U.S. DEPARTMENT OF THE TREASURY
THE STATE OF
LABOR MARKET
COMPETITION
March 7, 2022
U.S. DEPARTMENT OF THE TREASURY
THE STATE OF
LABOR MARKET
COMPETITION
b
EXECUTIVE SUMMARY
On July 9, 2021, President Biden signed an historic executive order on Promoting Competition in the American
Economy. That order underscored the importance of competition in the labor market, stating that “a competitive
marketplace creates more high-quality jobs and the economic freedom to switch jobs or negotiate a higher wage.”
The order tasked the Treasury Department, in consultation with the Department of Justice, the Department of
Labor, and the Federal Trade Commission to investigate the eects of a lack of labor market competition on the
United States labor market.
The purpose of the report is to summarize the prevalence and impact of uncompetitive firm behavior in labor
markets. In particular, the report catalogues the ways in which insuicient labor market competition hurts workers,
documents the proliferation of barriers to job mobility, and illustrates how a lack of labor market competition
can hold back the broader macroeconomy, while also providing an assessment of the degree to which lack of
competition lowers wages. This analysis is followed by a description of Biden Administration actions to
improve competition, including a commitment by the Department of Justice and Federal Trade Commission
to vigorously enforce antitrust laws in labor markets.
In discussing the market characteristics that enable monopsony power, this report describes how monopsony
power emerges when a single firm can restrain its hiring to lower wages and boost profits. While most labor
markets do not literally feature a single employer, a market with a small set of employers may mimic a monopsony
by each engaging in practices that give them market power over workers. Concentration in particular industries
and locations can lead to workers receiving less pay, fewer benefits, and worse conditions than what they would
under conditions of greater competition.
There is also increasing recognition that market power may be inherent in the firm-worker relationship. Much of
the theory of labor markets and wage setting is premised on the idea that individual workers and firms search
for one another, seek and find matches that maximize productivity and wages, and bargain over employment
terms. Workers oen find themselves at an informational disadvantage relative to firms, not knowing what other,
similarly placed workers earn, the competitive wages for their labor, or the existence of workplace problems like
discriminatory conduct or unsafe working conditions. Workers also may have a limited or no ability to switch
locations and occupations quickly and may lack the financial resources to support themselves while they search
for jobs that pay more and better match their skills and abilities. These conditions can enable firms to exert market
power, and consequently oer lower wages and worse working conditions, even in labor markets that are not
highly concentrated.
The report details the range of practices that firms use to restrain competition for workers, most clearly to lower
wages and benefits, but also potentially to negatively impact job characteristics beyond just compensation.Firms
can engage in tacit collusion by sharing wage information for dierent occupations, conspiring to fix wages, adopting
no-poach agreements where firms agree not to hire other firms’ workers, or forcing workers to sign non-compete
agreements that limit their ability to switch jobs. Non-disclosure agreements can be so broad as to eectively operate
as non-compete agreements. Mandatory arbitration agreements prevent workers from legal recourse to rectify
violations of labor laws, antitrust laws, or employment terms. Lack of pay transparency, from firms’ use of salary
history, pay secrecy, and punitive practices against workers sharing pay information, also restrains competition.
A growing literature in economics seeks to measure the labor market power exerted by firms over workers. As
David Card, the most recent recipient of the Nobel Prize in Economics, stated in his presidential address to the
American Economic Association, “I will try to make the case that the time has come to recognize that many—or
even most—firms have some wage-setting power.”
Measuring the extent of labor market power can be challenging, as it requires extensive insight into the
THE STATE OF LABOR MARKET COMPETITION
i
THE STATE OF LABOR MARKET COMPETITION
relationship between firms and workers that goes beyond standard measures collected. As this report highlights,
a careful review of credible academic studies places the decrease in wages at roughly 20 percent relative
to the level in a fully competitive market. In some industries and occupations, like manufacturing, estimates of
wage losses are even higher.
Wage-setting power is also evident in the large number of workers who are subject to rules and agreements that
limit their ability to switch jobs and occupations and, hence, their bargaining power. For example, a recent paper
estimates that one-in-five workers is currently subject to non-compete agreements and double that number report
having been bound by a non-compete agreement in the past. As the report discusses, many workers are also
subject to excessive occupational licensing requirements that impede their ability to switch jobs across states or
their ability to enter a new occupation.
The report also highlights the ways in which employers alter the structure of their own work relationships to lower
their labor costs and undercut competition at the expense of workers. The labor market has become “fissured,”
a wide variety of roles ranging from cafeteria workers and janitors to lawyers that were once “in-house” are now
contracted out. This domestic outsourcing is estimated to reduce wages from 4 percent to 24 percent in some
industries and occupations. Moreover, when firms misclassify workers, they oload labor costs and risks onto
workers—for example, by avoiding unemployment insurance taxes and workers’ compensation premiums—and
make it diicult for workers to organize or join a union and bargain collectively for better wages and conditions.
The decline in union density rates further weakens workers’ bargaining power, leaving them with less ability to
counterbalance firms’ wage setting power.
The impacts of insuicient labor market competition oen fall hardest on women and workers of color, who
make up a larger share of workers in lower-paid occupations. These workers oen have diminished bargaining
power because they lack the resources to easily switch jobs or occupations, to reject or negotiate against signing
restrictive employment agreements, or to seek legal recourse for violations of labor and employment law.
The report also highlights the ways in which a lack of labor market competition can impact the broader economy.
Lack of labor market competition contributes to high levels of income inequality, diminishes incentives for firms
to invest, inhibits the creation and expansion of new firms, and reduces productivity growth through lower
reallocation of labor across firms and industries.
The Biden Administration is committed to promoting robust competition in labor markets and has directed
a government-wide eort to support labor market competition. The Department of Justice and Federal
Trade Commission are committing to the vigorous enforcement of antitrust laws in labor markets, to combat
anticompetitive agreements, conduct, or mergers. The Administration has called on Congress to raise the
minimum wage and support increased worker power through increased organizing and collective bargaining
facilitated by the Protecting the Right to Organize Act and other legislation.
The President’s Task Force on Worker Organizing and Empowerment recommended 70 actions that executive
branch agencies and departments will implement to facilitate greater union organizing and collective bargaining.
As part of his Executive Order on competition, the President encouraged the Federal Trade Commission to consider
banning or limiting the use of non-compete agreements. The President’s Executive Order increasing the minimum
wage for federal employees and contractors raised wages for more than 300,000 private-sector employees and
70,000 federal employees.
Finally, in addition to education, compliance assistance, and enforcement of workplace laws, the Department of
Labor’s administrative actions include addressing worker misclassification, supporting worker organizing, and
working to improve job quality, including access to jobs with higher wages and better working conditions.
ii
THE STATE OF LABOR MARKET COMPETITION
Table of Contents
Introduction.................................................................................. 1
Theories of Labor Market Power ..................................................... 3
Pure Monopsony ......................................................................................................4
Monopsonistic Competition.....................................................................................5
Search and Matching Models of Labor Markets.......................................................5
Racial Inequality under Search and Matching Frictions ................................................ 7
Platforms/Regulatory Arbitrage...............................................................................8
Fissuring of the Workplace ......................................................................................10
Fissuring Considerations .................................................................................................. 10
Misclassication of Workers ............................................................................................ 11
Restrictive Employment Agreements and No-Poach Agreements........................... 13
Heterogeneity in Enforcement and Legality of Restrictive Agreements....................... 14
Theory ................................................................................................................................ 15
Mandatory Pre-Dispute Arbitration and Class Action Waivers ...............................18
Occupational Licensing............................................................................................19
Skill-Biased Technical Change and Job Polarization............................................... 21
Labor Market Power & Competition: Empirical Evidence .................. 23
Does Labor Market Power Suppress Wages, in Practice?......................................... 23
The Extent of Wage Losses due to Labor Market Power..........................................24
Changes in Labor Market Power and Concentration over Time...............................25
Alternative Perspectives on Market Concentration and Labor Market Power .........27
Restrictive Employment Agreements.......................................................................28
Trends in and Effects of Occupational Licensing ....................................................29
Variation in Licensing ....................................................................................................... 31
Wage Transparency.................................................................................................. 34
Decline in Department of Labor’s Labor Market Enforcement Actions...................36
Divergence Between Labor Compensation and Productivity ...................................37
Decline in Labor Share.............................................................................................39
Industry Examples ...................................................................................................41
Hospitals and Nurses ........................................................................................................ 41
Agriculture ......................................................................................................................... 42
Minor League Baseball ...................................................................................................... 45
iii
THE STATE OF LABOR MARKET COMPETITION
Implications Beyond the Labor Market............................................ 47
Rising Inequality, Low Interest Rate, and Aggregate Demand................................47
Impacts on Women and Workers of Color ...............................................................48
Declining Business Investment and Productivity Growth .......................................49
Firm Formation and Innovation ..............................................................................50
Declining Worker Mobility and Productivity Growth via Reallocation....................51
Biden Administration Proposals to Promote
Labor Market Competition.............................................................. 52
Proposed Legislation................................................................................................52
Antitrust Enforcement .............................................................................................53
DOJ Criminal Enforcement in Labor Markets.................................................................. 54
DOJ and FTC Civil Enforcement and Competition Advocacy....................................55
Research and Rulemaking................................................................................................. 57
Supporting Occupational Licensing Reform Efforts ....................................................... 57
Administrative Actions to Bolster Worker Power ........................................................... 58
Worker Organizing and Empowerment Task Force........................................................ 59
Reducing Job Lock and Boosting Mobility ....................................................................... 60
iv
1
THE STATE OF LABOR MARKET COMPETITION
INTRODUCTION
On July 9, 2021, President Biden signed a historic Executive Order on Promoting Competition in the American
Economy. The Order airms the importance of competition for workers, stating that “a competitive marketplace
creates more high-quality jobs and the economic freedom to switch jobs or negotiate a higher wage.” Yet, as the
Order explains, empirical evidence suggests that anti-competitive forces and practices have weakened workers’
bargaining positions and, consequently, worsened outcomes for workers. The Order outlined a whole-of-
government approach to addressing the excessive concentration of labor markets in the United States. As part of
this comprehensive approach, the Order directed the Secretary of the Treasury, in consultation with the Attorney
General, the Secretary of Labor, and the Chair of the FTC, to produce a report on the eects of lack of competition
on labor markets.
This report reairms the urgent need to promote competition in labor markets and increase workers’ bargaining
power. A central finding is that the American labor market is characterized by high levels of employer power.
Sources of this market power include natural labor market frictions, employer concentration, and anti-competitive
labor market practices. Employers exploit this market power by holding wages and certain non-wage benefits
beneath their competitive level. Simultaneously, the decline in unionization reduced worker bargaining power.
1
As
a result, workers are forced to accept lower wages and worse benefits than in a competitive market. These impacts
are oen disproportionately felt by socioeconomically vulnerable people, such as low-income workers, workers of
color, women, and immigrants. Problems stemming from lack of competition harm more than just the well-being
of workers and their families; it also holds back our entire economy, contributing to income inequality, inhibiting
innovation, and curbing economic growth.
Employer market power can manifest in forms beyond reductions in workers’ earnings that are challenging to
measure. Many of today’s jobs impose unpredictable just-in-time schedules, detailed on-the-job monitoring
coupled with demanding speed requirements and punitively short breaks, inadequate safety systems, and no
opportunity for advancement. While these determinants of job quality are harder to measure than wages, and
therefore less well studied, they also suggest that labor markets are not perfectly competitive.
First, this report begins by exploring some of the theoretical underpinnings of firm labor market power. We then
survey the empirical literature on many of the primary developments that have contributed to persistently low
labor market competition and worker bargaining power in recent decades. Topics surveyed include shiing firm
boundaries (fissuring of the workplace), restrictive employment agreements (e.g., non-compete agreements),
mandatory arbitration clauses, and occupational licensing. We also document the decline in worker mobility
and bargaining power and note the literature on the divergence between labor productivity and labor income,
labor’s share of overall income, and declining enforcement actions, among other things. We highlight how these
developments have impacted specific industries and sectors of the economy, including hospitals and nursing,
agricultural inputs and food processing, and minor league baseball.
Empirical studies of labor market power have proliferated recently, as academic interest in the topic enjoys a
renaissance. As papers address the empirical problem using a variety of methods, economists can increasingly
paint a nuanced picture of labor market power as it exists today. Considerable debate over details—big and small—
persist, but recent literature agrees on the broader picture: many employers exert market power when hiring
workers, and those workers are compensated less as a result.
We conclude the analysis portion of the paper by highlighting the implications of diminished labor market
competition on the broader economy. This includes growing income inequality, declining business investment and
Farber, Henry S., Daniel Herbst, Ilyana Kuziemko, and Suresh Naidu. 2021. “Unions and Inequality over the Twentieth
Century: New Evidence from Survey Data.” The Quarterly Journal of Economics 136 (3): 1325–1385.
1
THE STATE OF LABOR MARKET COMPETITION
productivity growth, declining worker mobility and productivity growth through less reallocation, and lower levels
of firm formation and innovation.
The extent to which this area has gained traction was demonstrated by an address by economist David Card,
the most recent recipient of the Nobel Prize in Economics, at the annual meeting of the American Economic
Association. In his address, Card calls on the field of economics to study the role of imperfect competition in labor
markets, while observing that widespread lack of competition has become the consensus view in economics. Card
concludes his address by noting:
One of the most exciting developments in the field today is the evidence of labor economists
taking questions about wage setting seriously. This eort began with Manning’s (2003) landmark
book: I hope that the growing body of work since then finds its way into the classroom and into the
textbooks soon. I also expect this work to lead to some re-thinking on policies such as minimum
wages, the regulation of trade unions, and anti-trust (see Longella and Manning 2021, and Naidu
and Posner 2022). Perhaps we may even see a re-evaluation of the widespread belief that excessive
wages are the root cause of many economic problems. Aer all, if your employer set your wage, it’s
hard to believe that it’s too high.
2
With a similar spirit, the Biden Administration has prioritized policies to restore labor market competition and
increase the relative bargaining power of workers. The report concludes with the Administration’s policies to
counteract the decline in labor market competition, including a policy favoring full enforcement of the antitrust
laws in labor markets, expanding opportunities for collective bargaining, raising the minimum wage, and
extending health insurance coverage to reduce job lock and boost mobility.
Card, David. 2022. “Who Set Your Wage?” National Bureau of Economic Research Working Paper 29583. 2
2
3
THE STATE OF LABOR MARKET COMPETITION
THEORIES OF LABOR MARKET POWER
Defined simply, the labor market matches workers and firms, creating jobs. Jobseekers oer their skills and time
to firms, which in turn oer pay and benefits. Simplicity, however, insuiciently describes the labor market. It
misses the pervasive variety: on one side of the market, each worker brings a unique set of skills, dispositions, and
circumstances to an employer. On the other side, there is an enormous variety of jobs in the United States.
3
In this
sense, labor markets are very dierent than some product markets, like commodity markets, where the product
is relatively homogenous, and buyers are usually indierent to who is selling and vice-versa. In the labor market,
both buyers (firms) and sellers (workers) take great interest in their counterpart’s characteristics.
In a strong and expanding economy, a well-functioning labor market typically delivers wage growth, low
unemployment rates, regular job switching, and improved job quality. This dynamic benefits society: when
workers and firms can easily match and separate, it increases the average productivity of each job. Over their
careers, workers find jobs that increasingly suit them, and employers find workers who best fit their needs.
However, “well-functioning” is not the default state of labor markets. The job search is beset by frictions, among
them time, information, diverse worker preferences, and geography. Alongside other factors, these frictions can
frequently generate market power for employers of all sizes, decreasing the market’s eiciency and reducing the
gains that would otherwise accrue to society.
We define “labor market power” (herein, “monopsony” or “market power”) as a firm’s power to reduce the
compensation it pays to its workers, paying less than an equivalent job would, in a hypothetical perfectly
competitive market. Market power allows a firm to decrease its compensation without losing its entire workforce,
where compensation refers to not just wages, but also benefits, job quality and working conditions.
4
Likewise,
the firm can expand its workforce by raising compensation.
5
Lower pay is the eective outcome of a labor market
characterized by “monopsony”—the situation when an individual firm has some control over the market and
thus can aect compensation. Still, monopsony does not imply a complete absence of market forces. So long as
workers have any alternatives, markets help dictate the extent of a monopsonist’s power.
Monopsony’s counterpart is perfect competition, an economic model in which both workers and firms take
wages as given—meaning they cannot raise or lower the prevailing wage. Under perfect competition, the residual
labor supply curve (or firm-specific labor supply curve) is flat, meaning each firm can hire whatever amount of
labor it wants but only at the market wage. Therein lies the key technical distinction between monopsony and
competition: an upward sloping versus flat residual labor supply curve. Note that in both contexts, the aggregate
(market-level) labor supply curve is typically upward-sloping.
To illustrate the contrast between competition and market power, consider this question: if an employer cut their
wages by 5 percent, what fraction of their workers would quit? In a perfectly competitive market, all workers would
leave. Yet, we know that this is not true in practice—indicating that many employers have some degree of market
power.
A labor market monopsonist leverages their position to pay their workers less than the competitive rate for a
given job. In a perfectly competitive labor market, each worker earns the market value of what they contribute
3 Not only are there 867 detailed occupations recognized by the Bureau of Labor Statistics’ (BLS) 2018 Standard Occupational
Classification, but there are plenty of dierences within those occupations. Further, similar jobs oer unique requirements
and benefits, which by itself is evidence of some level of monopsony in labor markets.
4 Throughout the paper, we intend compensation or wages to refer to not just money, but also benefits, job quality, and
working conditions.
5 In economist jargon, a firm that has an “upward-sloping residual labor supply curve” also has market power. The “residual”
part of that phrase distinguishes the firm-specific labor supply curve from the aggregate (market-level) labor supply curve.
THE STATE OF LABOR MARKET COMPETITION
to production—known as the “marginal revenue product of labor” (herein, MPL
R
). A labor-market monopsonist
instead sets its compensation below the MPL
R
, which reduces its cost of production and therefore raises profits.
Practically, the strength of a firm’s market power is indicated by the dierence between compensation and
MPL
R
. Throughout, we refer to this dierence—in eect, the amount by which a firm suppresses a worker’s
compensation—as either a firm’s “markdown,” or a worker’s “lost wages.” This is analogous to monopoly’s better-
known concept of a markup, where a firm charges a price for a good above the firm’s costs of production.
Broadly speaking, two distinct classes of economic theories might help explain the source of employers’ labor
market power. The first class is based on labor market structure: pure monopsony, monopsonistic competition,
and oligopsony. These are demand-side counterparts to the more familiar models of monopoly, monopolistic
competition, and oligopoly. If only one or a few firms are buying labor in a given labor market, they have the power
to set wages in that market and will keep wages below what workers might be able to charge in a competitive
market, so workers have nowhere else to turn.
The second class of theories stems from “search and matching” models of labor markets. Search and matching
models explicitly account for the frictions and opportunity costs inherent to job searches, both from the worker’s
and firm’s perspectives. In these models, employers account for the worker’s diiculties in finding a new job.
These diiculties include the direct costs of a job search (e.g., time), as well as indirect costs such as uncertainty
about the suitability of a new job, a lack of knowledge about wages or benefits oered by other firms, or foregone
pay during unemployment. It also encompasses the fact that jobs are more than just compensation to a worker,
who also values the nature of work, company culture, coworkers, managers, and commute times—and dierent
workers may value the same aspects of a job dierently. If one worker highly values a specific facet of a job, then
they would accept a lower wage than other workers for the same position. Consequently, the employer can reduce
its compensation and still maintain many of its workers. For the purposes of this report, both theories share the
same core outcome: they result in the reduction of worker compensation.
We now detail those theories and their implications.
Pure Monopsony
Pure monopsony describes a market with a single buyer. This is the mirror image of a monopoly model (a single
firm selling final goods and services), except a single firm is purchasing inputs (like labor). In the labor context,
monopsony exists if some workers have only one option for employment, such as a “company town” where there
is a single dominant employer in the community. As such, it is rarely the ideal model to describe U.S. labor markets.
Figure 1 - Pure Monopsony: Elasticity Drives the Scale of Wage Loss
4
5
THE STATE OF LABOR MARKET COMPETITION
The model nonetheless remains useful, both to help understand the market power problem in a simple context
and to establish nearly all the foundation for the more realistic model of monopsonistic competition.
Practically, an upward-sloping labor supply curve implies two costs to hiring a new worker: the first is the wages
directly paid to the new worker, and the second is the increased wages paid to workers already employed by the
firm. By the same logic, a monopsonist enjoys these two sources of reduced costs by constraining employment
below the competitive level.
In market structure models, the elasticity of labor supply lives at the heart of market power. Loosely defined, this
elasticity measures how strongly the workforce reacts when wages change. In turn, the elasticity of labor supply
dictates the markdown in wages. When the labor supply is highly elastic, a small decrease in wages results in a
large decrease in the number of workers who are willing to work for the firm. In this case, a monopsonist has little
to gain from markdowns since it stands to lose too much of its labor force. With lower elasticities, however, the
same decrease in wages prompts a weaker response from the workforce. This eectively grants the monopsonist
increased pricing power, as wage cuts induce fewer quits than in a higher-elasticity environment. Simply put, when
workers are prepared to walk away from a job, their employer has less power over them.
Monopsonistic Competition
At a national level, pure monopsony is clearly an inappropriate descriptor for labor markets. A more realistic model
of labor markets in the United States is that of monopsonistic competition.
Monopsonistic competition is similar to pure monopsony, except the firm faces a residual labor supply curve
rather than the aggregate supply curve. To reiterate, a firm’s residual labor supply curve is specific to that firm,
aer accounting for the labor supply curves facing the rest of the market. When wages fall economy-wide, workers
will more readily switch from a firm that lowers its wages than out of employment altogether. In other words,
residual labor supply curves are more elastic than aggregate labor supply curves. Taken further, the more similar
employment is between firms, the more readily workers will switch and the greater the elasticity of residual labor
supply curves.
An example of a monopsonistically competitive labor market might be a city with many restaurants. Though there
might be many restaurants employing chefs, they are not identical. A chef has skills that can be used in a multitude
of restaurants, but this does not mean the chef is indierent to where they are employed. Some restaurants may
provide a more suitable menu, have better or more predictable work schedules, or be more conveniently located.
In this case, the chef may be willing to accept a discounted wage to work at a particular restaurant, giving that
restaurant some degree of market power.
Search and Matching Models of Labor Markets
Search and matching models introduce important nuance to theories of labor market power.
6
Specifically, these
models provide conditions where all employers, to varying degrees, possess market power; but crucially, these
models also account for the frictions involved in job searches, among them time and considerable uncertainty.
Aware of these frictions, employers can discount wages while retaining their workforce and hiring new employees.
A worker will sometimes prefer to accept a job with a discounted wage than to continue a job search that may not
yield a better alternative quickly or at all.
A friction is any factor that makes job searches or switches more diicult than the theoretical ideal of switching
between two identical consumer goods, such as pantry staples. The job search process is also characterized by
For a reference and considered argument, see Manning, Alan. 2003. Monopsony in Motion: Imperfect Competition in Labor
Markets. Princeton: Princeton University Press.
6
THE STATE OF LABOR MARKET COMPETITION
considerable information gaps. For example, consumers can easily compare airfare prices on travel aggregator
websites, but it is typically impossible for workers to learn the compensation associated with every potential
employment opportunity. Real-world labor markets feature significantly more frictions than consumer markets.
Broadly speaking, there are two types of search and matching models relevant to labor market power. The first is
characterized by “ex-ante wage posting,” where employers announce their wages with every job oer. This is oen
applicable to lower wage jobs—think of a sign outside a fast-food restaurant that states, “Positions starting at $15.”
The second is characterized by “ex-post wage bargaining,” where the worker and firm negotiate wages and benefits
in the final stage of the hiring process. This is more typical in higher-paying jobs, where the job postings oen
include an ambiguous statement that the job pays “competitively.”
In an “ex-ante wage posting” model
7
, workers do not simply pick a job—they must be oered the job first, and
the oer comes at a known wage. Upon receiving a job oer, they can accept or decline, which they will do based
on their understanding of the rest of the market. If a worker thinks they are likely to receive a significantly better
oer elsewhere, they decline the current oer and keep searching. If the worker does not believe they are likely
to receive a better oer elsewhere (relative to the continued costs of job search and, for those not currently
employed, unemployment), then they will accept the oer. In this way, it is possible for firms oering the same
employment to oer dierent wages—a key characteristic of monopsony models. Firms can choose to raise their
wages and induce more workers to accept their oers while simultaneously keeping more of their existing workers,
displaying the key characteristic of a monopsonist: to face an upward-sloping residual labor supply curve. One
critical insight to these models is that a firm may be neither large nor dominant in its market but still exercise
market power.
In “ex-post bargaining” models, a jobseeker does not know the wage in advance. The worker and firm bargain over
the wage in the final stages of the hiring process. In these models, each job generates a “surplus,” defined as the
gap between the worker’s lowest acceptable wage (their “reservation wage”) and the highest wage an employer
can profitably pay (i.e., the worker’s MPL
R
). Firms and workers then bargain over how to allocate that surplus. The
share of this surplus going to firms represents profits, while the share accruing to workers represents wages above
their reservation wage. If labor markets were perfectly competitive, wages would simply be a function of worker
productivity (as wages would be competed upward to the maximum that firms could profitably pay)—meaning
workers would be paid their MPL
R
. Like the “ex-ante wage posting” models, job search frictions in “ex-post
bargaining” models give employers room to pay sub-competitive wages.
Various factors impact how firms and workers allocate the surplus of the worker’s employment. Generally, the
greater the bargaining power one side has, the larger a share of the surplus they can capture. The bargaining
power of employees largely rests on their alternative (“outside”) options and the degree to which they are
substitutable with other workers. For example, a worker who has unique and highly specialized training that
is valued by many other firms generally has greater bargaining power over their share of the surplus than an
employee that is relatively easily replaceable and has relatively non-transferable skills. On the other hand, a nurse
living in a rural town with only a single hospital within driving distance may have lower bargaining power because
that worker lacks alternative local employment options.
While some job search frictions arise naturally, employers can also actively take steps to increase frictions or
generate new ones. These frictions are the underlying source of market power in both types of search and
matching models, giving employers an incentive to increase frictions. Some frictions are “natural” in the sense that
they are not erected by the worker’s employers. For example, high costs of moving (including implicit costs like the
loss of access to one’s social network) may induce someone to stay in their current job despite better alternatives
elsewhere. Personal preferences are another natural factor that can give employers leverage. Insofar as a worker
For example, Burdett, Kenneth, and Dale T. Mortensen. 1998. “Wage Dierentials, Employer Size, and Unemployment.”
