportation up 4.4%. A 5.8 point swing in one year between the two biggest line items in most household budgets, pointing opposite ways.

Why this one is an Ottawa problem

Ottawa's price map is a commute gradient. Most of the price difference between the core and the edge is somebody paying less per square foot and paying for it in windshield time. Kanata, Barrhaven, Orleans, the rural belt east and west. That's the whole trade, and it just got repriced, not in the direction the listing photos suggest. The housing side of the ledger got about 1.5% cheaper in real terms. The driving side got 4.4% dearer, and gasoline on its own is up 19% in real terms once you strip out general inflation.

If you're weighing a house forty minutes out against something twenty minutes in, run the fuel number on your actual annual kilometres at today's pump price, not at what you paid in 2025. On two vehicles and a real commute, that is no longer a rounding error against the price gap. When you start comparing listings, put the drive in the spreadsheet.

I sell houses, not gasoline futures. Where oil goes from here depends on a Middle East conflict nobody can forecast, and I won't pretend otherwise.

What this sets up for October 28

The policy rate was 2.25% in January and it is 2.25% now. The next decision is October 28, with a full Monetary Policy Report attached. Governor Macklem speaks in Halifax today.

The September inflation number lands on October 19, nine days before that decision. That is the print that decides it.

The case for a cut: Ontario printed 2.4% in August, the lowest reading of any province in the country, while the spread ran all the way up to 5.1% in Nova Scotia. Ontario rent inflation was also 2.4%. Core is near target. Shelter is at 1.5%.

The case against: the headline has been stuck at 3% for months, the Bank has said outright that its upside risks have grown, new US tariffs and Canadian counter-measures are now in the mix, and August showed the non-gasoline basket firming rather than fading.

On the record so you can score me later. If ex-gasoline inflation prints 2.5% or higher on October 19, the Bank does not cut on October 28, whatever shelter does.

What I'd do with this

Selling this fall, stop treating a rate cut as the thing that rescues your price. It probably isn't arriving before you close, and the local demand gap got worse in August regardless. Price to the market in front of you.

Buying, the same logic cuts your way. Waiting on October 28 is a coin flip that currently looks weighted against you, and you will be bidding against everyone else who waited.

Renewing, the fixed and variable spread I went through earlier this month has not gotten friendlier. Take it to a mortgage broker. I'm not one.

Want this run against your actual numbers, your actual commute and your actual renewal date instead of the national average? Fifteen minutes on your situation beats another inflation summary in your inbox. Call me at 613-262-6545 or email fil@613realtor.ca.