Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Economics and Finance. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 02 February 2023

# Monopsony Power, Race, and Gender

• Aida FarmandAida FarmandDepartment of Economics and Schwartz Center for Policy Analysis, The New School for Social Research
•  and Teresa GhilarducciTeresa GhilarducciDepartment of Economics and Schwartz Center for Policy Analysis, The New School for Social Research

### Summary

Most monopsony research leaves out the employer as an active agent. The cause of monopsony rests solely on the workers: Their idiosyncratic preferences, their lack of information, and their geographical isolation create the monopsony conditions. Employers are viewed as mainly passive and only choose to exploit their monopsony potential when the conditions allow. The theoretical passivity of employers leaves out a whole class of behaviors necessary to identify and understand the persistence of monopsony.

For instance, the models consider gender as a monopsony factor because wives and mothers are presumed to have intensely inelastic labor supply functions. Women’s attachment to children and the children’s schools and to their husband’s locational decisions means women are less likely to leave a geographical area to pursue a competitor’s better offer. Again, it is the woman’s idiosyncratic choices that allow for monopsony exploitation.

However, it is likely employers consciously use race and gender stratification to segregate and divide workers to create differential labor supply elasticities and, thus, create monopsony conditions to the firm. A firm would maintain practices that use race to allocate jobs and separate men from women workers to maintain divisions among the workforce. Moreover, government policies that make it difficult for workers to unionize, keep minimum wages low, and subsidize low-paid work through the earned income tax credit help employers create and maintain monopsony power among subaltern groups, nonwhite workers, and women. Future research on monopsony should focus on specific employers’ practices that create monopsony conditions such as providing firm specific childcare, perpetuating occupational segregations, limiting opportunities of promotion for women and nonwhite workers, and lobbying for the wage subsidy programs such as the earned income tax credit.

### Subjects

• Economic History
• Labor and Demographic Economics

### Gaps in Monopsony Research

Monopsony research has not, until recently, focused on how the employer actively creates monopsony situations. In the nearly 100 years’ worth of monopsony literature, worker behavior drives monopsony conditions. From Joan Robinson in 1933 to Alan Manning in 2003 and beyond, scholars often place workers center stage. In monopsony models, workers’ idiosyncratic “preferences,” their lack of information, and/or geographical isolation create monopsony conditions, meanwhile employers are mainly passive. Monopsony models, famously Joan Robinson’s initial formulation, incorporate gender by assuming wives and mothers have exogenous preferences that make them less mobile than men and racial differences are barely considered. How employers use societal racism and gender stratification to create and extend monopsony power—mainly through occupational segregation and targeted employee benefits, for example, providing onsite childcare—is not generally present in the standard monopsony models. Exceptions have been the recent work on the degree of U.S. labor markets concentration that show employers act to create monopsonies by resisting unions and increases in minimum wages, conspiring with anti-poaching agreements and imposing non-compete contracts and zero-hours work contracts.

Government policies—supported by employers—also create monopsony conditions. In the last days of the Obama administration the Council of Economic Advisers brief on monopsonistic labor markets emphasized occupational licensing agreements and employer actions in creating monopsony, not race or gender stratification. What is missing is a more extensive study of the labor markets across different sectors and occupational groups, by gender, age, and race.

Future research on monopsony should focus on specific employers’ practices that create monopsony conditions such as providing firm-specific childcare, perpetuating occupational segregations, limiting opportunities of promotion for women and nonwhite workers, and lobbying for the earned income tax credit, which increases the supply of female labor (Eissa & Hoynes, 2006). Once women are in in the labor force, it may potentially lead to lower levels of labor supply elasticities for young women, especially mothers. Racial, gender, and age discrimination also make way for monopsony exploitation by limiting the rate at which a worker can receive job offers. This means a subaltern worker has lower firm-specific labor supply elasticity and the firm has increased level of potential monopsony power.