International Economic Review 39 (2): 257–273.
7
6
7
THE STATE OF LABOR MARKET COMPETITION
is willing to accept a lower wage to work for a given employer, for any personal reason, the firm has the potential
power to reduce that worker’s wage below MPL
R
and still retain the worker. This holds true even if the worker
knows that they can be paid more at a rival firm.
Information asymmetry regarding potential wages is another crucial friction. If workers underestimate the wages
paid by similar employers, then they will be less likely to actively search for a new employer. For workers, acquiring
information about outside options is oen more costly than for firms.
8
Recent evidence from Jäger et al. (2021)
suggests that worker beliefs about outside options are strongly and unduly influenced by their current wage,
which harms the lowest-paid workers the most.
9
They estimate that 10 percent of German jobs could not continue
at their current wages if workers had the correct understanding of their outside options. These “non-viable” jobs
were concentrated in lower-paid positions. Importantly, these asymmetries arise naturally, but employers can
increase them by concealing wages.
Employers can also act to decrease the value of a worker’s outside options. For example, restrictive employment
agreements that require workers to repay training costs if they leave the firm or non-compete agreements (both
discussed in greater detail below) reduce worker power by increasing the costs of leaving the firm. Those costs
are explicit in the case of training repayment programs but implicit in non-compete agreements. By preventing a
worker from accepting positions well-suited to their skills, firms decrease the expected gains from a worker’s job
search.
Finally, regulations can also increase the frictions in a job search. Occupational licensing is a notable example,
and one that is growing more common over time. These frictions are growing in several ways: the number
of occupations covered by licensing; the requirements, costs, and complexity of securing a license; and the
patchwork of licenses across states. With non-reciprocity in licensing, two states may have similar goals and
standards for a given occupation, yet it remains costly for a worker to move between states.
Licensing does benefit some workers, specifically incumbent workers, in many circumstances. By increasing
barriers to entry, licensing restricts the supply of new workers, thereby increasing incumbents’ bargaining power.
This comes at the expense of other workers who would like to take up the trade, as well as firms and consumers
in the form of higher prices. However, licensing can harm incumbents too: if licensure diers across states, then a
worker who is licensed in one state will find it costly to move, despite professional or personal reasons to want to
do so. Licensing can also protect public safety, help consumers distinguish high-quality from low-quality service,
and even play a role in ensuring a market for certain goods and services exists (as in Akerlof 1970).
10
On the other
hand, licensing can be misused to protect already powerful job occupations and incumbents.
Racial Inequality under Search and Matching Frictions
The frictions arising within search and matching models help explain the link between racial discrimination and
racial wage gaps. Models of racial discrimination in the style of Gary Becker’s Economics of Discrimination (1957)
apply within the classical monopsony framework, with the implication that if some employers discriminate based
on race, then market forces will eventually close the racial pay gaps that result from discrimination.
11
This sanguine
8 This is a result of economies of scale. Firms benefit from information when dealing with every worker they employ or
potentially employ. Workers only benefit from this information when it relates to themselves.
9 Jäger, Simon, Christopher Roth, Nina Rousille, and Benjamin Schoefer. 2021. “Worker Beliefs about Outside Options.”
National Bureau of Economic Research Working Paper 29623.
10 Akerlof, George. 1970. “The Market for “Lemons”: Quality Uncertainty and the Market Mechanism.” The Quarterly Journal of
Economics 84 (3): 488–500.
11 Becker, Gary S. 1957. The Economics of Discrimination. Chicago: University of Chicago Press. Note that the specific source of
discrimination is important. The model predicts that discrimination coming from consumers (or co-workers) results in a
wage gap that will not be rectified by market forces.
THE STATE OF LABOR MARKET COMPETITION
result does not hold within search and matching models, as shown by Black (1995).
12
Within a search and matching
model, discrimination by even a few employers has a market-wide impact. For example, if some employers
discriminate against Black workers, then Black workers face a worse set of potential outside wage oers than their
non-Black counterparts. As a direct result, the expected value of a job search is lower for Black workers than it is for
non-Black workers.
This lower expected value of search results in a lower average wage through two mechanisms. First, it decreases
the returns to a job search for Black workers, meaning they will dedicate fewer resources to search in equilibrium.
Second, if employers without proclivity towards discrimination know of the decreased expected returns to search,
then they also know they can oer Black workers lower wages than non-Black workers, all while maintaining an
equal chance that the oer is accepted.
We have considerable empirical evidence to document discrimination faced by Black workers searching for
a job. A substantial literature that has developed submits fake resumes to firms, en masse, with names that
are randomized to be “white-sounding” or “Black-sounding.”
13
The results consistently show that resumes
with stereotypically white names receive callbacks at higher rates than otherwise identical fake resumes with
stereotypically Black names. Bertrand and Mullainathan (2004), for example, find that “white-sounding” names
receive 50 percent more callbacks than “Black-sounding” names among applications submitted to Boston- and
Chicago-area newspapers. Though subsequent papers have typically found smaller eects, the direction of the
results have held consistently. To reiterate, this dynamic results in lower wages for Black applicants, all else equal.
In search and matching models, wages are a function of outside options—having fewer (or worse) outside options
leads to lower average wages, regardless of cause.
Platforms/Regulatory Arbitrage
Regulation is one tool to ameliorate the pernicious eects of monopsonistic power. In a standard example, a
judiciously determined price floor (minimum wage) can simultaneously increase wages and employment in the
basic monopsony model. For the same reasons, regulations on working conditions can potentially accomplish
desirable outcomes without job loss.
Regulatory arbitrage occurs when a company attempts to circumvent enforcement of regulations by availing
themselves of dierent regulatory schemes. Regulatory arbitrage oen comes about from ambiguities
(“loopholes”) in regulations that allow firms to operate in a type of grey space.
14
These ambiguities can weaken
regulatory action, including those meant to curb monopsonistic power.
The rise of e-commerce has created new opportunities for regulatory arbitrage as regulatory schemes of the
twentieth century meet twenty-first century innovations. Critics argue regulatory arbitrage is widespread in these
new markets and gives firms an unfair advantage over their competitors. For example, Amazon was essentially
exempt from sales taxes for the first 15 years of its existence, giving it an 8–10 percent price advantage over
competitors (Kahn 2017, footnote 204).
15
Such a large price advantage can allow a company to quickly gain
12 Black, Dan A. 1995. “Discrimination in an Equilibrium Search Model.” Journal of Labor Economics 13 (2): 309–334.
13 See, e.g., Bertrand, Marianne, and Sendhil Mullainathan. 2004. “Are Emily and Greg more employable than Lakisha and
Jamal? A field experiment on labor market discrimination." American Economic Review 94 (4): 991–1013; and Banerjee, Rupa,
Jerey G. Reitz, and Phil Oreopoulos. 2018. “Do large employers treat racial minorities more fairly? An analysis of Canadian
field experiment data." Canadian Public Policy 44 (1): 1–12.
14 See, e.g., Brief of the United States Department of Justice as Amicus Curiae at 4, The Atlanta Opera, Inc., 10-RC-276292
(NLRB Feb. 10, 2022). Cites potential for the National Labor Relations Board’s (NLRB) regulatory ambiguity to “creat[e]
opportunities for employers to undercut competition by misclassifying their own employees.”
15 Khan, Lina M. 2017. “Amazon’s Antitrust Paradox.” The Yale Law Journal 126 (3): 710–805. https://www.yalelawjournal.org/
note/amazons-antitrust-paradox.
8
9
THE STATE OF LABOR MARKET COMPETITION
dominance in the product space, which may contribute to the firm also gaining increased monopsony power.
Regulatory arbitrage can also weaken worker protections when a firm uses terminology to take advantage
of regulatory arbitrage in employment laws. For example, critics of ride-hailing companies, argue that these
companies engage in a type of regulatory arbitrage by claiming their drivers are independent contractors when
they may more aptly be classified as employees. This distinction, known as misclassification, is discussed in more
detail in the next section.
Fissuring of the Workplace
Changes in organizational structure of firms since the 1980s have dramatically reduced the bargaining power of
some workers. Prior to the 1980s, large corporations tended to directly employ workers across many occupations.
By the late 1980s, firms began to favor a management style that emphasized firms’ focus on the handful of
areas where their companies have a comparative advantage, known as their “core competencies.”
16
Accordingly,
firms began to shed workers by outsourcing, and in some cases oshoring, large parts of their employment,
particularly among jobs near the lower end of the income and skill distribution.
17
For example, instead of directly
hiring janitorial services, companies began to contract with janitorial service companies. David Weil, former
Administrator of the Wage and Hour Division at the Department of Labor (DOL), has termed this process of
outsourcing labor as the fissuring of the workplace.
18
Consequently, the modern large business looks more like a
“small solar system with a lead firm at its center and smaller workplaces orbiting around it” rather than a large
single entity (Weil 2014, 42). At the center of some of the biggest solar systems are firms that Autor et al. (2020)
have dubbed “superstar firms.”
19
Jobs that are fissured do not necessarily disappear—they are reorganized, although oen under very dierent
terms. Fissured jobs may be restructured in several ways, including sub-contracting, franchising, greater reliance
on temporary staing agencies, and classifying workers as independent contractors.
Fissuring Considerations
Fissuring potentially benefits firms and consumers. Contracting out areas of relative weakness can allow
management to focus on areas where they have a comparative advantage. Accordingly, firms are more productive
per retained worker, which could lead to lower prices for consumers and potentially more innovation. In certain
circumstances, fissuring can benefit smaller businesses as well. Very small firms may lack the funds to hire a full-
time custodial employee or accountant and contracting out such tasks could free up mental bandwidth for small
firms to focus on their core competencies.
Although it potentially benefits firms and consumers, fissuring can have a detrimental impact on workers. Fissuring
can, and empirically does, reduce labor’s share of surplus by weakening worker bargaining power and reducing
wages for outsourced workers. For example, Dube and Kaplan (2010) estimate that outsourcing among janitors
and guards reduced wages by 4–24 percent.
20
They also find substantially lower rates of non-wage benefits, such
16 Prahalad, C. K., and Gary Hamel. 1990. "The core competence of the corporation.” Harvard Business Review 68 (3): 79–91.
17 Oshoring is a special case of outsourcing. While both involve contracting out tasks or processes to a third party, oshoring
specifically refers to contracting out those tasks or processes to entities outside of the country.
18 Weil, David. 2014. The Fissured Workplace: Why Work Became So Bad for So Many and What Can Be Done to Improve It.
Cambridge: Harvard University Press.
19 Autor, David, David Dorn, Lawrence F. Katz, Christina Patterson, and John Van Reenen. 2020. "The Fall of the Labor Share
and the Rise of Superstar Firms." The Quarterly Journal of Economics 135 (2): 645–709.
20 Dube, Arindrajit, and Ethan Kaplan. 2010. “Does Outsourcing Reduce Wages in the Low-Wage Service Occupations?
Evidence from Janitors and Guards.” Industrial and Labor Relations Review 63 (2): 287–306.
THE STATE OF LABOR MARKET COMPETITION
as health insurance coverage, among outsourced workers. For some workers, this may underestimate the eect
of outsourcing if outsourced workers must spend more out-of-pocket to pay for equipment previously supplied by
their employer. Using German administrative data that allowed them to follow workers over time, Goldschmidt
and Schmieder (2017) show that wages fell by 10–15 percent among outsourced workers in the food, cleaning,
security, and logistics service industries compared to similar workers who did not experience outsourcing.
21
To some extent, the lower wages and decreased benefits that fissured workers receive are the point of fissuring
in the first place. Within a firm, a rising tide may li all boats, but when firms fissure their workforce, they exclude
certain people from that boat. Economists have long recognized that there are substantial wage dierences
between directly employed and outsourced workers doing similar work, even controlling for industry, work
environment, and, to some extent, unobserved skills.
22
These “wage premia” are regularly observed to be larger
in larger firms, although there is evidence that the scale of the large-firm wage premium may be decreasing over
time.
23
One reason for the higher wages paid to direct employees at some firms is that certain employers, especially
profitable ones, pay so-called “eiciency wages” (higher wages than their employees could likely earn elsewhere
in the market) to increase retention and worker productivity. Intra-firm dynamics and social norms can discourage
providing these higher eiciency wages to only a subset of the firm’s workers.
24
In this way, a janitor employed at a
large profitable firm may well earn above the market rate for their employment. Eiciency wages can also benefit
similarly situated workers in other firms by improving their outside options, thereby strengthening their bargaining
position with their current employer. These outside pressures, however, abate when firms contract out their “non-
core” workforce.
Aer having their jobs outsourced, fissured workers lose some of their bargaining power because they no longer
benefit from the larger workforce dynamics at that employer. Additionally, fissured workers likely miss out on
the internal career opportunities that would have been available if they were considered employees, which
compounds the impact of lost career opportunities for intra-firm mobility. Moreover, if workers from multiple firms
are outsourced to a single staing agency, the labor market in which those workers participate will have greater
employer market power.
It is diicult to assess exactly which occupations have been the most fissured; however, Weil (2019) provides
a compilation of industries where fissuring has been well documented and appears to be widespread.
25
Weil’s
compilation broadly suggests industries where fissuring has been most prevalent, including telecommunications
sub-industries (e.g., telephone call centers), food service industries (e.g., mobile food services), temporary help
services, construction subindustries (e.g., landscaping), janitorial services, security services (e.g., security guards),
and transportation subindustries (e.g., taxi and limousine services). In some of these industries, women and
21 Goldschmidt, Deborah, and Johannes F. Schmieder. 2017. “The Rise of Domestic Outsourcing and the Evolution of the
German Wage Structure.” The Quarterly Journal of Economics 132 (3): 1165–1217.
22 See e.g., Krueger, Alan B., and Lawrence H. Summers. 1988. "Eiciency wages and the inter-industry wage structure."
Econometrica 56 (2): 259–293. For a more general empirical view of the large-firm wage premia, see the large literature
starting with Abowd, John M., Francis Kramarz, and David N. Margolis. 1999. “High Wage Workers and High Wage Firms.”
Econometrica 67 (2): 251–333.
23 Bloom, Nicholas, Fatih Guvenen, Benjamin S. Smith, Jae Song, and Till von Wachter. 2018."The Disappearing Large-Firm
Wage Premium."American Economic Association Papers and Proceedings108 (May): 317–22.
24 See, e.g., Piketty, Thomas. 2014. Capital in the Twenty-First Century. Cambridge: Harvard University Press. For some firms,
especially unionized ones, this aversion to intra-firm income inequality may be mechanical: unions may require that the
top paid employee earns no more than some multiple of the lowest earning full-time employee. Fissuring can circumvent
such rules by no longer considering the lower-paid workers their employees (Dube and Kaplan 2010).
25 Weil, David. 2019. “Understanding the Present and Future of Work in the Fissured Workplace Context.” RSF: The Russell Sage
Foundation Journal of the Social Sciences 5 (5): 147–165.
10
11
THE STATE OF LABOR MARKET COMPETITION
minority groups are disproportionately represented.
26
For example, Hispanic workers make up roughly twice as
large a share of janitors and building custodians compared to their share of employment in the overall economy
(31.5 percent versus 18.0 percent).
27
Hispanic workers also make up a much greater share of construction laborers
than their share of employment (48.9 percent versus 18.0 percent). Similarly, women represent approximately
87 percent of registered nurses, even though they only represent about 47 percent of employment.
28
Although
fissuring is not exclusively a phenomenon among low-income workers, many of the industries where fissuring
appears widespread are industries with low average pay. For example, in November 2021, the average worker
in the overall economy earned about $1,100 a week, but telephone call center workers only earned an average
of about $775 a week and hotel and motel workers (except casinos) earned only about $650 a week.
29
Janitorial
service workers earned even less—about $575 a week.
30
Fissuring also reduces the power of collective action. By removing the immediate nexus between workers and the
firm for which they perform services, workers are prevented from bargaining directly with the entity that has the
economic power. Further, workers whose jobs are contracted out typically end up in a much more competitive pool
of relatively substitutable workers. As Kaplan and Dube (2010) explain, contracting reduces union power because
contracted workers can be permanently replaced by a switch in the contractor of record, even if they are unionized.
This reduces the incentive to try to unionize. In some cases, employers use new structures that make it diicult
to form unions. For example, in the janitorial services industry, workers are commonly considered independent
contractors (Weil 2014). Most worker protection laws, including the National Labor Relations Act, do not cover or
protect bona fide independent contractors, so these workers lack collective bargaining rights. Furthermore, they
face possible antitrust constraints when they try to act collectively in their economic interest.
The intra-firm dynamics highlighted above have a substantial impact on income inequality. Song et al. (2019)
notes that a third of the rise in income inequality from 1978 to 2013 occurred because of changes within firms (as
opposed to between them).
31
They further note that one of the two dominant explanations for this increase in
inequality within firms was that high-wage workers became more likely to work with each other, which is a natural
consequence of fissuring lower-wage workers from the firm.
32
Goldschmidt and Schmieder (2017) similarly find
domestic outsourcing deepened income inequality in Germany. We discuss the consequences of rising income
inequality later in the paper.
Misclassification of Workers
A firm misclassifies a worker when it treats a worker, who should be classified as an employee, as an independent
26 Weil (2019) reports employment figures by industry. These data do not report demographic data. The demographic data
reported below are based on a slightly dierent classification of employment (based on occupation rather than industry);
therefore, demographic decompositions do not perfectly correspond to subindustries identified in Weil (2019). However,
both sets of employment estimates originate from surveys conducted or sponsored by BLS.
27 U.S. Bureau of Labor Statistics. 2022. “Labor Force Statistics from the Current Population Survey: Household Data Annual
Averages: 11. Employed persons by detailed occupation, sex, race, and Hispanic or Latino ethnicity.” Last modified January
20, 2022. https://www.bls.gov/cps/cpsaat11.htm.
28 Nurses are much more likely to work for temporary staing agencies. See, e.g., Seo, Sukyong, and Joanne Spetz. 2013.
"Demand for temporary agency nurses and nursing shortages."INQUIRY: The Journal of Health Care Organization, Provision,
and Financing50 (3): 216–228.
29 U.S. Bureau of Labor Statistics. 2022. “Current Employment Statistics – CES (National): Employment and Earnings.” Table B-3a.
Last modified February 4, 2022. https://www.bls.gov/web/empsit/ceseeb3a.htm. Values rounded to the nearest $25 a week.
30 Id.
31 In this context “income inequality” is defined as the variance in the natural log of earnings. Song, Jae, David J. Price, Fatih
Guvenen, Nicholas Bloom, and Till von Wachter. 2019. "Firming Up Inequality." The Quarterly Journal of Economics 134 (1): 1–50.
32 The other explanation is the rise of sorting high-wage workers into high-wage firms.
THE STATE OF LABOR MARKET COMPETITION
contractor instead.
33
This concept is related to fissuring because both misclassification and fissuring describe a
process by which the purchasers of labor attempt to sever what would typically be considered an employee-employer
relationship. The employee-employer relationship has historically been the basis for worker protection laws, income
tax collection, social security collection, health insurance coverage, and other economic and social constructs.
Although fissuring is typically not a per se violation of the law, misclassifying a worker violates some laws.
34
Firms misclassify workers and outsource labor for similar reasons—it is cheaper and reduces their risk. For
example, assigning work to an independent contractor does not entail as many legal obligations, such as tax and
overtime obligations, as the hiring of an employee. Classifying workers as independent contractors can especially
reduce costs by shiing non-wage costs typically paid by employers (e.g., healthcare benefits) onto the employee.
35
These costs are non-trivial—approximately 30 percent of per-hour employer costs come from costs other than
wages and salaries.
36
Accordingly, a misclassified worker and a worker that is outsourced via fissuring face similarly
negative consequences. The ability of a firm to misclassify workers without successful pushback from employees
(who clearly would have an incentive to not be misclassified) can itself be viewed as a demonstration of the market
power firms have over workers.
The distinction between an employee and an independent contractor has developed over time and the legal
standards are not uniform. Fundamentally, the dierence depends on the nature of the work and the relationship
between the firm and worker. In some jurisdictions, courts determine whether a person should be classified as
an employee instead of an independent contractor using a three-part (“ABC”) test. Under this test, a worker is an
independent contractor only if their work relationship allows a “yes” answer to all of the following questions:
37
• Part A: The worker is free from the control and direction of the hiring entity in the performance of the work,
both under the contract for the performance of the work and in fact.
• Part B: The worker performs work that is outside the usual course of the hiring entity’s business.
• Part C: The worker is customarily engaged in an independently established trade, occupation, or business of
the same nature as the work performed for the hiring entity.
If the answer to any of these questions is “no,” the court should classify the worker as an employee. Although
dierent jurisdictions have adopted various exceptions to this ABC test, the test clarifies that, in general, workers
are only properly classified as independent contractors if their relationship with the business is suiciently arm’s
length and the worker maintains a large degree of autonomy. The ABC test is only one of several types of tests that
is used to determine whether a worker is misclassified and is not used under federal law. Other tests include the
common-law test and the economic realities test under the Fair Labor Standards Act.
33 Note, worker misclassification involves an incorrect statement by the firm, but does not necessarily imply (nor does it
legally need to imply) intentional misclassification on the part of the firm. They may genuinely consider their interaction
with a worker to be considered more like an independent contractor relationship than an employee relationship.
34 For example, in the District of Columbia, illegal misclassification is considered a form of payroll fraud. See, e.g., Belman,
Dale, and Aaron Sojourner. 2019. “Illegal Worker Misclassification: Payroll Fraud in the District’s Construction Industry.”
Oice of the Attorney General for the District of Columbia. https://oag.dc.gov/sites/default/files/2019-09/OAG-Illegal-
Worker-Misclassification-Report.pdf.
35 The true cost of such burden shiing largely depends on the sensitivity (elasticity) of each side (employer/employee) to the
costs. However, it is unlikely that employees are so sensitive as to eectively make it impossible for employers to reduce
employment costs by shiing nominal burdens to the employee.
36 U.S. Bureau of Labor Statistics. 2021. “National Compensation Survey – Employer Costs for Employee Compensation.”
Civilian workers dataset spreadsheet. Last modified December 16, 2021. https://www.bls.gov/web/ecec/ecec-civilian-
dataset.xlsx.
37 The language presented in this test come from Labor & Workforce Development Agency. 2022. “ABC Test.” State of
California. https://www.labor.ca.gov/employmentstatus/abctest/. Exact language and interpretation of the ABC test will
vary from state to state.
12
13
THE STATE OF LABOR MARKET COMPETITION
Worker misclassification has garnered particular attention around whether so-called “gig workers,” especially
people working for ride-sharing companies are properly classified. However, worker misclassification expands
way beyond gig workers and appears to be becoming more common. A 2018 study in Washington state found
that the proportion of employers that misclassify at least one of their workers almost tripled between 2008 and
2017 (from around 5 percent to 14.4 percent).
38
Among firms that misclassify at least some of their workers, they
tended to misclassify about 10–25 percent of their workforce. Using administrative data, that study estimated an
overall misclassification rate of a little over one percent between 2013 and 2017. Both the incidence and intensity
of misclassification varies widely by industry and occupation. The same report found that the industries with the
greatest incidence of misclassification were construction, clerical services, and hospitality (hotels and restaurants).
Worker misclassification has broader implications beyond its direct impact on the employee-employer dynamic.
Whereas employees’ income and Social Security taxes and employers’ payments of unemployment insurance
and other payroll taxes are managed by the employer, tax compliance among independent contractors, who
are required to file taxes on their own, is much lower.
39
Therefore, when an employer misclassifies an employee,
payments on that worker’s behalf are not made into social safety-net programs that otherwise would have if the
employers had properly classified their workers.
The Questionable Tax Employment Practices (QTEP) program, a joint state/federal program that audits tax data to
uncover tax non-compliance, has found large-scale misclassification.
40
Among the roughly 30,000 audits conducted
between (fiscal years) 2015 and 2020, the program reclassified more than 275,000 workers, resulting in the
reclassification of about $4 billion in wages.
41
Fissured workplaces may result in worker misclassification, and, in turn, worker misclassification impacts labor
market competition. Workers that are misclassified as independent contractors are deprived most methods by
which they can bargain for a greater share of labor market surplus. When the employer oloads the burdens of
labor costs on to the worker (including taxes, unemployment insurance, and social security), while continuing to
benefit from their productivity, the worker has very little recourse.
Restrictive Employment Agreements and No-Poach Agreements
Terms of employment contracts oen extend well beyond simply defining compensation from the employer and
job duties of the employee. Employers oen include a variety of clauses that restrict employees’ behavior, even
going so far as to dictate what they can do aer they leave the company.
42
As a result, workers are limited in their
ability to—or outright prohibited from—seeking higher-paying work in their field, which reduces their bargaining
and earning power. In some cases, such as no-poach agreements (in which employers agree not to solicit or hire
each other’s employees), employees are not even a party to the agreement.
38 Xu, Lisa, and Mark Erlich. 2019. “Economic Consequences of Misclassification in the State of Washington.” Harvard Law
School: Labor and Worklife Program. https://lwp.law.harvard.edu/files/lwp/files/wa_study_dec_2019_final.pdf.
39 See, e.g., Bruckner, Caroline, and Thomas L. Hungerford. 2019. “Failure to Contribute: An Estimate of the Consequences
of Non- and Underpayment of Self-Employment Taxes by Independent Contractors and On-Demand Workers on
Social Security.” Center for Retirement Research at Boston College Working Paper 2019-1. https://crr.bc.edu/wp-content/
uploads/2019/01/wp_2019-1.pdf.
40 See, e.g., Levine, Suzan G. 2021. “Questionable Employment Tax Practices (QETP) Program.” Employment and Training
Administration, U.S. Department of Labor Training and Employment Notice 3-21. https://wdr.doleta.gov/directives/attach/
TEN/TEN_03-21.pdf.
41 Note, QETP reclassifications include, but are not limited to, reclassifications due to worker misclassification. For example,
they also include reclassification due to the creation of shell companies to avoid tax payments.
42 In some instances, firms may even demand independent contractors to sign such agreements, although some courts may
find such clauses unenforceable on public policy grounds.
THE STATE OF LABOR MARKET COMPETITION
In practice, restrictive employment agreements can both result from and reinforce employer market power;
for example, an employer who has market power for other reasons, such as high market share, may be able to
increase its power over both employees and customers by requiring its employees to agree to restrictive clauses.
The potential relationship runs in reverse, as well: in a labor market characterized by pervasive use of restrictive
agreements, a merger that increases employer concentration may have greater detrimental eects on competition
than would otherwise be the case.
The table below outlines several types of restrictive employment agreements.
