The 91% That Disappeared

Picture two men at the same press.

Same plant. Same line. Same forty-ton press stamping out the same steel bracket it stamped out a generation ago. One of them is working in 1979. The other is working a shift today. Forty-five years apart, same job — except the man working today runs better tooling against a tighter clock, and he turns out more parts in an hour than the 1979 man ever did.

For that, he takes home almost exactly what the 1979 man took home.

Hold that picture, because it is the most important fact in American economics — and there is a graph that proves it.

From 1948 to 1979, two lines on that graph move together: productivity and the pay of a typical worker. When American industry produced more value per hour, American workers were paid more per hour. Not perfectly. Not without a fight. But the lines tracked each other. The rising tide really did lift most of the boats.

After 1979, the lines split. Productivity keeps climbing. Pay goes flat.

By 2024 the gap is staggering. Net productivity has grown 80.9% since 1979. Typical worker compensation has grown 26%. Productivity ran almost three times faster than pay.

That’s not a gap. That’s a heist conducted in broad daylight — documented in government data, published every year by the Bureau of Labor Statistics, and signed for by no one.

That’s the argument. The rest of this essay shows you the size of it, names where the money went, and proves it was taken on purpose.

Let’s put a dollar figure on it.

The Economic Policy Institute calculates that if pay had kept pace with productivity from 1979 to 2024, the typical worker would earn about $9.00 more an hour. For the man at the press, that’s roughly $18,700 a year. Every year. Money the pre-1979 economy would have put in his pocket, and the post-1979 economy put somewhere else.

Multiply $18,700 across the roughly 130 million production and nonsupervisory workers in this country — the bottom four-fifths of the workforce — and the yearly transfer runs into the trillions.

This isn’t theory. These are dollars his hours generated, dollars that in the old economy would have landed in his paycheck, that instead landed somewhere else.

Where?

Two places. Into the pay of the people at the very top — the top 10%, and especially the top 1% and the top tenth of one percent. And into profit — the returns that flow to shareholders and the owners of capital.

Pay piling up at the top. Income shifting from people who work to people who own. Those are the two engines of the inequality we live inside. Neither one is weather. Neither one fell out of the sky. Both were built, decision by decision.

The EPI has a name for it. Not wage stagnation. Wage suppression.

Stagnation says things just stopped moving on their own. Suppression says someone held them down.

And you can name the hands. After 1979, the policy turned, and it turned in one direction. Unemployment was allowed to run high to keep inflation down — which is a polite way of saying a worker’s leverage was cut on purpose. The federal minimum wage was left to rot in real terms. Labor law went slack while employers turned hard against unions, and membership fell from better than one in three workers in the 1950s to one in ten today. Top tax rates were slashed. Trucking, airlines, finance — deregulated, and the leverage moved from the people who worked there to the people who owned them.

None of it was an accident. None of it was fate. It was, in the EPI’s own words, intentional policy.

And it delivered exactly what it was built to deliver: growth that was both slower and far more lopsided.

Here is the detail that closes the case.

The gap didn’t open evenly. It opened in lockstep with how much leverage workers had. When the labor market was tight — 1996 to 2002, and again from 2014 to 2024, when there were more openings than bodies to fill them — pay climbed with productivity, about 1.7% a year. Every other year since 1979, raises for the typical worker ran near zero. For the lowest-paid, pay actually fell.

Read that again. When employers had to compete for the man at the press, the money found its way back to him. When they didn’t — when he was easy to replace — the money went up.

So the productivity–pay gap was never a market outcome. It was a power outcome. And for forty-five years the power was wired to flow one way.

Now let me get to the number in the title, because it’s exactly the kind of number the other side will try to pull apart. I want you holding the clean version when they do.

Start with what cannot be argued. Productivity up 80.9%. Typical pay up 26%. Of every dollar of productivity growth since 1979, the typical worker caught about a third. Two-thirds went past him. That figure — roughly 68% — comes straight from federal data and from the companies’ own output. It is bedrock.

Then there’s the second loss, the quiet one. Before 1979, productivity itself grew about 2.5% a year. After 1979, just 1.4%. The same policies that held his wages down also slowed the whole economy’s engine. So the man at the press got robbed twice: a smaller pie, and a thinner slice of it.

Put both losses together — measured against what the old social contract would actually have handed him — and the typical worker kept something like nine cents of every dollar that arrangement would have delivered. Ninety-one cents disappeared.

I’ll be straight about that 91%: it’s a modeled estimate, not a line item at the Bureau, and I’ll flag it as an estimate every time I use it. But the two-thirds underneath it is solid rock, and nobody serious disputes which direction the rest of it went.

Now connect this back to where the series began.

In the first thirteen essays, we proved that American industry buries the intelligence of its own frontline. Toyota pulls