This article contributes to the literature on monopsony models by moving away from their emphasis on exogenous factors—worker preferences, incomplete information, and barriers—affecting firm—specific labor supply elasticities and focusing on these factors as the main drivers of monopsony power. Employers have compelling profit reasons to create monopsony conditions and create labor market frictions. Using existing social stratifications and divisions give employers the ability to create and deepen monopsony conditions.

A comprehensive survey of the literature calls attention to the limitations in monopsony scholarship. The scholarship often stops short in exploring not only the active role of employers in creating monopsony conditions but also employers’ explicit use of race, gender, and other sources of stratification to gain advantage. This study aims to incorporates the economic stratification and the monopsony literature.

### Monopsony Basics

The monopsony model is a beloved model of all methodological traditions– neoclassical, institutional, and some radical economists—use monopsony models. Pro–free market economics use monopsony frameworks to oppose occupational licensing agreements—which often give more bargaining power to the worker—institutionalists who call for the end of non-compete and anti-poaching agreements, bolstering unions, and increasing minimum wages use monopsony as a justification. Radicals evoke monopsony to show markets persistently exploit workers.

At its core, the modern monopsony model rests on employers having more bargaining power than workers, but earlier economists focused more on the employer as actively aware of their power. Manning describes the history of thought in the first chapter of the breakthrough textbook and review of the literature, Monopsony in Motion (2003). Adam Smith in the Wealth of Nations observed, “in the long run the workman may be as necessary to his master as his master is to him; but the necessity is not so immediate” (Smith, 1970). The 19th-century economists in favor of worker collective action—Sidney Webb and Beatrice Webb—argued workers are typically more desperate for work than an employer is desperate for a particular worker. “The art of bargaining . . . forms a large part of the daily life of the entrepreneur. The manual worker, on the contrary, has the smallest experience of, and practically no training in, what is essentially one of the arts of the capitalist employer” (Webb & Webb, 1897 657-658). And Alfred Marshall in the Principles of Economics wrote, “labor is often sold under special disadvantages arising from the closely connected group of facts that labor power is ‘perishable,’ that the sellers of it are commonly poor and have no reserve fund, and that they cannot easily withhold it from the market” (Marshall, 1920 457).

Unlike economic models that presume discrimination cannot exist in a competitive labor market because employers must pay competitive wages to survive, the monopsony framework recognizes the unequal power relationship between employers and workers and the valuable ability of the employer to pay more or less than the competitive wage to maximize its objectives. Monopsony models are fertile ground for connecting employers’ actions to perpetuating systemic race and gender discriminations that serve their interests. But monopsony theory has not yet achieved this goal fully.

The company town with immobile residents was the exemplar of monopsony and therefore quaint and rare. Joan Robinson, in The Economics of Imperfect Competition, showed monopsony was a more general case of imperfect competition for three reasons. The firm will likely face an upward-sloping labor supply curve because “(a) workers’ inertia resulting in insensitivity to pay; (b) ‘preference or custom’ could attach workers to firms; and (c) ignorance may prevent workers from moving from one job to another in response to differences in the wages offered by the other firm” (Robinson, 1969, p. 296).

Robinson provided a service to economics by putting monopsony and imperfect competition center stage, but by doing so spawned decades of monopsony models that are easy on employers and hard on workers. Workers most affected by monopsony exploitation—women and nonwhite workers—are singled out for blame in bringing on the monopsony condition. For instance, Robinson notes women tend to stay with an employer for family reasons even if a higher paying job is elsewhere. Female workers – presumed to be married, wanting to be married, and more tied to elderly parents – have a more inelastic supply curve because they are tied to their family’s, husband’s and children’s geography. Geography, not employers are implicated, workers that face mobility barriers or geographical isolation have more inelastic supply responses.