Clause Description
Non-compete
agreements
Former employee cannot work for a competitor following
separation. Typically applies for a certain amount of time, over a
certain geographic area, and within a specific industry.
Non-solicitation
agreements
Employee agrees to not solicit a company's clients or customers
for their own benefit, or the benefit of a competitor, aer leaving
the company.
Non-recruitment
agreements
Employee or former employee is forbidden from recruiting
employers' employees away from employer for a period.
Training repayment
agreements
Employee must repay the cost of training provided by employer
if they leave employment prior to some period. Agreement is
typically pro-rated based on length of employment following
training.
Non-disclosure
agreements
Prevents employee or former employee from disclosing
information. Meant to protect information that is both
confidential and valuable.
No-poach agreements Two or more employers agree to not solicit or hire each other’s
current or former employees.
Heterogeneity in Enforcement and Legality of Restrictive Agreements
The mere statement of a restrictive term in an employment contract does not automatically make it enforceable.
Employment contracts are typically evaluated at the state-level pursuant to statute and case law. Therefore, the
degree to which courts will enforce such contract provisions varies between states. For example, Texas statutory
law allows for non-compete covenants but only “to the extent that it contains limitations as to time, geographical
area, and scope of activity to be restrained that are reasonable and do not impose a greater restraint than is
necessary.”
43
Enforceability also sometimes varies by occupation. For example, Texas places further conditions on
the enforceability of non-compete clauses in medical occupations.
44
California, in contrast, prohibits enforcement
of non-compete agreements.
45
Employers who illegally use restrictive covenants rarely face sanctions, such as monetary damages. Instead,
43 Tex. Bus. & Com. Code Ann. §15.50(a) (West 2021).
44 Id. at §15.50(b).
45 Cal. Bus. & Prof. Code § 16600 (West 2021).
14
15
THE STATE OF LABOR MARKET COMPETITION
courts normally either refuse to enforce the covenant or limit the breadth of overly expansive covenants. As such,
employers rarely face strong disincentives to including questionable restrictive covenants.
However, federal law has placed limitations on some restrictive employment agreements. For example, in 2016,
the Department of Justice (DOJ) and Federal Trade Commission (FTC) jointly issued guidance to human resource
professionals explaining (inter alia) that naked wage-fixing or no-poach agreements among competitors are per se
violations of the antitrust laws.
46
In early 2021, DOJ announced the first indictments charging naked no-poach or
wage-fixing conspiracies.
47
No-poach agreements are common in highly concentrated and high-skilled industries,
as well as in the franchise context, although some chains have ended them in recent years amidst legal and public
pressure.
48
No-poach agreements are also subject to challenge under state antitrust law.
Theory
Non-compete agreements are among the most common form of restrictive employment agreements, but many
of the lessons from that literature also apply to other forms of these agreements.
49
Non-compete agreements
(and other similar post-employment restrictive employment agreements) potentially solve a problem that would
otherwise limit a firm’s investments in their employees—namely, that workers would leave before a firm was able
to recoup the value they had invested in training a worker. At the same time, these agreements introduce frictions
into the labor market, weaken workers’ bargaining positions, and reduce competition over wages (McAdams 2019).
50
Non-compete agreements are also attractive to employers because employers typically cannot subject employees
to term contracts (i.e., a contract that requires the employee to work at a firm for a fixed period of time) because
courts refuse to issue injunctions compelling employees to stay in a job. The non-compete agreement indirectly
accomplishes this goal by depriving the employee of the most attractive alternative employment opportunities.
In theory, non-compete agreements can increase a firm’s investment in their employees by reducing the “hold-up”
eect, wherein firms face a disincentive to invest in their employees (including training, access to trade secrets,
client lists, etc.) for fear of employees quitting and appropriating the value of their investments before the firm can
recoup the lost investment value (Rubin and Shedd 1981).
51
This type of agreement could increase the probability
46 Department of Justice Antitrust Division and Federal Trade Commission. 2016. “Antitrust Guidance for Human Resource
Professionals.” Last accessed March 2, 2022. https://www.justice.gov/atr/file/903511/download. See also, Federal Trade
Commission. n.d. “Antitrust Red Flags for Employment Practices.” Last accessed March 2, 2022. https://www.c.gov/
system/files/documents/public_statements/992623/c-doj_hr_red_flags.pdf.
47 Department of Justice. 2021. “Health Care Company Indicted for Labor Market Collusion.” Press release 21-14. Last
modified March 4, 2021. https://www.justice.gov/opa/pr/health-care-company-indicted-labor-market-collusion; and
Department of Justice. 2021. “Health Care Staing Company and Executive Indicted for Colluding to Suppress Wages of
School Nurses.” Press release 21-284. Last modified March 30, 2021. https://www.justice.gov/opa/pr/health-care-staing-
company-and-executive-indicted-colluding-suppress-wages-school-nurses.
48 Starr, Evan. 2019. “The Use, Abuse, and Enforceability of Non-Compete and No-Poach Agreements: A Brief Review of the
Theory, Evidence, and Recent Reform Eorts.” Economic Innovation Group. https://eig.org/wp-content/uploads/2019/02/
Non-Competes-Brief.pdf. See also allegation contained in United States v. Adobe Systems Inc., et al., No. 1:10-cv-01629,
2011 U.S. Dist. (March 18, 2011). https://www.justice.gov/atr/case/us-v-adobe-systems-inc-et-al; Starr (2019); and
Abrams, Rachel. “8 Fast-Food Chains Will End ‘No-Poach’ Policies.” New York Times, August 20, 2018. https://www.nytimes.
com/2018/08/20/business/fast-food-wages-no-poach-franchisees.html.
49 Balasubramanian, Natarajan, Evan Starr, and Shotaro Yamaguchi. 2021a. “Bundling Employment Restrictions and Value
Capture from Employees.” SSRN, November 14, 2021. http://dx.doi.org/10.2139/ssrn.3814403. Nondisclosure agreements
are more common than non-compete agreements, but firm survey data suggest at least some employees have non-
compete agreements at approximately two-thirds of firms.
50 McAdams, John M. 2019. "Non-Compete Agreements: A Review of the Literature." SSRN, December 31, 2019. https://dx.doi.
org/10.2139/ssrn.3513639. Many of the papers cited in the following section were drawn from this literature review.
51 Rubin, Paul H., and Peter Shedd. 1981. "Human capital and covenants not to compete." The Journal of Legal Studies 10 (1):
93–110.
THE STATE OF LABOR MARKET COMPETITION
that an employer will be comfortable investing in employee human capital, even if those skills are transferable
to other firms, rather than simply relying on firm-specific training (Becker 1962).
52
Such training can be mutually
beneficial for both the employer and employee.
By design, non-compete agreements limit employees’ outside options, which, in turn, weakens workers’
bargaining power and raises hiring costs for other firms. The limits are typically within a geographic area for
a specific period and within a set of relatively similar occupations or industries but may be much broader.
Balasubramanian (2017) models the eects of non-competes to show how this narrowing of outside options
reduces employee bargaining power relative to their employer.
53
All else equal, this leads to what they call a “lock-
in” eect: lower worker mobility and longer tenure, as well as a flat or declining wage profile.
54
Both the mitigation of the “hold-up” eect and “lock-in” eect mentioned above can reduce worker mobility.
Lower worker mobility increases recruitment costs for all firms as fewer workers are seeking to switch jobs than
otherwise would, absent the post-employment restrictive employment agreement. The increases in recruitment
costs can lead to worse matches between employers and employees, lowering wages and aggregate productivity
(Javanovic 2015).
55
The “hold-up” and “lock-in” eects can coexist. The net eect of these two mechanisms on wages, tenure, and
mobility is theoretically ambiguous since the subset of employees who are aware of being asked to sign non-
compete agreements may demand higher wages in return (i.e., a compensating dierential). Additionally, since
mitigation of the “hold-up” channel can create mutually beneficial investments for both the employee and the
employer, longer tenure does not necessarily imply the employee is worse o.
However, the share of people who negotiate over a non-compete agreement appears to be quite small. Starr,
Prescott, and Bishara (2021) find only about 10 percent of employees negotiate over their non-compete
agreements.56 Therefore, it is unlikely that most employees demand (or receive) a compensating dierential from
signing a non-compete agreement. Furthermore, a worker with little bargaining power (e.g., low-income workers)
or who is unaware they are bound by a non-compete (which may be more likely for less-educated workers) is
unlikely to be able to secure a compensating dierential in exchange for signing a non-compete agreement. To the
extent that a compensating dierential requires an explicit negotiation, certain workers may be less willing or able
to do so—for example, Babcock and Laschever (2009) argue women are much less likely to negotiate during the
hiring process.
57
Accordingly, the share of workers whose wages increase as a result of non-compete agreements is
small.
While one of the main justifications for noncompete agreements (as well as other types of restrictive employment
agreements) is mitigation of the “hold up” eect, there are far less restrictive means of addressing this problem.
For workers with access to genuine trade secrets, there may be overlapping authority with trade secrecy laws,
irrespective of the existence of a noncompete agreement.
58
For the broader workforce, sectoral-based training may
52 Becker, Gary S. 1962. "Investment in human capital: A theoretical analysis." Journal of Political Economy 70 (5): 9–49.
53 Balasubramanian, Natarajan, Jin Woo Chang, Mariko Sakakibara, Jagadeesh Sivadasan, and Evan Starr. 2017 "Locked In?
The Enforceability of Covenants Not to Compete and the Careers of High-Tech Workers." Center for Economic Studies, U.S.
Census Bureau Working Paper CES-17-09. https://www.census.gov/library/working-papers/2017/adrm/ces-wp-17-09.html.
54 A person’s wage profile describes their wages over their lifetime. Typically, a person’s wages increase from their 20s through
their 60s until falling o as people cut back on work hours and transition into retirement. A flatter wage profile means the
increase in wages is slower than otherwise expected, which could have compounding eects on lifetime earnings.
55 Jovanovic, Boyan. 2015. "Matching, Turnover, and Unemployment." Journal of PoliticalEconomy 92 (1): 108–122.
56 Starr, Evan, J.J. Prescott, and Norman D. Bishara. 2021. “Noncompete agreements in the US Labor Force.” The Journal of
Law and Economics 64 (1): 53–84.
57 Babcock, Linda, and Sara Laschever.2009. Women don't ask. Princeton: Princeton University Press.
58 For example, 18 U.S. Code § 1832 criminalizes the of trade secrets (for use or intended for use in interstate or foreign
commerce) by an organization.
16
17
THE STATE OF LABOR MARKET COMPETITION
provide occupation-specific skills to workers without restricting their mobility.
59
These alternative arrangements
have the possibility of meeting a legitimate need of firms (to protect their intellectual property and have access to
skilled employees) without some of the detrimental eects noncompete agreements can have on workers.
Since non-compete agreements increase the bargaining power of employers relative to employees, they
potentially allow employers to capture a larger share of the surplus generated by the employee-employer match.
Johnson and Lipsitz (2020) argues this might be especially true for low-wage workers near the minimum wage
because employers are unable to capture additional surplus from oering lower wages but can nonetheless
benefit from non-compete agreements in other ways.
60
For example, requiring a worker to sign a non-compete
agreement could increase their tenure.
61
Likewise, a non-compete agreement may suiciently limit an employee’s
outside options to flatten their wage-tenure profile (that is, how much their wage goes up over time).
Restrictive employment agreements, including non-compete, non-solicitation, and non-recruitment agreements,
may reduce firm entry. In aggregate, this tends to lead to reduced demand and wage competition, leading to fewer
appealing outside options for similarly situated workers. Samila and Sorenson (2011) find that increases in supply
of venture capital funds has a stronger impact on firm start-ups, patent creation, and employment growth in states
that have weaker enforcement of non-compete agreements, suggesting non-compete agreements may reduce
certain types of entrepreneurial activity.
62
However, Carlino (2017) finds little evidence of this, at least in Michigan.
63
The reduction of firm entry could also reduce innovation and product variety because employees with new ideas
may be restrained from capitalizing on new ideas at their current firm in ways they would not be if they could start
their own business. On the other hand, this result is theoretically ambiguous since firms may be reluctant to invest
in research and development (R&D) if they fear employees can quit and appropriate that research for their own
business.
None of the mechanisms described above necessarily require restrictive employment agreements to be enforced,
or even enforceable, to have tangible labor market eects.
64
While guaranteed enforcement would strengthen
their eects, uncertainty over enforcement can nonetheless aect behavior (“in terrorem” eects). This is true
even if the actual probability of a contract being enforced is zero. So long as the perceived probability of an
employer attempting to enforce the contract is non-zero, restrictive employment agreements can create frictions.
65
Consistent with this, Starr, Prescott, and Bishara (2020) present survey evidence that workers with non-compete
clauses frequently decline job oers because of their preexisting non-compete agreement, even in states that do
not enforce such agreements.
66
Likewise, survey evidence also suggests that the incidence of non-compete clause
59 For an overview, see e.g., Holzer, Harry J. 2022. “Do sectoral training programs work? What the evidence on Project Quest
and Year Up really shows.” Brookings, January 12, 2022. https://www.brookings.edu/research/do-sectoral-training-
programs-work-what-the-evidence-on-project-quest-and-year-up-really-shows/.
60 Johnson, Matthew S., and Michael Lipsitz. 2020. "Why are low-wage workers signing noncompete agreements?"Journal of
Human Resources(May): 0619–10274R2.
61 This is beneficial to employers even in the relative absence of explicit training costs because recruitment costs are non-zero
and on-the-job learning makes high turnover less profitable (all else equal) relative to low turnover.
62 Samila, Sampsa, and Olav Sorenson. 2011. "Venture capital, entrepreneurship, and economic growth." The Review of
Economics and Statistics 93 (1): 338–349. See also Starr, Evan, Natarajan Balasubramanian, and Mariko Sakakibara.
2017. “Screening Spinouts? How Noncompete Enforceability Aects the Creation, Growth, and Survival of New Firms.”
Management Science 64 (2): 552–572.
63 Carlino, Gerald. 2017. “Do Non-Compete Covenants Influence State Startup Activity? Evidence from the Michigan
Experiment." Federal Reserve Bank of Philadelphia Working Paper 17–30.
64 See, e.g., Starr, Prescott, and Bishara (2021).
65 Because lawsuits can be lengthy, expensive, and mentally taxing, a rational employee may conclude it is not worth trying
to switch jobs, even if they are certain they would prevail in court against an attempted enforcement action by their former
employer.
66 Starr, Evan, J.J. Prescott, and Norman Bishara. 2020. “The Behavioral Eects of (Unenforceable) Contracts.”The Journal of
THE STATE OF LABOR MARKET COMPETITION
inclusion in employment contracts is not strongly correlated with enforceability of non-compete agreements,
which could suggest employers include such clauses even when they do not expect them to be enforceable.
67
This
partially occurs because people tend to be risk averse.
68
Therefore, even in places where non-compete contracts
are outlawed, the presence of unenforceable non-compete clauses can have a chilling eect on job-switching. The
eects may be particularly severe for lower-wage workers, who may have limited access to legal counsel.
Mandatory Pre-Dispute Arbitration and Class Action Waivers
Whereas restrictive employment agreements allow employers to limit how their employees can behave following
a separation, mandatory pre-dispute arbitration clauses and class action waivers in employment contracts reduce
the options employees or former employees have within the legal system.
Arbitration is a form of alternative dispute resolution in which a third-party, ostensibly neutral, arbitrator resolves
the dispute instead of the worker being free to bring a lawsuit through the judicial system. The decision of the
arbitrator is binding upon both parties and typically subject to strictly limited subsequent judicial review (i.e., the
substance of the decision is generally not appealable). Mandatory arbitration agreements require any dispute
ordinarily resolved through a judicial proceeding be, instead, addressed by arbitration, even before the worker has
raised any claim that a law has been violated.
Mandatory arbitration agreements are extremely common for non-unionized workers.
69
One recent report
estimated about 56.2 percent of non-union employees, or about 60 million workers, are subject to such
agreements.
70
The share of workers whose employment contracts contain mandatory arbitration procedures has
risen dramatically since the Supreme Court upheld their legality in 1991.
71
Mandatory arbitration is more common among large firms. Nearly two-thirds of workers at firms with at least
1,000 employees are subject to mandatory arbitration clauses. Likewise, mandatory arbitration clauses are more
prevalent in low-wage workplaces and industries disproportionately composed of women and Black workers
(Colvin 2018).
Class action waivers in mandatory arbitration agreements are clauses that bar employees from seeking legal
redress via collective legal action. The legality of such agreements has been strongly contested, but, in 2018, the
Supreme Court ruled that employers could legally require them.
72
Law, Economics, and Organization 36 (3): 633–687.https://doi.org/10.1093/jleo/ewaa018. See also Starr, Evan, J.J. Prescott,
and Norman D. Bishara. 2016. “The in Terrorem Eects of (Unenforceable) Contracts.” University of Michigan Law & Econ
Research Paper 16–032.
67 Prescott, J.J., Norman D. Bishara, and Evan Starr. 2016. “Understanding Noncompetition Agreements: The 2014
Noncompete Survey Project.” Michigan State Law Review 2016 (2): 369–464. Note, the weak correlation between the
inclusion of non-compete agreements and enforceability would also indicate weak salience of the enforceability of
non-compete agreements among employers. This may be especially true among smaller employers who do not have a
professional human resource or legal department to cra employment contracts.
68 For example, suppose a person is indierent between the amenities oered by a competitor relative to their current job. A
risk averse person would likely stay at their current job rather than switch to a new job if they were under a non-compete
agreement, even if they were highly confident (but not certain) that the non-compete clause was unenforceable. Instead,
they would require a premium to account for the possibility that their contract was enforced to their detriment.
69 Unionized employees usually have access to a collectively bargained grievance resolution process that culminates in
binding arbitration with the employee represented by the union.
70 Colvin, Alexander J.S. 2018. “The Growing Use of Mandatory Arbitration.” Economic Policy Institute, April 6, 2018. https://
www.epi.org/publication/the-growing-use-of-mandatory-arbitration-access-to-the-courts-is-now-barred-for-more-than-
60-million-american-workers/.
71 Gilmer v. Interstate/Johnson Lane Corp., No. 90-18, 500 U.S. 20 (1991).
72 Epic Systems Corp. v. Lewis, No. 16-285, 138 S. Ct. 1612 (2018). Justice Gorsuch strongly suggested in his opinion that
18
19
THE STATE OF LABOR MARKET COMPETITION
Proponents of mandatory arbitration generally argue the process is faster and less costly than traditional court
trials. Additionally, firms may find arbitration a less volatile, more private option than jury trials. Opponents of
mandatory arbitration argue arbitrators award smaller awards to employees on average and deprive them of due
process. Furthermore, they argue that arbitration is less transparent than traditional litigation. Not only are most
arbitration decisions non-public, but the mere existence of a decision is also rarely public, reducing awareness and
potential deterrence and compliance eects associated with public results.
73
These information asymmetries allow
firms to exert greater monopsonistic power by introducing additional search frictions for workers who may value
knowing a firm’s prior dispute history with workers (or alternatively, current workers who may update their priors
on the quality of their employee if they learned about disputes).
There is some evidence that employees are more likely to win in arbitration disputes than in court, though the
awards are lower on average.
74
Larger employers appear to win arbitration cases more oen, potentially owing
in part to repeat use of arbitrators that ruled favorably for them in the past.
75
Additionally, since employers are
more likely to be repeat players than employees, arbitrators may have an incentive to favor employers in order to
continue receiving their business.
Due to the lack of quality data on employer arbitration, an empirical analysis of their eect on the labor market is
diicult. However, much like non-disclosure agreements, the opaqueness of arbitration agreements can enable
employers’ continuing bad behavior as disputes and their resolutions are not made public. In this way, they make
it harder for jobseekers to identify the positions that are best suited to them or demand adequate compensation
for working in sub-par conditions, which can have the eect of ineiciently matching employees and employers.
Since class action lawsuits may lower the per-plainti cost of dispute resolution, mandatory arbitration
agreements with class action waivers tend to discourage employee-driven arbitration. This likewise has the eect
of reducing the ability of the dispute resolution system to deter future misconduct.
Occupational Licensing
Occupational licensing is a form of regulation that requires individuals who want to perform certain types of work
to obtain permission from the government.
76
Licensing occurs at all levels of government (federal, state, and local),
but licenses are primarily issued at the state level.
If markets were competitive, quality was freely observable, and poor (or high) quality imposed no negative
(or positive) externalities upon third parties, there would be little justification for occupational licensing.
77
In
such a world, consumers who highly valued quality would easily be able to dierentiate low- and high-quality
providers. Likewise, providers’ wages would be dierentiated based on their quality—with higher-quality workers
commanding greater wages because of their superior skill.
However, quality is typically not easily observable. Nor are the consequences of poor quality always self-evident.
Even when quality is observable, it can be costly to consumers in terms of time and resources to obtain such
information. This creates a moral hazard problem wherein low-quality workers asymmetrically know their quality,
but consumers do not. Low-quality workers, therefore, have an incentive to obfuscate their performance to
Congress had the ability to change the legality of class action waivers in lawsuits via new legislation.
73 Estlund, Cynthia L. 2018. “The Black Hole of Mandatory Arbitration.” North Carolina Law Review 96 (3): 679–710.
74 St. Antoine, Theodore J. 2008. "Mandatory Arbitration: Why It's Better Than It Looks." U. Mich. J. L. Reform 41 (4): 783–812
75 Colvin, Alex, and Mark Gough. 2015. “Individual Employment Rights Arbitration in the United States: Actors and Outcomes.”
Industrial and Labor Relations Review 68 (5): 1019–1042.
76 The focus of this section is occupational licensing, as opposed to certification. The primary dierence between the two is
that licensing involves government power whereas certification is typically done by a private actor, such as a non-profit
trade group.
77 Quality here is conceived of broadly to include safety.
THE STATE OF LABOR MARKET COMPETITION
consumers to extract greater wages than they would in a perfect information environment. Shapiro (1986) shows
how licensing that raises the minimum bar for professionals partially alleviates this moral hazard problem by
excluding the lowest-quality providers. They show that licensing benefits consumers who value high-quality at the
expense of those who do not.
78
Licensing can be welfare-enhancing if provider quality is not easily observable. Implicitly, this highlights how
the strongest theoretical justification for the benefits (to consumers) associated with occupational licensing
occur in occupations where quality meaningfully varies, dierences in quality are diicult to observe, and the
consequence of that variation matters. For example, the potential benefit of occupational licensing is likely higher
in an occupation like medicine (where quality could vary dramatically between providers, a layperson would have
diiculty in distinguishing between a high- and low-quality provider, and the consequences of being provided
poor medical treatment may be large) compared to an occupation like lawn mowing services (where quality may
not dier much, could relatively easily be observable by a lay person, and the consequences of poor service are
unlikely to be severe).
Note, Shapiro (1986) does not consider the possibility of spillover eects of quality. For example, if a low-quality
mechanic poorly fixes a car, that car may break down in the middle of the road. Even if the consumer is willing
to take that risk, a broken car in the middle of the road imposes additional costs on third parties. Likewise, a
low-quality healthcare provider may fail to properly diagnose a communicable disease, thereby increasing the
probability that unrelated third parties are infected (i.e., imposing a negative externality on the third party). In the
presence of such externalities, there is a stronger societal benefit to creating a quality floor.
However, gross benefits do not necessarily imply net benefits to consumers as there are potentially large trade-
os to occupational licensing. Licensing imposes barriers to entry into an occupation. Requirements such as
continuing or additional training and education, fees, exams, and paperwork can reduce labor supplied in the
licensed occupation. Workers who are liquidity constrained may be disproportionately excluded from entering a
licensed occupation if these barriers require large upfront investments, even though such training and education
would be worth it in the long run due to increased productivity.
Whether licensing enhances or reduces welfare depends not only on its impact on consumers, but workers as well.
While benefits of a reduction in labor supply due to licensing may accrue to practitioners in that occupation in the
form of higher wages, some or all of those rents may instead flow to licensing entities.
79
Thus, the economic benefit
to licensed workers is at least theoretically ambiguous, especially if workers must pay to become licensed.
Since most licensing is done at the state-level, dierences in licensing requirements impose inter-state barriers to
workplace mobility. That is, even if a worker benefits from licensure in one state, this can come at an implicit cost
of reduced mobility. Such restrictions to mobility can increase labor market frictions (i.e., require a much higher
oer to induce someone to leave their current work) and reduce search quality (i.e., a place may experience a
shortage of otherwise qualified workers simply because those workers live across state lines).
These restrictions to mobility imposed by occupational licensing can be particularly constraining on two-income
households facing a so-called “two-body problem” wherein partners of the same (target) household with highly
specialized occupations have diiculty in finding suitable work for both partners in the same geographic area. For
example, spouses of military members, who frequently move, may find it diicult to find gainful employment when
their spouse must relocate. Such barriers can exacerbate pre-existing inequities in household dynamics and lead
78 Shapiro, Carl. 1986. “Investment, Moral Hazard, and Occupational Licensing.” The Review of Economic Studies 53 (5): 843–862.
79 Department of the Treasury Oice of Economic Policy, Council of Economic Advisers, and Department of Labor. 2015.
“Occupational Licensing: A Framework for Policymakers.” https://obamawhitehouse.archives.gov/sites/default/files/docs/
licensing_report_final_nonembargo.pdf.
20
21
THE STATE OF LABOR MARKET COMPETITION
to worse average job searches, especially when both workers work in licensed occupations.
80
Once a government entity decides an occupation should be licensed, they must also determine the manner
of licensing. Too lax a licensing policy may not adequately screen out low-quality practitioners. This can harm
consumers who, believing licensing is an implicit governmental endorsement of quality, may unknowingly visit an
under-qualified practitioner. On the other hand, Shapiro (1986) noted licensing may benefit those who value high-
quality services, but it harms those who do not. Setting too high of a requirement to get licensed can overly restrict
the supply of labor to such a degree that very few consumers would benefit.
If quality is diicult to observe for consumers, it may also be diicult to observe for licensing entities. Therefore,
licensing requirements may imperfectly screen for quality, especially when the licensing process is relatively crude.
For example, a common requirement for licensing is to train for a certain number of hours before the worker can
partake in an occupation. During these trainings, which can take months for some occupations, workers are oen
unpaid and may even be required to pay for the training.
As mentioned above, these barriers may be infeasible for individuals with less financial resources, which
disproportionately includes people of color.
81
Furthermore, if licensing involves a professional examination, as it
oen does, those tests may reflect underlying biases of the test makers more than actual quality.
