What Happens When the Jobs Don't Come Back

Two days ago, the President of the United States stood in the House chamber and said: "More Americans are working today than at any time in the history of our country. The economy is roaring like never before."

The next day, Jack Dorsey cut 4,000 people from Block (formerly Square) — 40% of the company — because AI made their jobs unnecessary. The stock went up 22%. Dorsey said he was late.

Weeks before that, the Premier of Ontario poured a bottle of Crown Royal on the ground at a press conference. On live television.

"You hurt my people, and I'm going to hurt you," he told the camera, threatening to ban the whisky from provincial shelves because Diageo was closing a bottling plant in Amherstburg. Two hundred jobs.

Ford fought for months. Made headlines. Got a $23 million deal. Called it a victory.

The jobs still left.

This is the political playbook in 2026. A president declaring the economy is roaring while 2025 was the worst year for job creation since 2003. A premier pouring whisky on the ground to save 200 bottling jobs while AI is eliminating 4,000 knowledge-work jobs in a single afternoon, at a single company, to applause from Wall Street.

One is performing confidence. The other is performing outrage. Neither is talking about what's actually happening.

The Operating System

"Jobs" isn't just an economic metric. It's the foundational promise of Western democracy.

Every president. Every prime minister. Every premier. Every campaign, every rally, every State of the Union, every throne speech. At some point — always, without exception — they say the thing.

"I will create jobs."

"I will bring back jobs."

"I will fight for your jobs."

It doesn't matter if it's Trump or Carney, Conservative or Labour, Republican or Democrat. The promise is always the same. Jobs. More of them. Better ones. I will deliver them to you. That's why you should vote for me.

This is the operating system. Not a talking point. The operating system. Every leader is measured by jobs created. Every challenger attacks on jobs lost.

The monthly jobs report moves markets and news cycles in Washington. Employment figures dominate question period in Ottawa and PMQs in London. Unemployment rate is the scoreboard.

And for 80 years, the scoreboard has worked. Not because every leader actually created jobs. Because jobs always came back. After every war. Every recession. Every disruption. The economy shed jobs, and then it grew them back, and the leader in office took credit.

The system runs on one assumption: displacement is temporary. Jobs go down. Jobs come back. The government's role is to create conditions for the comeback.

What happens when they don't come back?

The Day After the Speech

Trump's SOTU was Tuesday night. His approval on the economy is 36%. Voters aren't buying it. Polls show Americans believe the president is out of touch with their daily reality — the cost of food, rent, childcare.

Democrats said as much in their response. Virginia Governor Spanberger called it a disconnect between the White House and kitchen tables.

But here's the thing nobody in either party said Tuesday night.

The problem isn't that the economy is weak. GDP grew. Inflation fell. Corporate earnings are strong. The stock market is near highs.

The problem is that the economy is growing and people are getting worse off at the same time. That those two things can be true simultaneously. And the political system has no language for that.

Trump says the economy is roaring. Democrats say affordability is the issue. Both are operating within the old framework — the economy is either good or bad, and the argument is over which party made it that way.

Neither is addressing the structural shift underneath.

Twenty-four hours after Trump told Congress the economy was roaring, Dorsey proved that a roaring economy is perfectly compatible with eliminating 40% of a company's workforce. Block's revenue is growing. Its profit outlook is improving. And it needs 4,000 fewer humans.

The economy is roaring. For companies. For shareholders. For AI infrastructure. For the people who own capital.

For the people who sell their labor — the people who vote, who show up at rallies, who every politician in every Western democracy is supposedly fighting for — the economy is compressing.

And no one on any stage, in any chamber, in any party, said that on Tuesday night.

The Scoreboard Is Lying

The West measures its economies — and its leaders — with a small set of numbers.

Unemployment rate. Nonfarm payrolls. Jobs created per month. These numbers run the news cycle. They move markets. They make or break midterm narratives in Washington and election calls in Ottawa.

The scoreboard was designed for an economy where displacement is temporary. A factory closes. Workers get rehired somewhere else, or retrained, or absorbed by a growing sector. Unemployment goes up, then comes back down. Cycle complete.

The scoreboard has survived everything. Stagflation. Recessions. The dot-com crash. 2008. COVID. Every time, same pattern: jobs down, jobs back, scoreboard resets, leader takes credit.

But the scoreboard has a blind spot that's about to become a crisis.

It doesn't measure what kind of jobs come back. It doesn't measure whether the person who lost a $150,000 product management role is now "employed" at $65,000 managing AI tool deployments.

It doesn't measure whether the consultant whose rates collapsed 70% is "self-employed" on paper and drowning in practice. It doesn't measure whether the graduate working retail with $80,000 in student debt counts as success.

All of those people are employed. The scoreboard counts them. The leader points at the scoreboard and says: "We've created 200,000 jobs this month."

And the voter looks at their bank account and thinks: you're lying.

This is the gap that's opening across the West. Not between left and right. Between the scoreboard and reality.

Trump says more Americans are working than ever. That may be technically true. It's also meaningless to the product manager who's now making half what she used to, the freelancer whose rates cratered, the 42.5% of recent graduates who are underemployed — the highest since tracking began.

Ford fought to save 200 bottling jobs. That's a real number with real people behind it. But while he was pouring whisky for cameras, AI was making thousands of knowledge-work roles redundant across the country.

No press conference. No one pouring anything. Just roles quietly not being refilled, contractors not being renewed, teams absorbing departures without backfilling.

The scoreboard doesn't capture any of that. And politicians are staring at the scoreboard.

The Toolkit That Doesn't Work

When jobs decline, Western politicians reach for the same tools they've always reached for. They'll reach for them again. None of it will work for this.

Retraining programs.

