Skip to main content

Many jobs, few employers

Why labor markets with few firms hurt workers

September 3, 2021

Authors

Andrew Goodman-Bacon Senior Research Economist, Institute
Lisa Camner McKay Senior Writer, Institute

Article Highlights

  • When few employers dominate a local labor market, they underpay their workers
  • Measuring degree of wage markdown is necessary to set local minimum wage
  • Concentrated labor markets lead productive firms to be too small
Many jobs, few employers
Illustration of workers trapped in the atrium of a building
Nick Shepherd for Minneapolis Fed

Matt Gies enjoyed his job fixing tractors at Faivre Implement, near his hometown of Berlin, Wis. Then Faivre sold its stores to a competitor who was buying up dealerships to take advantage of volume discounts from the country’s largest tractor maker, John Deere. Gies’ pay plateaued, while his hours increased. He had felt valued at Faivre, but after the sale, “I was just a number, an employee,” Gies told the New York Times.

When Gies looked for a new job as a tractor mechanic, he found that most of the dealerships within commuting distance of his home had been bought up by the same company he’d left. He wanted to stay in his hometown and in his industry, but the market for tractor mechanics in central Wisconsin meant that he couldn’t do both

The reality of many local labor markets is that employment in a specific industry is concentrated at only a few very large employers and many tiny (often single-person) ones.

At a time when employers say they can’t find workers, yet the percentage of the population that is employed is lower than it’s been in more than 20 years, Matt Gies’ story feels emblematic of the complex interactions playing out in U.S. labor markets. Workers feeling trapped and undervalued and employers feeling desperate is a situation that policymakers have linked to labor market concentration— when employment opportunities exist at only a small number of firms. How should policymakers decide what the best remedies are?

STUDY AUTHORS

DAVID BERGER, Associate Professor of Economics, Duke University; KYLE HERKENHOFF, Senior Research Economist, Federal Reserve Bank of Minneapolis, and Assistant Professor of Economics, University of Minnesota; SIMON MONGEY, Assistant Professor of Economics, University of Chicago

The first step in addressing labor market concentration is understanding how and how much it affects workers and the economy. This is the focus of David Berger, Kyle Herkenhoff, and Simon Mongey’s project in the Institute working paper “Labor Market Power.” What matters for workers, they explain, is the degree of market concentration at a local, not national, level: When farm equipment repair shops go to hire workers, they are competing for workers with repair shops across town, not across the country. And the reality of many local labor markets is that employment in a specific industry is concentrated at only a few very large employers and many small (often single- person) ones. This concentration has at least two important impacts on the labor market, the economists explain.

Labor market power and wages

First, powerful firms underpay their workers. To stop working for low pay at a company he didn’t like, Gies had to either move away from home or work in a different industry. Neither was what he wanted. In general, employers that face many competitors must pay workers what they are worth or risk losing them to firms that pay more, but firms only have to compete if their workers have a realistic way to leave. These two concepts— how difficult it is to change jobs and how difficult it is to change markets by moving to a different city—determine how much power firms have to “mark down” wages relative to a worker’s productivity. The number of firms is key here: With only a few options, it’s harder to switch employers, and so workers are less likely to do so.

Employers that face many competitors must pay workers what they are worth or risk losing them to firms that pay more, but employers only have to compete if their workers have a realistic way to leave.

Just how large is the wage markdown? The economists find that, in fact, “most firms in the economy are highly competitive, with narrow markdowns.” However, this doesn’t tell the whole story because most firms in the economy are tiny, employing a small fraction of the total workforce. When the economists weight their calculation by the size of a firm’s payroll, they find that the average markdown is around 25 percent—in other words, workers receive only 75 percent of what they would receive in a perfectly competitive economy.