82
Thus, even if
there is a benefit to screening out lower-quality practitioners, there is no guarantee that licensing entities can do
so eectively. Certain types of screening tools may be more eective than others and may thereby avoid some of
the limitations of licensing mentioned above. For example, employer-financed training can reduce the liquidity
constraints imposed by some licensing bodies. Likewise, union apprenticeships, wherein workers work alongside
a professional in preparation for becoming licensed may serve as a better screening mechanism than written tests,
where appropriate.
Skill-Biased Technical Change and Job Polarization
As mentioned above, worker bargaining power depends largely on their unique traits. If a firm can easily replace
a worker’s role in production at a similar cost (i.e., the worker is substitutable), then that worker has minimal
leverage during negotiations. Substitutes may come in dierent forms—for example, an equally qualified worker
who would accept the same job at the same wage or perhaps a machine or computer that can do the same work at
a similar or lower cost.
80 For example, ex-ante dierences in gender pay gaps (due to discrimination or otherwise) can be amplified because a
household seeking to maximize household earnings may elect to move to a state if the higher-earning member receives
a suiciently large pay increase from moving, even if the lower-earning member’s income suers. For a modeling
example, see, e.g., Rueda, Valeria, and Guillaume Wilemme. 2021. "Career Paths with a Two-Body Problem: Occupational
Specialization and Geographic Mobility." Upjohn Institute for Employment Research Working Paper 21-346.https://doi.
org/10.17848/wp21-346.
81 Even if capital markets allowed workers to borrow against their expected future earnings, most people are risk averse.
This risk aversion may make them hesitate to take on debt to finance training in an occupation with uncertain returns. The
net result remains the same: workers with fewer means are more likely to be screened out despite their underlying ability
relative to workers with greater means.
82 A test may poorly screen for quality, even if it is standardized. For example, the Scholastic Aptitude Test (SAT) is meant to
screen for college readiness, but it has long been recognized that it poorly screens students of color disproportionately
(see, e.g., Freedle, Roy. 2003. "Correcting the SAT's ethnic and social-class bias: A method for re-estimating SAT scores."
Harvard Educational Review 73 (1): 1-43.). Some evidence also suggests the SAT is a better predictor of family income than
college readiness (see, e.g., Goldfarb, Zachary A. 2014. “These four charts show how the SAT favors rich, educated families.”
Washington Post, March 5, 2014. https://www.washingtonpost.com/news/wonk/wp/2014/03/05/these-four-charts-show-
how-the-sat-favors-the-rich-educated-families/.).
THE STATE OF LABOR MARKET COMPETITION
As technology changes to develop better substitutes for lower-paid workers, workers see their bargaining
positions deteriorate relative to the firm. Whereas a cashier might once have been an indispensable employee
at a supermarket or fast-food restaurant, viable substitutes are now available. Intuitively, this limits the worker’s
bargaining power: if wages grow high enough, the employer may rather pay for kiosks than cashiers.
Many tasks once done by humans are now done by machines. “Skill-biased technical change” refers to changes
in technology or production that replace (or substitute) unskilled labor in favor of skilled labor since technology is
complementary to skilled labor.
83
This process has especially disrupted routine-based work (where automation is
easiest to implement) in occupations with relatively high-paying jobs. This has led to what some economists refer
to as job polarization, wherein the labor market is ever more segmented into a low-skilled, low-wage sector and
a high-skilled, high-wage sector. This process has contributed to both changes in the marginal product of labor
(which would lead to wage divergence under conditions of perfect competition) but also likely had dierential
impacts on bargaining power across the income distribution.
In this framework, the result is a low-wage sector is characterized by jobs that are not easily replaced by
technology (e.g., line cook), while the high-wage sector is characterized by jobs that are complementary to
technological advances (e.g. accountants utilizing spreadsheets to tackle more work in a day).
84
The term
“polarization” comes from the hypothesis that technology has replaced middle-skilled, middle-wage jobs (e.g.,
the cashiers mentioned above).
85
That said, both the existence of job polarization (especially aer the 1990s)
and its impact on income inequality remains hotly debated. For example, Michel et al. (2013) argue that the job
polarization found in Acemoglu and Autor (2011) is highly sensitive to measurement error problems, choice of
sample period, and empirical design.
86
Although work pertaining to skill-biased technical change originally focused on the role of education, recent
work by Acemoglu and Autor (2011) and others have focused more on the role of tasks, with machine automation
primarily able to replace routine non-cognitive based tasks.
87
Acemoglu (2020) built on this framework by modeling
not only tasks that are eectively automated away from humans, but also modeling new task formation that
flows from automation of older tasks.
88
In their model, the destruction of tasks via automation tends to increase
income inequality, but the creation of new tasks resulting from automation has an ambiguous impact on income
inequality.
83 This pattern of substitutability and complementarity does not always hold true. Examples of the converse pattern include
the power loom during the Industrial Revolution and GPS technology, which substitutes for a detailed knowledge of local
geography and traic routes.
84 Goos, Maarten, Alan Manning, and Anna Salomons. 2014. "Explaining job polarization: Routine-biased technological
change and oshoring." American Economic Review 104 (8): 2509–26.
85 Id.
86 Mishel, Lawrence, Heidi Shierholz, and John Schmitt. 2013. "Don’t blame the robots. Assessing the job polarization
explanation of growing wage inequality."Economic Policy Institute and Center for Economic Policy and Research Institute
working paper. See also Hunt, Jennifer, and Ryan Nunn.2019. “Is Employment Polarization Informative About Wage
Inequality and Is Employment Really Polarizing?” National Bureau of Economic Research Working Paper 26064.
87 Acemoglu, Daron, and David Autor. 2011. "Skills, tasks and technologies: Implications for employment and
earnings."Handbook of Labor Economics, edited by Orley Ashenfelter and David Card, vol. 4 (Part B), 1043–1171. Elsevier.
See also Frey, Carl Benedikt, and Michael A. Osborne. 2017. "The future of employment: How susceptible are jobs to
computerisation?"Technological Forecasting and Social Change114 (January): 254-280.
88 Acemoglu, Daron, and Pascual Restrepo. 2020. “Unpacking Skill Bias: Automation and New Tasks.” American Economic
Association Papers and Proceedings 110 (May): 356–361.
22
23
THE STATE OF LABOR MARKET COMPETITION
LABOR MARKET POWER & COMPETITION: EMPIRICAL EVIDENCE
Having discussed theories of labor market power and related issues, we now turn to the data. Depending on the
reader’s perspective, several dierent questions addressed in this section might be deemed ‘most important.’
Among those questions: how large are wage losses stemming from monopsonic power on average? Have those
losses increased or decreased over time? What are the sources of monopsony power, and how do employers exert
it in practice?
First, we address the question of causality: does labor market power exist, and does it suppress wages? We find
convincing evidence that both questions can be answered in the airmative. Further, we argue that evidence
suggests that this power derives more from labor market frictions than from market frictions. Second, we address
the scale of labor market power—on average, how large are the compensation losses which stem from it? We
argue that the highest quality estimates suggest wage losses of 15 percent, at minimum. Finally, we address the
incomplete evidence on time-trends in labor market power, as well as discussing some alternate perspectives on
the source of labor market power.
Does Labor Market Power Suppress Wages, in Practice?
Although theory predicts that labor market power will harm workers, the sources of labor market power oen
coincide with other market factors that might explain lower wages. For example, small rural communities with
a single large factory have both a single dominant employer (the factory) and low costs of living, which can also
partially explain low wages. Recent research has nevertheless demonstrated that labor market power causes lower
wages, though it is not the sole contributing factors. One set of papers, discussed in later sections, argues that
estimates of separation elasticities (how much workers respond to wage changes by separating with or joining a
firm) directly imply labor market power, a viewpoint which is consistent with the theory discussed above. However,
we focus on event-studies to directly address the question of the causal impact of labor market power on wages.
Prager and Schmitt (2021) oer some of the most compelling and nuanced evidence to address this question,
although the paper’s scope is restricted to hospital employment.
89
The paper studies the eect of employer
labor market power by examining the evolution of wages and employment following hospital mergers- mergers
that represent a potential source of increased labor market power. The empirical strategy is a “dierence-in-
dierences” framework, which compares changes in markets with one hospital merger from 2000 to 2010 to the
changes in markets without mergers during those years. In summarizing the paper, the authors write, “We find
evidence of wage slowdowns, but only following mergers that induce large increases in employer concentration,
and only for workers whose skills are industry specific.”
We highlight two findings from Prager and Schmitt (2021). First, it observes wage losses only in those hospital
occupations where skills are industry-specific (e.g., doctors, but not cafeteria workers), but only when market
concentration substantially increases. There are no detectable wage eects of mergers that only mildly increase
employer concentration, but the study does find evidence of slower wage growth following mergers that
meaningfully increase concentration. Among the most substantial mergers, the paper estimates a reduction in
annual wage growth of between 1.0 and 1.7 percentage points for workers with hospital-specific skills, roughly
one-quarter of these occupations’ typical wage growth rates. However, detectable wage slowdowns from hospital
consolidation are limited to occupations with health-care specific skills, even for the most substantial mergers. For
non-health-care specific occupations, those mergers have a less meaningful impact on the number of potential
employers and market concentration—leading to lesser or null wage eects. This suggests that occupational-level
89 Prager, Elena, and Matt Schmitt. 2021. “Employer Consolidation and Wages: Evidence from Hospitals.” American Economic
Review 111 (2): 397–427. This paper’s results are discussed in further depth below, in the Industry Examples subsection.
THE STATE OF LABOR MARKET COMPETITION
markets are more relevant than industry-level markets when analyzing labor market power, a suggestion that is
echoed in related papers.
Secondly, insofar as these mergers had detectable employment eects, they were positive.
90
This finding
carries particular importance since it is inconsistent with the classical theory of monopsony power, where the
monopsonist reduces wages by constricting labor demand, thereby decreasing the number of employees.
However, it is consistent with a search and matching framework of market power, which does not require a
decrease in jobs. Instead, this finding is consistent with a search and matching explanation for market power,
where frictions in the labor market shield employers from competition for workers, resulting in sub-competitive
wages. As mergers leave fewer potential employers, the employee believes that the benefits of job search are
lower, so they put less eort into their search.
Prager and Schmitt (2021) is useful for this report’s purposes, as it both (a) convincingly establishes a causal link
from mergers to increased labor market power, and (b) furnishes evidence that search and matching is the most
relevant framework for understanding monopsonistic power. Other recent papers estimate the wage eects of
mergers across a broader range of industries, showing that wage suppression as a result of labor market power
is not unique to the health care industry. Notably, Arnold (2021) finds similar eects across a wider range of
industries, along with a higher rate of job departures from recently merged employers (the data do not allow an
analysis of whether this is due to downsizing, quits, or other mechanisms).
91
However, the wages lost over the
course of this study are not meant to be estimates of the current level of average wage loss in the U.S. economy. We
next turn to papers more suited to estimate those wage losses.
The Extent of Wage Losses due to Labor Market Power
How large are wages losses stemming from the exercise of monopsonist power on average? Before turning to
the empirical estimates, it is worth restating a definition for “wage losses.”. The “loss” is relative to the wage in a
perfectly competitive and frictionless environment where workers would be paid a wage equal to the “marginal
revenue product of labor” (MRP
L
). Though a technical term, MRP
L
reflects a relatively simple idea. If a firm adds one
more worker, it can produce a little more of its product. When the firm sells that extra product, the total revenue
from that sale is the MRP
L
. Put dierently: a worker’s MRP
L
equals the revenue their employer would lose if they
were to quit.
Like with any complex question, studies oer a range of estimates regarding these wage losses. Among recent
empirical work, Yeh, Macaluso, and Hershbein (2022) estimate that workers at the average manufacturing plant
earn 65 percent of their MRP
L
, or 65 cents of every dollar they produce.
92
This is at the higher end of estimates
among our selected studies, yet it has plenty of supporting evidence. The paper adopts a direct approach
to estimating wage losses, marshalling detailed, plant-level Census data to do so. This is no small feat: due
to considerable technical hurdles, nearly all other eorts to estimate wage loss infer the values indirectly
by connecting wage loss to theoretically related statistics. One drawback to the paper is its industrial focus:
extrapolating the Yeh, Macaluso, and Hershbein (2022) estimate to non-manufacturing sectors is unwarranted,
therefore we do not say this is an economy-wide estimate. However, it remains a credible estimate pertaining to a
crucial sector of the U.S. economy.
The bulk of our selected studies estimate average wage losses to be on the order of 15–25 cents on the dollar
90 This positive estimate may well reflect pre-existing trends, rather than an actual eect. Aer including a linear time-trend in
their estimated regression, the employment eects are no longer statistically significant.
91 Arnold, David. 2021. "Mergers and acquisitions, local labor market concentration, and worker outcomes." Working Paper.
https://darnold199.github.io/madra.pdf.
92 Yeh, Chen, Claudia Macaluso, and Brad Hershbein. 2022. “Monopsony in the U.S. Labor Market.” American Economic Review,
Forthcoming: https://www.dropbox.com/s/3qpxons17tuk044/monopsony_dra_January2022.pdf?dl=0.
24
25
THE STATE OF LABOR MARKET COMPETITION
(alternately, workers earn between 75 and 85 cents for each dollar of value produced). Notable papers that
estimate wage losses in this range include Berger, Herkenho, and Mongey (2021), who study how competing firms
respond to changes in state taxes, leading to estimates of the scale of monopsony power in local labor markets.
93
This paper’s estimates suggest an average wage loss of 24 cents per dollar produced. Crucially, workers do not
suer a full 24 percent loss of welfare due to labor market power—a variety of mitigating factors lead to a still-
substantial average lifetime welfare loss of 4–9 percent.
Another estimate in this range comes from Bassier, Dube, and Naidu (2021), who study worker responses to
changes in firm-wide wage policies.
94
Their estimate of average wage loss is 19 cents on the dollar. This paper’s
estimates suggest that wage loss due to monopsony power is larger for lower-paid workers—the estimated loss for
the bottom quartile of wages is 26 cents on the dollar.
On the lower end of the spectrum, Azar, Berry, and Marinescu (2019) estimate wage losses of 15 cents on the
dollar.
95
Focusing on worker preferences between firms – rather than search frictions – Lamadon et al. (2022) also
find wage losses on the order of 15 cents on the dollar.
96
Notably, this paper supports the view that across-firm
dierences in non-pecuniary amenities are both a potential result of labor market power, and a potential source of
that power. Kro et al. (2021) arrive at a similar estimate.
97
Among our selected studies, this is the lower bound of
wage losses, meaning we believe the best available empirical evidence suggests that labor market power reduces
wages by at least 15 percent.
Changes in Labor Market Power and Concentration over Time
Whether labor market power has increased or decreased over the past 50 years remains an unresolved question.
Although concentration and market power are not necessarily linked, as argued throughout this report, we do
have stronger evidence regarding the trend in labor market concentration. Measured at the national level, the
concentration of employers in the labor market has increased since the 1980s. However, at the local level, which
is the relevant level for most workers, concentration has consistently decreased over that timeframe (Rinz 2018).
98
From the late 1970s through 2015, the average local labor market Herfindahl-Hirschman Index (HHI) fell by nearly
0.06 (equivalently 600 points).
99
Nevertheless, concentration remains high. Rinz (2018) finds the average concentration of local labor markets to be
around 1,500, the threshold at which DOJ may intervene to block a merger in goods markets. Using job postings
from a private jobs website, Azar, Marinescu and Steinbaum (2020) calculate an average HHI of 3,157.
100
Many other
93 Berger, David, Kyle Herkenho, and Simon Mongey. 2021. “Labor Market Power.” National Bureau of Economic Research
Working Paper 25719.
94 Bassier, Ihsaan, Arindrajit Dube, and Suresh Naidu. 2021. “Monopsony in Movers: The Elasticity of Labor Supply to Firm
Wage Policies.” The Journal of Human Resources, forthcoming.
95 Azar, José, Steven Berry, and Ioana Elena Marinescu. 2019. “Estimating Labor Market Power.” SSRN, September 18, 2019.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3456277.
96 Lamadon, Thibaut, Magne Mogstad, and Bradley Setzler. 2022. “Imperfect Competition, Compensating Dierentials,
and Rent Sharing in the US Labor Market.” American Economic Review 112 (1): 169–212. https://www.aeaweb.org/
articles?id=10.1257/aer.20190790
97
Kro, Kory, Yao Luo, Magne Mogstad, and Bradley Setzler. 2021. “Imperfect Competition and Rents in Labor and Product Markets:
The Case of the Construction Industry.” Working Paper. https://www.bradleysetzler.com/files/Kro-Luo-Mogstad-Setzler.pdf.
98 Rinz, Kevin. 2018. “Labor Market Concentration, Earnings Inequality, and Earnings Mobility.” Center for Administrative Records
Research and Applications, U.S. Census Bureau Working Paper 2018-10.
99 The HHI is defined as the sum of squared-market shares, for some defined market. Higher values of HHI indicate greater
market concentration.
100 Azar, José, Ioana Marinescu, and Marshall Steinbaum. 2020. “Labor Market Concentration.” Journal of Human Resources
(May): 1218–9914R1.
THE STATE OF LABOR MARKET COMPETITION
studies come to similar results, generally finding that wages are negatively correlated with concentration.
101
With market power, firms hire fewer workers than they would in a competitive environment. This reduction in
employment is more than a curiosity: it informs how we should measure the existence and extent of monopsony
power. In particular, it means that the exercise of market power decreases market concentration relative to a
competitive environment, if larger firms tend to have greater market power. This should give some pause to using
labor market concentration as a direct measure of market power. Theoretically, the markdown measures market
power most accurately, a point we return to in the empirical section. Unfortunately, markdowns are diicult to
measure.
On a related topic, this observation forms the theoretical foundation of how minimum wages can increase
aggregate employment.
102
Under monopsony’s lower wages, the economy sees fewer jobs than in competitive
equilibrium since lower wages mean fewer workers willing to accept jobs. Insofar as a minimum wage does not
exceed the competitive wage, it increases employment: more workers will accept employment at the increased
wage, while firms still find it profitable to employ all the willing workers.
Decreasing concentration does not necessarily mean increasing labor market competition: the relationship between
concentration and labor market power is theoretically ambiguous.
103
Indeed, many recent papers on the subject take
pains to point this out, including Yeh, Macaluso, and Hershbein (2022); Berger, Herkenho, and Mongey (2021); and
Bassier, Dube, and Naidu (2021). For example, Berger, Herkenho, and Mongey (2021) estimate that labor market
power has decreased over that time frame, thereby increasing labor’s share of income by 4 percentage points from
1977 to 2013. On the other hand, Yeh, Macaluso, and Hershbein (2022) argue that labor market power decreased from
1977 to 2002, then quickly rose over the ensuing decade. Figure 2 illustrates this secular trend.
Labor Market Power in U.S. Manufacturing
Markdown Indicies, Relative to 1977
Figure 2 - Labor Market Power in Manufacturing, Measured by Wage Markdowns (Yeh, Macaluso, and Hershbein (2022))
101 See, e.g., Benmelech, Efraim, Nittai K. Bergman, and Hyunseob Kim. 2020. "Strong employers and weak employees: How
does employer concentration aect wages?"Journal of Human Resources(December): 0119–10007R1.
102 While this white paper does not explicitly address the economics of minimum wages, questions of labor market power are
important subtext in the discussion of minimum wages and its potential dis-employment eects.
103 For an overview of the theory, see Syverson, Chad. 2019. “Macroeconomics and Market Power: Context, Implications, and
Open Questions.” Journal of Economic Perspectives 33 (3): 23–43.
26
27
THE STATE OF LABOR MARKET COMPETITION
Also illustrated by Figure 2 is the importance of index choice and aggregation method.
104
When we discuss
average labor market power at a national level, we are ultimately summing up the positions of many firms and
establishments into a single statistic. From the firm’s perspective, the plot illustrates that manufacturing labor
market power fell from 1977 to 2002, then increased back to roughly 1970s levels over the subsequent decade.
The same was not true from the manufacturing workers’ perspective, reflected by market-level aggregation (we
typically assume that the worker searches within a market, though that is not strictly true). From that perspective,
markdowns also fell through 2002, but then grew quickly over the past decade, well beyond the levels of the late
1970s.
Note, importantly, that this estimated increase in market power over the last decade was not associated with
an increase in concentration. In contrast to the market-level measure of markdown, local concentration in
manufacturing labor markets declined since 1977 and remained below the 1977 level all the way through 2012.
This observation, combined with observations in the other papers highlighted in this section, suggest that labor
market concentration is a flawed proxy for labor market power.
Alternative Perspectives on Market Concentration and Labor Market Power
The previous section featured papers arguing that labor market concentration and labor market power are
not necessarily correlated. However, a handful of recent studies have focused on concentration as not only an
indicator of market power, but also a cause of it. In a classical monopsony or oligopsony model, some degree of
concentration is a prerequisite for market power.
105
For example, Azar, Marinescu, and Steinbaum (2020) measure
concentration in local labor markets using data from postings on CareerBuilder.com, estimating that moving from
the twenty-fih to the seventy-fih percentile of concentration within U.S. local labor markets results in a 5–17
percent decrease in posted wages. Acknowledging that a correlation between concentration and posted wages
could be a confounded by productivity dierences, the paper uses an “instrument” for market concentration (a
common econometric strategy to address these kinds of concerns) of the inverse number of employers that make
job postings in the same occupation and quarter, but in dierent geographic markets. The crucial assumptions
are (a) occupation-level concentration in other geographic areas is correlated with local concentration, but (b) not
associated with local occupational wage postings in any other way. If workers commonly look outside their own
geographic area for a job, for example, then the second assumption would be violated.
Focusing on the manufacturing sector, Benmelech, Bergman, and Kim (2020) find that increasing local labor
market concentration from one standard deviation below the national mean to one standard deviation above the
national mean decreases wages between 9.1 percent and 14.4 percent. Notably, they also find that unionization,
which provides workers with countervailing market power, decrease how responsive wages are to local labor
market concentration by between 29 percent and 45 percent.
This white paper has argued that frictions are a more important source of labor market power than concentration.
However, it is important to stress that the two sources are not mutually exclusive. Evidence for one mechanism is
not necessarily evidence against the other.
104 Some notes on interpreting Figure 2: each of the three series are indexed to 1 in 1977, meaning that all points are relative
to that year. For example, the red “Market-Level” series for markdowns is roughly 1.1 in 2012, which can be interpreted
as markdowns that are 10 percent greater than they were in 1977. Only changes can be inferred from the figure itself; the
figure says nothing about the level of markdowns at any point.
105 In general, it is not true that concentration implies market power. Concentration is consistent with a competitive market
featuring dierences in productivity. In that context, the most productive firms are the largest employers, and this
allocation is eicient—any reallocation of workers would reduce wages.
THE STATE OF LABOR MARKET COMPETITION
Restrictive Employment Agreements
Both the exposure to and the eect of non-compete agreements and other types of post-employment agreements
dier by state, occupation, and workplace status (e.g., entry-level vs executive).
Twenty-one percent of workers in the top income quintile are covered by a non-compete agreement compared
to eight percent of workers in the bottom quintile of hourly wages.
106
However, this still leaves millions of workers
with minimal employer-specific training subject to non-compete agreements (Starr, Prescott, and Bishara 2021).
Top executives may be even more responsive to non-compete agreements. Garmaise (2011) finds that top
executives were 47 percent less likely to change jobs within industries as non-competes became more strictly
enforced and their tenure also increased by about 16 percent.
107
Additionally, Kini, Williams, and Yin (2021) show
that initial CEO compensation is higher when enforceability of non-competes is higher, suggesting CEOs demand
a compensating dierential in exchange for signing non-compete agreements.
108
The greater responsiveness of
compensation to noncompete agreements of top executives compared to lower-wage workers could be due to a
number of factors, including that top executives may be more likely to face increased coverage by a non-compete
agreement, a bigger relative loss in wages when switching jobs, and higher odds of enforcement of a non-compete
agreement.
109
Unlike higher income workers, lower wage workers likely lack suicient bargaining power to refuse a non-compete
agreement. As a result, whereas non-compete agreements may increase top-earner wages at the expense of
mobility, non-compete agreements appear to reduce both wages and mobility for lower-income earners. For
example, Lipsitz and Starr (2021) find that the ban on non-compete agreements for hourly workers (who tend
to be lower income) in Oregon increased overall hourly wages by 2–3 percent, with a stronger eect for female
workers.
110
Johnson, Lavetti, and Lipsitz (2021) likewise find stronger eects from enforcement of non-compete
agreements on income of women and people of color.
111
Young (2021) finds that a ban on non-compete clauses
for low-to-medium income workers in Austria modestly increased worker’s annual job-to-job mobility rate (a 0.27
percentage point increase against a base rate of 16 percent).
112
Non-compete agreements exist across occupations broadly, though their prevalence varies. For example, non-
compete agreements are relatively rare in agricultural occupations compared with sales and management related
occupations (Boesch, Lim, and Nunn 2021, fn. 1). Furthermore, employers with multiple locations are more likely to
have non-compete agreements (id.).
Balasubramanian, Starr, and Yamaguchi (2021b) show that employers oen bundle post-employment restrictive
covenants, which in addition to non-compete agreement include non-disclosure agreements, non-solicitation
106 Boesch, Tyler, Katherine Lim, and Ryan Nunn. 2021. “Non-compete contracts sideline low-wage workers.” Federal Reserve
Bank of Minneapolis, October 15, 2021. https://www.minneapolisfed.org/article/2021/non-compete-contracts-sideline-
low-wage-workers.
107 Garmaise, Mark J. 2011. "Ties that truly bind: Noncompetition agreements, executive compensation, and firm
investment." The Journal of Law, Economics, and Organization 27 (2): 376–425.
108 Kini, Omesh, Ryan Williams, and Sirui Yin. 2021. "CEO noncompete agreements, job risk, and compensation." The Review
of Financial Studies 34 (10): 4701–4744.
109 Id.
110 Lipsitz, Michael, and Evan Starr. 2021. “Low-Wage Workers and the Enforceability of Noncompete Agreements.”
Management Science 68 (1): 143–170. https://doi.org/10.1287/mnsc.2020.3918.
111 Johnson, Matthew, Kurt Lavetti, and Michael Lipsitz. 2020. “The Labor Market Eects of Legal Restrictions on Worker
Mobility.” SSRN, June 6, 2020.https://ssrn.com/abstract=3455381.
112 Young, Samuel G. 2021. “Noncompete Clauses, Job Mobility, and Job Quality: Evidence from a Low-Earning Noncompete
Ban in Austria.” SSRN, July 5, 2021. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3811459.
28
29
THE STATE OF LABOR MARKET COMPETITION
agreements, and non-recruitment agreements.
