What Milton Friedman actually argued about the minimum wage

Milton Friedman spent decades arguing that the minimum wage is one of the most misunderstood laws on the books. He did not argue that wages should be low. He did not argue that low-income workers deserved less. He argued something far more specific: that the minimum wage is a price control, and that like all price controls, its real costs fall on the very people the law is meant to help.

That argument made him unpopular — and made a particular kind of policy opponent easy to caricature. But Friedman's actual case is more careful than its reputation suggests. It deserves to be taken on its own terms, not as a slogan but as a theory with testable implications. This piece walks through what he said, how the argument fits together, and where the modern empirical literature has pushed back.

The claim, in his own terms

Friedman's sharpest formulation, repeated in lectures and interviews from the 1960s through the 2000s, was blunt:

"The minimum wage law is most properly described as a law saying employers must discriminate against people who have low skills. That's what the law says." — Milton Friedman, widely attributed

The language is deliberately provocative. But the claim underneath it is narrow and clear. Friedman's argument was that when the government sets a legal wage floor of, say, $15 an hour, it is simultaneously outlawing every employment arrangement below $15 — including the ones a low-skill worker might have voluntarily accepted in exchange for experience, training, or a foot in the door.

He was not claiming that a $15 job was bad. He was claiming that a $15 legal minimum eliminates every $10, $12, and $14 job that used to exist — and that the workers who held those jobs do not all get promoted to $15. Some get unemployed. Some get fewer hours. Some get replaced by automation or by slightly more experienced colleagues. The wage floor raises the pay of the workers who keep their jobs, and silently removes the jobs of the ones who don't.

The logic behind it

The economic theory is straightforward. Employers hire workers when the additional value a worker produces exceeds the cost of employing them. If a worker produces $12 of value per hour and the law requires $15 in wages, the employer loses $3 for every hour that worker is on payroll. Over time, that role stops existing — either through layoffs, reduced hours, hiring freezes, or substitution toward machines and software.

Friedman was careful to point out that the effect is not uniform. A wage floor of $8 in a high-wage city might affect almost no one, because market wages are already well above $8. But the same $8 floor in a low-wage town, or a $15 floor applied nationwide, would bind much more tightly. The damage concentrates on:

• Teenagers and young workers who are looking for their first job and have low observable productivity.
• Workers without credentials — no high-school diploma, no references, no prior employment — who need a first chance to prove themselves.
• Workers in low-cost regions where a national or state wage floor can exceed local productivity.
• Workers in low-margin industries — restaurants, retail, cleaning, home care — where small cost increases force quick adjustments.

Critics have long pointed out that the theory assumes a very clean model of the labor market — competitive, frictionless, with productivity that employers can measure precisely. Friedman was aware of those caveats. He argued that the general direction of the effect was robust, even if the exact size depended on local conditions.

Friedman's alternative

Friedman was not opposed to helping low-income workers. He argued that wage floors were simply the wrong tool. In his 1962 book Capitalism and Freedom, he proposed what he called the negative income tax: a direct cash transfer that phased in as earnings rose, topping up low incomes without interfering with the labor market.

That proposal — radical at the time — is the direct ancestor of today's Earned Income Tax Credit (EITC). Enacted in 1975 and expanded repeatedly under presidents of both parties, the EITC rewards work, scales with family size, and doesn't price anyone out of a job. It is, in its structure, a Friedman policy — and one of the most effective anti-poverty programs the United States has ever run.

The point is worth underlining. Friedman's opposition to the minimum wage was not an opposition to anti-poverty policy. It was an argument for replacing a blunt instrument with a targeted one. The structural difference between the two tools is covered in a separate article.

"The good things of life never have to be imposed on people by force." — Milton Friedman

The modern empirical debate

For decades, the economics profession largely agreed with Friedman's directional claim: moderate minimum wage increases produce small but real employment losses, concentrated among younger and lower-skill workers. Then, in 1994, David Card and Alan Krueger published a study comparing fast-food employment in New Jersey (which had just raised its minimum wage) to Pennsylvania (which had not). They found no employment decline. In some specifications, employment actually rose.

The Card-Krueger paper set off a thirty-year argument that is not over. A partial summary of where things stand:

• Some studies replicate the null result. Moderate wage floors, adopted gradually, in regions where they bind for a modest share of workers, often produce smaller employment effects than the basic theory predicts.
• Other studies find real effects. When wage floors are set high relative to local median wages (the "bite" ratio), researchers like David Neumark and William Wascher find meaningful disemployment effects, especially for teenagers and less-educated workers.
• Monopsony power matters. In labor markets where a few employers dominate, a moderate wage floor can raise wages without reducing employment — because market wages were already below the competitive level. This is a genuine theoretical concession from the competitive benchmark.
• Margins of adjustment are diverse. Even when employment headcounts don't fall, firms can respond by cutting hours, reducing benefits, raising prices, slowing hiring, or accelerating automation. The employment variable alone doesn't capture the whole story.

This is where the original argument holds up best. Not every prediction Friedman made in the 1960s survived empirical scrutiny unchanged — but the core insight that wage floors redistribute from the unemployed to the employed, and that their cost is hardest on the people at the bottom of the skill ladder, remains intellectually durable.

What to take from him today

Three things are worth carrying forward from Friedman's argument.

First, the language matters. Calling the minimum wage a "price control" is not rhetoric. It is literally what it is. Framing it that way forces the same questions we'd ask of any price control: who benefits, who loses, and where does the cost hide?

Second, the population of interest is the unemployed, not the employed. Minimum wage debates almost always center on workers who already have jobs. But the Friedman critique is about the workers who don't — the ones whose first jobs were quietly removed from the menu. That group is harder to see and harder to measure, which is exactly why the law's costs are politically invisible.

Third, the alternative exists and works. The EITC is not a theoretical proposal. It reaches roughly 25 million American households, lifts millions out of poverty, and does so without eliminating the first rung of the job ladder. If the goal is to raise incomes for low-income workers, Friedman's preferred tool is available, bipartisan, and effective. The case for it is here.

Friedman's minimum-wage argument is not an attack on low-wage workers. It is, at its core, a plea to use better policy to help them — and to stop mistaking a popular law for an effective one.