Understanding the claims about labor markets in debates on ‘Private Government‘

Download PDF

Unequal Power

Part of the Unequal Power project, an EPI initiative to reestablish the understanding in law, politics, economics, and philosophy, that equal bargaining power between workers and employers does not exist. Recognizing this inherent workplace inequality will bolster freedom, economic fairness, workplace protections and democracy.


by Elizabeth Anderson

I completed Private Government: How Employers Rule Our Lives (and Why We Don’t Talk about It) in 2017. In that work, I argued that under U.S. law, employers are dictators of their workplaces, empowered to exercise sweeping and virtually unaccountable power over their employees, even regarding their off-duty lives. This contradicts the dominant narrative—enshrined in public discourse, law, and much of economics—which represents the workplace as a domain of equals who have freely negotiated the terms of the employment contract so that it is tailored to each party’s interests. I argued that in reality, most workers don’t get the chance to negotiate terms, because the state already designed the default terms of the employment contract through the doctrine of employment at will. This and other legal doctrines hand virtually all the power cards over to the employer. Hence, workers have little choice but to acquiesce in an employment contract imposed by their employers, who have no interest in ceding the powers the state has granted them. The result is what I call “private government”—workers’ subjection to unaccountable authority.

Since I was poking holes in the dominant view of the workplace, I was not surprised by the criticisms Private Government received. My subtitle already alludes to the force of ideology in resisting my depictions of the workplace as a domain of subordination rather than freedom. Many criticisms appealed to economics. Hence, I am grateful for the work Chetan Cetty has done in providing empirical and theoretical support from economics for my arguments. Cetty shows why gig work does not offer the liberation for workers some have promised, and how monopsony offers an illuminating framework for understanding labor markets even where there are multiple employers. He argues that the lack of transparency in employment contracts reflects workers’ lack of power. He explains how coercion is present when employers extract workers’ consent to mandatory arbitration, why we can’t infer that workers are satisfied with their jobs from the fact that they don’t quit, and why the sky won’t fall if we limit the authority of employers over workers. Finally, Cetty demonstrates that economists’ cheery representations of work heavily depend on the assumption of full employment—a condition that rarely obtains for the average worker, and virtually never for Black and Hispanic workers.

In 2020–2021, the pandemic exposed many of the harsh realities of domination at work that I criticized in 2017. Front-line workers lionized in public opinion as “essential” to society were treated as disposable by their employers. Health care workers treating COVID-19 patients were fired for complaining about hospitals’ failure to supply personal protective equipment (Carville et al. 2020). Some were even forbidden from bringing in equipment they obtained at personal expense (Fadel 2020). Owners of meatpacking plants, a major source of rural COVID-19 hotspots, fended off closure by local public health authorities by lobbying the Trump administration to order them to stay open (Grabell and Yeung 2020). They even won the right to speed up their slaughter lines, even though this forced workers to crowd more closely, exposing them to higher risk of infection (Kindy, Mellnik, and Hernández 2021). Cetty shows why such abuses of employer power are not mere aberrations, but predictable outcomes of a sound economic analysis of American workplaces.

—Elizabeth Anderson is John Dewey Distinguished University Professor of Philosophy and Women’s and Gender Studies at the University of Michigan, Ann Arbor

Executive summary

In her book, Private Government: How Employers Rule Our Lives (and Why We Don’t Talk About It), political philosopher Elizabeth Anderson delivers a powerful critique of the disparate power relations between workers and their bosses. She argues that most people in the United States and other liberal societies spend their working lives—and hence a huge share of their lives overall—under the kind of autocratic rule that we would never tolerate in our public lives and that we normally associate with communist dictatorships. Workers are forced to perform in oppressive environments, are deprived of many freedoms, and are given practically no say over working conditions. Even in their nonworking lives workers are frequently subjected to employer scrutiny and sanction. And the legal framework and economic realities surrounding employment are such that exit is viable for only a small minority. Anderson calls such employer rule “private government,” by which she means “government that has arbitrary, unaccountable power over those it governs.”

Anderson’s critical examination of the contemporary workplace serves as a searing rebuke not only of the hegemony of monolithic firms over workers but also of the rhetoric of the free market as both a liberator and equalizer of workers’ ability to control their own economic lives. She aims to dispel the simplistic equation of the “state” with coercion and the “market” with freedom.

Her solution to the problem of private government is to make it public, that is, to render firms accountable to those over whom they rule. She proposes four different ways of doing so: (1) making exit a meaningfully viable option, (2) ensuring that the workplace is subject to the rule of law, (3) providing constitutional rights for workers, and (4) increasing worker voice through structures like a co-determination firm model. She is ultimately most partial to the last of these solutions, since the success of the first three ultimately depends on effective enforcement and, even if enforcement is effective, the changes only prevent employers from tyrannizing workers. They do not make employers accountable to workers.

How have Anderson’s arguments been received? Her work has generated its share of criticism from economists and political philosophers who question her view of the workplace, the labor market, worker bargaining power, firm monopsonist power, worker satisfaction, and managers’ authority. For example, these critics ask, doesn’t the gig economy give workers a way out? Can’t workers just quit if they’re dissatisfied? Can’t the terms of employment simply be made clear up front? Are workers really so unhappy? Don’t we consent to unequal relationships all the time? Aren’t her proposed reforms a hindrance to the efficient operation of firms and markets?

While this paper does not provide a complete assessment of these claims, it does lay the groundwork for what such an assessment would look like. In particular, this paper raises the pertinent questions that defenders of these claims would have to answer along with the types of empirical evidence they would have to provide to back up these answers. The goal here is to set the stage for further investigation into the veracity of these claims and, in turn, the overall merits of Anderson’s work.


In Private Government: How Employers Rule Our Lives (and Why We Don’t Talk About It), political philosopher Elizabeth Anderson advances a powerful critique of authoritarian workplaces and the disparate power relations between workers and their bosses. She argues that most people in the United States and other liberal societies spend their working lives under the kind of autocratic rule that we would normally associate with communist dictatorships. They are forced to work in oppressive environments, deprived of many freedoms, and given practically no say over working conditions. Even in their nonworking lives workers are frequently subjected to employer scrutiny and sanction. And the legal framework and economic realities surrounding employment are such that exit is viable for only a small minority. Anderson calls such employer rule “private government,” by which she means “government that has arbitrary, unaccountable power over those it governs” (Anderson 2017).

Anderson’s critical examination of the contemporary workplace, originally delivered in the form of two lectures and later compiled into the book (together with four critical responses and her reply), serves as a searing rebuke not only of the hegemony of monolithic firms over workers but also of the rhetoric of the free market as both a liberator and equalizer of workers’ ability to control their economic lives. It aims to dispel the simplistic equation of the “state” with coercion and the “market” with freedom. She proposes several strategies for combating the problem of private government, some of which involve state regulations.

Anderson’s work has generated great interest and, along with it, several criticisms that take exception to her observations, economic assumptions, and conclusions. Below is a summary of the objections discussed in this article along with the replies to them:

  • Doesn’t the gig economy offer a viable option for exiting oppressive work relationships? Not unlike craft workers in preindustrial times, gig workers are able to choose when and how long they work, and they aren’t restricted to a particular employer. But gig work continues to account for a tiny share of employment, and it is hardly free from exploitation.
  • Can’t oppressed workers just quit? The contention that workers are chained to their employers, critics claim, assumes a monopsonist model of firms that doesn’t hold true. And even if it did, big firms are still incentivized economically to give workers wide latitude because doing so allows the firm to pay lower wages. Yet dismissal of the existence of monopsony relies on an outdated conception of monopsony that ignores the modern evidence of employer power. And the contention that firms can trade off between worker freedom and wages relies on a model of equal employer–worker power that fails to reflect circumstances in the real world.
  • Can’t the terms of employment simply be made clearer up front? If the boundaries of employer control over workers were made clear from the start, workers could either consent or take a pass. But transparency can only address so much, e.g., the wages or the hours, and says little about the nature, pace, and intensity of the work. And are workers empowered enough to motivate their bosses to provide transparency, or to refuse changes in terms once on the job?
  • Are workers really so unhappy? The critics dispute Anderson’s statistics about worker discontent, saying that most workers are happy with their workplace conditions or are willing to put up with some authoritarian control in exchange for higher wages and flexible schedules. But this point assumes that workers have the bargaining power to effect this exchange, and the evidence points to declines in worker bargaining power. Moreover, surveys reveal high levels of worker dissatisfaction with the quality of their jobs.
  • Don’t we consent to enter unequal relationships all the time? We readily consent to sacrificing some liberties when we join a specific church, labor union, or book club. But an important distinguishing feature of an employment relationship is its centrality to our life and the ability to thrive.
  • Won’t the proposed reforms to employer control distort the smooth operation of the economy? Likely outcomes of fettering managers’ authority, the critics say, are greater inefficiencies in firm management, higher unemployment, a shift to black markets, and lower growth. But must managers have unfettered authority? It seems reasonable to assume that some limits could be placed on managers’ arbitrary decision-making power without making a dent in the smooth running of the firm.
  • Don’t labor markets empower workers to exit undesirable employment relationships? A majority of the criticisms of Anderson’s argument share a common feature: They assume away unemployment and take full employment as a given. But the economy has business cycles, and it is rarely at full employment.

This analysis begins with an overview of Anderson’s criticisms of the present structure of employer–employee relations. Each of the subsequent sections summarizes a distinct economic claim made to rebut Anderson’s arguments and assesses the kinds of questions we would need to answer in order to empirically verify those claims. The final section before the conclusion discusses the crucial economic assumption about full employment that underlies several of these economic claims.

Anderson’s argument in Private Government

Anderson delivers her critique of the contemporary workplace in two lectures. The first explains why we talk “as if workers are free at work, and that the only threats to individual liberty come from the state” (Anderson 2017, xxii). The second suggests how we can correct this way of talking about workers’ economic situation and the sources of oppression. Her overarching goal, she says, is ideology critique. By “ideology” she means a simplified picture that we use to comprehend and traverse our world. Anderson’s target is the free market ideology which holds that the state is the source of coercion and oppression and that markets liberate us from it. Lecture 1 begins by explaining the structure of the pre-industrial economy and how it led to us embracing this ideology, and then details why the Industrial Revolution rendered this model of the world obsolete. Lecture 2 offers an updated version of this model as well as solutions to the problems identified by it.

Anderson starts Lecture 1 by sketching the history of free market thinking. The early champions of free markets, Anderson notes, were egalitarian liberals, not the libertarians and classical liberals who are the free market’s most fervent defenders today. She explains this surprising fact by charting the development of pro-market thinking that emerged out of feudal practices prior to the Industrial Revolution. An important factor during this period was the rise of the Levellers, who fought to reform the English monarchy by calling for representative government. Their program included universal male suffrage and religious liberty, all revolutionary ideas at the time, and they saw property rights and markets as being equally crucial to liberating people from oppression. While this thinking is antithetical to the contemporary view that egalitarian justice should focus on redistribution, it was seen as integral to securing genuine freedom from hierarchy. As Anderson explains:

The personal independence of masterless men and women in matters of thought and religion depended on their independence in matters of property and trade. If the king held title to all property, then subjects with land were reduced to mere copyholders, whose customary property rights could be extinguished by laws made without their participation, such as those calling for enclosures and expulsions of residents from fens. If the church could fine dissenters in its own courts for violations of church decrees in restraint of trade, it would destroy their freedom of religion as well as their ways of making a living. (Anderson 2017, 14)

Along with the Levellers, Anderson focuses on two other egalitarian defenders of markets and property rights. The first, Adam Smith, saw the eventual liberalization of the market as crucial for securing freedom and security for all. As government-sanctioned workers’ guilds gave way to self-employment as the primary mode of employment, many peasants and domestic workers moved to cities to take up craft and tradesmanship, a transition that enabled them to escape the dominion of their landlords. The liberating effect of such self-owned forms of work is what led Smith to endorse markets. Once again, this thinking contrasts with the mainstream view of Smith today as an advocate of markets because of their efficiency and conduciveness to economic growth.

