Your weekend is a union victory. Productivity doesn’t pay wages — worker leverage does. The Editors of National Review skip that distinction in their Labor Day essay, “We Should Call It ‘Economic Growth Day’”, arguing the eight-hour day, the weekend, the minimum wage, the end of child labor, and the safer factory floor were not really union achievements. Productivity did the work, the editors claim, and unions showed up afterward to ratify what growth had already granted.

It’s a tidy story. It leaves out the only thing that actually pushes paychecks up.

Concede the first half, because it’s true: productivity is real, and the editors aren’t wrong to point at it. The Industrial Revolution did unleash an extraordinary expansion of what human labor could produce. Henry Ford did cut hours and raise wages at Highland Park. The rising cake is real. The trouble starts when they treat the rising cake as the same thing as a fair slice. Productivity is what the pie looks like before it’s cut. It tells you nothing about who gets the slice.

The nineteenth century runs the experiment for us. Productivity in Britain roughly doubled between 1800 and 1850. Real wages for ordinary workers grew far more slowly — and by some measures stayed flat for most of that stretch. Child labor persisted for decades after the productivity gains had arrived. The factories that made fortunes for their owners ran on exhausted twelve-year-olds. The 1833 Factory Act — passed after decades of agitation by the Ten Hours Movement, the Chartists, and the early trade unions — did what productivity alone had conspicuously failed to do. Productivity did not free those children. Politics, organized and patient, did.

The Ford story is the one the editors lean on hardest, and it works the same way once you actually look at it. Ford’s Highland Park plant had runaway labor turnover in the years before the $5 day — workers walked off the job by the thousands, year after year. The UAW was still a decade away, but the unrest was real, and Ford’s response was as much about staffing a line he could not keep as it was about enlightened management. He shared the gains because he had to. A workforce with no leverage is a workforce you cannot retain. And that “non-union” Highland Park required an armed Service Department that would later beat UAW organizers bloody at River Rouge in 1937.

Look at the American wage record. U.S. manufacturing output per worker roughly doubled between 1890 and 1930. Real wages for production workers barely moved in that period. The great wage gains for ordinary workers came after 1938, when union density climbed and the Fair Labor Standards Act passed. From 1945 to 1980, productivity and wages tracked each other almost exactly. After 1980, when union density collapsed from roughly a third of the workforce to around 10%, productivity kept rising while wages stalled. The labor share of national income fell from about 65 percent to about 57 percent. These are not three independent facts. They are the same fact, viewed from one end and the other.

On workplace safety, the editors say it’s “impossible to tell” when federal workplace safety rules were enacted. That’s misleading. Federal rules didn’t exist until 1970, and state rules were uneven for decades. What’s true is that fatality rates had been declining since the 1920s — but the early declines tracked industries where unions had won safety committees and grievance rights. The Triangle Shirtwaist fire killed 146 workers in 1911 partly because the garment shop was non-union. Within a decade, more than two dozen states had passed safety and compensation laws — most pushed by the same unions the editors dismiss. OSHA didn’t start the trend; it gave the trend federal teeth.

The phrase the editors reach for is “fair market wage.” It does more work than they acknowledge. A worker who can’t afford to lose this week’s paycheck, who has no savings, who lives in a town with one employer, who has no union and no realistic alternative — that worker is not bargaining in a free market. That worker is bargaining in a captive one. The fair market wage in a captive market is whatever the employer offers. That is what organized labor exists to change — not to defy markets, but to make the market for labor competitive again.

The country that has done this best in the last century is not the United States. It is Denmark, where workers can be fired easily, the safety net is generous, and wages are set sector by sector so that one company cannot undercut another by going non-union. Germany puts workers on the supervisory board of every large stock corporation, and German industry did not, in fact, collapse into the sea.

Productivity was the stage. Worker power was the play.

The working alternative is sitting right there: sectoral bargaining — the kind Germany, Denmark, and Sweden have run for ninety years, where wages and conditions are negotiated for whole industries at once and no single firm can gain by going non-union. The PRO Act, long stalled in Congress, would give American workers stronger tools to form unions and bargain industry by industry, even in the face of employer opposition. The state-level right-to-work laws passed over the last eighty years would have to be repealed alongside it, and the slow institution-building the Nordics took a century to finish would still take a generation. None of that is impossible. The economy is a set of choices, not the weather.

So this Labor Day, thank the productivity you’d like to. It doesn’t need thanks. It needs a counterparty.