One tool policymakers have to address wage markdowns is setting a minimum wage. Understanding the degree of local labor market concentration is useful for determining what that minimum should be. For example, the median wage for the Minneapolis/St. Paul metro area is $23.90 per hour. In Michigan’s Upper Peninsula, it’s $17.67. To understand what the minimum wage should be, the economists argue, we need to understand why those are the medians. Is it because prices are lower or workers are less productive in the Upper Peninsula? If so, then a higher minimum wage will hurt employment. Or is it because the labor market there is concentrated among only a few firms, causing large wage markdowns? If so, then a higher minimum wage may raise earnings and employment because firms are now attracting workers who were pushed away when wages were marked down.

Labor market power and productivity

The second major consequence of labor market concentration is that firms do not turn out to be the optimal size. Economists would generally agree that a company that is exceptionally good at what it does should grow. And workers want to work at those productive companies because those workers will be more productive too and, therefore, earn a higher wage. But when a company has labor market power, it marks down wages. As a result, some number of workers will choose to work elsewhere rather than accept this lower wage. The most efficient firms, therefore, are smaller (though more profitable) than they would be if they paid workers what they are worth. This effect, called “misallocation,” means that not enough production in the economy happens at the most productive firms, which shuts out workers from the best jobs.

The implication for policymakers is to look for policies that induce firms to behave competitively even when they are large—capturing the efficiency benefits from labor market concentration without its costs to workers.

The amount of misallocation that occurs is significant, for workers and for the economy. Employment levels at the most productive firms are in the vicinity of 35 to 40 percent lower than they would be in a perfectly competitive economy, Berger, Herkenhoff, and Mongey estimate. This calculation suggests that the economy would actually be more efficient, and workers better off, if the largest and most productive firms were bigger. This presents a catch-22: Efficient firms should be bigger, but bigger firms use their power to mark down wages. The implication for policymakers is to look for policies that induce firms to behave competitively even when they are large—capturing the efficiency benefits from labor market concentration without its costs to workers.

How much labor market power?

As the minimum wage example makes clear, implementing effective policy to redress the consequences of labor market concentration first requires identifying which markets are concentrated. One of the paper’s most helpful insights is its clear justification for measuring labor market power based on the concentration of payroll. Previous research has used the number of firms or the concentration of employment, but those are not the only patterns that matter for labor market power. “You would think that it matters if the largest or smallest firm pays the highest or lowest wage, right?” Herkenhoff said. If every firm pays the same wage, then it doesn’t really matter if some are big and some are small. If, instead, large employers offer not only the bulk of the job opportunities but also the only positions that pay livable wages, he added, “you would want that to be factored into your assessment of concentration.”

Using a measure of how concentrated payroll is in local areas challenges recent ideas about employer power in labor markets. First, the average local labor market is less concentrated by the payroll measure than other measures. The main reason for this is that the most concentrated markets—those with just one firm—are very small. They are usually in rural areas and account for less than one-fifth of 1 percent of wages nationwide. Rather, the concentration in the typical local labor market is equivalent to having nine equal-sized firms. Second, local labor market concentration has fallen since the 1970s, even though large employers have remained the norm at the national level.

These statistics help us understand the where and the when of labor market concentration in the United States, which complements the paper’s analysis of the harm that concentration does to workers through markdowns and misallocation. Ultimately, the combination of rigorous new measures of labor market power and improved estimates of its costs can provide important guidance to policymakers searching for solutions. “The potential gains from inducing firms to behave competitively in local labor markets are very, very large. So if you move the needle even 10 percent of the way, that can make people better off,” Herkenhoff said.


Andrew Goodman-Bacon
Senior Research Economist, Institute
Andrew Goodman-Bacon is a senior research economist with the Opportunity & Inclusive Growth Institute. He holds a Ph.D. in economics from the University of Michigan. Andrew’s research focuses on policy issues related to labor, demography, health, and public economics.
Lisa Camner McKay
Senior Writer, Institute

Lisa Camner McKay is a senior writer with the Opportunity & Inclusive Growth Institute at the Minneapolis Fed. In this role, she creates content for diverse audiences in support of the Institute’s policy and research work.