Manning (2003) inspired a conceptual shift in the literature by using the Burdett and Mortensen (1998) worker-search models to describe monopsony. Though Manning—as all microeconomists who recognize the widespread presence of monopsony—emphasizes that workers are not equally empowered agents compared to employers, he limits the employers’ autonomous actions to offering wages ex ante—rather than negotiating them—and benefiting from the “frictions” that exist exogenously (and not related to firm behavior). Workers buy houses, embed themselves in a community, do not gather information about job offers, all making separations and search expensive, and might have preferences for certain kinds of non-pay aspects of jobs. An example of the type of worker who is “naturally” reluctant to move and more unlikely to insist on a pay increase, even if a competitor would pay more, are women connected to the household because of disproportionate care burdens and the higher likelihood of being a secondary income earner whose job search is shaped by her husband (Stelzner & Bahn, 2021). But homeowners and people with families in the community are also more unlikely to move. None of these causes of monopsony are the employer’s doing.

Manning (2003) cites worker’s heterogeneous tastes, having mobility constraints, or being ignorant of the competition as the main causes of monopsony. The dominant view of what causes monopsony loses sight of how employer and policy forces and race and gender stratification create monopsony conditions.

### Methodological Issues in Measuring Monopsony Power

Key evidence of monopsony exploitation—workers not being paid the maximum the firm can afford before reducing employment—is measured in a mechanistic way (note, in a competitive market the elasticity is infinite), and if the researcher finds evidence that the wage elasticity of labor supply to a firm is not infinite, or is tiny, the researcher concludes there is evidence of monopsony. The main idea behind monopsony is that the labor supply curve to an individual employer is not infinitely elastic so that an employer that cuts wages by one cent may find it harder to recruit and retain workers but does not immediately lose all their existing workers to competitors as is predicted by the perfectly competitive model (Manning, 2021).

In modern and widely used models, monopsony is detected if firm expansions cause wage increases because, unlike employers in competitive markets, monopsonists can pay lower wages than the going rate and not lose anyone—employers only raise wages when expanding employment—a clear indication of inelastic supply curves to the firm. Economists estimate the degree of inelasticity to determine the magnitude of the monopsony power present in a labor market, with estimates of labor supply elasticity facing the firm ranging from 1.4 to 14 depending on the method of estimation (see Sokolova & Sorensen, 2021 meta-analysis of 53 papers). For instance, Hirsch et al. (2010) find—using a large, linked employer-employee German data set and comparing individuals across a wide range of jobs and employers—the average elasticity of labor supply to the firm is in the range of 2–4. Ransom and Sims (2010) find a firm-specific elasticity of labor supply of 3.7 for Missouri public school teachers. Ransom and Oaxaca (2010) find elasticities ranging from about 1.5 to 2 for women, and 2.5 to 3 for men. Women have lower measures, which is consistent with women facing more monopsony exploitation. Small elasticities suggest significant levels of employer market power (especially if they are long-run elasticities).

Omitted variables in estimating elasticities is a methodical concern. For example, in Ransom and Sims’s (2010) analysis, if the higher wages of some school districts represent compensating differences for difficult working conditions (as in teaching in tough urban schools), then the apparent differences in wages are greater than the real differences and the estimated elasticities will be small as well. However, higher wage jobs typically have better benefits, but, as in any labor market study, full compensation should always be used. Firm specific labor supply elasticity measures are biased if wages and employment are endogenous (Sokolova & Sorensen, 2021). Without exogenous variation in the wages provided by a single firm it is difficult to produce credible measures of firm-specific elasticities.

Recent literature using experimentally manipulated wages for online tasks—using Amazon Turk—typically find low labor supply elasticities (Dube et al., 2020). But the study does not consider variations by race or gender likely because of data constraints.

By randomly offering samples of male and female drivers of a national ride-share company wage increases, Caldwell and Oehlsen (2018) challenge the Robinson-type assumptions by finding that women have Frisch elasticities twice the size of men on both the intensive and extensive margins. Women seem no less willing than men to switch firms in response to offers in relative wages.

The studies mentioned assess the degree of firms’ wage-setting power by estimating the wage elasticity of labor supply to a firm, but they do not discuss how and to what extent firms can create or exercise monopsony power.

Structural and institutional factors impact how much of monopsony power can be exercised by employers. Eroding labor market institutions and declining degrees of market competition due to rising market entry barriers (e.g., non-compete clauses, occupational licensing) (Cunningham, 2019) create more conditions for monopsony power.