Like Smith, Thomas Paine saw self-employment as a way to protect against an oppressive state. He viewed the state as the source of all oppression and correspondingly endorsed a broadly libertarian economic view, rejecting, for instance, high taxation and state regulation of wages and limiting the role of the state primarily to protecting the “peace and safety” of private individuals.1 As Anderson explains, these reforms made sense given that taxation and wage regulation were used to keep people in poverty. The poor had to pay heavy taxes, and so removing the taxes would liberate the poor. As the dream of self-employment came to fruition in the late 18th century, endorsing state regulation of markets and property seemed unnecessary when many controlled their own labor.

However, once the Industrial Revolution came into full swing, the promise of markets and property rights vanished. Economies of scale allowed large factories and production houses to generate more at lower costs. As large firms ate up more of the labor-market share, opportunities for artisanal work that had flourished previously began disappearing. Anderson highlights a perhaps subtle but important effect of this phenomenon. Prior to the Industrial Revolution, independent contractors learned their trade by serving as apprentices, working alongside their masters, and performing the same work as their masters. This arrangement meant that masters were less likely to impose harsh work requirements. By contrast, factories employing thousands separated workers from their bosses, an arrangement that led to harsher working conditions. This transformation of the relationship between worker and boss completes the arc that connects the two quotes concerning the market with which Anderson begins the chapter, one by Smith and one by Karl Marx.2

Smith imagines the market as an equalizing force in which self-employed workers freely transact with consumers for mutual benefit. Marx’s focus is on the labor market. In his view, any appearance of equality and freedom in the labor market is a facade. As Anderson points out, “if one looks only at the conditions of entry into the labor contract and exit out of it, workers appear to meet their employers on terms of freedom and equality….[I]f one looks at the actual conditions experienced in the workers’ fulfilling the contract, the workers stand in a relation of profound subordination to their employer” (Anderson 2017, 35). The Marxist view of the market accurately reflected what the market had become as a result of the collapse of the ideology that initially fueled its rise under influences like the Levellers, Smith, and Paine.

Anderson’s second lecture supplies an alternative model to make sense of the effects of the market and property rights on workers’ lives. She begins with an evocative thought experiment that asks us to imagine life under a communist government that had extensive control over our lives, that imposed severe restrictions on both the conditions of our work as well as our lives outside of it, and that was not accountable to our choices and interests in any way. Moreover, most people could not really exit the dictatorship because the only feasible alternative would be to enter another one. She asks if the reader would consider persons in this situation to be free. She then reveals this scenario to reflect the lives of most workers in the U.S. If we would reject life under communist dictatorship as unfree, then we must accept the same verdict about the modern American workplace. This is the model Anderson wants us to adopt for how we should talk about the economic system for workers. According to this model, most workers today are subject to “private government,” that is, the arbitrary, unaccountable will of their bosses.

What makes such rule “private” is not that it functions in the private sector. Instead, Anderson uses public and private to distinguish between spheres that require accountability and those that do not. The private sphere is private because it is one in which we are not answerable to others. We can do as we wish. In the public sphere, however, our conduct can be held accountable to others. The state is traditionally viewed as public because state actors are accountable to the people—to those who are governed. However, in lording over their workers as they see fit without being answerable to them, most employers act as private governments that rule workers’ lives.

Anderson focuses on arbitrary, unaccountable control because it restricts the kind of freedom with which she takes egalitarians to be primarily concerned: republican freedom. Such freedom can be distinguished from two other popular conceptions of freedom, namely, positive and negative freedom. Positive freedom consists in the freedom to choose from a range of options, whereas negative freedom consists in freedom from interference. Republican freedom, by contrast, consists in freedom from the arbitrary, unaccountable will of others (or freedom as nondomination). All three forms of freedom are important, she notes. The private government regulating employees serves to restrict all three kinds of freedom, but more pertinently republican freedom. And just as we may impose restrictions on private property rights to ensure that the duties of noninterference (on third parties) that they entail do not impose too great a cost on employees’ negative freedom, we might similarly impose restrictions on private governments (i.e., large firms) by, say, providing legal protections for workers so they cannot be fired on the basis of their sexual preferences. Such regulation would enhance all three freedoms.

What is the evidence that firms exercise control akin to a communist dictatorship? Anderson discusses several examples. In the early 20th century, Ford paid its workers only on the condition that they kept their personal lives in order by complying with a variety of directives: “kept their homes clean, ate diets deemed healthy, abstained from drinking, used the bathtub appropriately, did not take in boarders, avoided spending too much on foreign relatives, and were assimilated to American cultural norms” (Anderson 2017, 49). Ford supervisors periodically visited workers’ homes unannounced to ensure compliance. The Affordable Care Act permits employers to impose a 30% premium penalty on workers if they fail to abide by firm-mandated fitness programs. Penn State University threatened to increase workers’ health insurance premiums by $100 per month if the workers did not complete a survey on sexual, marital, and reproductive choices. Or take the case of Amazon and its brutal treatment of warehouse workers, who are not allowed to take breaks, are expected to generate extremely unrealistic output numbers, are verbally abused and threatened on a regular basis, and are even forced to sign waivers absolving the company of work-related injuries. On top of all this, at least in nonunion environments, it is not uncommon for workers in today’s firms to be fired for being too attractive, for failing to show up at a political rally, or for engaging in certain social media activity. Only a minority are protected from such employer authority.

Anderson then discusses how our failure to update our assessment of the market economy and its relation to freedom has led to the creation of a legal framework that has ossified private government. For instance, the system of at-will employment means that no contract need be signed binding employers to particular terms. Such employer freedom may have made sense when most workers were self-employed farmers, craftspersons, and tradespersons and had a chance at a decent living. But in the current system of monolithic firm rule, employers are free to rule over workers as they please. Consequently, many employers require their workers to sign noncompete contracts, preventing them from working for a competitor firm in the same industry. Such agreements severely limit workers’ options if they wish to escape workplace dictatorship. Anderson also rebuts the suggestions from theorists of the firm that the wide latitude of control exercised over workers by employers is efficiency-enhancing by pointing to three things: (1) that this claim does not justify the extent of authoritarian control employers frequently exercise; (2) that it does not justify their oversight over workers’ nonwork lives; and (3) that Ronald Coase’s solution to the problem of private government, namely, the signing of a pre-work agreement clarifying and limiting employer control, does not obtain in actual employment contracts except in collective bargaining cases (which have declined greatly with the dismantling of unions in recent decades).

Anderson’s solution to the problem of private government is to make it public, that is, to render firms accountable to those over whom they rule. She proposes four ways of doing so: (1) making exit a meaningfully viable option, (2) ensuring that the workplace is subject to the rule of law, (3) providing constitutional rights for workers, and (4) increasing worker voice through structures like a co-determination firm model. She is ultimately most partial to the last of these solutions, since the success of the first three ultimately depends on effective enforcement and, even if enforcement is effective, the changes only prevent employers from tyrannizing workers. They do not make employers accountable to workers.

How have Anderson’s arguments been received? The next six sections will provide an overview of the claims made in opposition to her critique of the contemporary market order. While this paper does not provide a complete assessment of these claims, it does assess whether the analysis of and evidence offered for them are persuasive or not. In particular, each of these sections will raise the pertinent questions that defenders of these claims would have to answer along with the types of empirical evidence they would have to provide to back up these answers. The goal here is to set the stage for further investigation into the veracity of these claims and, in turn, the overall merits of Anderson’s work.

Is the gig economy a liberator from workplace oppression?

The first criticism of Anderson’s work focuses on the impact of the “gig economy” on workers’ economic freedom. The gig economy refers broadly to a labor market where workers are self-employed and hired for short-term, task-specific jobs, often through a digital marketplace.3 Unlike regular employment, gig workers are able to contract with employers for specific tasks, such as driving a passenger through Uber’s ridesharing services or completing a programming assignment for a tech company. They are not employed on a permanent basis or beholden to employment contracts. The nature of such employment is said to offer greater freedom to workers compared to more traditional employment forms, in which workers are hired on work schedules dictated by their employers; gig workers, at least in theory, are able to choose when and for how long they work. In addition, they are not restricted to any particular employer. Instead, they may freely choose to contract with any employer offering specific, short-term tasks. Flexibility in these areas is in stark contrast to the conditions that motivate Anderson’s criticism of more traditional forms of labor.

Presumably, as Sandrine Blanc notes in her review of Anderson’s book, the availability of gig work enables workers to choose something other than the exploitative form of employment Anderson identifies (Blanc 2018). Blanc points out that, in the present gig economy, working conditions seem “closer to preindustrial self-employment than employment in large firms” (Blanc 2018, 222). “Preindustrial self-employment,” of course, refers to the post-feudal economic forms that emerged in the 18th century. These modes of employment realized the Smithian ideal of a labor market populated by self-employed workers. As Anderson highlights, the Industrial Revolution shattered this regime, and the subsequent ossification of worker disempowerment led to the oppressive conditions facing most workers today. To some, the rise of the gig economy brings back the prospect of widespread self-employment. If workers now have the ability to take on jobs that grant them greater power relative to their employers, is Anderson’s critique still relevant? Blanc takes the emergence of the gig economy to problematize Anderson’s criticism of the current market order.4

How might we evaluate Blanc’s claims? If the gig economy is to liberate people from the dictatorship that exists within monolithic firms, then certain things must hold. But before we examine what these things are, we first need to determine what exactly constitutes the gig economy, what kind of scale it has, and whether such work is free from exploitation.

The term “gig work” has been used to describe both a very specific, narrow group of workers—those working through online platforms such as ride hailing transportation (Farrell, Greig, and Hamoudi 2018)—and anyone doing informal work, including and going beyond all self-employment.5 As Bureau of Labor Statistics (BLS) economists have written:

No clear consensus currently exists on what constitutes gig work. Most definitions include many self-employed workers, temporary workers, and independent contractors. Many definitions also include people who do gig work as their primary source of income as well as employed people who supplement their earnings with gig work. (Dorinda Allard and Polivka 2018)

Regarding the scale of the gig economy, Mary Dorinda Allard and Anne Polivka (2018) note that, contrary to the perception that self-employment now constitutes a sizable proportion of the entire workforce, “data from the Current Population Survey (CPS), the nation’s monthly labor market household survey, show that the percentage of workers who are self-employed has actually trended down over the past two decades.” In fact, a special survey BLS designed to measure nonstandard work shows that in 2017 the self-employed (on their main job) were just 6.9% of total employment, the same or lower share as in 2005 and 1995 (BLS 2018). To address the possibility that the questions used by the CPS to quantify self-employment are outdated and do not reflect how gig workers would classify themselves, Dorinda Allard and Polivka, examining data from the American Time Use Survey, found that gig work has not increased relative to traditional forms of employment.6 Evidence from tax records confirms that overall self-employment has grown modestly; according to Brett Collins et al. (2019, 3), the “share of earners participating in the 1099 workforce grew by 1.9 percentage points from 2000 to 2016, and now accounts for 11.8 percent of the workforce.” Moreover, the tax data show that “more than half of this increase occurred over 2013 to 2016 and can be attributed almost entirely to dramatic growth among gigs mediated through online labor platforms. We find that the rise in online platform work for labor is driven by earnings that are secondary and supplemental sources of income” (Collins et al. 2019, 3).