113
Consistent with previous studies, they find that below-median
income workers are more likely to be covered by none of these agreements compared to higher-income workers.
However, they also find should there be any post-employment restriction covenants low-income are about equally
likely as high-income workers to face the full bundle of restrictions. They suggest their estimates are consistent
with pure value capture (related to the “lock-in” eect mentioned above) being the dominant reason for bundling
agreements for average workers, whereas value creation (related to the “hold-up” eect mentioned above) is a
primary reason for top executives, like CEOs.
One type of restrictive employment agreement, the non-disclosure agreement (NDA), has garnered attention
recently. In the wake of the #MeToo movement, it was anecdotally argued that NDAs led to underreporting of
unlawful conduct resulting from fears of retaliation and lawsuits over breaching these agreements.
114
Sockin,
Sojourner, and Starr (2021b) show that changes in laws in three states (California, Illinois, and New Jersey), which
prohibited firms from using NDAs to restrict workers from sharing information about unlawful conduct, led to an
increase in negative reviews (5 percentage points greater share) on Glassdoor, especially pertaining to workplace
harassment (22 percent increase).
115
The authors argue that “by preventing outsiders from learning about
undesirable firm employment practices, over-broad NDAs impose potential negative externalities on job seekers
and competitor firms.”
Starr, Prescott, and Bishara (2021) find that the huge number of low-skill workers subject to non-competes
suggests that employers routinely apply them to workers who do not possess trade secrets or customer lists and
are not given specialized training. They cite as an example a large sandwich chain, which subjected its workers to
extremely broad non-competes. Though these non-competes are not likely enforceable under state law, they point
out that they may have an in terrorem eect that deters employees from obtaining jobs at competing employers.
Trends in and Effects of Occupational Licensing
The incidence of occupational licensing has grown dramatically since the 1950s, from about 5 percent to around 20
percent of workers by the mid-2010s.
116
In 2016, Treasury’s Oice of Economic Policy, in collaboration with the Council of Economic Advisers (CEA)
and DOL, released an extensive report documenting the eects of occupational licensing on labor markets
(Department of Treasury, Council of Economic Advisers, and Department of Labor 2015). The report, hereaer
referred to as UST (2016), examined dozens of studies on the eects of occupational licensing both broadly and
within specific industries.
UST (2016) found little evidence that marginal changes in occupational licensing typically increase quality,
113 Balasubramanian, Natarajan, Evan Starr, and Shotaro Yamaguchi. 2021b. "Bundling Postemployment Restrictive
Covenants: When, Why, and How It Matters." Economic Perspectives on Employment & Labor Law Journal (March).
Specifically, they look at non-disclosure, non-solicitation, non-recruitment, and non-compete agreements. Among these,
they find the most common clause that people are aware of is the non-disclosure agreement.
114 Sockin, Jason, Aaron Sojourner, and Evan Starr. 2021a. “What happens when states limit nondisclosure agreements?
Employees start to dish.” Washington Post, October 4, 2021. https://www.washingtonpost.com/outlook/2021/10/04/non-
disclosure-employee-reviews-study/.
115 Sockin, Jason, Aaron Sojourner, and Evan Starr. 2021b. “Externalities from Silence: Non-Disclosure Agreements
Distort Firm Reputation.” Institute of Labor Economics Working Paper. https://conference.iza.org/conference_files/
LaborMarkets_2021/sockin_j28322.pdf.
116 Kleiner, Morris M., and Alan B. Krueger. 2013. "Analyzing the extent and influence of occupational licensing on the labor
market." Journal of Labor Economics 31 (S1): S173–S202; and Kleiner, Morris M., and Evgeny S. Vorotnikov. 2018. “At What
Cost? State and National Estimates of the Economics Costs of Occupational Licensing.” Institute for Justice. https://ij.org/
wp-content/uploads/2018/11/Licensure_Report_WEB.pdf.
THE STATE OF LABOR MARKET COMPETITION
safety, or health. Evidence since then tends to corroborate these findings. For example, Kleiner et al. (2016) find
that “when nurse practitioners have more independence in their scope of practice, their wages are higher but
physicians’ wages are lower, which suggests some substitution between the occupations. Our analysis of insurance
claims data shows that more rigid regulations increase the price of a well-child visit by 3–16 percent. However, we
find no evidence that the changes in regulatory policy are reflected in outcomes that might be connected to the
quality and safety of health services.”
117
Bowblis and Smith (2021) study a federal staing provision that requires
skilled nursing facilities of a certain size to employ licensed social workers and find no evidence that the increase in
licensure improves patient care quality, patient quality of life, or quality of social services provided.
118
Meehan and
Stephenson (2020) find that changes in the number of hours of education required to become a certified public
accountant (CPA) from 150 hours to 120 hours did little to change pass rates or scores on the CPA exam, suggesting
the extra hours required had little impact on quality. However, the marginal changes do not necessarily correlate
to the overall eect of licensing—some degree of licensing may be welfare enhancing even if a study finds that
marginal changes to occupational licensing requirements reduces welfare.
119
Even if licensing does not objectively increase quality, the perception that it increases quality may nonetheless
impact market outcomes (e.g., price). However, it is unclear whether consumers notice or place much value on
licensure, especially when other methods for determining quality are available. For example, Farronato et al. (2020)
study a large online platform for residential home services and find that consumers are unresponsive to platform-
verified licensing status relative to review ratings and price. This suggests that consumers consider reviews from
other customers a better signal of quality than licensing (or at least verification of licensing).
120
Occupational licensing appears to restrict labor supply in some licensed professions (UST 2016). In some contexts,
licensing can disproportionately limit the labor supply for subsets of socioeconomically disadvantaged workers.
For example, Federman, Harrington, and Krynski (2006) find that state licensing requirements that require
proficiency in the English language tend to reduce the number of Vietnamese-American manicurists.
121
Cathles,
Harrington, and Krynski (2010) find that licensing laws requiring funeral directors to also be embalmers tended
to reduce the share of female funeral directors.
122
These disproportionate impacts on labor supply highlight how
the manner of licensing requirements (i.e., inclusion of English language requirements), not just the intensity of
licensing (e.g., required number of hours), can aect equity considerations. That said, evidence from Blair and
Chung (2018) suggests that occupational licensing may reduce prospective employers’ reliance on race and gender
during the hiring process, suggesting licensing can reduce racial and gender inequities in certain contexts.
123
117 Kleiner, Morris M., Allison Marier, Kyoung Won Park, and Coady Wing. 2016. "Relaxing Occupational Licensing
Requirements: Analyzing Wages and Prices for a Medical Service." The Journal of Law and Economics 59 (2): 261–291.
118 Bowblis, John R., and Austin C. Smith. 2021. "Occupational Licensing of Social services and Nursing Home Quality: A
Regression Discontinuity Approach." ILR Review 74 (1): 199–223.
119 For example, Meehan and Stephenson (2020) only identify the eects of a change in intensity (from 150 hours to 120
hours). Meehan, Brian, and E. Frank Stephenson. 2020. “Reducing a Barrier to Entry: The 120/150 CPA Licensing Rule.”
Journal of Labor Research 41 (December): 382–402. These studies cannot speak to the overall eects of occupational
licensing because requiring CPAs to be licensed may increase overall quality of CPAs, even if a reduction in the hours
required to obtain a CPA does not reduce quality. For example, this could be the case if 60 hours was suicient to screen
out unqualified candidates.
120 Farronato, Chiara, Andrey Fradkin, Bradley Larsen, and Erik Brynjolfsson. 2020. “Consumer Protection in an Online World:
An Analysis of Occupational Licensing.” National Bureau of Economic Research Working Paper 26601.
121 Federman, Maya N., David E. Harrington, and Kathy J. Krynski. 2006. “The Impact of State Licensing Regulations on Low-
Skilled Immigrants: The Case of Vietnamese Manicurists.” American Economic Review 96 (2): 237–241.
122 Cathles, Alison, David E. Harrington, and Kathy Krynski. 2010. “The Gender Gap in Funeral Directors: Burying Women with
Ready-to-Embalm Laws?” British Journal of Industrial Relations 48 (4): 688–705.
123 Blair, Peter Q., and Bobby W. Chung. 2018.“Job Market Signaling through Occupational Licensing.” National Bureau
of Economic Research Working Paper 24791. Specifically, they argue one of the main channels for this eect is that
30
31
THE STATE OF LABOR MARKET COMPETITION
Determining the impact of occupational licensing on wages is diicult. Though a restricted supply of labor can
increase wages for those who become licensed, if the most skilled workers are more likely to become licensed, they
may have earned more than their unlicensed counterparts even without becoming licensed. UST (2016) found the
size of the wage gap attributable to occupational licensing is sensitive to modeling choices. Studies that do not
control for underlying dierences (e.g., in educational attainment) between licensed and unlicensed workers tend
to find a large wage gap—on the order of 10–25 percent. However, studies that control for underlying dierences
typically find more modest eects of licensing on wages.
Variations in licensing requirements across states may discourage mobility and suppress the wages of licensed
workers. However, UST (2016) analysis using 2011 Survey of Income and Program Participation (SIPP) data found
weak evidence that licensed workers are less likely than unlicensed workers to move between states. Johnson
and Kleiner (2020) find stronger evidence of occupational licensing as a barrier to interstate migration.
124
They find
that the interstate migration rate for occupations with state-specific licensing exams are about a third lower than
other occupations. Importantly, they do not find similar results for occupations with national exams, highlighting
how synchronizing requirements and examinations can reduce mobility barriers created by licensing.
125
That said,
Johnson and Kleiner (2020) find that increases in occupational licensing only account for a very small share (about
2.5 percent) of the decline in interstate migration since 1980.
The impact of licensing on the prices of goods and services is clearer. In nine of the eleven studies UST (2016)
examined, more restrictive occupational licensing increased prices.
126
This eect increases earnings for licensed
workers at the expense of shutting some workers out of an occupation altogether. But the exact impact of licensing
on prices varies by occupation or even within individual studies of the same occupation. For example, Kleiner et
al. (2016)’s results imply that restricting nurse practitioners from conducting tasks without the supervision of a
physician tends to increase the cost of well-child exams by 3–16 percent (Kleiner et al. 2016).
Variation in Licensing
Occupational licensing is substantially more common in some occupations than others. Kleiner and Krueger
(2013), along with subsequent research, show that occupational licensing is very common in healthcare, legal
occupations, education, and protective services and less common in computer and mathematical, oice and
administrative support, and art and entertainment occupations.
occupational licensing sends a signal to employers of a worker’s non-felon status in occupations where only non-felons
may become licensed.
124 Johnson, Janna E., and Morris M. Kleiner. 2020. "Is occupational licensing a barrier to interstate migration?"American
Economic Journal: Economic Policy12 (3): 347–73.
125 Synchronizing licensing requirements and exams may be more diicult in some occupations than others, depending
on the portability of skills. For example, the knowledge and skillsets of lawyers are likely more state-specific than the
knowledge and skillsets of bus drivers.
126 However, many of the studies they examined were conducted at least three decades ago and by the same authors.
Accordingly, results may be highly correlated with each other.
THE STATE OF LABOR MARKET COMPETITION
Percent of Workers Licensed Within Occupations
Source: Kleiner and Krueger (2013), Westat data; UST and CEA calculations. Accessed via Department of Treasury, Council of
Economic Advisers, and Department of Labor (2015).
Occupational licensing is primarily determined at the state-level and varies considerably between states. For
example, Kleiner and Vorotnikov (2018) show that workers are substantially more likely to be licensed in some
states than others. For example, they find that Nevada (26.6 percent), Iowa (24.3 percent), and Maine (24.2 percent)
have the highest share of workers that are licensed, while Georgia (14.4 percent), Delaware (15.2 percent), and
Kansas (16.0 percent) have the lowest share of workers that are licensed (Kleiner and Vorotnikov 2018). While
much of the dierence between states can be explained by state policies, at least some is explained by underlying
dierences in the types of occupations within each state (e.g., greater presence of the gambling industry in Nevada
than other states).
Dierences between states result from dierences in both the extensive margin of licensing (who needs to be
licensed) and intensive margin of licensing (intensity of requirements to become licensed). For example, to obtain a
job as an “electrician,” 31 states (including the District of Columbia) require licensing, while 20 states do not. Alaska
and Hawaii both require licensing to become an electrician. However, Alaska requires 1,000 hours of training
(assuming no previous experience), while Hawaii only requires 240 hours (assuming no previous experience).
127
127 Herman, Zach. 2020. “The National Occupational Licensing Database.” National Conference of State Legislatures,
March 24, 2020. https://www.ncsl.org/research/labor-and-employment/occupational-licensing-statute-database.
32
33
THE STATE OF LABOR MARKET COMPETITION
While all states require licensing to become a “Nursing Home Administrator,” the cost of initial licensure is only
$100 in Indiana compared to over $3,500 in Oklahoma.
128
As the figure below shows, while the mean time to obtain
a license is about 220 days, there is enormous variation between occupations.
Mean Number of Days to Become Licensed
Among Licensed Workers
All Workers
Business and Financial Operations
Community and Social Service
Construction and Extraction
Educational Instruction and Library
Healthcare Practitioners and Technical
Healthcare Support
Installation, Maintenance, and Repair
Management
Office and Administrative Support
Personal Care and Service
Sales and Related
Transportation and Material Moving
0 200 400
600
800
1,000
Number of Days
Source: BLS Occupational Requirement Survey, 2020 release. Categories represent major occupation
groups.
Occupational licensing is not limited only to workers in high-income occupations. As the figure below shows, there
is little obvious correlation between the prevalence of occupational licensure and average income by occupation.
Occupational Licensing vs Annual Income
100
90
80
70
60
50
40
30
20
10
0
0 25 50 75 100 125
150
Mean Annual Income ($1000s)
Source: BLS (OES&ORS), authors calculations. Each circle represents an occupation. Size of circles
proportional to occupational employment. Note, scatterplot only includes detailed occupations
measured by both OES and ORS and with estimated employment of at least 100,000 or more.
Percent Licensed (%)
aspx#Database.
128 Id.
THE STATE OF LABOR MARKET COMPETITION
Wage Transparency
As discussed in the theory section, workers’ lack of information on potential outside oers creates an important
search friction. Using data they obtained from Denmark, Caldwell and Harmon (2019) find that changes in workers’
information about opportunities outside of their current firm spur mobility and wage growth.
129
When workers lack
information and are unable to easily find such information, they may stay in jobs they would otherwise leave or fail
to ask for a raise when they would otherwise have asked for one (see, e.g., Caldwell and Harmon 2019).
Employers know how much all their employees are compensated, but the converse is oen not true. While social
taboos around discussing compensation with coworkers plays a role, employer policies and practices play an
important role as well. A 2017–2018 survey by the Institute for Women’s Policy Research found that workers
reported employer policies that either discouraged (35.4 percent) or purported to prohibit (12.8 percent) discussing
pay with coworkers. Only about a quarter of workers reported their pay being publicly available, with shares
being much higher for public-sector and union workers in their sample.
130
These high rates of pay secrecy policies
persist despite legal protections in many jurisdictions for workers who discuss their pay, including anti-retaliation
protections, such as the National Labor Relations Act, Executive Order 13665, and 19 state anti-pay secrecy laws.
Moreover, while such laws provide important protections, they place the onus on individual workers or jobseekers
to seek information via employees or social and professional networks and to invoke legal protections if they
face retaliation. This may disadvantage individuals who may not have access to formal and informal professional
networks (e.g., those who grew up in low-income households).
Employers likewise oen have more information regarding workers’ outside options than the workers. Many
employers have access to non-public compensation surveys, giving them a better understanding of the wage
distribution for a given occupation and geography. Even when information is publicly available, HR departments of
firms are in a better position use the data than the typical worker—HR departments have institutional knowledge
and a stronger incentive to know where vacancies are posted than a time-constrained worker. Firms can also
benefit from asking about applicants’ employment and compensation history (where permitted). In contrast,
workers very oen do not even know what their peers at the same establishment make. For example, Biasi and
Sarsons (2021) show that many teachers in Wisconsin did not know how much their colleagues were paid.
131
In
their survey, they also found that compared with men, women were 11 percentage points less likely to know how
much their colleagues earned (30 percent for women vs 41 percent for men). This highlights how informational
asymmetries can have disproportionate impacts on women (Biasi and Sarsons 2021).
There are many ways to mandate greater pay transparency. Some approaches might include: 1) requiring
disclosure of aggregated income statistics to workers, applicants, or the public, which might be broken out by
worker characteristics, like gender; 2) requiring individual income disclosure, oen only for subsets of workers (e.g.,
high-paid government workers or managers); and 3) requiring employers to disclose prospective pay ranges in job
postings.
Consistent with the logic that pay secrecy exacerbates gender pay gaps, empirical research suggests that pay
transparency reduces wage gaps between women and men. For instance, using Canadian administrative data,
129 Caldwell, Sydnee, and Nikolaj Harmon. 2019. “Outside Options, Bargaining, and Wages: Evidence from Coworker
Networks.” Working Paper. https://sydneec.github.io/Website/Caldwell_Harmon.pdf.
130 Sun, Shengwei, Jake Rosenfeld, and Patrick Denice. 2021. “On the Books, O the Record: Examining the Eectiveness
of Pay Secrecy Laws in the U.S.” Institute for Women’s Policy Research Policy Brief C494. https://iwpr.org/wp-content/
uploads/2021/01/Pay-Secrecy-Policy-Brief-v4.pdf. Note: sample sizes for government and union workers are much
smaller than the overall sample, so interpret point estimates cautiously.
131 Biasi, Barbara, and Heather Sarsons. 2021. "Information, Confidence, and the Gender Gap in Bargaining." American
Economic Association Papers and Proceedings 111 (May): 174–78.
34
35
THE STATE OF LABOR MARKET COMPETITION
Baker et al. (2019) finds that a public sector salary disclosure law, enabling the public to access salaries of
individual faculty, reduced the gender pay gap between male and female full-time faculty at Canadian universities
by about 20–40 percent.
132
Bennedsen et al. (2019) examine a 2006 Danish law requiring private firms with more
than 35 employees to provide salary statistics by gender to an employee representative.
133
Although they find the
policy reduced the within-firm gender pay gap by about two percentage points (13 percent relative to the pre-
legislation mean), it primarily did so by slowing wage growth for male employees. Using data from Glassdoor,
Sockin and Sockin (2019) likewise find that changes in pay transparency laws in the United States reduce the
gender pay gap by about 2 percentage points for base earnings, though they detect no change for variable pay
(e.g., bonuses and commissions).
134
There is also evidence that wage transparency can reduce the gender wage
gap. Roussille (2022) show that when Hired.com started pre-filling job searchers’ salary ask with the median oer
tendered to applicants with similar qualifications, it resulted in an elimination of the wage ask gap with no impact
on the number of oers women received or the likelihood that they receive an oer.
135
Wage transparency can increase job search and job-to-job transitions. Using a change in pay disclosure laws
in California, Mas (2017) finds that a 2010 mandate requiring the online posting of salaries for top municipal
managers led to a large (about 75 percent) increase in the quits as well as a 7 percent decline in average
compensation for top managers.
136
Using a randomized treatment in access to individual peer-income information
for employees at the University of California, Card et al. (2012) find that information about peer pay for workers
in their pay unit (specific faculty and sta departments) and occupation increased job searching among workers
earning below the median income for their occupation and pay unit (but not for those above the median for their
occupation and pay unit).
137
Though wage transparency may increase job searching and transitions, it plausibly does so partly because it can
decrease (current) job satisfaction and overall happiness, at least in the short run, for some workers (especially
among relatively lower-paid workers). For example, Card et al. (2012) find that workers above the median income
for their occupation and pay unit reported no change in job satisfaction, but workers below the median income for
their occupation and pay unit reported lower job satisfaction. More broadly, Perez-Truglia (2020) present evidence
that a 2001 law enacted in Norway making individuals’ tax records publicly accessible online led to a deepening
of the rich-poor (self-reported) happiness and life satisfaction gaps.
138
The author argues the widening of the gap
was both a consequence of higher reported happiness and satisfaction among higher-income workers and lower
reported happiness and satisfaction among lower-income workers, suggesting the results are driven by income-
132 Baker, Michael, Yosh Halberstam, Kory Kro, Alexandre Mas, and Derek Messacar. 2019. “Pay Transparency and the Gender
Gap.” National Bureau of Economic Research Working Paper 25834.
133 Bennedsen, Morten, Elena Simintzi, Margarita Tsoutsoura, and Daniel Wolfenzon. 2019. “Do Firms Respond to Gender Pay
Gap Transparency?” National Bureau of Economic Research Working Paper 25435.
134 Sockin, Jason, and Sockin, Michael. 2019. “A Pay Scale of Their Own: Gender Dierences in Variable Pay.” SSRN, December
16, 2019. http://dx.doi.org/10.2139/ssrn.3512598. Note: Sockin and Sockin (2019) lump together salary history bans and
wage transparency laws into one indicator variable; therefore, their results do not distinguish between the eect of a
salary history ban (discussed later) and a change in a wage transparency law.
135 Roussille, Nina. 2022. "The Central Role of the Ask Gap in Gender Pay Inequality." Working Paper. https://ninaroussille.
github.io/files/Roussille_askgap.pdf.
136 Mas, Alexandre. 2017. "Does transparency lead to pay compression?" Journal of Political Economy 125 (5): 1683–1721.
Note, top managers in a public-sector job are unlikely to be representative of rank-and-file workers both because wage
determination in the public sector diers from the private sector and because top-paid managers are more likely to be
near the top of the income distribution.
137 Card, David, Alexandre Mas, Enrico Moretti, and Emmanuel Saez. 2012. “Inequality at Work: The Eect of Peer Salaries on
Job Satisfaction.” American Economic Review 102 (6): 2981–3003.
138 Perez-Truglia, Ricardo. 2020. "The eects of income transparency on well-being: Evidence from a natural experiment."
American Economic Review 110 (4): 1019–54.
THE STATE OF LABOR MARKET COMPETITION
comparison eects. To be clear, decreased job and life satisfaction in the short run may well be more than oset
in the longer run for workers who are induced to switch jobs to one that pays them better (or provides a more
favorable bundle of non-wage amenities) or successfully press for better pay at their current job. Nonetheless,
some workers, especially those who feel they cannot switch jobs or renegotiate their income, may be made worse
o by wage transparency.
139
Prohibiting employers from asking applicants’ compensation history (salary history bans) can also reduce the
employer’s information advantage and increase workers’ bargaining power.
140
In a survey of new hires, Hall
and Krueger (2012) find that “about half of all workers reported that their employers had learned their pay in
their earlier jobs before making the oer that led to the current job.”
141
Employers may use such pay history to
refine their wage oer. If employers do so by oering whatever the employee made in their previous job plus a
moderate raise, reliance on pay history can perpetuate existing income inequalities among workers who have
historically been paid less (e.g., women and people of color). Barach and Horton (2021) present some empirical
evidence that suggests banning the collection of pay history could lead to employers to “take a chance” on lower-
waged and less-experienced workers. Using field evidence from an online labor market, they find that employers
tended to hire workers with about 13 percent lower past average wages than the control group that had access to
compensation history.
142
Decline in Department of Labor’s Labor Market Enforcement Actions
All else equal, a reduction in the probability of being inspected reduces a firm’s incentives to comply with
the workplace regulations and standards. Likewise, it aects employee bargaining power because the threat
of reporting bad behavior is less credible if the enforcement agency lacks the ability to respond quickly and
eectively with inspections and sanctions. Conversely, when workers know their employer’s bad behavior is likely
to be punished, they gain bargaining power against their employer to improve working conditions.
Labor market enforcement action by government agencies can reduce actions of bad actors directly and indirectly.
The direct approach is through actual enforcement actions (inspections, penalties, etc.). However, it is far beyond
the ability of any agency to fully monitor all covered workplaces within its purview at any given time. Therefore,
the Occupational Safety and Health Administration (OSHA) and similar agencies rely primarily on deterrence
actions to enforce workplace standards and regulations.
From an employer’s perspective, the cost of being caught failing to comply is weighed against the benefits of
not complying. Non-compliance risks fines, penalties, and reputational damage. Firms make this tradeo by
evaluating the likelihood and potential costs of being caught against the potential savings associated with non-
139 Both studies reported above involve individual-level income disclosures. It is possible that the (dis)satisfaction
eects reported in these studies would be less severe under a policy of only releasing aggregate statistics instead of
individualized income disclosures. This could be the case, for example, if decreased job satisfaction and happiness comes
not only from knowledge that a worker earns less than their peers, but knowledge that their peers now know they make
more than that worker.
140 Several states and localities have enacted laws that require employers to post salary range information for applicants.
Some of these localities include Colorado, Connecticut, Nevada, New York City, Rhode Island, and Washington. Exact
details on each of these laws vary—some, such as Rhode Island’s law, have been passed but not yet gone into eect.
141 Hall, Robert E., and Alan B. Krueger. 2012. "Evidence on the incidence of wage posting, wage bargaining, and on-the-job
search." American Economic Journal: Macroeconomics 4 (4): 56–67.
142 Barach, Moshe A., and John J. Horton. 2021. "How do employers use compensation history? Evidence from a field
experiment."Journal of Labor Economics39 (1): 193–218. Note, Barach and Horton (2021)’s estimates are based on a
“partial equilibrium” approach, i.e., their estimates would likely change if all employers were subject to the types of bans
the treated group was subjected to in the experiment.
36
37
THE STATE OF LABOR MARKET COMPETITION
compliance. If firms think the cost or likelihood of being caught in non-compliance is high relative to the benefits,
they may comply even absent actual inspections or oversight.
In recent years, the probability of a firm being inspected has decreased sharply. Numerous agencies are responsible
for inspections and enforcement actions. However, as an example of how enforcement and inspections have
declined, OSHA commenced the largest number of workplace inspections in 1984, at 140,000 inspections. The
COVID-19 pandemic sharply reduced the number of inspections conducted in 2020. Even before the pandemic, the
number of inspections was much lower than in the 1980s. In 2019, OSHA inspected about 81,000 workplaces, or
40 percent less than it conducted in 1984. From 2013 to 2021, OSHA experienced a 13 percent reduction in Federal
enforcement personnel due to reduced budget availability. The workforce is now larger than it was in the mid-1980s,
and the nature of workplaces has changed during this time period. With fewer enforcement personnel and a larger
workforce, it is increasingly diicult for enforcement actions to reach the same portion of workplaces.
Divergence Between Labor Compensation and Productivity
This section and the next highlights important aggregate trends in wages and labor income. The precise
contribution of firm labor market power to these trends remains an open question, but we highlight some of the
links established in the literature.
