Canadian inflation didn’t get worse—unless you live in the majority of provinces. Statistics Canada (StatCan) data shows inflation stalled in August, with slowing gas prices helping to calm growth. While national growth stalled at the Bank of Canada’s critical upper threshold, that would be considered slow in most of the country. A provincial breakdown shows most provinces are significantly above the national average, especially in Atlantic Canada.
Canadian Headline CPI Stalls, Remains At BoC Upper Tolerance
CPI annual growth rate.
Source: StatCan.
Headline inflation remained flat at 3.0% annual growth reported in August, unchanged from July. Most of the climb in recent months has been attributed to gas prices, and the slight deceleration helped to slow the march beyond 3.0%. However, CPI excluding gasoline’s annual growth accelerated to 2.4% in August from 2.2% a month before. That’s still relatively tame growth, but it indicates the upward pressure on prices is broadening.
The main driver of CPI’s annual growth continues to be gasoline (+22.8%), where growth remains lofty despite the slowdown. It was followed by Travel Tours (+26.1%), and rent (+2.8%), rounding out the top 3. Rent was also the top contributor to the monthly move, rising 0.8% in August alone. In other words, the third-biggest driver of inflation over the past year saw more than a quarter (28.6%) of its growth in a single-month. That doesn’t speak well for a country that’s already struggling to contain rising housing costs.
Keeping inflation in check over the past year has been the cost of owning a home. According to the agency, homeowners’ replacement costs (-1.9%) was the biggest downward pressure, based largely on new home prices. It was followed by household appliances (-5.5%), and other owned accommodation expenses (-1.6%) rounding out the top downward pressures in the past year.
Shelter is the driver of inflationary and deflationary pressures. As a result, its distribution follows the diverging provincial real estate markets.
Atlantic Canada Leads Inflation Growth, Most Provinces Accelerated
Annual inflation growth by province and the national average.
Source: StatCan; Better Dwelling.
Most provinces may be a little skeptical of the national stability, and with good reason—inflation is accelerating. Annual growth accelerated in 6 of 10 provinces despite Canada’s headline average being flat. Ontario (+2.4%) was the only province below the 3.0% national rate, while B.C. (+3.0%) matched it exactly. The other 8 provinces exceeded the national rate, suggesting the national numbers are masking the pain felt across most of the country—especially in Atlantic Canada.
Headline CPI is rising much faster in Atlantic Canada, where CPI is rising at a level that would be seen as a crisis if it were national. Nova Scotia (+5.1%) has seen the highest growth across the country, followed by New Brunswick (+4.6%), and PEI (+4.4%), tied with Manitoba (+4.4%) for the third-biggest move.
To put that in context, purchasing power is cut in half every 35 years at the Bank of Canada’s 2% target rate. At Nova Scotia’s current rate, purchasing power would be slashed in half in just over 14 years.
Circling back to rents suddenly returning to one of the highest drivers of growth is interesting. Atlantic provinces also happened to lead in quarterly rental price growth, representing virtually all of the upward pressure on the Q2 2026 report.
It’s funny double digit home price increases didn’t affect cpi in 2023, but now they help? It’s like this is all a game by the govt to convince us things are ok….
It’s all going to blow up once Carney’s trade war hits. I read a really dumb article that the 100k job losses are ok as oil prices are high? Someone is really straining those rose colored glasses
It’s Trump’s trade war. And literally no one is trying to “convince us things are ok.”
I don’t really understand how someone can even suggest this is a Carney’s trade war. This is Trump’s trade war. This exact mindset is holding back Conservatives in the polls: the party is stuck in the old mental model of blaming Liberals and avoiding held Trump accountable. Trump brought a structural change towards nationalism into the US, so their old relationship with Canada is over, and we are all suffering the consequences. I wish Conservatives would listen more to Stephen Harper, who right now doesn’t fully align with Carney but understands and articulates very well the ongoing structural change, and the unavoidable trade war with the US, despite Harper himself being the more enthusiastic, pro-US Prime Minister ever on his own time. Conservatives would do great for the party and the country if they could elaborate policies for this new world order, instead of living in a past where a Trump-initiated trade war could be averted. That past does not exist anymore. For conservatives, looking ahead starts with listening to policy thinkers like Harper and building a strategic roadmap towards this new world. For Canada, it means looking ahead towards diversifying trade partners and mitigate exposure to a single customer. US is not and won’t be a reliable partner again in the foreseeable future.