The scale of the online demand platform workforce in 2016 was fairly small, according to the tax data, as “the share of workers with labor OPE [online platform economy] income…was approximately 1 percentage point of the workforce” (Collins et al. 2019, 3). These data sources also indicate that the percentage of workers earning their main living through self-employment has not grown in several decades and that the smaller group of online platform workers has been growing but remains far too small—less than 2% of employment—to provide a meaningful alternative to private government. This is especially so since most of the online demand workers are seeking income to supplement another job: “a majority of participants only derive small amounts of income from labor OPE—fewer than half earned more than $2,500 in 2016” (Collins et al. 2019, 3–4). It seems unlikely that such work will soon become extensive enough to liberate workers.

But even if it constitutes only a small share of the economy, is gig work free from exploitation? Does it provide high-quality employment preferable to current employment opportunities? We have already seen that much of gig work is primarily part time and thus not an alternative to regular full-time work. And the analysis of tax data showed that there is no evidence that traditional work arrangements are being supplanted by independent contract arrangements reported in 1099s. An analysis of Uber driver earnings, taking into account the commissions Uber deducts, the extra payroll taxes that the self-employed must pay, auto and other driver expenses, and the need to supply one’s own pension, health, and worker compensation benefits, indicates that earnings are very low, less than $10 an hour, an amount that “falls below the mandated minimum wage in the majority of major Uber urban markets” covered by the study (Mishel 2018, 2). Uber recognizes it provides low-wage work; as it stated in its public offering, “we aim to provide an earnings opportunity comparable to that available in retail, wholesale, or restaurant services”—all sectors that provide below-average wages. Finally, it is not certain that Uber and other gig employers provide independent contractor opportunities or rather just simply misclassify a W-2 employment situation (Mishel and McNicholas 2019). Uber has been able to provide an alternative to some jobs and it has attracted millions of workers, enough to offset the very high turnover (Uber drivers last, on average, about three months) and to expand. Nevertheless, it is far from clear that the gig economy, in which transportation and ridesharing are predominant, provides an alternative to the traditional labor market for middle-wage or high-wage workers.

If the gig economy is to be a liberator for workers, then it seems that the following should hold: As a proportion of the entire economy, the gig economy must be reasonably large, for otherwise only a very small percentage of those working today would be able to avail themselves of the putative liberties of being a gig worker. Alternatively, there must at least be evidence of the gig economy expanding such that it will predictably constitute a reasonably large percentage of the labor market in the near future. However, the data suggest otherwise.

Is exit a deterrent to workplace dictatorship?

A second objection to Anderson focuses on her claim that the lack of power employees have in the workplace is in part due to their absence of viable exit options. Were it easier for workers to quit, they could leverage their right to exit to convince employers to improve working conditions. However, for most workers the alternative to leaving one “communist dictatorship,” she argues, is joining another, and so no alternative at all. Furthermore, contractual restrictions such as noncompete clauses, which prohibit employees from working with another employer within the same industry, make exiting not merely pointless but costly, since the only possible way to find alternative employment would be through significant retraining and a switch of career paths, both of which are infeasible for most workers. These two factors reduce the ability of workers to use their exit rights to pressure employers to provide greater workplace freedoms. Tyler Cowen, in his response to Anderson, contends that this point is overblown because (1) the “monopsonistic” control that Anderson alleges large firms have over the labor market is empirically unsubstantiated, and (2) large firms have profit-related incentives to provide greater workplace freedom to employees even if they are monopsonies. Let’s look at each objection in turn.

In Cowen’s first objection, concerning the suggestion that workers are effectively unable to leave their firms, he understands Anderson to be painting a monopsonistic model of the labor market. A monopsony occurs when there are many sellers but only one buyer (it is the inverse of a monopoly). If a firm exercised a monopsony, then workers would have few alternatives regarding to whom they can sell their labor, and thus exiting their firm would not be a viable option. However, Cowen argues that Anderson provides insufficient evidence to show that the monopsony model holds.7 For instance, he claims that, despite being the largest employer in the private sector, Walmart does not have monopsony power except in a few small parts of the U.S.8 As a general matter, he suggests that “most economists assign [the monopsony model] only a secondary status in explaining labor markets” (Cowen 2017, 109) and supports this claim by citing a short book review by Peter Kuhn (Kuhn 2004). Anderson, he says, does not provide sufficient evidence to counter this general perception, and without adequate evidence we have little reason to assume the monopsony model or, by extension, the view that exit is unavailable as a means of securing greater workplace freedom.

Yet even if large firms like Walmart functioned as monopsonies in the labor market, Cowen doubts that they would be incentivized to assert authoritarian control over workers. His justification is both theoretical and empirical. The theoretical claim concerns what firms have reason to do given their profit motive. It is in the financial interest of firms to provide workers with greater freedoms since doing so can allow the firms to lower wages.9 For instance, “an employer who would like to lure in more workers, but without bidding up wages for all workers, would be well-served…[by]…offering employees selective workplace freedoms” (Cowen 2017, 110). Even if large firms were monopsonies, it would be rational for them to increase rather than decrease workplace freedom, and so the existence of monopsonies is not a threat to greater workplace freedom.

Cowen thinks that this theoretical point against Anderson’s monopsony model is strengthened by the fact that it is backed up by empirical evidence. In reality, he contends, large firms do provide workers with more freedom, partly to protect the firm’s reputation among consumers. Many firms, like McDonald’s, are willing to abstain from discriminating against historically oppressed groups to avoid bad publicity. This public relations aspect aside, granting greater workplace freedom is a way to “wage discriminate” when employing new staff; what we actually see, he suggests, is “employers catering to the job-quality preferences of the incumbents, rather than the marginal new hires, really quite often” (Cowen 2017, 110).

To be sure, Cowen does concede to Anderson that the cost of exit for most workers is very high. However, he views the cause of this problem to be “bad government decisions rather than…markets or the nature of the corporate employment relationship” (Cowen 2017, 111). In particular, he argues that legal reform decoupling employment from health insurance, retirement benefits, and one’s immigration status would help make exit less costly, freeing workers to leave if they desired. Large corporations and the structure of the labor market, in his view, do not impede the feasibility of exit and might even improve worker conditions, making the need for exit less important. The monolithic firm structure, thus, does not contribute to the problem of private government.

Cowen’s defense of exit as a liberating force hinges on two claims: that monopsony does not exist, and that employers are incentivized to offer freedoms in the workplace because doing so allows them to pay lower wages. What would it take for these claims to hold true? First, the contention that today’s workers are at the mercy of large monolithic firms would need to be shown to be false. Cowen’s first claim that monopsony does not exist seems to lean heavily on a very narrow, outdated conception of monopsony, or employer power, and ignores the modern evidence on employer power. Monopsony power is a broader concept than the “labor concentration” model—few firms in a labor market—to which Cowen refers. Alan Manning’s important work, Monopsony in Motion: Imperfect Competition in Labor Markets, reestablished the broader view of monopsony and employer power (Manning 2003). Manning explains in the Handbook of Labor Economics that “from the worker perspective, it takes time and/or money to find another employer who is a perfect substitute for the current one and that, from an employer perspective, it is costly to find another worker who is a perfect substitute for the current one” (Manning 2011, 976). Therefore, dynamic monopsony can exist even where there is no labor concentration. Manning argues that “imperfect competition is pervasive in labour markets” (Manning 2011, 975), and he concludes that “one’s views of the likely effects of labour market regulation should be substantially altered once one recognizes the existence of imperfect competition. All labor economists should take imperfect competition seriously” (Manning 2011, 1031). Similarly, a recent paper examining the economic literature on monopsonies concludes that “significant monopsony power can explain a number of empirical puzzles, such as bunching in wages…or wage dispersion” (Sokolova and Sorensen 2018, 2). The paper finds that “overall, the literature provides strong evidence for monopsonistic competition and implies sizable markdowns in wages”—meaning that employers obtain a share of workers’ wages (Sokolova and Sorensen 2018, 31).

As for Cowen’s second claim—that employers have every incentive to offer workers freedoms in the workplace because doing so allows them to pay lower wages—there is no evidence offered for this proposition, and it appears to rely on the assumption of a perfectly competitive market with equal employer and employer power. In such a market, compensating wage differentials would provide higher wages if job attributes were negative (workers are provided fewer freedoms) or lower wages if job attributes were positive. Cowen’s contention essentially assumes away the problem of private government since, in his view, employers have no power, and they are incentivized to supply greater worker freedoms in order to obtain a lower-wage workforce. Critics have challenged whether compensating differentials observed in the labor market align with Cowen’s theoretical assertions. Peter Dorman and Les Boden demonstrate that the workers with the weakest bargaining power face the most risk for injury and fatality and receive the least compensating differentials—an observation confirmed by the experience of essential workers during the pandemic (Dorman and Boden 2021). A recent comprehensive examination of compensating differentials for workplace attributes indicates a more widespread failure of this theory: “While the theory on the relationship between job characteristics and wages is clear…the empirical literature documenting the existence and magnitude of such trade-offs has lagged behind, often finding the opposite relationship that theory would predict” (Maestas et al. 2018). Certainly, more evidence is required than the simple assertion that employers have incentives to supply workers freedoms in the workplace.

Cowen also neglects to acknowledge the pervasiveness and insidious impact of noncompete agreements. Noncompete agreements threaten to effectively consign workers to long-term unemployment if they are prevented through such agreements from finding work with other employers in the same industry. Short of going through significant retraining or downgrading to a lower-paying job, the affected worker would be forced to put up with bad working conditions, however oppressive they might be. What does the evidence say about the prevalence and nature of noncompete agreements? A study surveying 11,505 respondents found “38.1% of the sample report agreeing to a noncompete at some point in their lives, while 18.1%, or roughly 28 million individuals, report currently working under one” (Starr, Prescott, and Bishara 2020, 5). More importantly, the authors also observe the following:

61% of individuals with noncompetes first learned they would be asked to agree to the provision before accepting their job offer while more than 30% first learned they would be asked to agree only after they had already accepted their offer (but not with a promotion or change in responsibilities). This late notice appears to matter to employees. In a follow-up question for those who received late notice, 26% report that if they had known about the noncompete upfront, they would have reconsidered accepting the offer in the first place. (Starr, Prescott, and Bishara 2020, 8)

Workers also generally do not get a say over whether to accept the noncompete agreements. A 2019 report found: “Roughly half, 49.4%, of responding establishments indicated that at least some employees in their establishment were required to enter into a noncompete agreement. Nearly a third, 31.8%, of responding establishments indicated that all employees in their establishment were required to enter into a noncompete agreement, regardless of pay or job duties” (Colvin and Shierholz 2019, 1). Moreover, noncompete agreements are not imposed merely on high-skilled workers, contrary to what most economists assume. As Alan Krueger and Eric Posner point out, “a remarkably high 21 percent of workers who earn less than the median salary are currently or have been restricted by a non-compete agreement” (Krueger and Posner, 2018, 8).10 Thus, noncompete agreements plague a substantial proportion of workers, coercing them into staying in jobs regardless of how oppressive those jobs might be.

Can greater transparency help workers avoid dictatorial traps?