During the first part of the post-World War II period, productivity and average compensation largely moved in
tandem. That is, when workers were more productive for each hour they worked, their pay proportionately
increased, on average. During this period, gains in productivity appeared to be proportionately dispersed among
the compensation distribution.
50
100
150
200
Index (1973=100)
1940 1960 1980 2000 2020
Net Productivity per Hour Worked
Average Hourly Compensation, Overall
Average Hourly Compensation, Production/Nonsupervisory Workers
Median Hourly Compensation, Overall
Source: Economic Policy Institute, BEA, BLS, author's calculations. Productivity is output per hour worked.
Net productivity adjusts for depreciation.
Divergence Between Productivity
and Labor Compensation
However, as the figure above shows, starting around 1980, a divergence in productivity and wages started to
emerge, particularly for the lower end of the compensation distribution.
143
This divergence between productivity
143 Note, the figure reports net-productivity rather than gross productivity. Not accounting for accelerated depreciation
THE STATE OF LABOR MARKET COMPETITION
and compensation, particularly among lower-income and non-management workers, has been the subject of
considerable debate.
144
Some have noted that part of the divergence may be attributable to dierences in how
productivity and compensation are adjusted for inflation, possibly due to dierences in how the dierent series
account for changes in technological products.
145
However, Stansbury and Summers (2018) argue that some
deviation has occurred even aer accounting for such measurement issues.
146
Bivens and Shierholz (2018) argue the dierence between typical (median) worker compensation and productivity
can be decomposed into two components—declining labor share and income inequality.
147
Using a back-of-
the-envelope calculation, they estimate approximately five-sixths of the decline is attributable to rising income
inequality and only a sixth attributable to declining overall labor share. The fall in the share of labor, discussed
in greater detail in the next section, is partly captured in the above figure as the divergence between average
compensation and productivity, especially since 2001. Rising income inequality is reflected in the above figure
as the split between mean and median compensation. This divergence suggests that higher-income and
supervisory workers have captured a greater share of income over time. A similar schism between compensation
of nonsupervisory workers and overall compensation has occurred, likely for similar reasons.
The increase in the share of productivity gains captured by higher-income workers is hotly debated and touches
upon the larger debate regarding the causes for the rise in income inequality since the 1980s. In principle, the
disparity could be the result of significant increases in productivity among management and stagnation in
productivity among lower-income workers. For example, changes in technology could make management
substantially more eicient. However, this does not appear to be supported in the literature. Stansbury and
Lawrence (2018) argue that a technological change-driven explanation would imply greater divergence during
periods of higher productivity gains, however it does not find empirical evidence supporting that implication.
Evidence suggests that declining competition in the labor market coupled with loss of bargaining power among
lower-wage workers contributes to income inequality. For example, Furman and Orszag (2018) argue that declining
competition for labor has decoupled wage growth from productivity gains as workers face fewer choices and
decreased mobility.
148
Consistent with this finding, Benmelech, Bergman, and Kim (2020) use manufacturing
plant-level data from 1978 to 2016 to show that wages are noticeably lower in local labor markets that have
in recent decades tends to overstate the divergence between output per hour worked and compensation. For a
critical review, see Lawrence, Robert Z. 2016. “Does Productivity Still Determine Worker Compensation? Domestic and
International Evidence.” In The US Labor Market: Questions and Challenges for Public Policy, edited by Michael R. Strain, 42–
62. Washington: American Enterprise Institute.” Even here, Lawrence finds declines in labor share post 2000.
144 This debate includes the proper way to account for prices. For details, see discussion in Mishel, Lawrence. 2021. “Growing
Inequalities, Reflecting Growing Employer Power, Have Generated a Productivity–Pay Gap Since 1979.” Economic Policy
Institute, September 2, 2021. https://www.epi.org/blog/growing-inequalities-reflecting-growing-employer-power-have-
generated-a-productivity-pay-gap-since-1979-productivity-has-grown-3-5-times-as-much-as-pay-for-the-typical-worker/.
145 See Fleck, Susan, John Glaser, and Shawn Sprague. 2011. “The compensation-productivity gap: a visual essay.” U.S.
Bureau of Labor Statistics Monthly Labor Review (January): 57–69. https://www.bls.gov/opub/mlr/2011/01/art3full.pdf. See
also, Brill, Michael, Corey Holman, Chris Morris, Ronjoy Raichoudhary, and Noah Yosif. 2017. “Understanding the labor
productivity and compensation gap.” U.S. Bureau of Labor Statistics Beyond the Numbers: Productivity 6 (6): 1–14. https://
www.bls.gov/opub/btn/volume-6/pdf/understanding-the-labor-productivity-and-compensation-gap.pdf.
146 Stansbury, Anna, and Lawrence H. Summers. June 2018. “Productivity and Pay: Is the Link Broken?” Peterson Institute for
International Economics Working Paper 18-5. https://www.piie.com/system/files/documents/wp18-5.pdf.
147 Bivens, Josh, and Heidi Shierholz. 2018. “What labor market changes have generated inequality and wage suppression?”
Economic Policy Institute, December 12, 2018. https://www.epi.org/publication/what-labor-market-changes-have-
generated-inequality-and-wage-suppression-employer-power-is-significant-but-largely-constant-whereas-workers-power-
-has-been-eroded-by-policy-actions/.
148 Furman, Jason, and Peter Orszag. 2018. “Slower Productivity and Higher Inequality: Are They Related?” Peterson Institute
for International Economics Working Paper 18-4.
38
39
THE STATE OF LABOR MARKET COMPETITION
higher employer concentration. Their results also show that this correlation is even more pronounced in areas
with low levels of unionization. In a vein like Autor et al. (2020)’s concept of “superstar firms,” a 2018 paper by the
Organisation for Economic Co-operation and Development (OECD) also noted that the divergence of wages and
productivity “at the technological frontier has been accompanied by increasing market shares of frontier firms.”
149
Decline in Labor Share
Economists decompose an economy’s aggregate income into that which is attributable to labor (wages and other
compensation for work) and capital (i.e., interest, rent, and dividend payments). For decades, labor’s share of
income was estimated at slightly less than two-thirds.
150
However, starting around the 1980s, this share began to
decline not only in the United States, but around the world.
151
Source: Elsby (2016) using data from Bureau of Labor Statistics and Bureau of Economic Analysis.
Numerous theories have been oered for why labor’s share of income has declined. Elsby, Hobjin, and Sahin
(2016) suggest oshoring of the labor-intensive portion of the United States’ supply chain is a leading potential
149 Organisation for Economic Co-operation and Development. 2018. “Decoupling of Wages from Productivity: What
Implications for Public Policies?” OECD Economic Outlook 2018 (2): 51–65. https://www.oecd.org/economy/decoupling-of-
wages-from-productivity/.
150 Kaldor, Nicholas. 1961. "Capital Accumulation and Economic Growth." In The Theory of Capital, edited by D.C. Hague,
177–222. London: Palgrave Macmillan.
151 See Karabarbounis, Loukas, and Brent Neiman. 2014. "The global decline of the labor share." The Quarterly Journal of
Economics 129 (1): 61–103. There is debate whether the share of labor has fallen or the observed changes are due to
changes in measurement, such as an increase in self-employment, business owners taking capital instead of labor
income, etc. See Autor (2020) for a skeptical overview.
THE STATE OF LABOR MARKET COMPETITION
cause, and note that measurement issues account for a quarter of the observed decline.
152
Karabarbounis and
Neiman (2014) suggest rapidly falling prices, especially of capital, may have played a part. Still others, like Weil
(2014), suggest fissuring has played a role by decreasing the relative bargaining position of labor. The relative
contributions of measurement, technology change, changes in industry composition, and firm wage setting power
remain issues of study.
The declining share of labor might also be a result of increasing employer product market power. De Loecker,
Eeckhout, and Unger (2020) document how markups in product markets have risen nearly three-fold since 1980.
153
They show that this increase primarily came from the very upper end of the markup distribution, i.e., large firms
within industries increasing their size, margins, and profitability. Their modeling suggests labor share is inversely
proportional to markups, so an increase in markups naturally leads to a decline in the share of labor.
154
As De Loecker, Eeckhout, and Unger (2020) explain, a natural consequence of increased market power and
markups is a decrease in aggregate output.
155
This corresponds with decreases in labor demand, which places
downward pressure on wages. The reduction in output also mechanically corresponds to an increase in output
price, implying a decrease in real wages (since the same dollars of wages buy fewer goods).
In a related work, Autor et al. (2020) argue that the decline in labor share might be attributable to a rise of what
they term “superstar firms” that dominate a particular market and have high markups and low labor share.
Using microdata from the U.S. Census Bureau, they document that across many industries, sales are increasingly
concentrated among a few firms and industries where this concentration rises most tend to see the largest declines
in labor share. The rise of such superstar firms also drives the decline in labor’s share of income, even if it does not
occur among most firms (which is consistent with the observation of De Loecker, Eeckhout, and Unger (2020) that
median markups have not changed much even as the top of the mark-up distribution has increased dramatically).
Autor et al. (2020) argue that the rise of superstar firms could be driven by several factors. They note that the
increase could be driven by persistent incumbent dominance. Persistent dominance could be explained by a
variety of factors. For example, superstar firms tend to be more productive. To the extent that incumbent firms are
more innovative, they could remain dominant because customers prefer their products. Alternatively, persistent
dominance can be due to anticompetitive business practices. The authors acknowledge that arguments such as
the weakening of antitrust enforcement advocated by Gutierrez and Philippon (2018) could plausibly explain some
of their results.
156
While the increase in market concentration has occurred across numerous industries, the explanation for the rise
of superstar firms in each industry need not be the same. The welfare implications of a rise of a superstar firm
because of being more innovative compared to one that has engaged in regulatory capture or simply evaded anti-
trust enforcement are quite dierent.
152 Elsby, Michael W.L., Bart Hobijn, and Ayegül ahin. 2013. "The Decline of the US Labor Share." Brookings Papers on
Economic Activity (Fall): 1–63.
153 De Loecker, Jan, Jan Eeckhout, and Gabriel Unger. 2020. “The Rise of Market Power and the Macroeconomic Implications.”
The Quarterly Journal of Economics 135 (2): 561–644.
154 De Loeker, Eeckhout, and Unger (2020) model an economy with imperfect output markets, allowing for firms to extract
economic profits. Accordingly, they find that not only does their model imply the share of labor decreases with increased
markups, but so does the capital share since profits increase with increased markups.
155 This is a natural consequence because firms can increase their markups/profit by restricting output so long as demand
is not perfectly elastic. Intuitively, firms with market power are willing to lose some customers in exchange for charging
more per item. Thus, a firm with market power would avoid decreasing output only if consumers did not respond to
higher prices.
156 Gutierrez, German, and Thomas Philippon. 2018. "How EU Markets Became More Competitive than US Markets: A Study of
Institutional Dri." SSRN CEPR Discussion Paper DP12983, June 2018.
40
41
THE STATE OF LABOR MARKET COMPETITION
Industry Examples
The following subsections highlight the various ways in which developments in example labor markets have
harmed workers in their respective occupations or industries. In the hospital and nursing subsection, we show
consolidation in the product market (hospitals) can negatively impact workers (nurses). In the agricultural sector,
both tacit and explicit collusion between employers has led to highly concentrated markets where workers have
little to no bargaining power. In minor league baseball, lobbying eorts, coupled with Supreme Court precedent,
have weakened worker pay protections, allowing the monopsonist to extract rents and exert extraordinary control
over their worker’s mobility.
Hospitals and Nurses
The hospital industry has consolidated in recent decades. Despite a growing population, the number of hospitals
decreased from 7,156 hospitals in 1975 to only 6,093 hospitals in 2021.
157
Empirical evidence suggests these
consolidations have increased the prices of hospital services with no evidence of quality improvement.
158
Consolidation also impacts the input market. As hospitals consolidate, they gain monopsony power. When the
hospital industry consolidates by closing hospitals, it increases monopsony power mechanically by increasing the
cost among nurses to finding work elsewhere (i.e., longer commutes). Even when consolidation does not reduce
the number of hospitals (e.g., through a merger of hospital systems) it can increase monopsony power by reducing
competition among the remaining firms. Krueger (2018) notes that consolidation also increases monopsony power
even if hospitals do not have a literal monopoly because fewer players in a market increase the probability of
collusion, tacit or otherwise.
159
Even before the recent wave of hospital consolidation, there was evidence that hospitals exerted considerable
monopsony power over healthcare workers. Using changes in wages at Veterans Aairs hospitals, Staiger, Spetz,
and Phibbs (2010) found that labor supply to individual hospitals is quite inelastic.
160
Their results imply that a 10
percent decline in the wages of nurses only decrease employment by about 1 percent in the short run, which is a
much smaller change in employment than one would expect in a perfectly competitive market where hospitals
had little market power. The recent wave of consolidation has likely only increased hospital monopsony power.
Prager and Schmitt (2021), supra, present evidence that certain types of hospital mergers causally decrease wages
for certain healthcare workers. They find that mergers that cause the largest increases in hospital concentration
(those in the top quartile of increases in the HHI) cause wage growth among skilled workers and nursing and
pharmacy workers to slow, particularly among nurses and pharmacy workers. Importantly, they fail to find
negative eects on wage growth from smaller mergers (i.e., those that do not increase market concentration
much), which suggests the eects they find among larger mergers are caused by the increase in hospital
157 National Center for Health Statistics, Centers for Disease Control and Prevention. 2017. “Hospitals, beds, and occupancy
rates, by type of ownership and size of hospital: United States, selected years 1975–2015.” Table 89. https://www.cdc.gov/
nchs/data/hus/2017/089.pdf; American Hospital Association. 2022. “Fast Facts on U.S. Hospitals, 2022.” Last modified
January 2022. https://www.aha.org/system/files/media/file/2022/01/fast-facts-on-US-hospitals-2022.pdf. Note: The
initial dates are from the Centers for Disease Control and Prevention, while the latest value is from the American Hospital
Association. Estimates may not be directly comparable.
158 Beaulieu, Nancy D., Leemore S. Dafny, Bruce E. Landon, and Jesse B. Dalton. 2020. "Changes in Quality of Care aer
Hospital Mergers and Acquisitions." New England Journal of Medicine 382 (January): 51-59. https://www.nejm.org/doi/
full/10.1056/NEJMsa1901383.
159 See references in, e.g., Krueger, Alan. 2018. “Reflections on Dwindling Worker Bargaining Power and Monetary Policy.”
Luncheon address to FRB Kansas City’s Jackson Hole Symposium, August 24, 2018. https://www.kansascityfed.org/
documents/6984/Lunch_JH2018.pdf.
160 Staiger, Douglas O., Joanne Spetz, and Ciaran S. Phibbs. 2010. "Is there monopsony in the labor market? Evidence from a
natural experiment." Journal of Labor Economics 28 (2): 211–236.
THE STATE OF LABOR MARKET COMPETITION
monopsony power post-merger rather than factors common to most mergers.
161
Prager and Schmitt (2021) also fail to find that mergers decrease wage growth among hospital workers in jobs
requiring little training—which is consistent with these workers having closer employment substitutes outside
hospitals, thereby reducing the ability of hospitals to exert monopsonistic power over their wages.
162
The paper
does not examine the eects of mergers specifically increasing concentration in the relevant labor markets for
these workers in jobs with little hospital-specific skill.
While the antitrust agencies have the authority to challenge hospital mergers,
163
such enforcement eorts are
resource-intensive and not always successful.
164
In addition, states may grant Certificates of Public Advantage
(COPA), which have the eect of immunizing certain hospital mergers from federal antitrust law.
165
These
state COPA laws purport to supplant federal antitrust laws with a regulatory scheme that allows for hospital
consolidation even in highly concentrated markets, thereby hindering the ability of the antitrust agencies to
challenge anticompetitive mergers. This, in turn, may lead to consolidation among hospital employers that
depresses wages and raise health care costs to the public.
166
For instance, while evaluating a proposed merger of two Texas hospitals that applied for a COPA, FTC sta
conducted a labor market analysis and concluded that the merger would likely reduce hospital competition and
depress wage growth for registered nurses.
167
The FTC is currently conducting a study of the impact of COPA on
competition in healthcare markets, including possible labor monopsony eects.
168
Agriculture
Food processing is highly concentrated nationally, but its employment is also geographically concentrated. Food
processing tends to occur away from urban centers and is more concentrated in low-density areas. For example,
161 Prager and Schmitt (2021) also show that their results cannot be explained by pre-merger trends—such as poor local
economic conditions, which may induce a merger to begin with—explaining why wages decline for nurses and pharmacy
workers post-merger.
162 For example, janitorial sta at a hospital may be able to find comparable work outside of a hospital environment, while a
nurse has fewer options outside of the hospital industry that would not entail a large pay cut.
163 The federal antimerger law, the Clayton Act, applies to mergers involving non-profits, and the antitrust agencies have
opposed several mergers involving non-profit hospitals. See, e.g., Federal Trade Commission v. OSF Healthcare System,
852 F. Supp. 2d 1069, 1081 (N.D. Ill. 2012); United States v. Rockford Memorial Corp., 898 F.2d 1278, 1284-87 (7th Cir. 1990);
and Hospital Corp. of America v. Federal Trade Commission, 807 F.2d 1381, 1390-91 (7th Cir. 1986).
164 Federal Trade Commission v. Thomas Jeerson University, No. 20-1113 (E.D. Pa. 2020). https://www.c.gov/enforcement/
cases-proceedings/181-0128/thomas-jeerson-university-matter.
165 See, e.g., Federal Trade Commission. 2016. “FTC Sta Provides Public Comment and Testimony in Tennessee Opposing
Certificate of Public Advantage Application.” Press release, November 23, 2016. https://www.c.gov/news-events/press-
releases/2016/11/c-sta-provides-public-comment-testimony-tennessee-opposing.
166 See, e.g., Gaynor, Martin, Kate Ho, and Robert J. Town. 2015. “The Industrial Organization of Health-Care Markets.” Journal
of Economic Literature 53 (2): 236; Gaynor, Martin, and Robert J. Town. 2012. “The Impact of Hospital Consolidation –
Update.” Robert Wood Johnson Foundation and the Synthesis Project Policy Brief 9; and Baicker, Katherine, and Amitabh
Chandra. 2006. “The Labor Market Eects of Rising Health Insurance Premiums.” Journal of Labor Economics 24 (3): 609–
634.
167 See Conner, Ian, Andrew Sweeting, and Bilal Sayyed. 2020. “Federal Trade Commission Sta Submission to Texas
Health and Human Services Commission Regarding the Certificate of Public Advantage Applications of Hendrick Health
System and Shannon Health System.” Federal Trade Commission, September 11, 2020. https://www.c.gov/system/files/
documents/advocacy_documents/c-sta-comment-texas-health-human-services-commission-regarding-certificate-
public-advantage/20100902010119texashhsccopacomment.pdf.
168 See Federal Trade Commission. 2019. “FTC to Study the Impact of COPAs.” Press release, October 21, 2019. https://www.
c.gov/news-events/press-releases/2019/10/c-study-impact-copas.
42
43
THE STATE OF LABOR MARKET COMPETITION
as of the first quarter of 2021, Alabama, Nebraska, Arkansas, and Iowa each employed more animal slaughtering
and processing workers than the state of California even though California has approximately three times as many
people as those four states combined.
169
In the agricultural input sector, the use of temporary agricultural workers through the H-2A visa program has
received attention because of its increased use in recent years. From 2010 to 2021, the use of this program
quadrupled—from about 79,000 jobs certified annually in 2010 to over 317,600 in 2021.
170
Governed by 8 U.S.C. § 1188 and 20 C.F.R. § 655, Subpart B, the H-2A program is an employer-sponsored temporary
visa program that allows agricultural employers to employ nonimmigrant foreign workers to perform agricultural
labor or services, as defined by Congress, on a temporary or seasonal basis, typically lasting 10 months or less.
While the number of workers that can be admitted and issued an H-2A visa is not capped by Congress, the program
does require an employer to oer and provide numerous employment guarantees and protections to H-2A workers
and any U.S. workers performing the same work. For example, employers must show that hiring foreign workers
will have no “adverse eect” on the wages and working conditions of U.S. workers similarly employed. Employers
must provide workers with housing, meals or kitchen facilities for workers to prepare meals, and transportation,
and must pay petition and certification fees.
171
Both employers and workers rights advocates have criticized the H-2A program. Employers have argued the
program is too bureaucratic, complex, and expensive. For example, they argue that the requirement that workers
obtain visas to enter the United States, which was not a requirement under H-2A’s predecessor program, is
expensive (about $200 per application). They also oen argue that they are required to guarantee a wage rate that
is, in their view, artificially high.
172
169 Based on Q1 2021 data comparing statewide average employment data for North American Industry Classification System
(NAICS) code 3116 in the Quarterly Census of Employment and Wages to 2020 Census population estimates.
170 2010 and 2021 data are from Oice of Foreign Labor Certification, U.S. Department of Labor. “Performance Data.”
Historical Case Disclosure Data for the H-2A Program (file name for FY2021 data: H-2A_Disclosure_Data_FY2021.xlsx; file
name for FY2010 data: H-2A_FY2010.xlsx). Last accessed March 4, 2022. https://www.dol.gov/agencies/eta/foreign-labor/
performance.
171 Wage and Hour Division. 2010. “Fact Sheet #26: Section H-2A of the Immigration and Nationality Act (INA).” U.S.
Department of Labor, February 2010. https://www.dol.gov/agencies/whd/fact-sheets/26-H2A.
172 Per 20 CFR § 655.120, employers must generally oer and pay a wage that is at least the highest of “the AEWR [(Adverse
Eect Wage Rate)] the prevailing hourly wage or piece rate, the agreed-upon collective bargaining wage, or the Federal
or State minimum wage.” The AEWR is set by DOL as a rate that ensures wages of similarly employed U.S. workers are not
adversely aected. Typically, the AEWR is the wage that binds, if any.
THE STATE OF LABOR MARKET COMPETITION
Workers’ rights advocates argue H-2A restricts competition in unfair ways and is rife with employer abuse.
173
Importantly, one way the H-2A program plausibly restricts competition is by allowing employers to coordinate
hiring eorts through professional associations, including wage decisions.
174
While such associations allow
employers to take advantage of economies of scale in bringing over foreign workers, they also, almost by
definition, concentrate labor demand. These associations can account for a large share of hiring by occupation.
For example, a recent lawsuit, Llacua v. Western Range Association, alleges that two trade associations accounted
for the hiring of approximately 91 percent of all shepherds.
175
When in conflict, courts appear to favor the
interpretation of immigration law (which is permissive of such monopsony power) over anti-trust law (which, at
least in principle, is less permissive of monopsony power) (Riviere 2021, 1581).
173 See, e.g., Farmworker Justice. n.d. “No Way to Treat a Guest: Why the H-2A Agricultural Visa Program Fails U.S. and Foreign
Workers.” Accessed March 3, 2022. https://www.farmworkerjustice.org/resource/no-way-to-treat-a-guest-why-the-h-
2a-agricultural-visa-program-fails-u-s-and-foreign-workers/; National Farm Worker Ministry. n.d. “H-2A Guest Worker
Program.” Accessed March 3, 2022. https://nfwm.org/farm-workers/farm-worker-issues/h-2a-guest-worker-program/;
Lahoud, Raymond G. 2021. “Human Traicking Indictment Uncovers H-2A Abuses.” National Law Review 11 (350); and
Mississippi Center for Justice. 2021. “Black Farmworkers Sue Mississippi Farm for Racial Discrimination, Lost Wages, and
Abuse of Immigration System to Deny U.S. Workers of Jobs.” Press release, September 8, 2021. https://mscenterforjustice.
org/black-farmworkers-sue/.
174 See 8 U.S.C. § 1188(d); and Riviere, Candice Yandam. 2021. “The Legal Causes of Labor Market Power in the U.S.
Agricultural Sector.” University of Chicago Law Review 88 (6): 1555–1594. https://lawreview.uchicago.edu/sites/lawreview.
uchicago.edu/files/Yandam_LaborMarketPower_88UCLR1555.pdf.
175 Llacua v. Western Range Association, No. 17-1113, 930 F.3d 1161 (10th Cir. 2019). Note, this is an outlier example of
concentration, even among H-2A jobs; furthermore, sheep and goat herders account for a small share (about 1 percent) of
H-2A certified jobs.
44
45
THE STATE OF LABOR MARKET COMPETITION
Minor League Baseball
Although it directly impacts a relatively small share of the workforce, minor league baseball provides a useful case
study of how a true monopsonist can restrict worker mobility, pay, and even successfully lobby for legislation that
further solidifies their dominance over their employees.
In 2014, minor league baseball players brought a class-action lawsuit against Major League Baseball (MLB), the
organizer of Minor League Baseball (MiLB), alleging that MiLB’s wages and labor practices violate minimum wage
laws and overtimes rules set forth in the Fair Labor Standards Act of 1938.
176
The players alleged, among other
things, that they routinely worked sixty or more hours in a week but were not paid overtime pay and did not
receive pay for certain types of activities that the players considered work-related.
In an apparent attempt to preempt litigation, the MLB lobbied Congress in 2018 to include the Save America’s
Pastime Act (SAPA) as part of a $1.3 trillion dollar spending package.
177
SAPA explicitly exempts workers in MiLB
from minimum wage requirements under FLSA. Furthermore, SAPA purports to be retrospective, applying not only
to future MiLB work, but all past work as well. This legislation adds an additional challenge that minor leaguers
would have to overcome to prevail on federal employment-law claims.
178
As of February 2021, MiLB underwent a major reorganization in which 40 minor league teams were cut but wages
were raised. Although the percentage raise was significant for many players, absolute salaries remain quite low –
players in the highest category are expected to earn approximately $14,700 a season.
179
Players in the lowest tier
experienced the largest relative benefit from this restructuring, with their minimum salary increasing by over 70
percent relative to 2019, up to $10,500.
180
MLB also restructured teams to be more geographic-centric, which will
hopefully reduce travel burdens.
181
The MLB still exerts tremendous monopsony power over minor league baseball players, due in part to an
“aberrational,” judicially-created doctrine that the Supreme Court has called “something that looks a bit like an
antitrust exemption for professional baseball,” which was first announced by the Supreme Court in 1922.
182
While
Congress passed legislation in 1998 to clarify that conduct related to major league baseball players is subject
to antitrust laws, the legislation did not address minor leaguers’ employment.
183
There are pending lawsuits
addressing whether the MLB so-called baseball exemption continues to apply to restraints on minor league players
in light of subsequent developments undermining its foundations.
Minor league players are typically unable to receive unemployment insurance (UI) benefits during the o-season
176 Complaint,Senne et al. v. Oice of the Commissioner of Baseball, et al., No. 3:14-cv-00608-JCS (N.D. Cal. Feb. 7, 2014), ECF No. 1.
