Our current economic expansion has lasted almost nine years, yet wages have hardly budged, especially for less skilled workers. Inflation-adjusted wages for the average worker have risen only by 3 percent since the 1970s — and have actually declined for the bottom fifth.
For a long time, the conventional wisdom was that wage growth had slowed because of rising competition from low-paid workers in foreign countries (globalization), as well as the replacement of workers with machinery, including robots (automation). But in recent years, economists have discovered another source: the growth of the labor market power of employers — namely, their power to dictate, and hence suppress, wages.
This new wisdom has displaced a longstanding assumption among economists that labor markets are competitive. In a competitive labor market, employers must vie for workers; they try to lure workers from other firms by offering them more generous compensation. As employers bid for workers, wages and benefits rise. An employer gains by hiring a worker whenever the worker’s wage is less than the revenue the worker will generate for the employer; for this reason, the process of competition among employers for workers ought to result in workers receiving a substantial portion of the output they contribute to.
And as the economy grows over time — which has historically been the case in the United States — this dynamic should naturally lead to a steady increase in compensation for workers.
It turns out, however, that labor markets are often uncompetitive: Employers have the power to hold down wages by a host of methods and for numerous reasons. And new academic studies suggest the markets have been growing ever more uncompetitive over time.
The return of the “company town,” in different form
The company town is a familiar historical example of a situation in which employers hold all the cards when it comes to setting wages. In the late 19th century, companies like Pullman, a manufacturer of sleeping cars for trains, established such towns adjacent to their factories, even providing housing and collecting rent. Since such towns had one employer, the workers couldn’t leave for better pay without uprooting their families, which they tended not to want to do.
Few company towns exist today. Still, a variation of the company town effect exists in some regions, at least for certain occupations. A nurse or doctor who lives in a small town or rural area can choose only among a handful of medical institutions within driving distance of his or her home, for example.
And in many areas of rural America, the best jobs are in chicken processing plants, private prisons, agribusinesses, and other large-scale employers that dominate their local economies. Workers can either choose to take the jobs on offer or incur the turmoil of moving elsewhere. Companies can and do take advantage of this leverage.
Yet another source of labor market power are so-called noncompete agreements, which are far more prevalent than many Americans realize. These agreements prohibit workers who leave a job from working for a competitor of their former employer.
Almost a quarter of all workers report that their current employer or a former employer forced them to sign a noncompete clause. (Jimmy John’s, the sandwich franchise, famously asked its “sandwich artists” to sign covenants forbidding them from taking jobs with Jimmy John’s competitors.) Relatedly, Apple and several other high-tech firms were caught entering into collusive “no poach” agreements so they didn’t have to worry about losing engineers to each other, and settled with the Justice Department.
But the practice continues in many sectors of the economy — including fast-food franchises. No-poaching agreements, like noncompete clauses, enhance employers’ labor market power by depriving workers of the threat to quit if wages fall or stagnate.
There are other, more subtle, ways that employers gain labor market power. Different employers offer different working hours, leave policies, and workplace conditions, and workers tend to choose employers whose conditions suit their personal and family situations. If such an employer cuts wages, a worker may be unwilling to move to another employer that asks her to work different hours — or to be on call during “off” hours.
Developing a specific set of skills can be a double-edged sword too, opening doors yet limiting mobility. An expert welder working for the only manufacturer in town may not find it easy to leave that job and find an equally well-paying job (in, say, nursing) because the skill sets are so different.
The “match” problem is exacerbated by the time and energy that job searches demand; it can be hard to hold a job while also seeking a job. This factor, too, gives employers the power to hold down worker wages without fear of losing too many workers.
Unions and regulation once kept employers’ labor market power in check
While employers have taken advantage of labor market power throughout modern economic history, a worldwide social movement at the end of the 19th century moderated the worst excesses. Workers organized labor unions, which enabled them to oppose employers’ market power with the threat to shut down plants. A powerful legal regime was put in place that supported unions and protected workers with health, safety, minimum wage, and maximum-hour regulations.
Such laws, along with union rules, helped standardize work requirements, which made jobs more interchangeable and thereby allowed workers to more easily quit a workplace if the employer abused its power. These reforms helped spur broadly shared wage growth during the 30 years following World War II.
But the good times ended in the 1970s. Globalization, changes in workplace technology, and the rise of a more heterogeneous workforce put strains on unions. A conservative reaction to technocratic liberalism, led by Ronald Reagan and Margaret Thatcher, eroded support for labor and employment law. A wave of mergers produced larger corporations with even greater labor market power.