So Canada can have tariffs on the US but the minute the US says they are tired of being used now its Trumps tariff war? How come Canada doesn’t impose these same types of tariffs on China?
This was a good summary of the August CPI update from Daniel. The Bank of Canada would take consolation in the core measures that define their operational guide showing inflation on target (CPI-median at 2.0%) or slightly below target (CPI-trim at 1.9%). Lately, CPI-trim has been consistently either showing the same or a slightly lower inflation rate than CPI-median. The Governing Council is probably already planning to make a pitch to the Department of Finance to make CPI-trim the one core measure that rules them all when the 2026 renewal of the inflation-control agreement is published later this fall.
But these are the very lowest of the 11 or 12 core measures published by the Bank of Canada. The highest is CPI-common with an inflation rate of 2.6%, down from 2.7% in July. This is, of course, the cuckoo in the nest of the three core inflation measures that defined the operational guide of the Bank of Canada starting in 2017 following the 2016 renewal of the inflation-control agreement, a decision that was reaffirmed with the 2021 renewal agreement. The Governing Council of the Bank of Canada itself recognized it as the cuckoo in the next, giving very different inflation rates from the other two measures, and having embarrassingly large revisions in its estimates, which is why it was dropped from the operational guide in 2017, almost at the start of the new five-year mandate for which it had been re-affirmed. Just the same it is strange that the Bank of Canada would now ignore this measure when it gives a much different reading from the new operational guide, when it previously put so much faith in it.
The core measure with the next highest inflation rate was CPIW at 2.5%, down from 2.6% in July. It has never been part of the operational guide, but has been published every month since May 2001, at the same time as the CPIX. Its biggest drawback is that unlike most of the core measures, it does not exclude changes in indirect taxes. This is something that should be changed with the renewal of the inflation-control agreement to be announced later this fall, but it’s unlikely that it will be. The Bank of Canada is too busy studying how to improve Canadian productivity, an endeavour that grossly exceeds its authority to make such obvious improvements in its core measures.
The next highest inflation rate for a core measure was the CPIX, which was the operational guide until it was plutoed by the 2016 renewal of the inflation-control agreement. It went from 2.3% in July to 2.4% in August. The CPIX is the only core inflation measure which excludes mortgage interest change, although really they all should. This is another thing that should get fixed with the 2026 renewal but almost certainly won’t.
In a sense, the CPI excluding indirect taxes is also a core inflation measure, and its inflation rate actually exceeds the official CPI rate, going from 3.3% in July to 3.6% in August.
In summary, if one looks at the full range of core inflation measures published by the Bank of Canada, one doesn’t get the warm fuzzy feeling one gets if one looks only at the operational guide.
Air transport was the fourth largest contributor to the upward 12-month movement in the CPI, increasing from 12.1% in July to 15.0% in August. Statcan ignored this component in its commentary, fortunately perhaps, as the rise in the annual rate of change was entirely due to an exit effect, which it would have called a base year effect. There was a 5.1% decline in air transport fares from July 2025 to August 2025. Since there was a 19.0% increase in fares from August 2025 to July 2025, partly due to an aviation fuel surcharge, the exit effect was boosted from 5.1% to 6.8%. The actual increase in the annual inflation rate was less than half of that only because there was an entry effect of -4.0%, reflecting a July 2026 to August 2026 price change of -4.0%.
Like travel tours, air transport is a service category that should really be seasonally weighted. While winter travel to sunspot destinations in winter months is most popular, for many households travel back to Europe to visit family in the summer months is the trip of choice. This was certainly the case in my family when Miroslav was still going to school, and my wife and son and sometimes me, always made a trip to Serbia sometime between late June and September, except during the COVID pandemic. We never travelled in winter when people were going to Miami. There were lots of families like ours. Only a monthly-weighted price index provides a representative measure of international passenger fares for consumers in the Canadian case.