Other factors, such as occupational segregation, can increase employers’ ability to exercise monopsony power by reducing the rate of job offers arrivals for subaltern groups while simultaneously increasing their ability to exercise this power by lowering the likelihood of subaltern workers quitting.

Note there are three non-inclusionary ways to discuss monopsony power—causes of monopsonistic conditions, causes of employers’ ability to exercise their monopsony power, and causes of employers creating monopsony power. For example, discriminatory pay and promotion practices reduce job arrival rates and predicted wage rates that in turn affect the intensity of monopsonistic conditions. Declining union power and the real value of the minimum wage help employers exercise monopsony power. Both result in lower wages, but due to structures versus employers’ rent-sharing behaviors.

Often employers’ ability to manipulate wages is perceived as exogenous—lucky conditions for the employer—because of market concentration (Azar et al., 2020; Benmelech et al., 2020; Webber, 2015), restrictions to mobility (Naidu et al., 2016; Naidu & Yuchtman, 2013), and moving costs (Ransom, 2021). But there are nuanced distinctions between sources of monopsony and how they are exercised by the employers. Labor market frictions, workers’ lack of information about outside options, and mobility barriers can be sources of monopsony. Employers take advantage of these labor market frictions to exploit workers by paying lower wages and reducing the overall level of employment. But a more fulsome model also theorizes that employers are not only recipients of their monopsony power but are also creators of monopsony power. The key and often overlooked argument here is that employers have an active role in creating monopsonistic labor markets and can decide the degree to which they want to expose different groups of workers to exploitation. Subsequent discussion about employers’ potential monopsony power will highlight tactics employers may use directly to create, intensify, and maintain monopsony power.

### Employers’ (Potential) Monopsony Power Can Explain Discrimination

One of the insights missed in the modern literature is Martin Bronfenbrenner’s 1956 paper that argues every employer is a potential monopsonist who pays slightly higher wages than the minimum to attract more applicants than needed. Employers always have the potential to be monopsonists because they can create excess supply. This raises a familiar monopsony question: Will the employee leave if offered a better package and, more poignantly, will that worker leave if the current employer lowers total compensation? Relevant to how racial and gender stratification helps a firm maintain monopsony power is that a worker facing racial and gender discrimination has a higher cost of separation and lower returns to skill. Everything else held constant, it is more difficult for women and nonwhites to receive a competitive wage offer as indicated by the well-documented gender and racial wage gaps. Therefore, it is more unlikely, everything else held constant, that a subaltern worker would leave a current employer and face the risk of encountering a bigoted, potential employer. Research shows that Black and female workers are aware of these obstacles and adjust their employment decisions in response to labor market discrimination. Heckman (1998) presents evidence that Black workers will self-select and remain in segments of the labor market where their skills will be better rewarded. While women self-select college majors and occupational tracks based on gendered considerations including discrimination (Pager & Pedulla, 2015). More research is needed to understand how variation in wage offers by gender and race affect labor supply elasticity of different groups.

Table 1 suggests women and nonwhites are paid less for the same work. Black and Hispanic workers and women with similar educational credentials and hours of work earn lower wages than white men. Comparing full-time workers across different industries with the same education levels, a Black female worker with a college degree earns median weekly earnings of $1,115 compared to their white male counterparts$1,634. In female-dominated industries—education and healthcare—the wage gap between white and nonwhite female workers is larger compared to all industries, indicating that occupational segregation has a bigger impact on lowering wages for nonwhite workers.

Racial disparities also exist between men. Across different industries, full-time Hispanic male workers with a college degree have median weekly earnings of $1,364, while white male workers earn$1,634. Nonwhite and female workers have lower weekly earnings across the board.