For a time, economists believed that labor markets were nonetheless competitive. But that conventional wisdom was vaporized by a series of empirical studies that suggest that labor market power is real and significant. A number of studies, summarized here, have found, for example, that when wages fall by 1 percent, only about 2 to 3 percent of workers leave, at most.
If labor markets were really competitive, we might expect the figure to be closer to 9 or 10 percent. Other studies have found that employer concentration has been increasing over time and that this concentration is associated with lower wages across labor markets.
The costs of employer power
It is sometimes mistakenly thought that wage suppression, even as it hurts workers, at least benefits consumers, who pay lower prices for goods and services (since the cost of production is lower for companies). In fact, that’s not the case: Employer market power, sometimes called “monopsony,” harms economic growth and raises prices. (Monopsony is the concept of monopoly, or dominance of a market for a given good, applied to the “buy side” — namely, the inputs that firms purchase, including labor and materials.)
Monopsony harms growth and raises prices because it works much like monopoly: by reducing production. To increase its profits, the monopolist raises prices and thus lowers production (because fewer consumers are willing to pay these inflated prices).
Similarly, to raise its profits, a monopsonist lowers wages below the value of the workers to the employer. Because not all workers are willing to work at these depressed wages, monopsony leads some workers to quit.
Firms bear the loss in workers (and resulting lowered sales) in exchange for the higher profits made off the workers who do not quit. The resulting group of workers looking for jobs are what Marx called the “reserve army of the unemployed.”
Employer labor market power thus reduces employment, raises prices, and depresses the economy. Those sound a lot like the harms that conservative economists have long attributed to excessive taxation. And that’s no coincidence. Wage suppression is just like a tax: a tax on the labor of workers.
But unlike most taxes, the proceeds do not fund public services or redistribution that benefits the vulnerable. Instead, they fund corporate profits and cause the share of income accruing to workers to fall. (That share has fallen almost 10 percent in the US in the past decade). This fall in labor income and rise in profits have fueled the remarkable rise in the incomes of the top 1 percent of earners about which so much has been written.
To make matters worse, because the “monopsony tax” drives workers out of the labor force, it simultaneously reduces tax revenue and increases social welfare payouts to the unemployed and destitute.
This one phenomenon explains many of our economic woes
Thus far, however, all of this discussion has been purely theoretical. How much of the decline in labor’s share, or the fall in employment, is attributable to the rise of monopsony or labor market power?
Answering this question precisely will take years of empirical research. But by combining standard economic models with recent evidence about the prevalence of monopsony power and other crucial economic parameters, we can get a back-of-the-envelope sense of the drag of the monopsony tax. (In a recent working paper, you can find a fuller account of our analysis and assumptions.) The answer, as you will see, is simple: huge.
Our focus is the degree of employer labor market power that prevails throughout the economy. To represent this phenomenon, we use a parameter that ranges from 0 (representing perfect competition in the labor market) to 1 (if there were only a single employer in the whole economy).
This parameter can be roughly thought of as the effective number of employment options a typical worker enjoys. If a worker has a very high number of options (if the number is closer to 0), then she will quit and take another job if her wages decline. If she doesn’t (so the number is closer to 1), then she will stay in her job despite a wage decline — or exit the labor force altogether.
Figures 1 and 2, below, show the results of our analysis. At the left side of the figures, labor market power is zero: Labor markets are competitive, and workers have many options. As you move from left to right, labor market power increases to 1, where pure monopsony prevails and workers have only one reasonable option.
Most work in economics has assumed that employer labor market power is close to zero. But recent empirical work has suggested that, on average, labor market power ranges from 0.1 to 0.6, the shaded area.
You can see that in that range, economic output (the blue solid line) is considerably less than it would be if markets were competitive — from 8.5 percent less to as much as 26 percent less. That’s a huge dead weight on economic output.
The crucial point here is how little the model of employer-employee relations needs to diverge from the assumption of perfect competition in order for there to be massive effects on the economy.
Where did that output go? Economic theory tells us that employers suppress wages by underemploying workers. The blue solid line in Figure 2 shows the extent of that wage suppression:*
In our working paper, we take a first cut at estimating the effects of monopsony on both employment rates and wages. Employment, we calculate, is 5 to 18 percent less than it would be in a competitive market. (Here is Marx’s reserve army of the unemployed.) This effect can explain all of the decline in employment rates among prime-age men observed by labor economists.