Food purchased in restaurants was the fifth largest upward contributor at 3.1%, up from 2.9% in July, when it was the third largest contributor. Statcan made the point that groceries had an inflation rate of 2.8% in August, the first time they were below their inflation rate was below the overall rate since July 2024. But restaurant meal prices, with a slightly lower inflation rate than the overall index in July, now have a slightly higher one, and are threatening the livelihoods of a lot of people who work in restaurants.
I wish that StatCan published detailed tables of component contributions to percent change for annual and monthly inflation rates like the US BLS. It has them, but it doesn’t publish them, even for higher level subaggregates. The idea of showing tables of upward and downward contributors comes from Destatis, and has to be one of Canada’s more dubious German imports.
The August CPI inflation rate was identical with July’s to the second decimal place (3.03%). It shows you the depths of economic debate in this country that Opposition Leader Pierre Poilievre was denounced as a liar, liar, pants on fire last month for saying that the inflation rate was above the Bank of Canada’s upper bound of 3.0%, even though technically, he was right. Now, whether you think it was above the upper bound, or right on it, for two months in a row, it is not a good look on the government or our central bank.
The main contributor to upward inflation was motor gasoline, with a price increase of 22.8% as opposed to 25.7% in July. It has become de rigueur to note that high gasoline prices are mostly due to the American war on Iran and are external shocks, but this rather exonerates PM Carney from his performance as a pom pom boy in cheering on Trump immediately after the strikes on Iran were announced. He has admittedly, true to his reputation as an unreliable boyfriend, somewhat changed his tone since then.
The second most important contributor to upward inflation was travel tours (StatCan’s label for what the ONS calls package holiday trips), which were the fifth largest contributor in July. According to The Daily:
”Year over year, prices for travel tours rose at a faster pace in August (+26.1%) compared with July (+15.2%), partly due to a base-year effect (sic). Canadian travel to the United States declined sharply in 2025, putting downward pressure on prices for airfares and travel tours. Given that this decline has now stopped affecting the 12-month price movement, along with the introduction of fuel surcharges amid higher prices for jet fuel, upward pressure has been put on the travel tours index.”
The unstated monthly price change for August 2025 that exited the annual inflation rate in August 2026 was -11.3%. The negative of this, 11.3% is larger than the increase of 10.9 percentage points in the annual inflation rate from July to August. There was a huge 29.1% increase in the travel tours index in the 11 months between August 2025 and July 2026, which boosted the exit effect to 14.7%. The Daily release is misleading as it implies that there was a positive entry too, when there was not. The monthly change from July to August 2026 was -2.9%, which was boosted to an entry effect of -3.7%.
As an aside, the travel tours is notoriously one of the volatile components in the CPI. It would be so whatever methodology was chosen, but StatCan made a big mistake when they went from pricing only from January to March to pricing in all months of the year. This happened with the September 2013 update. This would have been the perfect opportunity to switch to monthly expenditure weights for items in the index. Travel tours to sunspot destinations are less popular or non-existent in some cases, while travel tours to Europe are then in high swing. At the same time, the volume of transactions is less then, so they should have less influence on the overall index. This just doesn’t happen in an index with fixed annual expenditure weights.
Rent was the third largest contributor in August, but its inflation rate actually increased from 2.5% in July, when it was the second largest contributor, to 2.8% in August. The labour force survey in Canada, as in the UK, has seen a decline in its response rates during the COVID epidemic, and over-the-top immigration levels in the current decade have also made it difficult to keep response rates high. There is also the more basic problem that the LFS rent module only obtains rent-comparisons for five months of the year from any survey recipient when the average flat only changes rent once a year. Since 2019 StatCan uses a hedonic model to calculate the rent indices, but it is being asked to do more than a hedonic model can really be expected to do. There is no question that rent increases have been highly inflationary in recent years. Whether they will now on a consistent basis be lower than the overall CPI inflation rate is open to question, and if they are, can the numbers be believed?