Will a Raise Just Bring Robots? (What Data Says)

Every proposed raise meets the same objection: employers will swap workers for machines. The minimum wage automation argument sounds like common sense, and in some cases it is correct. A kiosk replaces a cashier. A scheduling algorithm trims hours. But the argument treats the wage as the only reason a machine shows up. The record is more complicated, and the people making the case deserve to see the evidence on both sides.

What is the minimum wage automation argument?

It says that if labor gets more expensive, employers will replace labor with capital. Economists accept the basic logic. Higher relative labor costs raise the payoff from automating a task, all else equal. Where they disagree is on size and timing.

The CBO gives a useful anchor. Its 2021 analysis of raising the federal minimum wage to $15 by 2025 projected a central estimate of about 1.4 million fewer jobs. The same analysis projected that about 17 million workers would earn more and that about 900,000 people would rise above the poverty line (CBO, 2021). The agency also gave a wide range around its job-loss figure, from near zero to about 2.7 million, which signals how uncertain the estimate is. Both the gain and the loss are real parts of the picture.

Do machines arrive because of wages or because of technology?

Mostly technology. A restaurant kiosk needs hardware that works, software that handles a messy order, and customers who accept it. When those conditions hold, adoption tends to spread whatever the local wage is. Self-checkout lanes appeared in states with low minimums as well as high ones. Wage levels shift the timing and the mix of tasks that get automated, but the underlying capability does the heavy lifting.

The bank teller offers a long-running test. Economist James Bessen has documented that teller employment kept rising for years after ATMs spread, because machines lowered the cost of running a branch, banks opened more branches, and tellers moved toward sales and service. Automation changed the job more than it eliminated it. That pattern does not guarantee the same outcome in every sector, but it breaks the idea that machines always mean fewer workers.

Which jobs are actually at risk?

Routine, narrowly defined tasks. Order-taking, cashiering, and some warehouse picking sit at the high end of exposure. Care work, skilled trades, cleaning, and any job that demands physical improvisation sit at the low end. Many of the lowest-paid jobs fall in that second group. A home health aide or a child care worker cannot be replaced by a screen, and those workers are often among the lowest paid.

Sources: researchers differ on exact exposure rankings; the table summarizes the general consensus in labor economics.

What does the evidence say about the trade-off?

The honest answer is mixed. Studies of local minimum wage increases find job effects ranging from small negative to roughly zero for moderate raises, with larger risks when a wage floor jumps far above local typical pay. We cover that research directly in pros and cons of the minimum wage and the benefits of raising it.

Raises also create offsetting effects. Higher pay cuts turnover. A Center for American Progress analysis from 2012 estimated that replacing a lower-wage worker costs an employer roughly 16% of annual pay, so retention is worth real money. Workers with more income spend more at local businesses, which supports demand.

So should we skip the raise?

That conclusion does not follow. If the fear is that the machines are coming either way, the answer is to pair wage floors with retraining, phase-in schedules, and support for displaced workers, not to freeze the pay of workers who are on the job today. A wage frozen at $7.25 since 2009 has not protected anyone from automation. It has protected the price of labor.

International comparisons help here. Several wealthy countries run higher floors than the U.S. with low unemployment, a point developed in countries with high minimum wages. For the fiscal side of the question, see what low wages cost taxpayers. The broader frame is in the American Dream is broken.

What happened when cities raised wages and machines were available?

Real-world cases give a better guide than theory. Restaurants and retailers in cities with higher local minimums have adopted ordering kiosks, but so have those in low-wage states. Researchers who study local increases usually find that firms adjust through several channels at once: slightly higher prices, reduced hours, slower hiring, lower turnover, and some capital investment. Machines are one tool among many, and employers use whichever one fits their business.

Studies on fast food are instructive. Many find that menu prices rose by a small amount after wage increases, which means customers shared the cost. The size of the price effect is typically modest relative to the wage gain, since labor is only part of a restaurant's costs. That pattern suggests that a raise is often absorbed through multiple small adjustments rather than a single decision to replace workers.

The harder cases are routine back-office and warehouse tasks, where robotics and software improve every year. There, the wage floor affects timing, not direction. Workers in those roles face the same risk whether the wage is $7.25 or $15, which is an argument for investing in retraining and public support. It is not an argument for holding pay down.

One more point deserves attention. Automation changes who has bargaining power. When pay is higher, workers have a stronger claim on the gains from new machines, whether through raises or shorter hours. When pay is stuck at the floor, owners keep the savings. The question is not machines versus workers. It is who shares in the productivity those machines create.

Automation is a decision about who captures productivity gains. When output per worker rises and the wage floor does not, the gain goes to owners. A living wage is a way to share it, and the machines were always going to arrive.