The Manufacturing Renaissance Is a Lie. Here’s What’s Actually Being Built.

You’ve seen the headlines. “US Manufacturing Boom!” “America Is Back!” Charts going vertical. Politicians cutting ribbons. It feels like a comeback story straight out of a Hollywood script.

But pull up the actual construction spending data from the St. Louis Fed, and a different picture emerges — one that nobody in Washington wants you to look at too closely.

We’re not rebuilding American manufacturing. We’re building two industries on life support and calling it a renaissance.

Total construction spending on manufacturing in the United States has gone nuclear. We’re talking about a line that used to crawl along the bottom of the chart suddenly rocketing upward like it was shot out of a cannon. In the span of a couple of years, annual spending more than doubled. That’s not a trend — that’s a shock.

But here’s where most analysts stop thinking. They see the line go up and declare victory. They write think pieces about how America is reshoring, decoupling from China, reclaiming its industrial crown. It’s a great story. It’s also mostly fiction.

Let me walk you through what’s actually happening.

The surge isn’t broad-based. It’s not thousands of small factories reopening across the Midwest. It’s a handful of mega-projects — semiconductor fabs and electric vehicle plants — fueled by federal subsidies so massive they distort the entire dataset. CHIPS Act. Inflation Reduction Act. These aren’t market signals. They’re government checks.

When the government is your biggest customer, you don’t have an industry. You have a dependency.

Think about what happened with TSMC’s Arizona fab. The project was announced with tremendous fanfare. America would build advanced chips on American soil. Then reality hit: labor shortages, cost overruns, cultural clashes with Taiwanese engineers, delays. The fab that was supposed to be operational by 2024 got pushed back again and again. This is the flagship project of the “renaissance,” and it can’t even open on time.

Or look at the EV battery plants sprouting across the South. Billions in construction spending. Thousands of jobs promised. But who’s buying the EVs? Demand growth has stuttered. Ford lost billions on its EV division. GM pulled back. These factories are being built for a market that hasn’t fully materialized — and may not, at the pace policymakers assumed.

Here’s the tension that keeps me up at night: we are pouring concrete and steel into facilities that may not have enough customers or enough workers to justify their existence. That’s not industrial policy. That’s a bet — a very expensive bet — that demand will materialize and that we can somehow conjure a skilled manufacturing workforce out of thin air.

You can subsidize a factory. You cannot subsidize the person willing to work inside one.

The labor problem is the silent killer of this entire narrative. The American manufacturing workforce aged out. The skills gap is real. Trade schools were gutted. A whole generation was told that factory work was beneath them, and now we’re surprised that we can’t find people to run CNC machines and manage cleanrooms. Construction spending can hit record highs, but if you can’t staff the building once it’s done, you’ve built a very expensive monument to wishful thinking.

And then there’s the overcapacity risk. If global demand falters — if the EV transition slows, if semiconductor orders soften, if a recession bites — these mega-factories become albatrosses. We’ve seen this movie before. The 2000s housing boom looked unstoppable too. Everyone had a reason why “this time is different.” It wasn’t.

Now, let me be clear about where I stand. I’m not anti-manufacturing. I’m not anti-reshoring. I think producing critical goods domestically is strategically vital. But strategic necessity doesn’t automatically translate to economic sustainability, and pretending otherwise is how bubbles get blown.

The difference between a renaissance and a bubble is whether the thing survives without the subsidy.

What would a real manufacturing comeback look like? It would be broad-based — chemicals, machinery, textiles, food processing, all rising together. It would be driven by market demand, not political deadlines. It would be accompanied by workforce development that started a decade ago, not workforce promises made at press conferences. It would look boring and incremental, because real industrial capacity is built over generations, not election cycles.

Instead, we have a spending chart that looks spectacular and a foundation that looks fragile. We have concentration risk masquerading as diversification. We have two industries — semiconductors and EVs — carrying the entire “America is back” narrative on their shoulders, and both of them are deeply entangled with government money.

So here’s what I’d tell any business leader looking at that Fed data and feeling optimistic: enjoy the chart, but read the footnotes. The construction spending is real. The renaissance is not — not yet. The next five years will determine whether we built durable industrial capacity or a collection of subsidized monuments that looked great in press releases and crumbled under market pressure.

Concrete cures in 28 days. Industrial ecosystems take decades. Don’t confuse the two.