#### Table 1. Full-Time Black and Hispanic Workers With a College Degree Have Lower Weekly Earnings

All industries

Male-White

Male-Black

Male-Hispanic

Female-White

Female-Black

Female-Hispanic

Median weekly earnings (2020)

$1,634$1,250

$1,346$1,211

$1,115$1,092

Education industrya

Male-White

Male-Black

Male-Hispanic

Female-White

Female-Black

Female-Hispanic

Median weekly earnings (2020)

$1,600$1,211

$1,269$1,250

$1,086$1,084

Healthcare industryb

Male-White

Male-Black

Male-Hispanic

Female-White

Female-Black

Female-Hispanic

Median weekly earning (2020)

$1,600$1,250

$1,442$1,250

$1,070$1,160

Note: Sample is restricted to full-time wage and salary workers with a college degree.

a Education industry includes employees working in elementary and secondary schools, colleges, and universities (including junior colleges); business, technical, and trade schools and training; other schools and instruction; and educational support services.

b Healthcare industry includes employees working in offices of physicians, dentists, chiropractors, optometrists, and other healthcare practitioners; outpatient care centers; home healthcare services; general medical and surgical hospitals, and specialty hospitals; psychiatric and substance abuse hospitals; nursing care facilities (skilled nursing facilities); residential care facilities; individual and family services; community food and housing, and emergency services; vocational rehabilitation services and child daycare services.

Source: IPUMS-CPS Outgoing Rotation Group-2020 (Flood et al., 2020).

Monopsony can explain persistent wage discrimination not based on productivity. Firms gain latent monopsony power from racism, sexism, and other forms of social discrimination, like undocumented status, because such practices and institutions make the worker uncommonly attached and beholden to a particular employer. It should be noted that all employers could benefit from the explicit discriminatory actions of a portion of employers. Because white workers do not differentiate between working for a discriminatory or nondiscriminatory firm, their labor supply function will be independent of firm type. It is therefore evident that discriminatory and nondiscriminatory firms will both choose the same wage offer for white workers. Prejudiced firms refuse to hire nonwhite workers at any positive wage. However, presence of prejudiced firms increases the monopsonistic power of unprejudiced firms toward nonwhite workers, and the wages offered by unprejudiced firms would also be lower for nonwhite workers than for white workers (for details refer to Lang & Lehmann, 2012).

Stelzner and Bahn’s (2021) insight that workers with less wealth have less bargaining power, and thus the racial and gender wealth gap creates monopsony power, needs extending to explore how employers take advantage of nonwhite and gender wealth gaps. Wealth matters in monopsony because non-labor income strengthens a worker’s fallback position and can help a worker afford a move to change jobs. A worker who can leave is more likely to ask and receive a raise. Stelzner and Bahn (2021) point out that wealth inequality creates inequality in fallback positions that in turn motivates employers to act on these reservation wage inequalities. Because Black and Latinx families have less wealth and women have more care and home responsibilities, the intersection of being a woman and nonwhite causes even lower wages for the same productivity than the addition of the negative separate effects of being nonwhite and being a woman (Stelzner & Bahn, 2021).

Also, employers may not be pure profit maximizers. Thus, employers can use personal prejudices to make decisions about who to hire and how much to pay. Racial and gender discrimination can persist if the typical employer is not striving to maximize anything, not even short-run monopsony profit. Institutional labor economics have always recognized employer behavior diverging from profit maximization, which helps explain even more racial and gender discrimination. When a firm is ‘satisficing,’ a firm strives for an aspiration that falls significantly short of maximizing behavior (Robinson, 1969). Or if a firm manager is an optimizer, profits may not be the only thing in the objective function—other goals could include control over the enterprise and public relations with consumers, workers, governmental bodies’ (Bronfenbrenner, 1956). Decades of labor economics and industrial organization scholarship recognizes principal agent problems that convincingly find profit maximizing as too-crude an explanation of a typical firm’s labor policies. Dube et al. (2018) argue that monopsony power blunts the cost of mispricing labor, hence removing the need for firms to stay on the knife-edge profit-maximization requirements. Instead of a firm using their monopsony power to minimize wage rates, the firm uses their monopsony power to raise employment standards or indulge in their prejudices. This is what Bronfenbrenner defines as potential monopsony power.