As mentioned, the problem of private government is the problem of arbitrary, unaccountable control that employers exercise over workers. A big reason why bosses are able to wield such arbitrary power is that workers’ terms of employment are drawn relatively vaguely and are, for the most part, difficult to specify in advance. What workers can be commanded to do is not precisely defined at the point of their employment, nor in their employment contracts (assuming they signed one). New instructions might be given after the fact and, worse, sanctions might be imposed for actions that were not mentioned in the terms of employment. Anderson details some of the most egregious of these sanctions, including punishments meted out for workers’ social media activity, choice of romantic partner, political affiliation, physical fitness levels, and consensual sexual activity in off-work hours.

To be sure, Anderson concedes that firm efficiency requires that employers wield some control over workers. The alternative to this system would be one in which firms contract with independent contractors to work on capital and with each other. Anderson writes that when factories in New England experimented with such a system in the second half of the 19th century, it turned out to be very inefficient, due to factors such as the hoarding of information by contractors, delayed innovations, insufficient coordination between different contractors, and greater wear and tear of machinery (Anderson 2017, 52). The hierarchical management structure of modern firms solves these problems through the open-ended authority granted to managers, who, as a result, are able to “redeploy workers’ efforts as needed to implement innovations, replace absentees, and deal with unforeseen difficulties” (Anderson 2017, 52). Despite this system’s benefits, what Anderson considers unacceptable is the extremely wide latitude of employer authority over workers in places like the United States. The hierarchical control that the theory of the firm justifies, she argues, is distinct from the private government that most employees are subject to.

In his video review and discussion with Anderson, Mike Munger suggests that the problem of employers exercising such wide, arbitrary control over workers can be addressed by increasing transparency about working conditions prior to the start of employment (Public Square 2017). While he agrees with Anderson that the various forms of control bosses exercise over workers are deeply problematic, he thinks that this stems from the fact that the boundaries of employers’ authority over workers is left unspecified. Had the boundaries been clearly defined at the start of the employer–employee relationship, then workers would be able to decide for themselves if they wished to consent to those terms. The implication here is that greater transparency about one’s working conditions would eliminate concerns about dictatorial practices, since workers would have consented to them: “workers have a lot of freedom,” he says, “before they sign an employment contract and enter into an employment relationship” (Public Square 2017). If greater transparency were introduced, then the imposition of any new rules not initially agreed to would constitute fraud.

Munger takes this solution to the problem of private government to be an extension of Anderson’s own call for the application of due process when workers are treated wrongfully. As mentioned, one of Anderson’s solutions to the problem of private government is better enforcement of the rule of law in the workplace to protect workers from unjust employer treatment. Munger does not elaborate on precisely how greater transparency would help secure due process, but a plausible interpretation is that, were employers required to make explicit the terms of employment, including the specific tasks required of workers, any later imposition of new work requirements would constitute a violation of the terms of the contract.

We may look to the work of Cynthia Estlund to expand upon Munger’s proposal here. Estlund focuses on whether the “general case for workplace transparency extends to information about wages and salaries” (Estlund 2014, 782). She concludes that “there is a fairly strong though not uncomplicated case to be made that mandatory disclosure of meaningful salary information would tend to produce less discrimination, less favoritism, and probably somewhat lower disparities overall” (Estlund 2014, 783). Along the way, she refutes claims to the contrary, namely, that increasing transparency in this area would (1) publicize vital trade secrets and proprietary information, (2) harm employee morale and productivity, (3) violate employees’ own preference for pay secrecy, and (4) promote employer collusion.11

More transparency would certainly seem worthwhile and would make workplaces conform to the assumption in labor economics that both parties are completely and perfectly informed. Which dimensions of the employer–employee bargain can be made more transparent and which must be left unspecified? Unlike with other commodities, the inherent nature of the employment relationship means that many important matters will inevitably be left undefined. As Julia Tomassetti notes:

Employment is an agreement to work for another. But what does it mean to work for another? Work is simply the use of one’s voluntary faculties. Working for another means deploying those under the direction of another. So, an employee agrees to agree to the employer’s commands when entering the relationship. Employment is a special kind of “agreement to agree”—it is an agreement to obey. What essential terms does employment leave open and permit the employer to determine after the parties enter the purported contractual agreement? The employer gets to determine the terms regarding “What, and for how much?” Take the example of an employee paid an hourly wage: The employer determines the quantity term—how much work the employee provides—by commanding the nature, pace, and intensity of the work. In agreeing to an employment relationship, the employee agrees to provide an indefinite amount of labor—whatever the employer can extract for an hour—in return for a definite amount of pay. (Tomassetti 2020)

The transaction over the nature, pace, and intensity of work will necessarily reflect the balance of power in the workplace, and the essential bargain cannot be specified in advance, suggesting that transparency cannot eliminate exploitation about these matters, at least not by itself.

That said, transparency can address the “how much” question in work agreements, the terms of employment frequently described as wages, hours, and working conditions. Employees would certainly benefit from such transparency. However, it is worth considering why employees are so ill-informed of their terms of employment as well as their rights under the whole range of employment laws (Freeman and Rogers 2006). This lack of awareness, in fact, may reflect the lack of power workers face in both the workplace, where they are unable to adequately assert their preferred employment terms or form organizations such as unions, as well as in public life, where they are unable to pressure schools to educate everyone about their employment rights and their employers to make sufficiently transparent their employment rights. Perhaps the kinds of arguments Estlund rebuts in her case for greater pay transparency are precisely the reasons why the current legal structure, in which employers have disproportionately greater influence, has not required such transparency.

It is worth considering why the law, at present, does not require transparency in contractual agreements. Transparency seems integral to contracts of other kinds, such as mortgage agreements. What explains the difference? We certainly do not see conservatives or libertarians, or employer trade associations, campaigning to require transparency of employment terms at the time of hire. In fact, we see employers opposing transparency. A requirement by the National Labor Relations Board (NLRB) under the Obama administration that employee labor rights be clearly and visibly posted in the workplace (Labor-Management Standards Office 2010) was challenged by the National Association of Manufacturers, and the Court of Appeals ruled in 2013 that such postings violated the employer’s freedom of speech (National Association of Manufacturers et al. v National Labor Relations Board et al. 2013). No outcry was heard from employers, conservatives, or libertarians. Oddly enough, the NLRB under President Trump in 2020 mandated that employers post information about NLRB procedures, and employers did not find this a constraint on their free speech. The mandated posting, however, served the employers’ interests, as it informed employees of their right to decertify—get rid of—a union. Specifically, the NLRB mandated a “voluntary recognition bar,so that, when an employer voluntarily recognized a union (such as through a card check), “unit employees must receive notice that voluntary recognition has been granted and are given a 45-day open period within which to file an election petition” (National Labor Relations Board 2020). This requirement was portrayed by the NLRB as an effort to “better protect employees’ statutory right of free choice on questions concerning representation.”

This contradictory adherence to transparency aside, Munger’s proposal is potentially tautological. If transparency is a remedy for worker disempowerment, then how can workers motivate their employers to provide it? The only two ways in which they can do this is by either directly pressuring their bosses or by pushing for legislative change. But their present lack of power undermines both prospects, and so we are back to asking how we can increase the power of workers relative to their employers.

Another foundational issue concerns how Munger’s solution relates to the actual application of employment law. Currently, the law recognizes “contracts” formed between workers and their employers both prior to and after employment. But how do workers consent to new contracts formed after accepting a job? In a recent Supreme Court case looking at whether forced arbitration provisions (in which employees are forced to submit claims to private arbitration and to do so as individuals, not as a class) violated the National Labor Relations Act, the majority opinion contended that the employees freely agreed to the arbitration terms in a market transaction. But in fact, as noted in Justice Ginsburg’s dissent, the employees were informed about the terms of the forced arbitration agreement after already employed:

Petitioner Epic Systems Corporation e-mailed its employees an arbitration agreement requiring resolution of wage and hours claims by individual arbitration. The agreement provided that if the employees “continue[d] to work at Epic,” they would “be deemed to have accepted th[e] Agreement.”…Young similarly e-mailed its employees an arbitration agreement, which stated that the employees’ continued employment would indicate their assent to the agreement’s terms….Young’s employees thus faced a Hobson’s choice: accept arbitration on their employer’s terms or give up their jobs. (Epic Systems Corp. v. Lewis 2018, 138:1636).

Essentially, then, showing up to work legally constitutes consenting to any change in the terms of employment made by one’s employer. An agreement between employers and employees is treated as a spot market, continuously being renegotiated. Why should this be considered consent by workers in any meaningful sense? Further, how can transparency requirements overcome the unilateral imposition of such terms once already employed? Interestingly, libertarian philosophers are loath to consider tacit consent as being a legitimate form of consent when it comes to state coercion.12 They often argue that the mere fact that we comply with laws should not be taken to mean that we have consented to them, for we might have little choice but to go along.13 Consistency would seem to require that libertarians also not consider cases like the one above as constituting employee consent. And if such cases don’t, then it seems that legitimate consent to new employment contracts would require, as a precondition, that employees be meaningfully able to refuse them without significant cost to themselves (such as being fired or being forced to relocate after being employed). How would this be possible without first improving the worker’s bargaining position relative to the employer’s, say through better union representation, mandated legal requirements, or some of the measures that Anderson suggests? Thus, it seems that a plausible interpretation of Munger’s proposal would require rather than replace Anderson’s proposed solutions to the problem of private government.

Is worker discontent simply overblown?

In her argument, Anderson does not merely imply that workers wish to remedy their lack of freedom. Rather, she uses several statistics to demonstrate this point. Two in particular are important: (1) “25 percent of employees…understand that they are subject to dictatorship at work”; (2) “55 percent or so… are neither securely self-employed nor upper-level managers, nor the tiny elite tier of nonmanagerial stars (athletes, entertainers, superstar academics) who have the power to dictate employment contracts to their specification, nor even the ever-shrinking class of workers under ever-retrenching collective bargaining agreements” (Anderson 2017, 63). Members of the former group are aware of their oppressed conditions, whereas the latter is “only one arbitrary and oppressive managerial decision away from realizing what the 25 percent already know” (Anderson 2017, 63). The upshot of these two statistics is that the vast majority of workers are subject to the problem of private government and are either aware of it or likely to be so.

Tyler Cowen, Mike Munger, and Thomas Ferretti object to this empirical claim about worker unhappiness. They contend that many workers are either happy with their working conditions or are willing to put up with the authoritarian control they are subjected to in exchange for higher wages. Either way, they argue, it seems unlikely that there is widespread unhappiness about workplace dictatorship. A few of Cowen’s claims detailed above apply here. Specifically, firms have financial incentives to grant their workers greater workplace freedoms since granting such freedoms is cheaper than raising wages. In fact, this trade-off helps employers justify lowering wages. And this, Cowen contends, is what many firms actually do. Relatedly, firms have reputational reasons to not constrict worker’s liberties, as the aforementioned McDonald’s example illustrates. These reasons aside, the “desire to attract and keep talent is the single biggest reason why companies try to create pleasant and tolerant atmospheres for their workers, and why it is rare for businesses to fire workers for their political views or their (nondestructive) off-premises activities” (Cowen 2017, 114). Shifting to the employees’ perspective, most workplaces, he argues, are “sources of human dignity and fulfillment” (Cowen 2017, 114). Workers generally get too much satisfaction and fulfillment from their workplaces to work from home, despite the conformity the workplace might require.