177 It is unclear how much Congressional support SAPA had, as the two-page bill was included on page 1,967 of the 2,323-
page spending package.
178 Pannullo, Robert. 2020. “The Struggle for Labor Equality in Minor League Baseball: Exploring Unionization.” American Bar
Association Journal of Labor & Employment 34 (3): 443–476.
179 Blum, Ronald. 2021. “Minor Leagues Get a Reset with 120-Team Regional Alignment.” AP News, February 12, 2021. https://
apnews.com/article/sports-mlb-baseball-rob-manfred-coronavirus-pandemic-f8a0f1c09161e83db87bca8e78219725. A
typical season lasts about five months.
180 These reflect minimum salaries. Actual compensation, including bonuses, may be significantly larger, especially for
higher-tiered players. Furthermore, these values reflect first-time contracts—second contracts tend to be significantly
larger. For more, see Fagan, Ryan. 2021. “Even aer overdue salary bump, baseball's minor leaguers still paid far below
NBA, NHL counterparts.” Sporting News, February 12, 2021. https://www.sportingnews.com/us/mlb/news/even-aer-
overdue-salary-bump-baseballs-minor-leaguers-still-paid-far-below-nba-nhl-counterparts/1gpql94asy7a10uo5nvc3yp4k.
181 Janes, Chelsea. 2021. “MLB overhauled the minors this season. Some advocates say it hasn’t been enough.” Washington Post, July
16, 2021. https://www.washingtonpost.com/sports/2021/07/16/minor-league-baseball-advocacy-mlb-overhaul/.
182 National Collegiate Athletic Association v. Alston, 141 S. Ct. 2141, 2159 (2021).
183 Curt Flood Act of 1998, Pub. L. No. 105-297. 15 U.S.C. Sec. 26b.
THE STATE OF LABOR MARKET COMPETITION
because they are classified as seasonal workers. The logic in denying seasonal workers UI benefits is that the end
of their employment is predictable and therefore they could plan other job opportunities around the seasonality of
their work. Still, some have argued this is unfair, especially since the start of the COVID-19 pandemic.
184
Ordinarily collective action through unionization can provide a counterbalance to employer power. While major
league baseball players have been unionized for decades by the Major League Baseball Players Association, minor
league players have no union. A primary reason for union hesitation among the players is a fear of retaliation
by MLB (see Pannullo 2020). Additional factors include high turnover of MiLB players, geographic dispersion of
MiLB players, and low salaries that discourage existing unions from expanding their membership to include MiLB
players.
185
184 Baccellieri, Emma. 2020. “Minor Leaguers and the Fight to Claim Unemployment.” Sports Illustrated, June 12, 2020. https://
www.si.com/mlb/2020/06/12/minor-league-baseball-players-unemployment.
185 Broshuis, Garrett R. 2013. “Touching Baseball’s Untouchables: The Eects of Collective Bargaining on Minor League
Baseball Players.” Harvard Journal of Sports & Entertainment 4 (June): 51–103. https://harvardjsel.com/wp-content/
uploads/sites/9/2013/06/Broshius.pdf.
46
47
THE STATE OF LABOR MARKET COMPETITION
IMPLICATIONS BEYOND THE LABOR MARKET
A decline in the competitiveness of labor markets lowers worker wages, may decouple wages from productivity,
and likely diminishes the relative share of income that goes to workers. Moreover, actions of the firm such as
requiring workers to sign non-compete agreements and limiting workers’ access to information diminishes worker
mobility, implicitly reducing workers’ bargaining power relative to employers.
These direct eects on workers’ wages, employment, and mobility have important broader negative impacts
on the economy. Higher inequality likely makes it more diicult to sustain suicient aggregate demand. Lower
wages disproportionately impact women and workers of color. A large pool of low-priced labor likely weakens firm
incentives to invest and improve productivity, while lower mobility diminishes productivity growth by hindering
the reallocation of labor to more productive firms and industries. Non-compete agreements may prevent workers
from starting their own businesses and discourage innovation. In short, a growing body of evidence suggests that
declining labor market competition may stymie the drivers of U.S. economic growth. To be clear, labor market
competition is unlikely to be the only or even primary driver of broader macroeconomic trends, but, on the margin,
likely contributes and exacerbates some drivers of slower economic growth.
Rising Inequality, Low Interest Rate, and Aggregate Demand
Over the last several decades, income inequality has risen sharply. As documented by Piketty and Saez (2003)
and Saez and Zucman (2020), the share of income earned by the top 1 percent has risen since 1980 and now
approaches levels last seen in the 1920s; the top 1 percent collects nearly one-fih of national income.
186
Average
income growth from 1980 of the top 1 percent has surged at rates well above 2 percent per year, while overall
income growth averages just 1.4 percent over the same period and is lower for the bottom 85 percent of the U.S.
income distribution.
186 Piketty, Thomas, and Emmanuel Saez. 2003. “Income Inequality in the United States, 1913–1998.” The Quarterly Journal of
Economics 118 (1): 1–41; and Saez, Emmanuel, and Gabriel Zucman. 2020. “The Rise of Income and Wealth Inequality in
America: Evidence from Distributional Macroeconomic Accounts.” Journal of Economic Perspectives 34 (4): 3–26.
THE STATE OF LABOR MARKET COMPETITION
Income inequality has several causes; however, inequality in income from labor and slow growth in wages plays
an important role in driving overall income inequality. To control for demographic changes that possibly increases
in income volatility, Guvenen et al. (2021) measure inequality in male lifetime earnings using Social Security
data.
187
They find that median lifetime earnings fell 10–19 percent for men entering the workforce in 1983 versus
men entering the workforce in 1957. Put another way, the realized lifetime real income for the typical male worker
in 1983 was substantially lower than their 1957 counterparts. For cohorts entering aer 1983 (and still working),
they find evidence of continued stagnation of income for the median worker and increasing inequality in lifetime
earnings. Similar stagnation in lifetime earnings has also been observed for currently working cohorts (gains for
female cohorts prior to 1983 came o a very low base).
A growing body of research suggests that rising income inequality carries important implications for the
macroeconomy. The secular stagnation hypothesis posits that the natural rate of interest, the interest rate needed
to achieve full employment, has been falling for several decades. Several distinct drivers of low interest rates
have been suggested, including rising income inequality.
188
As the secular stagnation literature emphasizes, an
excessively low natural rate of interest complicates the conduct of monetary policy. In recessions, interest rates
must fall to stimulate the investment and maintain aggregate demand. Central banks are generally unable to lower
short-term interest rates below zero; when interest rates need to be kept low to sustain full employment, monetary
policy can face an inability to lower interest rates suiciently in recessions before hitting the zero-lower bound.
Since 2000, the zero lower bound has posed an increasing challenge for using monetary policy to boost demand.
In the wake of the 2008 financial crisis, the Federal Reserve, European Central Bank, and other central banks had
to keep interest rates close to zero for an unprecedented duration to sustain an economic recovery. Prior to the
2020 pandemic, U.S. short-term rates were just 2 percent and, absent an unprecedented increase in fiscal support,
appeared insuicient to oset the pandemic’s eect on aggregate demand.
While the precise contribution of lower labor market competition to income inequality is open to debate, the rise in
inequality has been stark and pronounced. And the link to low interest rates has increasing support as a theoretical
mechanism and in empirical evidence. To the extent that increases in labor market competition boost wages and
labor share, this would likely imply raised demand and a higher natural rate of interest.
Impacts on Women and Workers of Color
Evidence suggests that the burden of lower worker power fall disproportionately on women and workers of
color. Rosenfeld and Kleykamp (2012) estimate that declines in private-sector unionization have contributed to
substantial racial wage gaps—up to 30 percent for Black women.
189
Lower rates of unionization may have also le
women workers and workers of color more vulnerable to wage the and other workplace violations (i.e., Bernhardt
et al. 2009).
190
Continued labor market power can allow racial discrimination in hiring to persist; Quillian et al.
187 Guvenen, Fatih, Greg Kaplan, Jae Song, and Justin Weidner. 2021. “Lifetime Earnings in the United States over Six
Decades.” Becker Friedman Institute Working Paper 2021-60. https://bfi.uchicago.edu/working-paper/2021-60/.
188 Eggertsson and Mehrotra (2014) show how low interest rates could contribute to low or negative natural rates of interest.
Eggertsson, Gauti B., and Neil R. Mehrotra. 2014. “A Model of Secular Stagnation.” National Bureau of Economic Research
Working Paper 20574. Mian, Sufi, and Straub (2020) show how bequest motives may explain why wealthier households
save a larger portion of their income; therefore, higher income inequality lowers the natural rate. Mian, Atif, Ludwig Straub,
and Amir Sufi. 2021. “Indebted Demand.” The Quarterly Journal of Economics 136 (4): 2243–2307.
189 Rosenfeld, Jake, and Meredith Kleykamp. 2012. “Organized Labor and Racial Wage Inequality in the United States.”
American Journal of Sociology 117 (5): 1460–1502.
190 Bernhardt, Annette, Ruth Milkman, Nik Theodore, Douglas Heckathorn, Mirabai Auer, James DeFilippis, Ana Luz González,
et al. 2009. “Broken Laws, Unprotected Workers: Violations of Employment and Labor Laws in America’s Cities.” Center
for Urban Economic Development, National Employment Law Project, and UCLA Institute for Research on Labor and
48
49
THE STATE OF LABOR MARKET COMPETITION
(2017), for example, find no evidence of decreasing discrimination in hiring against Black workers.
191
More generally, lower wage growth and a declining labor share have had a greater eect on lower- and middle-
income workers than high wage workers and business owners. As a result, wage stagnation has a disproportionate
impact on women and workers of color who, in any case, receive lower wages than men or white workers. Gould
(2020, Table 3) shows that stagnation of wage growth among the lower 90 percent of earners was accompanied
by increased within-group wage inequality—wages grew by less within each decile for Black workers.
192
Hispanic
workers fared somewhat better, with their wages rising relative to white workers between 2000 and 2019 but
earned generally 25 percent less than white workers at every decile.
While the gender wage gap continues to narrow, progress in closing the dierence in men and women’s earnings
has slowed in the last two decades compared to the 1980s and 1990s when female educational attainment
improved and wages in male dominated industries faced weaker labor demand. In 2020, the typical woman
working full-time, year-round earned only 83 cents for every dollar earned by the typical man working full-time,
year-round. And the wage gap is much wider for most women of color, contributing significantly to economic
inequality.
193
For example, Hispanic women earned 57 cents and Black women earned 64 cents compared to every
dollar earned by white, non-Hispanic men in 2020.
194
The persistent gender wage gap is also tied to increased wage
dispersion as wage growth has slowed for all lower and middle wage workers, relative to top earners.
Wage stagnation also has a disproportionate negative impact for minorities because these households derive less
income from other sources. Black and Hispanic workers have a much lower homeownership rate than whites—
approximately 40 percent and 50 percent respectively compared to over 70 percent for whites. The dramatic
wage stagnation aer 2000 coincided with the 2008 housing bust that decimated the largest source of wealth for
most Americans. The wave of foreclosures in the wake of the 2008 housing crises dramatically lowered minority
homeownership rates, meaning that these households are unlikely to have benefited from the recent increase in
house prices. Reduced frequency of homeownership leads to less generational wealth, increasing the dependency
of Black and Hispanic Americans on wage growth to build income and wealth.
Declining Business Investment and Productivity Growth
Lower employment is a consequence of decreased labor market competition, as discussed in the section on
monopsonistic theory. So long as capital and labor are complementary, which they oen are, lower employment
also results in lower investment. Considered in a dierent way, the exercise of monopsony power behaves as if
it were a tax on labor as an input. This ‘tax’ leads to lower production and deadweight loss, and therefore lower
investment in capital.
More generally, business investment has been relatively weak in recent decades despite a rising profit share and
repeated reductions in corporate taxation. Weak wage growth and a large pool of low-priced labor likely dampen
business incentives to invest in tangible capital. In a tight labor market, firms would need to find ways to utilize
scarce labor more productively and would likely boost investment to make workers more productive.
Employment. https://www.nelp.org/wp-content/uploads/2015/03/BrokenLawsReport2009.pdf.
191 Quillian, Lincoln, Devah Pager, Ole Hexel, and Arnfinn H. Midtbøen. 2017. “Meta-analysis of field experiments shows no
change in racial discrimination in hiring over time.” Proceedings of the National Academy of Sciences 114 (41): 10870–10875.
192 Gould, Elise. 2020. “State of Working America Wages 2019.” Economic Policy Institute, February 20, 2020. https://www.epi.
org/publication/swa-wages-2019/.
193 Bleiweis, Robin, Jocelyn Frye, and Rose Khattar. 2021. “Women of Color and the Wage Gap.” Center for American Progress,
November 17, 2021. https://www.americanprogress.org/article/women-of-color-and-the-wage-gap/.
194 U.S. Bureau of Labor Statistics. 2021. “Highlights of Women’s Earnings in 2020.” BLS Reports Report 1094. https://www.bls.
gov/opub/reports/womens-earnings/2020/home.htm. See also, Bleiweis, Frye, and Khattar (2021).
THE STATE OF LABOR MARKET COMPETITION
Monopsony power can also decrease aggregate productivity, provided that firm-level productivity and market
power are correlated, as Mertens (2020) argues.
195
Given that correlation, higher-productivity firms reduce
their output disproportionately, relative to lower-productivity firms. Naturally, this increases low-productivity
firms’ share of national production, resulting in decreased aggregate productivity. As Gutierrez and Philippon
(2017, 2020) show, the largest firms, which hold an increasing share of employment and sales, have stagnant
investment rates, and a decreasing relative contribution to aggregate productivity growth.
196
Thus, the largest
firms are becoming more profitable while investing less and generating less productivity growth. To be clear, a
causal link from lower labor market competition to decreases in investment and productivity growth has yet to be
established. However, increased concentration does appear to be a driver of weak investment, low productivity
growth, and high profits and likely contributes to lower labor market competition.
Firm Formation and Innovation
Business formation and exits have both declined since the early 1980s. As a share of the total number of firms,
about 20 percent fewer firms were created in 2018, compared with 1982.
197
Over the same period, the share of
payroll attributable to firms less than 5 years old with at least one employee on payroll declined by almost a
quarter, from 38 percent in 1982 to about 29 percent in 2018 (Congressional Budget Oice 2020). Accordingly, firms
today are, on average, older than they were in the past.
The decline in business formation is likely driven by several factors. In their 2020 analysis, the Congressional
Budget Oice (CBO) pointed to the aging domestic workforce as a key factor, though they note immigration
(especially high skilled immigration) has oset some of that decline. Cyclical factors (e.g., recessions) play a role as
well. Moreover, the shi in economic activity to larger and older firms may not necessarily have a negative impact
on welfare (Autor et al. 2020).
However, the decline in business formation is potentially troubling because it could suggest that dominant
firms maintain their lead status by erecting barriers to entry rather than maintaining their dominance through
innovation. Gutierrez and Philippon (2019) provide evidence to this eect, showing that firm entry has become less
sensitive to market valuations over time (i.e., high profits do not lead to increasing firm entry). The authors provide
evidence that large firms have been able to erect hurdles to the entry of new firms.
198
As Aghion, Akcigit, and Howitt (2015) note, more intense competition tends to encourage innovation in “frontier”
firms (firms that are in sectors at the cutting-edge of technology), whereas barriers to entry become increasingly
detrimental to growth as a country approaches the technological frontier.
199
Using a structural model, Akcigit and
Ates (2019) find that declines in firm entry and worker reallocation towards new firms reflects slower knowledge
195 Mertens, Matthias. 2020. “Labour Market Power and Between-Firm Wage (In)Equality.” Leibniz Institute for Economic
Research Halle Discussion paper 13/2020.
196 Gutierrez, German, and Thomas Philippon. 2017. “Declining Competition and Investment in the U.S.” National Bureau
of Economic Research Working Paper 23583; and Gutierrez, German, and Thomas Philippon. 2020. “Some Facts about
Dominant Firms.” National Bureau of Economic Research Working Paper 27985.
197 Congressional Budget Oice. 2020. “Federal Policies in Response to Declining Entrepreneurship.” December 2020. https://
www.cbo.gov/system/files/2020-12/56906-entrepreneurship.pdf.
198 Gutierrez, German, and Thomas Philippon. 2019. “The Failure of Free Entry.” National Bureau of Economic Research Working
Paper 26001.
199 Aghion, Philippe, Ufuk Akcigit, and Peter Howitt. 2015. "Lessons from Schumpeterian Growth Theory."American Economic
Review105 (5): 94–99. Intuitively, the reason why barriers to entry discourage growth in a “Schumpeterian growth” model
is because new firms innovate to gain market share, thus threatening incumbent firms and forcing them to innovate as
well. With barriers to entry, incumbent firms face fewer incentives to innovate and, instead, extract monopoly rents from
their dominant position.
50
51
THE STATE OF LABOR MARKET COMPETITION
diusion from frontier firms to new entrants,
200
which could reflect impediments to worker mobility.
. The use of non-compete clauses, especially among internet-based commerce firms, could be discouraging
firm entry (Congressional Budget Oice 2020). For instance, Marx, Strumsky, and Fleming (2009) finds that
an unintended change in Michigan law boosting the enforceability of non-compete agreements led to sharp
declines in the mobility of patent holders.
201
Restricting the use of non-compete agreements and other restrictive
employment agreements could allow for new firm creation, as workers at incumbent firms could leave the firm to
pursue new ideas, thereby forcing incumbent firms to innovate to stay dominant.
Declining Worker Mobility and Productivity Growth via Reallocation
The reallocation of workers across firms is a key driver of firm-level and overall productivity growth. Workers quit
their jobs and search for new jobs that better fit their skills, while firms are seeking the right mix of workers to
improve their productivity. A large economic literature provides both theoretical and empirical evidence for linking
the pace of reallocation to aggregate productivity growth.
Pre-pandemic, job reallocation (the creation and destruction of new jobs) and worker reallocation (workers
quitting and finding new work) have been declining steadily over several decades.
202
Worker mobility across
space has also declined over time.
203
Like the literature on declining firm entry rates, demographic factors or the
changing industrial composition of the economy may explain some of the decline in reallocation and spatial
mobility. Akicigit and Ates (2019) link declining job and worker reallocation to slower diusion of ideas from market
leading firms to new entrants. Davis and Haltiwanger (2014) argue that factors inhibiting competition, including
specifically occupational licensing, may account for declining labor market dynamism.
204
They find a particularly
large decrease in worker reallocation among younger workers and workers with lower educational attainment.
Kleiner and Krueger (2013) also document increasing prevalence of occupational licensing that may inhibit worker
switching across occupations and space. It is also likely that restrictive employment agreements are contributing
to lower levels of worker mobility.
200 Akcigit, Ufuk, and Sina T. Ates. 2019. “Ten Facts on Declining Business Dynamism and Lessons from Endogenous Growth
Theory.” National Bureau of Economic Research Working Paper 25755.
201 Marx, Matt, Deborah Strumsky, and Lee Fleming. 2009. “Mobility, Skills, and the Michigan Non-Compete Experiment.”
Management Science 55 (6): 875–889.
202 Decker, Haltiwanger, Jarmin and Miranda (2020) and Haltiwanger (2015) summarize the decline in job and worker
reallocation since 1980 and its implications for aggregate productivity growth. Decker, Ryan A., John Haltiwanger, Ron
S. Jarmin, and Javier Miranda. 2020. “Changing Business Dynamism and Productivity: Shocks versus Responsiveness.”
American Economic Review 110 (12): 3952–3990; and Haltiwanger, John. 2015. “Job Creation, Job Destruction, and
Productivity Growth: The Role of Young Businesses.” Annual Review of Economics 7 (August): 341–358.
203 Molloy, Raven, Christopher L. Smith, and Abigail K. Wozniak. 2014. “Declining Migration within the U.S.: The Role of the
Labor Market.” National Bureau of Economic Research Working Paper 20065.
204 Davis, Steven J., and John Haltiwanger. 2014. “Labor Market Fluidity and Economic Performance.” National Bureau of
Economic Research Working Paper 20479.
THE STATE OF LABOR MARKET COMPETITION
BIDEN ADMINISTRATION PROPOSALS TO PROMOTE
LABOR MARKET COMPETITION
As this report makes clear, insuicient labor market competition has harmful eects on workers and the economy
and worsens inequality. In response, President Biden issued an Executive Order on Promoting Competition in the
American Economy, establishing a whole-of-government eort to reduce the trend of corporate consolidation
and improve competition for American workers, consumers, and small businesses. Pursuant to this Order, federal
agencies are acting to develop and implement several proposals to promote competition in labor markets. Robust
labor market competition requires careful maintenance and is a critical component to promoting economic
growth, spurring innovation, and addressing economic inequality. The following initiatives and policy proposals
will bolster labor market competition and increase workers’ bargaining power.
Proposed Legislation
The President is calling on Congress to pass proposed legislation that would promote increased competition in
labor markets by improving workers’ ability to negotiate fair wages and a larger share of income. The legislative
proposals discussed below would greatly enhance the negotiating power of workers and mitigate the decline
in wages that have contributed to a historic rise in income inequality. By restoring balance to the labor market,
the proposed legislation would force employers to compete for workers on a level playing field and ensure that
workers get their fair share of the value they create.
Increasing union representation can help increase workers bargaining power and raise wages. Recent survey
data suggests that roughly half of nonunion workers would vote for a union if they had the opportunity and the
percent of Americans who support labor unions stands at 68 percent, the highest since the early 2000s. Despite
this support, private-sector unionization stood at just 6.1 percent in 2021.
205
Current labor law is a major obstacle
to unionization as workers face multiple hurdles and employers can intimidate and coerce workers, oen incurring
no penalties for retaliatory actions against workers or interfering with union election processes.
Protecting and Expanding Workers’ Right to Organize: The President and Vice-President have called for Congress
to pass the Richard L. Trumka Protecting the Right to Organize Act and the Public Service Freedom to Negotiate
Act (PSFNA). These bills would ensure more private- and public-sector workers nationwide have a genuine right
to organize and bargain collectively. They would also promote racial income equality by shrinking the Black-white
wage gap by boosting worker power. The PSFNA would establish minimum standards for collective bargaining by
state and local public service workers; these workers lack formal bargaining in half of the states. President Biden
and Vice-President Harris also have endorsed several proposals to expand labor rights to more workers (especially
workers of color, women, and immigrants) and help counteract monopsony power in sectors not covered by
current labor laws. These include guaranteeing labor rights to farmworkers and domestic workers—two segments
of the labor force excluded from the protections of the National Labor Relations Act. For example, the National
Domestic Workers’ Bill of Rights, which Vice President Harris championed in the Senate and the President has
endorsed, would expand federal labor law to domestic workers and create a new wage and standards board for
regulating working conditions in the sector.
Raising the Federal Minimum Wage: Raising the minimum wage is a straightforward approach to addressing lower
wages under monopsony and can help increase employment. However, the federal minimum wage has remained
unchanged since 2009,
206
during which time inflation has eroded the purchasing power of the minimum wage.
205 Bureau of Labor Statistics. 2022. “Union Members Summary.” Last modified January 20, 2022. https://www.bls.gov/news.
release/union2.nr0.htm.
206 The Economic Policy Institute. 2022. “Minimum Wage Tracker.” Last modified January 1, 2022. https://www.epi.org/
52
53
THE STATE OF LABOR MARKET COMPETITION
Workers in states that have not enacted meaningful increases to the state’s minimum wage have been le behind
as a result of this decline in purchasing power. President Biden has endorsed raising the federal minimum wage to
$15 per hour, indexing future increases of the federal minimum wage, phasing out the tipped minimum wage, and
eliminating the subminimum wage for teen workers and workers with disabilities.
207
Raising the federal minimum
wage would give nearly 32 million Americans a raise and would boost the purchasing power of low-income families
allowing them the opportunity to more fully participate in the growing economy.
208
Restricting the Use of Mandatory Arbitration and Class Action Waivers: Legislation restricting the use of mandatory
arbitration and limits on class actions would prevent employers from forcing employees into forfeiting the
opportunity to have their case heard by a judge and jury or their right to join together in a collective action
to remedy collective harms. Congress has already taken a first step to limit the enforceability of mandatory
arbitration and class waivers by enacting the Ending Forced Arbitration of Sexual Assault and Sexual Harassment
Act of 2021, which makes mandatory arbitration and class waiver provisions invalid and unenforceable in court
for claims involving sexual assault or harassment. President Biden signed the Ending Forced Arbitration of Sexual
Assault and Sexual Harassment Act into law on March 3, 2022.
Mandatory arbitration agreements undercut labor market competition by eectively reducing wages paid to
employees by arbitrarily imposing liability costs on employees. When workers are unable to negotiate for higher
pay and are forced into arbitration, their real wage rate is too low, preventing the labor market from functioning
eiciently. President Biden supports banning employers’ use of forced arbitration and class waivers to restore
worker rights and impose accountability on employers. Mandatory arbitration and class action waivers can distort
labor markets by insulating businesses from the full costs of doing business, primarily by limiting liability and
public exposure. DOL is prioritizing enforcement against employers that employ mandatory arbitration or class
action waivers as a check against employers’ abuse of their market power. Recently, a court held that DOL’s ability
to enforce laws through the courts was not limited by an arbitration agreement between an employee and their
employer.
209
Criminal Antitrust Anti-Retaliation Act of 2019: OSHA’s Whistleblower Protection Program is implementing the
Criminal Antitrust Anti-Retaliation Act of 2019 (CAARA). The law provides legal protections for employees who
blow the whistle on criminal antitrust violations by prohibiting employers from taking punitive actions against
whistleblowers for reporting these violations to their employer or assisting a federal government investigation into
a criminal antitrust violation. In addition to OSHA’s ongoing enforcement and outreach, OSHA plans to publish in
May 2022 an Interim Final Rule promulgating procedures for the handling and investigation of CAARA claims.
Antitrust Enforcement
In recent years, the federal antitrust agencies—the Antitrust Division of DOJ (“DOJ” or “Antitrust Division”) and
FTC—have prioritized competition enforcement and advocacy in labor markets by increasing their institutional
capacity for labor market enforcement, bringing expertise in-house, and reviewing and, where appropriate,
reforming enforcement practices agency-wide to respond to the challenges raised by the modern economy. By
leveraging their civil, research, and rulemaking powers, the Antitrust Division and FTC have a significant role
to play in improving competitive conditions in labor markets by, among other things, reducing concentration
minimum-wage-tracker/.