A potential monopsonist employer could indulge the monopsony exploitation by insisting the worker be white or male. That women and nonwhites are last to be hired (though equally qualified) is also consistent with potential monopsony. Employers can create monopsony power by offering a pay package that exceeds the minimum compensation package necessary to attract the mobile worker. By creating a surplus labor supply, employers can choose among equally qualified dental hygienists, purchasing managers, and so on, and, therefore, rather than setting a wage equal to an equilibrium wage just low enough to clear the labor market—such as a reservation wage or a competitive wage—the firm sets rates a little higher so that hiring officers have flexibility to pick and choose among applicants and store an excess inventory of labor in case demand increases (Bronfenbrenner, 1956). Rather than the sweatshop wages required for profit maximization under the active monopsony model, the employer chooses a higher wage and creates an excess of applicants. The rejected workers presumably remain unemployed or take less desirable jobs.

Much of labor market theory establishes that firms set hiring standards based on their interpretation of factors that predicts productivity (Spence, 1973). These expectations and judgements lead to a great degree of biases connected with gender and race. And, as expected, an employer’s taste for white men would lead to decreased bargaining power for the group of subaltern workers (Agesa & Hamilton, 2004).

But it is known from occupational segregation that discrimination also means reserving the more desirable and better paying jobs for the favored group and relegating the less desirable and lower paying jobs to the other group. A full monopsony model should consider occupational segregation and further research should explore an empirical analysis of how occupational segregation affects labor supply elasticities. The idea is that occupational segregation reduces job arrival rates for the subaltern group thereby reducing their reservation wage. Another avenue of exploration would be developing a model of monopsonistic discrimination based on the standard Burdett and Mortensen (1998) models of job search and equilibrium wage distribution to estimate gender- and/or racial-specific elasticities of labor supply.

Under conditions of occupational segregation, nonwhite workers receive lower wages and do dirtier jobs than white counterparts even though the nonwhite labor supply elasticity might be higher than that of white labor. In other words, monopsony is one among many factors suppressing wages for equivalent labor. Wage (and conditions of work) repression comes about in the case of occupational segregation because subaltern labor is used as a reserve army of labor. Low wages and poor working conditions can exacerbate labor market churn, which, in turn, affects workers’ wage growth by interrupting the accumulation of firm-specific and occupation-specific human capital.

There are obvious analogies between firms that have potential monopsony power and firms that are in monopolistic competition. Firms in noncompetitive markets vary the product rather than prices, similarly employers in noncompetitive labor markets vary the type of worker hired, not the wage package. Bronfenbrenner notes that in monopoly situations consumers resent price increases so the firm avoids the resentment by reducing the quality or service at a constant price (Bronfenbrenner, 1956). Something similar happens in noncompetitive labor markets— hiring standards are often altered instead of changing wages.

Bronfenbrenner’s insight about the importance of potential monopsony provides a framework to explain how employers use racism and patriarchy to create surplus supplies of applicants and reserve armies of labor.

The potential monopsony wage rate is higher than the minimum wage necessary to attract the minimally qualified worker so the employer can pick and choose. The exploitation is still there, it just does not show up in the wage. For instance, using employee records of the Ford Motor Company from 1918–1947, Foote et al. (2003) find that Ford profited from other firms’ racial bigotry by offering Black workers more than their competitors and the same as white people. Thus, there were minimal racial wage gaps inside Ford. But Black workers faced more monopsony exploitation because they quit Ford jobs less often than white workers. Profits came from exploiting this non-wage aspect of the jobs, as Ford placed Black workers in hot, dangerous foundry jobs where quit rates were generally high (Foote et al., 2003). A modern-day version of the faux non-bigoted and authentically secure monopsonist employer is one that promotes a diversity-friendly environment with onsite childcare and nonwhite friendly programs.

Whole industries could be built on this type of exploitation and these hiring and pay strategies. The firms have a lot of reasons to keep wages high and use its exploitation power in ways other than the active monopsony model. When a monopsonistic employer raises wages to get more labor, the employer can also attract labor in waiting. The surplus labor group may be made larger by making wages higher than normal and providing greater selectivity for employment.