Ferretti and Munger expand on Cowen’s comments, arguing that, to the extent that workers are displeased with their working conditions, they are willing to put up with them given the benefits these conditions bring. For instance, tolerating workplace dictatorship might be preferable to some workers “in exchange for other benefits such as higher wages, more flexible schedules, less responsibility and less risk” (Ferretti 2018, 280). Similarly, Munger claims that many workers prefer to receive higher pay than greater workplace freedom so they can “create lives outside of work that have meaning….if workers value personal voice over pay, some firms would be working on that margin already” (Munger 2018, 629). Where Munger diverges from Anderson is in focusing on the meaning workers might find not within but outside of work. In Munger’s view, Anderson errs by ignoring this fact and its impact on how workers weigh the concomitant effects of authoritarian employer rule. Nonetheless, Munger, like Cowen, takes Anderson’s concern about worker dissatisfaction with such control to be overstated given the motivations and practices of many employers on the one hand and the preferences of workers on the other.

We have earlier examined the theoretical claim that employers have incentives to provide workplace freedoms that would yield lower wages, and we found that the “compensating differential” claims of such competitive, equal power, labor market speculations do not fit the facts on compensating differentials for risk or other major work attributes. How should we evaluate these claims about worker dissatisfaction (or the lack thereof)? Two questions immediately arise. First, how much bargaining power do workers have relative to their employers and how has it changed over time? If workers have no way of pushing for their demands with their employers, then employers have no reason to please workers in any way, whether by granting greater workplace freedoms or higher wages. Second, what alternatives do workers have to accepting the terms offered to them by their employers? If they have no viable alternatives, then they have no choice but to accept those terms. Regarding the first question, empirical data suggest that worker power has declined over the past four decades. For instance, a 2016 report by the Council of Economic Advisers details how firms have sought to suppress workers’ wages through a variety of means such as collusion, which is only made possible by the fact that they have monopsony power over the labor market (Council of Economic Advisers 2016). Alan Krueger backs up the charge that firms’ monopsony power allows them to suppress workers’ wages. He notes that forces that previously kept such power in check, such as union representation for workers, collective bargaining, and better minimum wage laws, have fallen by the wayside, leaving firms free to reduce wages as they please (Krueger 2018). This power shift seems to help explain why “only the highest earners have seen steady wage gains; for most workers, wage growth has been sluggish and has failed to keep pace with gains in productivity” (Council of Economic Advisers 2016, 1; see also Bivens and Mishel 2015, 3). The discipline of economics is increasingly focused on the deleterious effects of monopsonies and employer power on workers’ wages (Smith 2019).

To be sure, we might wonder whether workers have been compensated for their increased productivity in other ways, say, through better employee benefits. But the evidence suggests otherwise. Josh Bivens and Lawrence Mishel point out that “benefits have grown far less than most people realize, rising from 18.3 percent of compensation in 1979 to just 19.7 percent of compensation in 2014” (Bivens and Mishel 2015, 14), and the hourly compensation, both wages and benefits, of the vast majority of workers has risen very modestly for four decades and far less than productivity growth. Increased employer power over the terms of work can readily explain this phenomenon. A recent paper by Anna Stansbury and Lawrence Summers goes even further, arguing that it is the loss of worker power that best explains the perplexing combination of economic trends over the past several decades, namely (1) increased profitability of firms, (2) sluggish wage growth and a falling labor share of income, and (3) low unemployment and inflation (Stansbury and Summers 2020). Richard Freeman and Joel Rogers’ earlier work also notes the existence of “substantial gaps between the importance of rights to workers and employer performance, with wide variation among the categories covered in the survey” (Freeman and Rogers 2006, 9).14 Such evidence seems to undermine the idea that workers have much bargaining power relative to their employers and that worker satisfaction is widespread. It is instructive that both major political parties, since 2016, have highlighted wage stagnation as a serious problem.15

What about workers’ alternatives to accepting the terms offered to them? Perhaps workers can easily leave, in which case those who choose to stay do so willingly (and maybe because they find their workplaces to be sources of dignity and fulfillment, as Cowen suggests). But even here, the evidence seems to point in the other direction. For instance, as Anderson mentions, firms have often colluded to suppress wages through methods such as noncompete or no-poaching agreements. Moreover, “agreements among independent firms to refrain from hiring each other’s workers or to set pay increases at a common level” are illegal but still occur (Krueger 2018, 5). And even if workers stayed in their jobs, they have little choice but to accept their fates. Krueger notes how legal changes permitting mandatory arbitration have undermined workers’ ability to file lawsuits against their bosses (Krueger 2018, 4). We have already noted the decline of both unions and collective bargaining weakening the ability of workers to shape their wages, hours, and working conditions. All this evidence combined suggests that workers’ decisions to stay in their jobs may be due more to powerlessness than to job satisfaction.

In response to Cowen, a third question is also worth asking to evaluate claims about worker dissatisfaction: In what sense are the detrimental psychological effects of unemployment indicative of workers’ satisfaction with their working conditions? Such evidence seems to suggest only that workers would prefer to work than not work, but it’s not clear how this is relevant to Anderson’s critique. After all, workers could both prefer to work than not work and still want better working conditions. Anderson’s arguments merely concern the latter. Perhaps the fear of unemployment, in combination with the plethora of considerations just highlighted, force workers to stay in their jobs, but this hardly shows that they are happy as a result.

Finally, what does empirical data say about current levels of worker happiness? The Gallup 2019 Great Jobs Survey concludes that:

40% of American workers are currently in “good jobs,” defined by high levels of satisfaction with the job characteristics workers care about most. Meanwhile, 16% of workers are in “bad jobs,” which are rated poorly across most key characteristics. The rest of the U.S. workforce—44% of employed Americans—are in “mediocre jobs,” which warrant satisfaction on some but not all dimensions….[F]ewer than half of workers tell Gallup that they are completely satisfied with their jobs, and even fewer express complete satisfaction with such key dimensions as pay, health insurance and retirement benefits. (Rothwell and Crabtree 2019, 6, 10)16

The study also found that “the largest sources of job quality disappointment (the gap between satisfaction and importance) are in pay and benefits, factors commonly rated as important but for which satisfaction ratings are especially low. Just 54% of workers overall are satisfied with their current pay level” (Rothwell and Crabtree 2019, 3). Another recent paper, by David Howell, finds a widespread decline in job quality: “there has been an astonishing decline in the number of decent jobs generated per dollar of GDP [gross domestic product] since the 1980s, particularly for young workers without a college degree, but it also appears for those with at least a college degree” (Howell 2019, 49). Importantly, Howell argues that globalization, computerization, and insufficient education funding cannot explain this trend. Instead, he points to “major shifts in institutions, policies and employer human resource strategies that have undermined worker bargaining power” (Howell 2019, 2) as the real cause.17 All in all, it seems clear that Cowen needs to provide better evidence to support his claims that workers are satisfied.

Even if unequal, isn’t employment just another contractual relationship?

Another criticism of Anderson’s work concerns the parallel she draws between state dictatorship and the dictatorship of the workplace. Anderson argues that what makes authoritarian control over workers morally equivalent to living under a communist dictatorship is that both deprive their members of republican freedom—the freedom from arbitrary, unaccountable control by others. In their comments on Anderson’s work, Jacob Levy and Jessica Flanigan argue that there is an important difference between state control and control by employers that renders the latter at least less morally problematic than the former (Levy 2019; Flanigan 2019). Specifically, the relationship between the worker and the employer is entered into voluntarily. By contrast, most people do not get to choose whether they are subjected to the state’s coercive authority. They must obey its commands whether they endorse its rule or not. Indeed, Levy and Flanigan argue that this distinction renders the employer–employee relationship equivalent to the ones we form in our private lives, and thus no more problematic.

In his comments, Levy levels this criticism using the metaphor of “islands of unfreedom in a sea of freedom.” By granting us our civil and political liberties, the liberal society allows us to live in a sea of freedom. However, within this space, we are free to form relationships that involve hierarchical structures and constitute “islands of unfreedom.” For instance, “churches are islands of internal religious orthodoxy and intolerance in a sea of religious liberty.” Similarly, “marriages are (aspirationally) islands of sexual exclusivity in a sea of free sexual choice among consenting adults” (Levy 2019). Other examples he mentions include families, households, universities, and labor unions. In all these cases, we often form relationships that are distinctly inegalitarian, and part of what it means to be free to form consensual relationships is that we may choose to form unequal relationships. Despite the power differential in many of the relationships we form with others, we ultimately do not consider them to be illegitimate. This tolerance in the personal sphere is in stark contrast to how we evaluate hierarchical control in state–citizen relations. We do not think it is permissible for the state to exercise control in the same stringent ways we might tolerate on our islands. To be sure, Levy does not claim that islands of unfreedom cannot be subject to criticism. His point is simply that we generally accept such a dualistic picture, that is, “general liberty in public and from the perspective of the state to do as we like with the freedom to form local islands of commitments to doing one thing rather than another” (Levy 2019).

Levy uses this dualistic picture to contest the parallel Anderson draws between state and workplace dictatorship. If the firm is just another island of unfreedom, then it is misleading to apply to it the standard we would apply to state–citizen relations. The two are importantly distinct, he argues. And yet, “the deliberate rhetorical force of Private Government is to blur that boundary, not only in the ‘communist dictatorship’ metaphor but in the turn to standards like republicanism and in a sense of surprise appealed to by the titular metaphor itself” (Levy 2019). Anderson’s argument that the power structures in the modern firm deprive workers of republican freedom thus overlooks the fact that we might choose to enter such associations with our employers.

Flanigan echoes Levy’s claims, asserting that so long as consent and rights to exit are present, there is nothing problematic about any relationship we enter into with others. She predicts Anderson’s likely reply to be that there are various legal barriers (such as noncompete clauses) that make it difficult for workers to effectively exercise their exit rights. She notes that Anderson’s preferred solution to this problem is legal reform to remove such barriers that disadvantage workers. Flanigan, however, favors a basic income system. She argues that if everyone were provided with a basic income, then the cost of leaving work would not be so costly. Indeed, she contends that “a basic income would make seemingly dictatorial workplaces look more like the voluntary patriarchal marriage I described, with genuine consent and rights of exit” (Flanigan 2019). Moreover, a basic income system would be fairer to employers, and this fairness, she argues, is important since employers benefit the economically needy more than other citizens do by providing them with jobs.

Given the fact that universal basic income is only beginning to gain political traction and is far from likely in the near future, the most obvious question raised by Flanigan and Levy’s responses are, what do we do in the here and now, given that we do not live under a more ideal world where all workers receive basic income? Flanigan rejects Anderson’s proposed fixes on the grounds that they will lead to bad economic effects, a point we will discuss in the next section. Thus, it seems that the success of her basic income solution to the problem of private government hinges on Anderson’s solutions being economically worse. For if they are not, then both sets of solutions would be on par (at least on economic grounds), and we would need some other way of adjudicating between them.

In the absence of a basic income solution, we need to determine whether workers are truly free to enter and exit the “island of unfreedom” that is the workplace. In the preceding sections, we investigated a variety of empirical considerations that appear to suggest that workers both lack a meaningful right to exit and, when they do exit, have little option but to end up in another unfree island. These restricted options aside, it is also unclear why we should think that most people can opt out of entering the workforce. For unlike a country club or one’s church, bare economic necessity forces one to obtain a job.18 Until this is no longer the case, it seems clear that we are forced to grapple with the hurdles facing an employee’s ability to exercise his or her autonomy in the workplace.

A different issue we might also raise concerns the plausibility of the parity Levy and Flanigan attempt to draw between the employer–employee and other relationships. An important distinguishing feature of an employment relationship is its centrality to our lives and overall identity. People tend to spend most of their waking lives at work, and it is, as Cowen suggests, a source of meaning and self-worth for many. This aside, our employment contract also determines a whole range of benefits we may seek, such as health insurance coverage, retirement security, time for holidays, sick leave, and of course our wages and income. Work determines the very ability to stay alive and be healthy. Thus, unlike the vast majority of relationships we might form, consensually or otherwise, our employment relationship is crucial to our ability to further our most fundamental interests. As a result, it seems all the more important that people are genuinely free to enter and exit workplace relationships. This importance justifies the close examination of and proposed reforms to workplace dictatorship. Not all islands are equal in moral significance in the sea of liberty.

