Labor’s Share of Revenue

Technology has never handed its gains to the people who use it. It’s forced to, or it doesn't.

A hundred circles, one company's revenue. In the industrial economy, roughly eighty of them lit up as wages, money that left the company and landed in a paycheck, a mortgage, a kid's braces. Run it forward to the platform economy and the light drains out. Big Tech pays under one percent of its revenue to the people who work there. One circle in this illustration and maybe not even that.

01Labor’s Share · U.S. Revenue

Where the Money Stopped Going

80% of revenue paid out as wages

20% kept as profit and rent.

Industrial era: most of the revenue went out as wages.

Two cited points: the industrial era (about 80%) and today’s platform era (under 1%). The path between them is illustrative — it shows the proportion changing, not a year-by-year measurement.

Source: Yanis Varoufakis, Technofeudalism: What Killed Capitalism (2024), reported in Tamim Mobayed, Psychology Today · wages as a share of company income: GE, ExxonMobil & GM (~80%) vs. Big Tech (under 1%).

Labor’s Share

Labor's share of national income (the slice of everything the country produces that actually reaches workers) has been on a designed slide for decades. The shrinking share percentage doesn’t show that the missing money didn't evaporate, rather it changed owner. What used to be payroll became rent.

The 80-to-1 figure comes from Yanis Varoufakis, and it's a provocation, not a clean measurement. You can poke holes in the comparison. But the direction it points is real, boring, and well documented.

When you sell on Amazon, or build an app for the App Store, or pour your day into the feed, you aren't trading in a market. You're working on someone's land and paying for the privilege. The platform owns the road, sets the toll, reads the traffic, and takes a cut of everything that moves. The squares that went dark didn't go nowhere. They went up, to whoever owns the thing in the middle.

When a genuinely new technology arrives, it opens a gap between what gets produced and what ordinary people get to keep. During the Industrial Revolution, British output per head climbed almost fifty percent across sixty years while real wages barely moved. Economists call it Engels' Pause. The electric dynamo sat in factories for forty years before it showed up in the productivity numbers, because the gains went to the people who owned the rebuild, not the people running the machines.

The gap closed eventually, both times. It never closed on its own. It closed because people organized, because public schooling spread, because governments wrote rules that forced the gains to circulate. The technology didn't distribute anything. Institutions did. That's the part the optimists skip when they tell you to be patient and wait for the tide.

Here's what worries me. Every time the gap closed before, the thing that closed it was a force focused on the technology. It’s been a union, a school, a regulator, a court. Something you could turn against the owners. That's getting harder. The companies that own the technology increasingly own the rules too. They shape the regulation, sit inside the public systems, and set the terms they'd theoretically be corrected by. When the lock and the locksmith are starting to look like the same people, the non-owners need new ways to exhibit influence on the owners.

Throughout a career of designing and watching people adopt new systems, the thing that decides who comes out ahead is only partially about who owns the technology.

Real influence is based on who can actually use the tech. Leverage can come from owning the cloud, most of us never will. The non-owners’ leverage comes from being fluent enough with the technology to bargain with it instead of being processed by it. That fluency lives at the interface, that is, the place where you meet the system and either gain a little power or hand a little away. Designing that interface so it returns power to the human on the other end is the whole game. It's also, increasingly, not what the owners are paid to build.

Drag the slider all the way to now. The grid goes nearly black. A single circle remains.

That circle is you, the unpaid work you do every time you feed the thing, train it, teach it your habits for free. The question this illustration leaves open is whether we design our way back to some leverage, or finish the job of designing ourselves out of the loop.

I don't think that's decided yet. I think it gets decided in the next few years, in the rooms where the interfaces and the rules get built. Much more on who's in those rooms soon. In the meantime, see more about the presentation I gave at the 2025 American Psychological Association conference “Mapping the AI Landscape - How Psychological Expertise Can Guide Humanity-centered AI”.

development and use

Resources