In the Bronfenbrenner potential monopsony framework, a firm with a perpetual excess supply of applicants can expand without having to raise wages. The firm can engage in quality dilution rather than raising wages. In addition, wages may even be lowered by keeping the “labor in waiting” small and hiring practically any one willing to work at that wage (Bronfenbrenner, 1956).

In the post–World War expansion, the sudden establishment of four steel and defense plants caused 35 out of 70 firms in New Jersey to lower their hiring standards for new employees while only 12 firms raised starting wages (Lester, 1955). The conventional, active monopsony model would have predicted wage increases for all firms given the short-run fixed labor supply. After all most firms were monopsonists, and the unemployed, unchosen workers, women and nonwhite men, faced unemployment and were hired with lower wages when employment expanded.

But imagine – and it is not implausible – that employers are aware of and thus utilize the existence of different supply elasticities from different groups. Suppose two or more groups of workers, that is, preferred employees and subaltern employees, have different supply patterns distinguishable by a monopsonistic employer. Would the employer ignore the differences and pay identical wages to all groups? Conventional monopsony theory solved this problem. The group with the lower elasticity of supply will gain in both wages and employment if discrimination is removed, while the group with the higher elasticity of supply will lose on both counts, and so will the employer. This is to say, that given monopsony, the high-elasticity group gains from discrimination, as does the employer (Bronfenbrenner, 1956). This is the classic Robinson finding. But what this framework fails to recognize is that this is not the only avenue of exploitation available to the employer. Discrimination and exploitation do not only take the form of unequal pay for identical work; it can manifest in nonwhites and women having the same pay but worse job stability and working conditions. The more common form of discrimination is reserving the more desirable and better paying jobs for the preferred employees.

For instance, Storer et al. (2020) find that white workers in the retail and food service sectors are significantly advantaged in terms of job quality. White workers are less likely to work consecutive closing and opening shifts, be forced into part-time work, or have employers cancel their shifts and require them to work “on-call.” Also, white workers have less difficulty being satisfied when they request for time off work (Storer et al., 2020). Because most workers in these occupations earn incomes near the minimum wage, the racial wage gaps are not as pronounced as other occupations; still, nonwhite workers are subject to exploitation in the form of more precarious working schedules.

Bronfenbrenner (1956, 579) argues that the common trait of potential monopsony or monopoly is the use of “rules, rituals, and routines,” but he did not elaborate on how employers could use the rules, rituals, and routines of patriarchy and race peculiar to the firm or common to its industry to attract and retain workers, which in theory are solved by the market mechanism in a competitive economy. How firms use and extend patriarchy and racism in personal practices are vital to understanding the persistence of monopsony power.

### Monopsony Power and the Intersection of Race, Age, and Gender

Though many firms are not hiring from within and appear to use contingent and temporary labor, many firms are still relying on their internal labor market, and for their employees, seniority means occupational promotion. Promotion from within is still the general rule. Under these circumstances, a person is hired not for one job at one wage rate but for a roughly envisioned sequence of jobs at an equally roughly estimated sequence of wage rates. This is perhaps truer for clerical and management workers than for manual workers. Older workers have given the “best years of their working lives to their employers,” and when they are let go, Bronfenbrenner (1956, p. 588) calls it “unfair” or even “exploitative” because they do not get their surplus back. In Marxian terms but with a modern twist, these workers have deposited surplus value with the employer and are deprived of the chance to take it back because of early retirement. This idea of the loyal worker being more prey to exploitation is supported by research on age discrimination.

While monopsony impacts the gender wage gap overall, age discrimination affects older women in particular. Age discrimination adds to the constraints older women face in the labor market and makes monopsonistic exploitation even more likely. Neumark et al. (2019) found evidence for prevalent age discrimination against older women. Job applicants near retirement age receive callbacks at a 35% lower rate than prime-age workers, and most of this bias is experienced by older women workers. As older women have more difficulty receiving job offers, employers can offer them lower wages or more difficult working conditions, which leads to exploiting this cohort of workers.