Won’t the proposed solutions have distortionary economic effects?

The final objection to Anderson’s argument concerns the impact of her proposed solutions on the economy. A few of the aforementioned critics argue that the reforms she calls for, in particular, the application of the rule of law at the workplace, greater constitutional protections, and greater voice for workers in firm decision-making, are likely to have distortionary effects on markets and generate bad economic outcomes. Specifically, reforms to liberate workers from workplace dictatorship, they argue, are likely to lead to (1) inefficiencies in the running of firms, (2) increased unemployment, (3) shifting of labor to unregulated black markets, and (4) reduced economic growth.

The first economic cost of Anderson’s proposed reforms concerns efficiency. Cowen argues that “employer discretion” is needed to ensure the smooth running of firms. If employers are not able to dictate who gets fired and under what conditions, then they would not be able to address the effects of worker misconduct and noncompliance (Cowen 2017, 112). This powerlessness, in his view, is due to lack of legal or contractual means to delimit worker conduct in specific-enough ways. Employer discretion ensures that costly behavior on the part of workers does not harm others in the firm, including fellow workers. For instance, he points out that “a lot of workers put racist, sexist, or otherwise discomforting comments and photos into their Facebook pages.” When their bosses fire them, they might be doing it “to protect some notion of the freedom of the other workers” (Cowen 2017, 112). In their efforts to ensure the smooth running of the firm, employers often consider the overall preferences of workers, and might enforce such measures to preserve overall worker happiness. Anderson’s proposals to protect workers from such arbitrary firings, however, would prevent employers from doing what is necessary to keep workers generally happy. In this sense, her solutions might end up impairing worker autonomy rather than enhancing it. On the whole, Cowen argues, giving employers discretion in firing matters is more beneficial for workers than costly, and so worth preserving.19 Of course, this discretion might not be beneficial if employers misuse it, but, as we saw, Cowen rejects this suggestion. He contends that, while it is true “at the margins, that employer discretion leads to abuses…those abuses are relatively few in number, and the gains for workers and customers from the firing discretion—not just the gains for bosses—outweigh those costs” (Cowen 2017, 112–13).20 Since employers generally do not abuse the wide latitude in control they have over workers, and workers receive a net benefit from such control, it is worth preserving.

Cowen’s concerns about firm inefficiency feed into Levy and Flanagan’s contention that Anderson’s proposals are likely to lead to the second cost: higher unemployment. If firms suffer economic costs due to the new regulations, then they might be forced to fire workers to balance the books. Thus, Anderson’s attempts to protect the vast majority of workers she argues are vulnerable to authoritarian control would, in his view, lead to higher unemployment overall.

In his discussion with Anderson, Munger argues that prohibiting firms from exercising wide discretion over worker management is likely to lead to the third of the economic costs mentioned above: the creation of unregulated black markets (Public Square 2017). Though he raises this objection in a cursory manner, we can develop the line of thought here further. Consider the following possibility. Regulations of the forms Anderson defends cause employers to hire fewer workers after the employers deem such regulations to be detrimental to their efficiency and, in turn, profit margins. However, since workers would rather be employed than not, they search for work in unregulated markets, namely, in the underground economy. Employment here might involve both more demeaning and dangerous work, for lower pay, and with no legal protections. Munger worries that Anderson’s favored regulations are likely to push many workers into such a situation. In turn, he argues for alternative solutions to the problem of private government, such as the transparency-in-contract reform mentioned above.

The final cost, highlighted by Flanigan and Levy, concerns the sum effect of the previous three. A reduction in firm efficiency ultimately reduces both the profitability of firms along with their ability to expand and innovate. Likewise, increased unemployment means lower tax revenue and reduced spending on goods and services.21 Relatedly, black markets reduce government revenue (since black markets aren’t taxed) and are not reported in the national estimates on employment and the country’s GDP.

The plethora of economic claims made by Anderson’s critics here call for scrutiny on multiple fronts. First, we need to ascertain in precisely what way does limiting employers’ unlimited managerial authority lead to increased inefficiency. Why must a manager wield unlimited authority over workers? To be sure, managers have uncontested control over a wide variety of firm decisions, including pricing, marketing strategies, and investment plans, that employment legislation and collective bargaining do not constrain whatsoever. None of these domains are the target of Anderson’s criticism. Instead, she contests managers’ unfettered ability to decide whom a firm hires and fires and for what reason and how workers are paid. Relatedly, the debate seems to focus on whether managers should be able to unilaterally dictate employee benefits, including wages, sick pay, and family leave. There is also the matter of working conditions, such as safety and health standards and scheduling practices, all of which impact a worker’s overall well-being. These policies are central to ensuring worker freedom. Why should we think that managers’ need unconstrained authority in these realms?

This question aside, what evidence is there to suggest that limiting such authority, even partially, would lead to inefficiency? At least in theory, it seems that we could place some limits on the arbitrary decision-making powers managers have without making a meaningful dent in the smooth running of the firm. Anderson discusses the German co-determination model, under which decision-making powers in the firm are shared between shareholders and workers. Among her critics, only Cowen provides any evidence against this alternative to unlimited managerial control, and even if the single study he cites is correct in claiming that “this organizational form costs about 26 percent of shareholder value because of lower productivity” (Cowen 2017, 116), we should ask why, as Anderson points out, “this isn’t seen as a point in favor of the Germans, in that the people who actually do the work enjoy greater shares of the pie” (Anderson 2017, 142). Anderson also cites research suggesting that evidence about the productivity effects of co-determination models is mixed at best. A new study of the German co-determination model conducted by the National Bureau of Economic Research provides further grounds for dismissing Cowen’s claim. According to the researchers:

In sharp contrast to the canonical hold-up hypothesis—that increasing labor’s power reduces owners’ capital investment—we find that granting formal control rights to workers raises capital formation. The capital stock, the capital-labor ratio, and the capital share all increase. (Jäger, Schoefer, and Heining 2019)22

This aside, it is worth emphasizing that we already limit employers’ ability to discriminate according to race, gender, sexuality, disability, age, and other factors. Firms are still able to run very efficiently in spite of these restrictions. There has been extensive research on the impact of job-flexibility limitations, higher minimum wages, family leave, and sick pay, and their impact on economic performance such as unemployment and productivity levels.23 These claims cannot be settled based on assertions about the presumed need for unlimited managerial authority. Examining the empirical evidence concerning the kinds of policies that Anderson and others propose to protect worker freedoms seems central.

Assuming away unemployment

The majority of the criticisms discussed above share a common feature: They assume a certain idealized picture of labor market conditions and their impact on workers. For instance, the idea that employees possess equal bargaining power because of their ability to quit assumes that alternative sources of employment are readily available to them. In other words, it assumes that we are operating under full employment. Similarly, if unemployment rates are high, then many workers have fewer opportunities to avoid oppressive conditions at work. That they do not complain about their situation does not then imply that they are content with it or that they consider the workplace to be a source of dignity and fulfillment. The success of such criticisms is dependent on these idealizations mapping onto our economic reality. A satisfactory assessment of the merits of these criticisms of Anderson’s arguments thus requires an examination of what these idealizations consist in and how they undergird these criticisms.

The first assumption, as mentioned, is that we are operating under conditions of full employment. While economists disagree about how we should define full employment, a simple definition that suffices for the present discussion is the following: the situation in which anyone who wants to work is able to find work. To be sure, this definition should not be understood as being equivalent to an unemployment rate of zero. Even in the best of times, unemployment is never zero. For instance, during the period between when workers leave their jobs and find new ones, they are unemployed. Likewise, workers and employers might take time before deciding whether to contract with each other (formally or informally). These cases are instances of “frictional unemployment,” which occurs when workers switch jobs. In any case, the question at hand is whether we are ever at full employment such that finding work is feasible and making a transition to another job is not costly. The bottom line in the context of our discussion is whether the presumed situation is as competitive labor market theory assumes, that it is as easy for an employer to replace a worker as it is for a worker to quit and find a comparable or better job.

In reality, our economy has business cycles, we are rarely at full employment, and workers are far more likely to struggle finding employment than employers are to find workers. Indeed, these circumstances are policy choices, and the U.S. for the last several decades has chosen monetary and fiscal policies that tolerated, or propelled, excessive unemployment. As Josh Bivens and Ben Zipperer point out, “the Fed and other macroeconomic policymakers tolerated too high unemployment for most of the time between 1979 and 2007” (Bivens and Zipperer 2018, 13). We also know that the failure to maintain full employment has suppressed wage growth for workers, with low-wage and minority workers most adversely affected and middle-wage workers impacted more than high-wage workers.24 For instance, Bivens and Zipperer observe the following:

Tight labor markets can narrow not just absolute racial employment gaps, but also relative racial employment gaps (or ratios). We find that tighter aggregate labor markets lead to disproportionate gains in African American employment (as measured by employment-to-population ratios, or EPOPs) and average hours worked for African American households relative to white workers. (Bivens and Zipperer 2018, 20)

A recent paper quantifies the wage impact of lower unemployment:

Research indicates that a 1 percentage point drop in unemployment results in annual wage growth 0.5–1.5 percentage points faster for workers at the 10th percentile. For example, if annual real wage growth is 1%, then a 1 percentage point fall in unemployment would result in annual real wage growth rising to 1.5% to 2.5%. For workers near the median of the wage distribution, wage growth is faster by 0.4–0.9 percentage points, and for workers at the 90th percentile it is 0.3–0.5 percentage points faster. These estimates indicate that excessive unemployment generates increases in both the 50/10 and 90/50 wage gaps. (Mishel and Bivens 2021)

Black workers face twice the rate of unemployment every year as workers overall, whether unemployment is high or low and regardless of education level, including college (Bernstein and Jones 2020, 20). New research by Lawrence Mishel synthesizes all these points:

The pattern of unemployment over the 1979 to 2019 period amply demonstrates how excessive unemployment has been the norm in our economy. Consider that the unemployment rate averaged 6.2%, above a conservative estimate of full employment (5%) and far above the 3.7% rate achieved at the end of 2019 (without inflation becoming problematic). Over this period, however, Black and Hispanic workers faced recession-level rates of unemployment, respectively, of 11.9% and 8.6%. Digging further, combining race and “class,” we can examine the average unemployment of a relatively educated Black worker, someone with at least some years of college or an associate college degree but not a bachelor’s degree or more: about a third of Black workers had both more and less education over the last four decades. A Black worker with some college education faced an unemployment rate exceeding 10% for 43.1% of the months over 1979-2019 and enjoyed unemployment below 5% for only nine months of the forty-one years encompassing the 1979-2019 period. (Mishel 2020)25

To add to all this, we also know that when unemployment is high, employees are less likely to quit.26

In light of all this evidence, it is simply not reasonable to discuss workplace freedoms and power without taking the prevalence of excessive unemployment and underemployment (such as involuntarily working a part-time job when full-time work is desired) into account. The general employment level and its impact on people’s economic opportunities therefore seem particularly relevant to any discussion of Anderson’s assessment of worker power.