207 House Committee on Education and Labor. n.d. “Raise the Wage Fact Sheet.” Accessed March 3, 2022. https://edlabor.
house.gov/download/hr-603-raise-the-wage-act-fact-sheet
208 Ibid.
209 Department of Labor. 2021. “Court Airms US Department of Labor’s Independent Authority to Recover Unpaid Wages,
Damages in Court for Employees Who Signed Private Arbitration Agreements.” News release, September 23, 2021. https://
www.dol.gov/newsroom/releases/sol/sol20210923
THE STATE OF LABOR MARKET COMPETITION
and disciplining the use and abuse of restrictive employment agreements, including non-compete agreements,
forced arbitration clauses, non-solicitation clauses, and other covenants that exacerbate bargaining asymmetries
between workers and employers. Both agencies can clarify public guidance to bolster labor market competition,
and challenge civil action mergers and unilateral conduct that harm labor markets. The Antitrust Division has sole
jurisdiction to criminally prosecute conspiracies and other collusive agreements among employers.
DOJ Criminal Enforcement in Labor Markets
The Antitrust Division has both civil and criminal enforcement authority. In particular, the Antitrust Division
prosecutes criminal conspiracies among competitors, including price fixing, bid rigging, and market allocation.
This includes agreements among employers to fix wages, which is price fixing in the labor market, and to allocate
labor markets using no-poach agreements.
210
In recent years, the Antitrust Division’s criminal program has become
increasingly central to its eorts to prosecute and deter wage fixing and no-poach agreements, which steal from
workers by depriving them of competitive wages, benefits, and other terms of employment.
Beginning in October 2016, the Antitrust Division made a series of public statements indicating that it intended to
criminally prosecute “naked” no-poach and wage-fixing conspiracies.
211
That decision followed from longstanding
caselaw establishing that these restraints are equivalent to agreements to fix product prices and allocate product
markets—conduct that the Antitrust Division has prosecuted for over 100 years. Indeed, the Supreme Court
held long ago that the Sherman Act applies equally to all industries and markets, including labor markets, and
the conduct of employers is not entitled to special treatment under U.S. antitrust laws, except in the context of
legitimate collective bargaining and other labor union activities.
212
The Antitrust Division views rooting out criminal
collusion in labor markets as part of its overall mission to deter, detect, and prosecute cartels.
Over the last several years, the Division has continued to invest substantial time and resources to ensure vigorous
competition in labor markets. These eorts, which included substantial public engagement and awareness
building, led to a notable increase in the number of citizens who reported alleged conspiracies to the Antitrust
Division since October 2016. Over the same period, labor market investigations have comprised a growing portion
of the Antitrust Division’s docket. Between December 2020 and December 2021, the Antitrust Division charged five
criminal cases for alleged collusion in labor markets, including four companies and nine individuals.
213
In January
2022, the Antitrust Division filed a further indictment charging four managers of home health care agencies with
participating in a conspiracy to suppress the wages and restrict the job mobility of essential workers during the
COVID-19 pandemic.
214
The Antitrust Division’s criminal enforcement program has led to the prosecution of long-
running employer conspiracies against workers in multiple critical markets, including physical therapy, dialysis
nursing, home health care services, and aerospace, with more active labor market investigations currently
underway.
Remedial measures are another important tool for the Antitrust Division in protecting competition for workers.
In particular, the Division may require provisions regarding labor market competition in corporate criminal
210 See United States v. Knorr-Bremse AG, et al., 18-747 (D.D.C.) (April 4, 2018).
211 Renata B. Hesse. “The Measure of Success: Criminal Antitrust Enforcement During the Obama Administration.” Remarks at
the 26th Annual Golden State Antitrust, UCL and Privacy Law Institute, November 3, 2016. Department of Justice, https://
www.justice.gov/opa/speech/file/908301/download.
212 See Final Judgement, United Mine Workers of Am. v. Pennington, 381 U.S. 676, 85 S. Ct. 1607, 14 L. Ed. 2d 626 (1965).
213 Indictment,United States v. Jindal, No. 4:20-cr-00358 (E.D. Tex. Dec. 9, 2020); Indictment,United States v. Surgical Care
Ailiates, LLC,No. 3-21-CR0011-L (N.D. Tex.Jan. 5, 2021);Indictment,United States v. Hee et al., No. 2:21-cr-00098-RFB-
BNW (D. Nev. Mar. 30, 2021); Indictment,United States v. DaVita, Inc., No. 21-cr-00229-RBJ (D. Colo. July 14, 2021); and
Indictment, United States v. Patel et al., No. 3:21-cr-00220-VAB (D. CT. Dec. 15, 2021). See also United States v. Jindal, No.
4:20-cr-00358 (E.D. Tex. Nov. 29, 2021) (denying defendants’ motion to dismiss).
214 Indictment, United States v. Manahe et al., No. 22-cr-0013-JAW (D. Maine, January 27, 2022).
54
55
THE STATE OF LABOR MARKET COMPETITION
resolutions where the charged conduct restrained or impacted worker mobility.
At its core, the Antitrust Division is committed to prosecuting naked conspiracies in labor markets because they
rob workers of competitive wages, benefits, and other terms of employment. They also deprive honest businesses
of talented workers who contribute substantially to the products and services on which Americans rely. While this
work is principally criminal enforcement, it also reflects a commitment to ensuring free market competition for
workers’ labor.
DOJ and FTC Civil Enforcement and Competition Advocacy
Civil enforcement represents an equally important, and in some respects even more expansive, toolset for
enforcers to improve labor market competitiveness because it reaches a broader swath of competition concerns.
The antitrust agencies are currently committed to using their civil authorities to detect, investigate, and challenge
anticompetitive non-compete agreements, mergers that create or enhance monopsony power in labor markets,
the anticompetitive exercise of monopsony power, and information sharing by employers. To aid these eorts, the
Antitrust Division and the FTC have issued public guidance that reflects the importance the U.S. antitrust agencies
place on protecting competition in labor markets and may update that guidance to reflect improved information
about market dynamics and competition analysis.
As part of their respective competition advocacy programs, the Antitrust Division and FTC have recently filed
statements of interest and amicus briefs in multiple significant labor market matters. In March 2021, the agencies
filed an amicus brief inNCAA v. Alstonon behalf of college athletes.
215
A unanimous Supreme Court decided in the
athletes’ favor that colleges could not agree to limit the education-related benefits oered to students, rejecting an
argument that these limits preserved amateurism and widened consumer choice by providing a unique product—
amateur college sports as distinct from professional sports.
216
BeforeNCAA v. Alston, the Antitrust Division filed a
number of amicus briefs and statements of interest urging courts to uphold the per se rule for naked restraints in
labor markets, includingIn re Railway Employee No-Poach Antitrust Litigation,Seaman v. Duke University,andAya
v. AMN Healthcare.
217
In April 2020, the agencies warned employers, staing companies, and recruiters that despite
the need for unprecedented cooperation among public and private organizations to respond to the spread of
COVID-19, the agencies would be closely monitoring labor markets to challenge any anticompetitive conduct that
harms workers.
218
In February 2022, the DOJ filed an amicus brief before the National Labor Relations Board (NLRB)
highlighting the potential impacts of misclassification on labor market competition and supporting the NLRB in
its eorts to create a “sound, up-to-date, consistent approach to worker classification that adequately protects
workers’ rights to organize.” DOJ also filed a statement of interest in a private non-compete case in Nevada arguing
that competition-suppressing agreements should be subject to strict antitrust scrutiny, especially where (as
alleged in the pleadings) the eect of enforcement would be to prevent health care workers from earning a living or
serving patients in their home metro area.
219
215 Brief for the United States as Amicus Curiae Supporting Respondents, NCAA v. Alston, 141 S. Ct. 2141, 210 L. Ed. 2d 314
(2021).
216 Final Judgement, National Collegiate Athletic Association v. Alston, 141 S. Ct. 2141, 210 L. Ed. 2d 314 (2021).
217 Statement of Interest of the United States,In Re: Railway Industry Employee No-Poach Antitrust Litigation, 395 F. Supp. 3d
464 (W.D. Pa. 2019) (No. 2:18-mc-00798-JFC); Statement of Interest of the United States,Seaman v. Duke Univ., No. 1:15-
CV-462, 2019 WL 4674758(M.D.N.C. 2019); and Brief of Amicus United States of America in Support of Neither Party,Aya
Healthcare Serv., Inc. v. AMN Healthcare, Inc. et al., 9 F.4th 1102 (9th Cir. 2021).
218 Department of Justice and Federal Trade Commission. 2020. “Joint Antitrust Statement Regarding COVID-19 and
Competition in Labor Markets.” Press release, April 2020.
219 Brief of the United States Department of Justice as Amicus Curiae at 9, The Atlanta Opera, Inc., 10-RC-276292 (NLRB
February 10, 2022); and Statement of Interest of the United States, Beck v. Pickert Medical Group, P.C., No. CV21-02092
(Nev. 2nd. Jud. Dist. Ct. February 28, 2022).
THE STATE OF LABOR MARKET COMPETITION
Consistent with the DOJ’s recent filing before the NLRB, the agencies intend to continue to seek opportunities to
provide guidance to courts in cases that implicate the proper scope of the antitrust exemptions that protect labor
organizing. Although multiple federal statutes exempt labor organizing from the antitrust laws’ purview, federal
courts have held that these protections apply only to workers formally classified as employees.
220
As a result,
collective action and organizing by certain workers—including those who have the terms of their work dictated by
a firm yet are classified as non-employees—may be susceptible to an antitrust lawsuit, including by private parties.
When appropriate, the agencies may consider providing guidance on how they interpret the antitrust laws with
respect to organizing activities that are exempt from antitrust prosecution.
In addition to these case-specific interventions, the Antitrust Division and FTC are considering updates to
their guidance, particularly in areas where changes in the economy may have led some people to incorrectly
interpret the agencies’ past guidance in ways that are insuiciently protective of workers’ access to robust,
competitive labor markets. Currently, the Antitrust Division and FTC are working to revise their joint Antitrust
Guidance for Human Resource Professionals, which was published in 2016 to help human resources professionals
“implement safeguards to prevent inappropriate discussions or agreements with other firms” (Department
of Justice Antitrust Division and Federal Trade Commission 2016).This guidance was primarily intended to
educate and inform business and human resource professionals about how the antitrust laws apply to hiring and
compensation decisions. However, due to recent case experience and research that have shown that information-
sharing, particularly in concentrated markets, may have potentially significant anticompetitive eects even
when purportedly anonymized, the agencies are in the process of updating this guidance to reflect this new
information.
221
Similarly, the agencies believe that the principles for addressing and preventing concentration embodied in the
Horizontal Merger Guidelines apply just as much to labor markets as to any other market. In January 2022, the
agencies announced a joint eort to solicit updated public input on the Horizontal Merger Guidelines in order to
better detect and prevent illegal, anticompetitive deals in today’s modern markets, including labor markets.
222
As part of this eort, some commentators have suggested that the applicability of antitrust principles to labor
markets should be more explicitly articulated, and the Antitrust Division and FTC are considering this feedback as
they review the Horizontal Merger Guidelines. The agencies are also considering commentators’ contention that
labor markets may become subject to market power at more moderate levels of employer concentration than
product markets, due to the employee-side search frictions that characterize labor markets.
The agencies’ work on the Horizontal Merger Guidelines will reflect lessons learned from multiple recent merger
cases brought by the agencies that implicated the rights of workers. In November 2021, the Antitrust Division filed
to stop a proposed merger between Penguin Random House and Simon & Schuster, two large book publishers,
primarily on the grounds that it would harm competition for author labor by giving Penguin Random House,
currently the largest of the five remaining traditional publishers, outsized control over publication opportunities
and lead to reduced pay for authors.
223
In 2017, the D.C. Circuit airmed the Division’s successful challenge of
220 15 U.S.C. § 17.; L.A. Meat & Provision Drivers Union, Local 626 v. United States, 371 U.S. 94 (1962); United States v. Women’s
Sportswear Mfg. Ass’n, 336 U.S. 460 (1949); and Columbia River Packers Ass’n v. Hinton, 315 U.S. 143 (1942).
221 See Complaint, Kra Heinz Foods Co. v. Amick Farms et al., 20-cv-02278 (N.D. Ill. April 11, 2020) (alleging use of 3rd party
agricultural information to “enabling Defendants to monitor what each producer was doing in furtherance of … concerted
action among the producers.”); and Complaint, U.S. v. Sinclair Broadcast Group, Inc., 18-cv-02609 (D.D.C. November
13, 2018) (alleging “information exchanges [that] distorted the normal price-setting mechanism in the spot advertising
market and harmed the competitive process.”).
222 Federal Trade Commission. 2022. “Federal Trade Commission and Justice Department Seek to Strengthen Enforcement
Against Illegal Mergers.” Press release, January 18, 2022. https://www.c.gov/news-events/press-releases/2022/01/c-
and-justice-department-seek-to-strengthen-enforcement-against-illegal-mergers.
223 Complaint, United States v. Bertelsmann SE & Co. et al., 16-cv-02886 (D.D.C. November 21, 2021).
56
57
THE STATE OF LABOR MARKET COMPETITION
Anthem’s proposed acquisition of Cigna, a merger of two significant health insurers that would have reduced
reimbursement rates for physicians in multiple markets.
224
In that case, the labor harms were alleged alongside
product-market harms, underscoring the notion that antitrust enforcement in labor markets can complement
enforcement in product markets. Similarly, two private duty nursing providers called o their proposed merger
aer the FTC raised concerns about potential eects on competition for nursing services and for employing nurses
in local markets across the country.
225
The agencies also will be attentive to the over-broad use of non-compete clauses against employees in conjunction
with mergers, as they can raise barriers to entry in markets where workers are a key input to eective competition.
For instance, the FTC recently issued an order against a national chain of dialysis clinics to remedy concerns that
its acquisition of additional clinics would reduce competition for outpatient dialysis services in Provo, Utah. In
addition to requiring divestitures, the FTC’s order prohibits the company from entering or enforcing any non-
compete agreements with physicians that would restrict their ability to work for a competitor.
226
Research and Rulemaking
To establish a foundation for future eorts to protect workers, in December 2021, the Antitrust Division and FTC
concluded a two-day public workshop on the subject, entitled “Making Competition Work: Promoting Competition
in Labor Markets.” The workshop convened lawyers, economists, academics, policy experts, labor groups, and
workers, and covered recent developments at the intersection of antitrust and labor, as well as implications
for eorts to protect and empower workers through competition enforcement and rulemaking. Feedback and
comments obtained from the workshop will be incorporated into the agencies’ eorts going forward, including
with respect to enforcement, guidelines, and rulemaking aecting labor market antitrust enforcement.
In addition to its authority to bring law suits to prohibit unfair methods of competition, the FTC Act gives the
FTC authority to identify and prohibit unfair methods of competition through a rulemaking process that follows
the Administrative Procedure Act.
227
The FTC held a workshop in 2020 to discuss how it could use its rulemaking
authority to address the overuse of non-compete clauses, and several organizations, including a group of 19
state attorneys general, have petitioned the agency to initiate a rulemaking to limit their use.
228
As suggested in
the President’s Executive Order on Competition, the Chair of the FTC is encouraged to work with the rest of the
Commission to exercise the FTC’s statutory rulemaking authority to curtail the use of non-compete clauses and
other clauses that may unfairly limit worker mobility.
Supporting Occupational Licensing Reform Eorts
To better understand and reduce the impacts of ineicient licensing requirements, the DOL has previously
awarded several grants for states to review the licensing requirements for various occupations and reduce the
224 Complaint, United States et al. v. Anthem Inc. et al., 16-cv-01493 (D.D.C. July 21, 2016).
225 Federal Trade Commission. 2022. “Statement of FTC Chairman Regarding Announcement that Aveanna Healthcare and
Maxim Healthcare Services Terminated Their Acquisition Agreement.” Press release, January 30, 2020. https://www.c.
gov/news-events/press-releases/2020/01/statement-c-chairman-regarding-announcement-aveanna-healthcare.
226 In re DaVita Inc., FTC File No. 21-10013 (October 25, 2021).
227 National Petroleum Refiners Association v. FTC, 482 F.2d 672 (D.C. Cir. 1973).
228 Federal Trade Commission. 2020. “Workshop: Non-Competes in the Workplace: Examining Antitrust and Consumer
Protection Issues” News release, January 9, 2020. https://www.c.gov/news-events/events-calendar/non-competes-
workplace-examining-antitrust-consumer-protection-issues; See Petition for Rulemaking to Prohibit Worker Non-
Compete Clauses by Open Markets Institute, et al., https://static1.squarespace.com/static/5e449c8c3ef68d752f3e70dc/t/
5eaa0486252116d1dd04c1/1588200595775/Petition-for-Rulemaking-to-Prohibit-Worker-Non-Compete-Clauses.pdf; and
Oice of the Attorney General of the District of Columbia. 2020. “Public Comments of 19 State Attorneys General.” March
2020. https://oag.dc.gov/sites/default/files/2020-03/FTC-Comment-Letter-Non-Compete-Clauses-Workplace.pdf.
THE STATE OF LABOR MARKET COMPETITION
barriers to entry into excessively consolidated occupations. These grants were also intended to improve labor
mobility in licensed occupations with an emphasis on transitioning veterans to licensed civilian occupations and
improving portability for military spouses. These investments yielded tangible results including a searchable
database of licensing requirements for 48 occupations,
229
and comprehensive reports on the barriers facing
vulnerable communities, including veterans and military spouses, justice-involved individuals, and immigrants
with work authorization. These grants laid a foundation from which to launch future reform eorts.
Several of these grants have since expired; two grants, one to the National Council of State Legislatures and one
to the Council of State Governments are set to expire in 2022. These grants have helped reveal the substantial
diiculties inherent to occupational licensing reform. Many states are reticent to attempt reforms and, even
when reforms are considered, they are occupation specific and not as broad as might be ideal.
230
The federal
government, in support of this Executive Order, will do more to support state eorts at reforms, including elevating
and disseminating best practices from current and past demonstration investments, directing support for workers
pursuing occupational licensing, exploring funding and support that has been shown to be eective in the
adoption of meaningful license reforms, and improving labor market competition by increasing worker mobility.
The Department of Defense also has a grant to the Council of State Governments to work with states to promote
and expand participation in interstate licensing compacts, another major way to increase license portability. The
Licensure Portability Grant Program of the Oice for the Advancement of Telehealth, Health Resources & Services
Administration, has also supported the development of many interstate licensure portability compacts.
231
A silver
lining of the COVID-19 pandemic is that the need to rapidly and safely deploy health care professionals to areas in
need has greatly increased support for compacts and other portability initiatives. These initiatives can streamline
the process of authorizing practitioners to work across state lines, potentially increasing the supply of practitioners
in underserved areas and increasing competition. Accordingly, this is an opportune time for federal support of
portability measures, especially in health care.
Administrative Actions to Bolster Worker Power
The Administration has taken steps to increase the level of competition in labor markets, raise the minimum wage
for workers involved in federal contracting, protect workers’ rights, and incentivize employers not to unlawfully
shi costs onto workers and thereby gain unfair competitive benefits. Taken together, these changes will make
labor markets more competitive, improve worker negotiating positions, protect workers’ rights, and address
discriminatory wages.
On April 27, 2021, President Biden issued an Executive Order setting the minimum wage at $15 per hour by January
30, 2022, for workers participating on or in connection with federal contracts. This order also continues the practice
229 National Conference of State Legislatures. 2020. “The National Occupational Licensing Database.” Last modified March 24,
2020. https://www.ncsl.org/research/labor-and-employment/occupational-licensing-statute-database.aspx
230 Nunn, Ryan. 2019. “Eliminating the Anti-Competitive Eects of Occupational Licensing.” Brookings, January 17, 2019.
https://www.brookings.edu/opinions/eliminating-the-anti-competitive-eects-of-occupational-licensing/; Avery,
Beth, Maurice Emsellem, and Phil Hernandez. 2018. “Fair Change Licensing Reform Takes Hold in the States.” National
Employment Law Project, May 15, 2018. https://www.nelp.org/publication/fair-chance-licensing-reform-takes-hold-states/;
Kleiner, Morris. 2015. “Reforming Occupational Licensing Policies.” The Hamilton Project, Brookings Discussion Paper
2015-01. https://www.brookings.edu/wp-content/uploads/2016/06/THP_KleinerDiscPaper_final.pdf; and The Captured
Economy. n.d. “Occupational Licensing.” Last accessed March 3, 2022. https://capturedeconomy.com/occupational-
licensing/.
231 Health Resources and Services Administration. 2021. “Oice for the Advancement of Telehealth.” Last modified December
2021. https://www.hrsa.gov/rural-health/telehealth; and Goldman, Karen A.. 2018. “Policy Perspectives: Options to
Enhance Occupational License Portability.” Federal Trade Commission, September 2018. https://www.c.gov/system/
files/documents/reports/options-enhance-occupational-license-portability/license_portability_policy_paper_0.pdf.
58
59
THE STATE OF LABOR MARKET COMPETITION
of indexing the contractor minimum wage to inflation, phases out the tipped contractor minimum wage by 2024,
ensures at least a $15 minimum wage for federal contract workers with disabilities, and restores protections to
guides operating on federal land.
On January 21, 2022, Secretary Walsh also announced the DOL’s Good Jobs Initiative (GJI), which
providescriticalinformationto workers, employers, and government agencies as they work to improve job
qualityandcreateaccess to good jobsfree from discrimination and harassmentforallworking people. The eorts
undertaken through the GJI, together with the other actions advancing the recommendations of the White House
Task Force on Worker Organizing and Empowerment, will help strengthen workers’ bargaining power and help
mitigate employer power in labor markets. The GJI focuses on empoweringworking peopleby:
1) Providing easily accessible information to workers about theirrights including the right to bargain
collectively and form a union;
2) Engaging employerstakeholdersas partners in improving job qualityand workforce pathways to good jobs;
and
3) Supportingpartnerships acrossfederal agenciesand providingtechnical assistance ongrants, contracts,
and otherinvestmentsdesigned to improve job quality.
The GJI coordinates work done since the beginning of this administration (and oen for decades before) under
one umbrella to promote good jobs and, consistent with applicable legal authority, ensure that other agencies
continue to have access to these resources in building job quality standards and equitable pathways to those jobs.
The DOL also announced a final rule, which came into eect on December 28, 2021, placing reasonable limits
on when an employer can take credit against its minimum wage obligations, such as when a tipped employee
performs non-tipped work. This rule enhances the DOL Wage and Hour Division’s capacity to protect the rights
aorded to these essential workers, more than half of whom are women, people of color, and immigrants.
With regard to independent contractors, the DOL has withdrawn the Trump Administration’s “Independent
Contractor Rule” that inappropriately narrowed the interpretation of the Fair Labor Standards Act’s coverage and
thereby risked excluding workers from minimum wage and overtime protections.
232
As discussed in detail above,
misclassification of employees as independent contractors oen leaves employees without the benefits and labor
protections they are aorded by labor, employment, and tax laws. The National Economic Council has created
an interagency policy committee to address worker misclassification (including through legislative solutions) as
endorsed in the President’s FY 2022 budget proposal. The Wage and Hour Division also has conducted agency-wide
training to support eorts to combat misclassification and is partnering with local, state, and federal agencies
to identify and address misclassification. Additionally, DOL will conduct research into the impacts related to re-
classification on workers, an important step in understanding how misclassification aects the competitiveness of
the labor market.
Worker Organizing and Empowerment Task Force
Empowering workers to advocate for better wages and working conditions, as well as enabling them to collectively
bargain without fear of reprisal, is a worker-first approach to promoting labor market competition.
Recognizing this, President Biden issued an Executive Order creating the Task Force on Worker Organizing and
Empowerment. This Executive Order established the first-ever all-of-government approach to finding ways
that executive branch agencies can use their existing authority to facilitate worker organizing and collective
232 Wage and Hour Division. 2021. “Independent Contractor Status under the Fair Labor Standards Act: Withdrawal.”
Department of Labor, May 6, 2021. https://www.regulations.gov/document/WHD-2020-0007-4330.
bargaining.
233
The Task Force report to the President was published February 7, 2022, and set forth nearly 70
recommended actions for agencies to take to reduce barriers and promote worker organizing among both private
and public sector employees.
234
The President approved the recommendations, and the report was released to
the public in February 2022. When implemented, the Task Force recommendations should help increase worker
organizing and collective bargaining, which will give workers more collective power vis-à-vis their employers.
Reducing Job Lock and Boosting Mobility
As already noted, factors that limit worker mobility diminish bargaining power and limit the eective degree
of labor market competition. The ability of workers to quit their job for a better option, move to new locations,
or start their own business can strengthen their bargaining power and support fair wages, while fears about
inadequate access to childcare and housing can tie workers to locations, boosting the eective monopsony power
of firms. Therefore, factors that help workers move freely can be an important component of raising labor market
competition and boosting wages.
For many workers, health insurance is provided through their employer, playing an important role in any decision
to switch employers or start a business. The passage of the Aordable Care Act in 2010 greatly strengthened the
individual health insurance market, providing subsidies for households to purchase insurance and guarantee
standards of coverage. By eliminating job lock associated with health insurance, the CBO projected at the time that
some workers would start their own businesses or leave their jobs, leading to increased wages.
The American Rescue Plan provided larger tax credits for those purchasing coverage on health insurance
exchanges. The Administration proposals—if adopted—would extend these credits to make coverage more
aordable and accessible, thus further reducing job lock due to insurance coverage and strengthening worker
mobility and bargaining power.
Worker mobility can also be enhanced by better access to childcare and lower housing costs. Though many non-
economic factors impact households’ decisions of where to live, these decisions are impacted by the general
cost of housing and, for parents of young children, proximity to their parents or other caregivers. Investments in
aordable housing, childcare support, and universal pre-kindergarten provision can mitigate job lock for housing
cost or childcare reasons. These eects are likely to be modest and diicult to quantify, but, even on the margin,
higher worker mobility improves bargaining power and raises wages.
233 The White House. 2021. “Fact Sheet: Executive Order Establishing the White House Task Force on Worker Organizing
and Empowerment.” News release, April 26, 2021. https://www.whitehouse.gov/briefing-room/statements-
releases/2021/04/26/fact-sheet-executive-order-establishing-the-white-house-task-force-on-worker-organizing-and-
empowerment/.
234 The White House. 2022. “White House Task Force on Worker Organizing and Empowerment.” Report to the President,
February 7, 2022. https://www.whitehouse.gov/briefing-room/statements-releases/2022/02/07/white-house-task-
force-on-worker-organizing-and-empowerment-report/#:~:text=Today%2C%20the%20Task%20Force%20on,and%20
collective%20bargaining%20for%20federal.
60
THE STATE OF LABOR MARKET COMPETITION