The results for wage rates are even more disturbing. Given the way our economy works historically, labor’s share of economic output should be about 74 percent if labor markets were perfectly competitive. Because of employers’ power to drive down wages, labor’s share of economic output falls to somewhere between 51 and 64 percent. This transfer significantly increases income inequality.
To put this into more concrete terms, consider the market for nurses. The median wage for a nurse is about $68,000. Given what we know about the labor market power of medical institutions, the true competitive wage for a nurse would be at least $90,000, possibly as much as $200,000.
However, because most areas have few hospitals, they can suppress nurses’ wages without worrying that nurses will move to a rival hospital. Some nurses will drop out of the labor market entirely, but the hospital still earns a greater profit by shrinking its operation and cutting wages dramatically.
For the labor market as a whole, the median annual compensation is $30,500. If markets were competitive, we estimate that this amount could rise to $41,000, and possibly to as much as $92,000.
If labor market power reduces employment and wages, then it must also reduce government’s revenue from taxes. True, government will obtain more tax revenue from the owners of firms, who benefit from paying lower wages. But because tax rates on labor income are higher than on capital income, and because of the overall loss in output, our model finds that revenue falls as well. Our calculations suggest that revenue declines by 20 to 58 percent as a result of labor market power.
In sum, growing labor market power may well be a significant explanation of the host of maladies that have beset wealthy countries, notably the United States, in the past few decades: declining growth rates, falling labor share of corporate earnings, rising inequality, falling employment of prime-age men, and persistent and growing government fiscal deficits. It’s remarkable how well labor market power alone can simultaneously explain all these trends.
Many conservative economists blame high taxes for these problems. But inordinately high taxes cannot explain these trends, because tax rates have been cut several times during this period. Nor can globalization and automation. Globalization and automation can help explain why inequality has increased but not why economic growth rates have stagnated: On the contrary, globalization and automation should have increased economic growth (by expanding markets and by reducing the cost of production), not reduced it.
The power corporations wield over labor markets is no longer a theoretical curiosity. We think it’s clear it’s a major source of our economic malaise. But what can be done about it?
The law already provides resources, but they’re underused. First, workers can bring antitrust lawsuits against firms that obtain labor market power by merging and colluding. While federal antitrust authorities have historically given little attention to labor market power, that began to change during the Obama administration.
The Obama Justice Department began to crack down on no-poaching agreements, and Trump’s Justice Department has begun criminal investigation of no-poaching agreements. Workers have enjoyed relatively few successes in antitrust actions, but as the economic wisdom grows, they should succeed more often.
Second, workers can organize, relying on union representation to help them counter their employers’ labor market power. Indeed, the recent public teachers strikes in red states can be seen as bargaining tactics against the biggest monopsonist around: a Republican-controlled state government insistent on lowering public sector wages in order to deliver tax cuts. Competition for teacher labor is limited by the few schools in most jurisdictions, as well as credentialing differences across states, suggesting that unions can be a necessary counterweight even in the public sector.
Bringing back unions after decades of decline will take a major shift in both policy and popular opinion. That will likely not take place until the Democrats win power and, if recent history is any guide, not even then.
And, third, we can insist that governments expand and enforce traditional employment law protections — including minimum wage laws. Here, there’s room for optimism. Many local jurisdictions, even deeply conservative ones, have raised the minimum wage in recent years, and several states have passed or are considering laws that restrict noncompetes.
Democrats have tried to place antitrust on the agenda. Last year, they announced a group of proposals in a document titled A Better Deal, which acknowledged the problem of corporate concentration and called for stronger antitrust laws, higher minimum wages, and more labor rights. So far, their proposals have gained little political traction. We suspect part of the problem is that the political groundwork for these proposals has not been established.
Americans are not inclined to blame large corporations for the ills of the economy the way they were back in the late 19th century when anti-monopoly social movements gained considerable support. But as research continues to map out the extent of the problem, we suspect this might change.
*CORRECTION, 4/12: This passage originally misidentified the lines in Figures 1 and 2 that show economic output and wage suppression.
Suresh Naidu is an associate professor of economics and public affairs at Columbia University and a contributor to the CORE project. Eric Posner is a professor at the University of Chicago Law School. Glen Weyl is a principal researcher at Microsoft Research New England, a visiting senior research scholar at Yale’s economics department and law school, and author, with Posner, of the forthcoming book Radical Markets: Uprooting Capitalism and Democracy for a Just Society. Find Weyl on Twitter @glenweyl.
The Big Idea is Vox’s home for smart discussion of the most important issues and ideas in politics, science, and culture — typically by outside contributors. If you have an idea for a piece, pitch us at email@example.com.