The persistent excessive unemployment that workers face, and which certain groups of workers face disproportionately, provides essential context for assessing Levy and Flanigan’s claim of parity between the employer–employee relationship and the many other relationships we form in our private lives. To recall, the claim here involves first recognizing that while many of our private relationships are unequal in nature (such as that between a religious leader and a member of a parish), we generally do not regard them to be problematic so long as they are entered into consensually. The next step is to argue that workplace relationships embody a similar structure, that is, they might be unequal in nature but are unproblematic if entered into consensually. This last conditional is where the truth value of the idealization concerning full employment matters. If we do not in fact operate under full employment conditions, then maintaining an economic relationship with our employer may be dictated by necessity and is probably not consensual. Again, high or even moderate unemployment levels raise the opportunity costs of not maintaining such a relationship. If workers are facing such pressures, then, as Anderson notes, the choice of whether to enter into an employment agreement with an employer is analogous to the choice that Victorian women had regarding whether to enter a marriage, and quite unlike the choice of whether to join a book club.


Ultimately, Anderson’s work serves as a call for a comprehensive reevaluation of the way we view work today in the United States and beyond, and it shines a spotlight on the subjugation of workers under a system that paradoxically views them as equal members in the economic bargain with their employers. Private Government highlights the myriad ways in which the market oppresses in much the same ways the state did. The various criticisms discussed in this essay only serve to reinforce the need for a deeper study into the current market order and how that order might be reformed. Relatedly, the assumptions that seem to underlie many of these criticisms mandate going beyond mere theoretical assertions to engage in deeper economic analysis of the actual economic circumstances facing workers.


1. “Government is nothing more than a national association; and the object of this association is the good of all, as well individually as collectively. Every man wishes to pursue his occupation, and to enjoy the fruits of his labours, and the produce of his property in peace and safety, and with the least possible expence. When these things are accomplished, all the objects for which government ought to be established are answered” (Philp 1998).

2. Smith claims that “When an animal wants to obtain something either of a man or of another animal, it has no other means of persuasion but to gain the favour of those whose service it requires. A…spaniel endeavours by a thousand attractions to engage the attention of its master who is at dinner, when it wants to be fed by him. Man sometimes uses the same arts with his brethren, and…endeavours by every servile and fawning attention to obtain their good will….But man has almost constant occasion for the help of his brethren, and it is in vain for him to expect it from their benevolence only. He will be more likely to prevail if he can interest their self-love in his favour, and shew them that it is for their own advantage to do for him what he requires of them. Whoever offers to another a bargain of any kind, proposes to do this. Give me that which I want, and you shall have this which you want, is the meaning of every such offer….It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest. We address ourselves, not to their humanity but to their self-love….Nobody but a beggar chuses to depend chiefly upon the benevolence of his fellow-citizens” (Smith 1981). Marx asserts that “[The] sphere…within whose boundaries the sale and purchase of labour-power goes on, is in fact a very Eden of the innate rights of man. There alone rule Freedom, Equality, Property and Bentham. Freedom, because both buyer and seller of a commodity, say of labour-power, are constrained only by their own free will. They contract as free agents, and the agreement they come to, is but the form in which they give legal expression to their common will. Equality, because each enters into relation with the other, as with a simple owner of commodities, and they exchange equivalent for equivalent. Property, because each disposes only of what is his own. And Bentham, because each looks only to himself….On leaving this sphere of simple circulation or of exchange of commodities, which furnishes the “Free-trader Vulgaris” with his views and ideas, and with the standard by which he judges a society based on capital and wages, we think we can perceive a change in the physiognomy of our dramatis personae. He, who before was the moneyowner, now strides in front as capitalist; the possessor of labour-power follows as his labourer. The one with an air of importance, smirking, intent on business; the other, timid and holding back, like one who is bringing his own hide to market and has nothing to expect but—a hiding” (Marx 1912).

3. A report by the Congressional Research Service defines the gig economy as “the collection of markets that match providers to consumers on a gig (or job) basis in support of on-demand commerce. In the basic model, gig workers enter into formal agreements with on-demand companies (e.g., Uber, TaskRabbit) to provide services to the company’s clients.” See Donovan, Bradley, and Shimabukuru 2016.

4. Cynthia Estlund’s review of Anderson is also relevant here. She notes that in recent decades major technological changes, including “faster and cheaper ways to reliably transport and monitor goods, services, and information…[have greatly reduced]…the operational advantages of direct hierarchical control over employees” (Estlund 2018, 820). This development has driven firms to outsource the process of hiring workers to third-party labor-supplying firms, a process sometimes referred to as fissuring (Weil 2019). Firms that would previously hire workers and retain them in a long-term capacity now prefer contracting with workers for task-specific jobs. As a result, “many workers today find themselves plagued less by employer domination than by the insecurity of having no employer at all” (Estlund 2018, 821). To be sure, Estlund thinks that the problem of private government persists despite this change, and that the problem is perhaps more serious for those employed by such supplier firms. Interestingly, this problem arises even though many workers are not operating under the monolithic firm structure that is the target of Anderson’s critique.

5. “The gig economy, with independent workers and short-term contracts, can also be a source of employment and income. Here, gig work covers three types of non-traditional activities: offline service activities, such as child care or house cleaning; offline sales, such as selling items at flea markets or thrift stores; and online services or sales, such as driving using a ride-sharing app or selling items online. This definition of gig work, encompassing both online and offline activities, takes a broad view of the gig economy and underscores the fact that such supplemental work predates the internet. Gig work is largely done in addition to a main job, so this is often distinct from those who work as contractors in their main job” (Board of Governors of the Federal Reserve System 2018, 18–19).

6. For instance, the reviewers point out that “about 1 percent of the population engaged in income-generating activities on an average day in 2003–07, and those who engaged in these activities spent 2.7 hours doing so. These estimates are similar to those for 2012–16. They do not support the claim that official estimates of employment have missed a large increase in gig or informal work that has occurred over the past decade.” Likewise, “In 2012–16, 1 percent of the employed, 2 percent of the unemployed, and 1 percent of those not in the labor force engaged in income-generating activities on an average day. Although these numbers indicate that there may indeed be misclassification of labor force status, they suggest the effect on overall levels of employment is small” (Dorinda Allard and Polivka 2018).

7. He says, “I am worried she, like others, doesn’t offer much evidence to back up her portrait, save for one footnote to an adequate but not very influential book” (Cowen 2017, 109).

8. Cowen cites Bonanno and Lopez 2009. Cowen says this is “the best study I know” on monopsonies (Cowen 2017, 109).

9. As he states, “the monopsony model does not itself predict workers will enjoy less freedom or fewer perks in the workplace. This sounds counterintuitive, as we associate monopsony with lower bargaining power for workers and thus inferior working conditions. But rest assured, I am offering the correct reading of theory” (Cowen 2017, 110).

10. For another report on noncompetes, see Federal Trade Commission 2020.

11. For instance, in response to the first objection, she notes, “conventional economic analysis suggests that ‘information disparities’ between contracting parties—for example, between an employer and employee regarding salary patterns within the firm—are an impediment to efficient bargaining” (Estlund 2014, 787).

12. “Tacit consent” here is to be contrasted with “explicit consent,” where the latter involves an overt action to signal one’s willingness to accept a situation or set of terms offered.

13. It is generally accepted, by philosophers and beyond, that one cannot be said to consent if one could not have chosen to do otherwise. This makes sense of why obeying a robber threatening to kill us if we do not hand over our wallets does not constitute consent. In much the same way, citizens who are effectively unable to leave their state are not in a position to refuse the state’s commands. Their continued residence within the state then cannot constitute their consent to its commands. Libertarian philosopher Robert Nozick famously contended that one cannot be bound through tacit consent since “tacit consent isn’t worth the paper it’s not written on” (Nozick 1977, 287). For a more developed libertarian refutation of the idea of tacit consent, see Huemer 2013, 22–28. Among other things, Huemer argues that tacit consent counts as a legitimate form of consent only if one has a reasonable way of opting out of a contract.

14. The authors used 2001 Hart poll survey results to arrive at this conclusion, and note that while “differences in questions make it hard to assess whether the Hart results show larger or smaller gaps than those in the WRPS [Worker Representation and Participation Survey]…the Hart survey finds large gaps, as did the WRPS” (Freeman and Rogers 2006, 9).

15. As the 2016 Republican Party platform asserts, “our economy has become unnecessarily weak with stagnant wages. People living paycheck to paycheck are struggling, sacrificing, and suffering” (Republican Party 2016, i). The 2016 Democratic Party platform, likewise, states that “too many Americans have been left out and left behind. They are working longer hours with less security. Wages have barely budged and the racial wealth gap remains wide, while the cost of everything from childcare to a college education has continued to rise” (Democratic Party 2016, 1).

16. The authors differentiated “good,” “mediocre,” and “bad” jobs by having workers “rate the importance of 10 dimensions of job quality, including [level of pay, stable and predictable pay, stable and predictable hours, control over hours and/or location, job security, employee benefits, career advancement opportunities, enjoying your day-to-day work, having a sense of purpose and dignity in your work, having the power to change things about your job that you’re not satisfied with]….[B]ased on those answers, workers are asked to rate their level of satisfaction with each dimension….[A] good job has an importance-weighted average score of ‘4’ or above; a mediocre job has an importance-weighted score that is less than ‘4’ but above ‘3.’ A bad job has a score at or below ‘3’” (Rothwell and Crabtree 2019, 9).

17. Howell’s paper employs “Current Population Survey data to document changes in job quality for 1979-2017 with measures of decent-, low- and lousy-wage jobs for groups defined by age, gender, education, race and nativity. These indicators are defined by two wage thresholds chosen to reflect the wage a full-time worker would need for a basic-needs budget: 2/3 of the mean wage for full-time, prime-age workers ($17.50 in 2017), which marks the cutoff between decent- and low-wage jobs; and 2/3 of the median full-time wage ($13.33), the boundary between lousy- and other low-wage jobs” (Howell 2019, 2).

18. We could in fact extend this point in the opposite direction to what Levy suggests. We could argue that few of the examples he brings up are really islands that we can opt out of entering. One’s family, one’s local religious community, school, and even the set of friends one has, are all islands that one is often compelled to enter given one’s geographical situation. Can one really choose to not associate with one’s classmates or the kids at the local playground? Granted that one possesses some degree of autonomy in such situations, there seems to be a marked difference between choosing to enter them and choosing to enter a choir. All of this is to say that there might not exist a clear distinction between the islands we can enter into freely and those we cannot.

19. He also argues that “economists in fact have a pretty good but not perfect explanation of why employers often have so much discretionary authority over workers. The employers (often, not always) have a more unique contribution to the value of the capital goods, and thus they own the property rights to that capital….Ultimately, most workers benefit from this arrangement, if only in their role as consumers; most people don’t want their co-workers in charge of the ultimate disposition of the capital goods” (Cowen 2017, 113).

20. The italics are Cowen’s.

21. As Flanigan says, “like many libertarians, I suspect that Cowen is skeptical because he thinks that to the extent that any feasible policies would curb workplace dictatorship, they also threaten economic growth or cause unemployment, and the costs associated with forgone growth and employment aren’t worth the costs of workers’ empowerment” (Flanigan 2019).

22. The researchers “studied a natural experiment in form of a 1994 reform in Germany that preserved worker representation on supervisory boards for some cohorts of corporations, while abolishing it for their slightly younger peers incorporated on or after August 10, 1994” (Jäger, Schoefer, and Heining 2019, 41).

23. For instance, see Dube 2019; OECD 2006.

24. Likewise, a report on the economic impact of Covid-19 on minority workers finds that “on every indicator of economic wellbeing, racial and ethnic minorities have larger-than-average responses to overall indicators; this means that indicators of labor-market slack (like unemployment or employment rates) deteriorate faster and further for people of color and, depending on the strength of the recovery, can take longer to make up lost ground” (Bernstein and Jones 2020, 1).