The default answer in every capital. "We'll retrain displaced workers for the jobs of tomorrow." The track record is terrible. Federal retraining programs in the US have historically placed 0-15% of participants in related jobs.

Canada's record isn't better — the federal government's own evaluations show modest impact at best. And those programs were designed for a world where the "jobs of tomorrow" were identifiable.

Retrain a displaced knowledge worker for what, exactly, when the entire category of knowledge work is compressing?

Tax incentives to create jobs.

The political classic. Give companies breaks to hire. But in an AI displacement, companies don't have a hiring problem. They have a headcount-reduction incentive.

You can give Block every tax break in existence and they're not rehiring those 4,000 people. The work doesn't require them. The market paid 22% to prove it.

Infrastructure spending.

Build roads. Build bridges. Build data centers. This creates real jobs — trades and construction. But it doesn't help the displaced project manager or the laid-off financial analyst.

The professional class being downshifted isn't moving into construction. Not because they're above it. Because the skills don't transfer, the pay is lower, and the entire premise ignores that these people spent a decade building expertise that was supposed to matter.

Protecting specific jobs.

Ford's playbook. Fight publicly for 200 jobs at a bottling plant. Get a deal. Declare victory. The jobs leave anyway.

This is the political equivalent of bailing water out of a boat with a hole in the hull. It's visible. It's dramatic. And it doesn't address the structural problem.

Every tool in the Western policy toolkit was built for cyclical unemployment. For temporary disruption. For a world where the economy sheds jobs and grows them back in a different sector.

This is structural. The jobs aren't moving to a different sector. They're being absorbed by software. And no policy toolkit in Washington, Ottawa, London, or Brussels has an answer for that.

The Promise Gap

At some point — probably 2027 in the US, maybe sooner in countries with weaker safety nets — the political system will feel this. Not because the scoreboard says so. Because voters say so.

Because the gap between "the economy is strong" and "I can't find a job that pays what I used to make" becomes too wide to perform around.

And when that happens, a leader will walk to a podium and make the promise.

"I will bring back the good jobs."

It will sound right. It will get applause.

And it will be a lie.

Not a cynical lie. Most of them will believe it. They'll believe that the right policy mix — some retraining, some tax incentives, some regulation of AI — can restore the economy they grew up in.

The economy where a degree led to a career led to a house and the whole thing worked if you showed up and tried.

That economy is ending. Not because of policy failure. Because a technology emerged that substitutes for human intelligence itself, and the economics of replacing a $150,000 worker with a $100/month subscription are too compelling for any company — in any country — to resist.

"Bring back the good jobs" is the 2030s version of "bring back coal." The leader who says it isn't evil. They're promising something that no longer exists. And the voter who believes it isn't stupid.

They're holding on to the only framework they've been given for how life is supposed to work.

The framework is breaking. And nobody on any stage, in any chamber, in any party, in any country, is willing to say it out loud.

The Anger That Comes Later

If history is any guide, the political reaction won't arrive when the displacement happens.

It'll arrive a decade later.

The China Shock — the wave of manufacturing job losses caused by China's entry into the WTO — hit American communities in the early 2000s. The economic damage was immediate. The political expression of that damage didn't arrive until 2016. Fifteen years.

An entire generation of displaced workers. Rising mortality rates. Collapsing communities. Opioid epidemics. And the political establishment spent fifteen years saying the economy was strong because the scoreboard said so.

Those communities never recovered. The jobs didn't come back. The retraining didn't work. The "new economy" didn't arrive in those zip codes.

The same pattern is starting now. But the displaced class is different.

This isn't factory workers in Ohio. It's product managers in Austin. Analysts in New York. Consultants in Chicago. Developers in San Francisco. Middle managers in Toronto. Financial services professionals in London.

The professional class. The people who vote in every election. Who donate to campaigns. Who write op-eds and call their representatives. Who have degrees and expectations and the vocabulary to express their anger articulately.

When the political anger from AI displacement arrives — and it will — it won't look like the quiet decline of a Rust Belt town. It'll be louder. More organized. More politically sophisticated.

And more dangerous. Because the people experiencing it have the tools and the platforms to channel it.

The question is what they channel it toward.

What I Don't Know

I usually end these essays with a thesis. A position. Something I believe strongly enough to put a date on.

I don't have that here.

I know the jobs aren't coming back. Dorsey said he was late. The market paid him 22% for cutting. Salesforce cut 4,000. Amazon cut 30,000. The pattern is clear and accelerating. No Western leader has a policy tool that reverses structural AI displacement.

But what comes next politically — I don't know. And I don't think anyone does.

Does the system adapt? Does some leader emerge who's honest enough to say "the old economy is over, here's what we build instead"? Does the West find a new social contract — some framework where the promise isn't "I'll give you a job" but something else, something that accounts for a world where human intelligence isn't the scarce input anymore?

Or does the system do what it did with the China Shock — ignore the damage for a decade, let the anger compound, and act surprised when it explodes into politics nobody predicted?

I watched the State of the Union on Tuesday. I watched a president tell a country the economy is roaring. I watched a premier pour whisky on the ground for 200 bottling jobs. And the next morning I watched a CEO fire 4,000 people, make $2 billion, and tell everyone he should have done it sooner.

Three performances. Three versions of a political system that doesn't have the language for what's happening.

The promise that has held Western democracy together for 80 years — "I will create jobs, and those jobs will give you a good life" — depends on jobs existing. Depends on the economy needing human workers at scale. Depends on the conveyor belt running.

The conveyor belt is slowing. And the people at the podium are still pointing at the scoreboard.

What fills the gap — between what leaders promise and what the economy actually delivers — is the most important political question of the next decade.

I don't have the answer.

But I think we need to start asking the question before the anger answers it for us.