Honda Spent $11 Billion Figuring Out What the Market Already Knew
Killing the Canadian EV factory isn't a retreat — it's an overdue admission that scale without demand is just an expensive theory.

Photo · InsideEVs - Articles
There's a version of ambition that looks like courage until the numbers come back. Honda's reported decision to cancel its $11 billion electric vehicle factory in Canada — after already putting it on hold for two years — is that version, fully developed.
A Nikkei report says the plant is dead in the water. Not paused. Not restructured. Gone. And the interesting thing isn't the money, though $11 billion is not a rounding error. The interesting thing is what the sequence reveals: hold it, study it, kill it. That's not a pivot. That's a conclusion arrived at slowly, in stages, the way you arrive at conclusions you didn't want to reach.
The Timeline Was Always the Fiction
The EV transition was never going to move at the speed the press releases implied. What's happening now — with Honda, and with others before them — is that automakers are finally saying out loud what supply chain people and dealership networks have been saying in hallways for two years: demand isn't tracking the infrastructure, and infrastructure isn't tracking the mandate, and the mandate was written by people who hadn't stress-tested the math at scale.
Building a factory of that size requires certainty about what you'll be selling in a decade. Honda didn't have that certainty. Nobody does, honestly — but Honda is the one who just admitted it in the most expensive way available.
The two-year hold was the tell. A hold is what you do when you believe the conditions will change but you're not sure when. A cancellation is what you do when you've stopped believing. That shift — from uncertainty to conclusion — is the real story here, and it's one the industry is going to be reckoning with for a while.
Scale Is Not a Strategy
There's a logic that took hold in the early part of this decade: build the capacity, and the buyers will follow. It worked for the original EV pioneers because they were building from zero and any sales were proof of concept. It does not work the same way for legacy automakers trying to retool at industrial scale while simultaneously managing existing combustion lineups, franchise dealer relationships, and consumer bases that are not moving as fast as the announcements suggested they would.
The Canadian factory was a bet that scale itself would create momentum. That if Honda committed hard enough — $11 billion hard — the demand side would feel the gravity and respond. What the cancellation says is that the demand side did not respond on schedule, and holding a half-built thesis for two years while the market sends clear signals is not a viable long-term position.
This isn't a story about Honda failing. It's a story about the gap between what an industry needed to be true and what turned out to actually be true — and what it costs, in dollars and in credibility, to finally close that gap.
The factories that get built next will be smaller, later, and far more conditional. That's probably the right answer. It's just an expensive place to land.
Keep reading cars.

Run It to Zero and See Who Blinks
A writer pushed the 2026 Chevy Bolt until the battery died. What they found says more about trust than range.

Coronado Didn't Simulate Chaos. It Just Let It In.
Loose manhole covers, a 25-car pileup, a drunk fan at the window — NASCAR's new street course in San Diego didn't promise disorder. It delivered it.

The Truck Costs $100K. The Restraint Cost Nothing. Guess Which One They Skipped.
New insurance data puts the most expensive EVs at the top of the worst-behavior charts — and it says something uncomfortable about what performance ownership has become.
From the other desks.

Forty Years, and Peak Performance Went Back to Ask the Mountain
A heritage brand's best argument isn't a campaign. It's the conditions it was built for.

Giannis Is Available. The Cap Isn't.
Boston wants the league's best player. The salary cap wants a word first.

Goats Built a Neural Network. Nobody Called Them Sentient.
A Microsoft researcher ran Age of Empires II livestock through a neural network architecture to make a point the AI industry keeps refusing to hear.