25. Mishel notes that “the racial categories are non-Hispanic Black, non-Hispanic White and Hispanic, mutually exclusive categories.” Mishel’s analysis of the monthly unemployment rate is based on data provided by the Economic Policy Institute State of Working America data library (Economic Policy Institute 2019). These data are themselves taken from the Bureau of Labor Statistics’ Current Population Survey, which is the basis of the regular monthly unemployment report.

26. “In the aggregate, quits, which are dramatically procyclical [meaning that they are higher during economic expansion and lower during a recession], comove nearly one to one with hires and job openings” (Mercan and Schoefer 2020, 101).


Anderson, Elizabeth. 2017. Private Government: How Employers Rule Our Lives (and Why We Don’t Talk About It). Reprint edition. Princeton, N.J.: Princeton University Press.

Bernstein, Jared, and Janelle Jones. 2020. The Impact of the COVID19 Recession on the Jobs and Incomes of Persons of Color. Center on Budget and Policy Priorities. June.

Bivens, Josh, and Lawrence Mishel. 2015. Understanding the Historic Divergence Between Productivity and a Typical Worker’s Pay: Why It Matters and Why It’s Real. Economic Policy Institute. September.

Bivens, Josh, and Ben Zipperer. 2018. The Importance of Locking in Full Employment for the Long Haul. Economic Policy Institute. August.

Blanc, Sandrine. 2018. “Book Review: Private Government. How Employers Rule Our Lives (and Why We Don’t Talk About It).” Business Ethics Quarterly: The Journal of the Society for Business Ethics 28, no. 2: 219–23.

BLS (Bureau of Labor Statistics). 2018. “Contingent and Alternative Employment Arrangements” (economic news release). June 7.

Board of Governors of the Federal Reserve System. 2018. Report on the Economic Well-Being of U.S. Households in 2017. May.

Bonanno, Alessandro, and Rigoberto A. Lopez. 2009. “Is Wal-Mart a Monopsony? Evidence from Local Labor Markets.” Paper presented at the International Association of Agricultural Economists conference, Beijing, China, August 16-22.

Carville, Olivia, Emma Court, Kristen Brown, and Bloomberg. 2020. “Hospitals Tell Doctors They’ll Be Fired If They Speak Out About Lack of Gear,” Fortune, March 31, 2020.

Collins, Brett, Andrew Garin, Emilie Jackson, Dmitri Koustas, and Mark Payne. 2019. “Is Gig Work Replacing Traditional Employment? Evidence from Two Decades of Tax Returns.” Unpublished paper, IRS SOI Joint Statistical Research Program.

Colvin, Alexander J.S., and Heidi Shierholz. 2019. Noncompete Agreements: Ubiquitous, Harmful to Wages and to Competition, and Part of a Growing Trend of Employers Requiring Workers to Sign Away Their Rights. Economic Policy Institute. December.

Council of Economic Advisers. 2016. “Labor Market Monopsony: Trends, Consequences, and Policy Responses.” October.

Cowen, Tyler. 2017. “Work Isn’t So Bad After All.” In Elizabeth Anderson, Private Government: How Employers Rule Our Lives (and Why We Don’t Talk About It). Reprint edition. Princeton, N.J.: Princeton University Press.

Democratic Party. 2016. “2016 Democratic Party Platform.”

Donovan, Sarah A., David H. Bradley, and Jon O. Shimabukuru. 2016. What Does the Gig Economy Mean for Workers? Congressional Research Service. February.

Dorinda Allard, Mary, and Anne E. Polivka. 2018. “Measuring Labor Market Activity Today: Are the Words Work and Job Too Limiting for Surveys?” Monthly Labor Review 12, no. 6: 1–19.

Dorman, Peter, and Les Boden. 2021. Risk Without Reward: The Myth of Wage Compensation for Hazardous Work. Economic Policy Institute. April.

Dube, Arindrajit. 2019. Impacts of Minimum Wages: Review of the International Evidence. University of Massachusetts, National Bureau of Economic Research, and IZA Institute of Labor Economics. November.

Economic Policy Institute. 2019. State of Working America Data Library.

Epic Systems Corp. v. Lewis. 2018. 138 S. Ct. 1612. U.S.

Estlund, Cynthia. 2014. “Extending the Case for Workplace Transparency to Information About Pay.” UC Irvine Law Review 4: 781.

Estlund, Cynthia L. 2018. “Book Review: Rethinking Autocracy at Work.” Harvard Law Review 131, no. 3: 795–826.

Fadel, Leila. 2020. “Doctors Say Hospitals Are Stopping Them From Wearing Masks.” NPR, April 2, 2020. 

Farrell, Diana, Fiona Greig, and Amar Hamoudi. 2018. The Online Platform Economy in 2018: Drivers, Workers, Sellers, and Lessors” JPMorgan Chase Institute.

Federal Trade Commission. 2020. “Non-Competes in the Workplace: Examining Antitrust and Consumer Protection Issues.”

Ferretti, Thomas. 2018. “Book Review: Private Government: How Employers Rule Our Lives (and Why We Don’t Talk About It).” Economics and Philosophy 34, no. 2: 275–82.

Flanigan, Jessica. 2019. “Author Meets Critics: Elizabeth Anderson’s ‘Private Government.’” Presentation at the American Philosophical Association Eastern Division Meeting, New York, N.Y., January 8.

Freeman, Richard B., and Joel Rogers. 2006. “What Workers Want.” ILR Press Book Samples.

Grabell, Michael, and Bernice Yeung. 2020. “Emails Show the Meatpacking Industry Drafted an Executive Order to Keep Plants Open.,” ProPublica, September 14, 2020.

Howell, D. 2019. From Decent to Lousy Jobs: New Evidence on the Decline in American Job Quality, 1979–2017. Washington Center for Equitable Growth. August.

Huemer, Michael. 2013. The Problem of Political Authority: An Examination of the Right to Coerce and the Duty to Obey. London, England: Palgrave Macmillan.

Jäger, Simon, Benjamin Schoefer, and Jörg Heining. 2019. “Labor in the Boardroom.” Working Paper Series. National Bureau of Economic Research. doi:10.3386/w26519.

Kindy, Kimberly, Ted Mellnik, and Arelis Hernández. 2021. “The Trump Administration Approved Faster Line Speeds at Chicken Plants. Those Facilities Are More Likely to Have Covid-19 Cases.” Washington Post, January 3, 2021. 

Krueger, Alan B. 2018. “Reflections on Dwindling Worker Bargaining Power and Monetary Policy.” In Luncheon Address at the Jackson Hole Economic Symposium. Vol. 24.

Krueger, Alan B., and Eric Posner. 2018. A Proposal for Protecting Low-Income Workers from Monopsony and Collusion. The Hamilton Project. February.

Kuhn, Peter. 2004. “Is Monopsony the Right Way to Model Labor Markets? A Review of Alan Manning’s Monopsony in Motion.International Journal of the Economics of Business 11, no. 3: 369–78.

Labor-Management Standards Office. 2010. “Notification of Employee Rights Under Federal Labor Laws.” Federal Register 75 (May): 28367–402.

Levy, Jacob. 2019. “Author Meets Critics: Elizabeth Anderson, Private Government: How Employers Rule Our Lives (And Why We Don’t Talk About It).” Presentation at the American Philosophical Association Eastern Division Meeting, New York, N.Y., January 8.

Maestas, Nicole, Kathleen Mullen, David Powell, Till Von Wachter, and Jeffrey B. Wenger. 2018. “The Value of Working Conditions in the United States and Implications for the Structure of Wages.” Working Paper w25204. National Bureau of Economic Research.

Manning, Alan. 2003. Monopsony in Motion: Imperfect Competition in Labor Markets. Princeton, N.J.: Princeton University Press.

Manning, Alan. 2011. “Imperfect Competition in the Labor Market.” In David Card and Orley Ashenfelter, eds., Handbook of Labor Economics. London, England: Elsevier.

Marx, Karl. 1912. Capital: A Critique of Political Economy. Edited by Frederick Engels, translated by Samuel Moore and Edward Aveling. Chicago, Ill.: Charles H. Kerr, 195–96.

Mercan, Yusuf, and Benjamin Schoefer. 2020. “Jobs and Matches: Quits, Replacement Hiring, and Vacancy Chains.” American Economic Review: Insights 2, no. 1: 101–24.

Mishel, Lawrence. 2018. Uber and the Labor Market: Uber Drivers’ Compensation, Wages, and the Scale of Uber and the Gig Economy. Economic Policy Institute. May.

Mishel, Lawrence. 2020. Letter to Chetan Cetty. October 2.

Mishel, Lawrence and Josh Bivens. 2021. Identifying the Policy Levers Generating Wage Suppression and Wage Inequality. Economic Policy Institute.

Mishel, Lawrence, and Celine McNicholas. 2019. Uber Drivers Are Not Entrepreneurs: NLRB General Counsel Ignores the Realities of Driving for Uber. Economic Policy Institute, September.

Munger, Michael. 2018. “Book Review: Private Government: How Employers Rule Our Lives (and Why We Don’t Talk About It).” Independent Review 22, no. 4: 627–29.

National Association of Manufacturers et al. v National Labor Relations Board et al. 2013. United States Court of Appeals.

National Labor Relations Board. 2020. “Representation-Case Procedures: Election Bars; Proof of Majority Support in Construction-Industry Collective-Bargaining Relationships.” Federal Register 85 (April): 18366–400.

Nozick, Robert. 1977. Anarchy, State, and Utopia. New York, N.Y.: Basic Books.

OECD (Organisation for Economic Co-operation and Development). 2006. “Boosting Jobs and Incomes: Policy Lessons from Reassessing the OECD Jobs Strategy.” Address to the meeting of the G8 Employment and Labour Ministers, Moscow, October 9–10, 2006.

Philp, Mark, ed. 1998. Thomas Paine: Collected Writings. Oxford, England: Oxford University Press.

Public Square. 2017. “Private Government: How Employers Rule Our Lives (and Why We Don’t Talk About It).” YouTube video. Published November 4, 2017.

Republican Party. 2016. “The 2016 Republican Party Platform.”

Rothwell, Jonathan, and Steve Crabtree. 2019. Not Just a Job: New Evidence on the Quality of Work in the United States. Gallup, Lumina Foundation, Bill & Melinda Gates Foundation, Omidyar Network. October.

Smith, Adam. 1981. An Inquiry into the Nature and Causes of the Wealth of Nations, Vol. 1., Glasgow Edition of the Works and Correspondence of Adam Smith. Indianapolis, Ind.: Liberty Fund.

Smith, Noah. 2019. “Economists Get Serious About the Harm from Monopolies.” Bloomberg Opinion, January 11.

Sokolova, Anna, and Todd Sorensen. 2018. Monopsony in Labor Markets: A Meta-Analysis. IZA Institute of Labor Economics. November.

Stansbury, Anna, and Lawrence H. Summers. 2020. “Declining Worker Power and American Economic Performance.” Brookings Papers on Economic Activity.

Starr, Evan, J.J. Prescott, and Norman Bishara. 2020. Noncompetes in the U.S. Labor Force. University of Michigan Law School. doi:10.2139/ssrn.2625714.

Tomassetti, Julia. 2021. Power in the Employment Relationship: Why Contract Law Should Not Govern At-Will Employment. Economic Policy Institute.

Weil, David. 2019. “Understanding the Present and Future of Work in the Fissured Workplace Context.” Russell Sage Foundation. Journal of the Social Sciences 5, no. 5: 147–65.

See more work by Chetan Cetty