As noted in the section “Gaps in Monopsony Research,” women and men have different separation rates elasticities, that is, women are not as mobile as men. A classic test of Joan Robinson’s insight comes from any study that shows women have lower firm-specific labor supply than men. Therefore, a classic sign of monopsony power and exploitation is that women are more tied to a firm than men are.

As noted, in Germany women’s attachment to their employer is stronger than men’s; their elasticities with respect to wage are small (Hirsch et al., 2010). Vick (2017) also takes gender differences in job mobility as given and shows that in Brazil because women do not move as readily as men, they have lower earnings than equivalently qualified men. Women’s lower elasticity of supply to a firm confirms the monopsony theory’s explanation of the pay gap. Vick (2017) finds male elasticity ranges from 1.6 to 2.2 while female elasticity is much lower, from 1.2 to 1.5, and concludes that women suffer a wage penalty of 11.4%–20.5%. compared to an actual gender wage difference of 16.4%. Similarly, using matched employer-employee data from Norway, researchers find women had smaller wage elastic labor supply facing each establishment than men (Barth & Dale-Olsen, 2009).

The monopsony model and the estimates imply a significant portion of the pay gap, about 33%, between men and women is explained by women’s greater reluctance to leave their current employer for a better job. Employers can pay a woman less who does the same job and as well as a male employee without consequences, such behavior is consistent with wage discrimination by profit-maximizing monopsonistic employers (Boal & Ransom, 1997).

The language explaining the reasons for women’s smaller elasticities are tightly focused on gender roles, on personal decisions of women and men, and on forces outside the firm. Women’s labor supply is less elastic than men’s at the firm level because women have different preferences about job characteristics (they want family-friendlier working conditions and will take less pay to get them) than men and women are less mobile than men (Vick, 2017). The soft implication is that it is the woman’s fault, culture’s fault, and the arc of human history’s fault, but not it is not fault of the firm’s behavior under the dynamics of capitalism.

Nonwhite workers are underrepresented in high-paying industries and occupations (Carnevale et al., 2019), leading to racial wage gaps that are especially large at the top of the distribution. Employers often argue that this pattern is due to their stated perception that there is a shortage of qualified nonwhite workers rather than to biases in their policies. But as Gerard et al. (2021) show—using Brazil as a case study—assortative matching accounts for about two thirds of the underrepresentation gap. The remainder reflects an unexplained preference for white workers at higher paying establishments. The wage losses associated with unexplained sorting and differential wage setting are largest for nonwhites with the highest levels of general skills.

Besides the research discussed, the monopsonistic framework has paid little to no attention on racial disparities.

### Future Research is Needed on Employers and Policies Actively Creating Monopsony

Employer action is sometimes noted in the monopsony literature. The existence of anti-poaching agreements (see Marinescu & Posner, 2019) and other employer-initiated devices are signs an employer is creating a monopsony. There also have been some governmental policy responses to curb employers’ exploitative practices. Nineteen states and 22 localities have adopted laws and regulations that prohibit employers from requesting salary history from job applicants (Washienko, 2021). The laws are aimed at ending racial and gender pay discrimination. This practice will help curtail employers’ discriminatory power because it discourages them from discovering workers’ reservation wages and reduces wage premiums for privileged groups of workers (Sinha, 2019). Knowing the reservation wage of the employee puts women and nonwhite workers at a disadvantage because they usually have lower reservation wages (Caliendo et al., 2017) and allows employers to share the rents with privileged groups.

Moreover, some jurisdictions prohibit an employer from taking disciplinary action against employees who discuss pay with coworkers. Keeping salaries secret reinforces discrimination by making it harder to uncover gender and race-based pay disparities. This lack of transparency allows employers to practice their monopsony power by offering higher wages to their preferred employees.

But even when nonwhite workers discuss pay and ask for raises, they face more obstacles to receiving higher salaries than white workers. Research has shown that when Black job applicants negotiated their starting salaries, evaluators viewed them as more aggressive than white job applicants, which eventually led to Black job applicants receiving lower starting salaries as a result of the bargaining process (Hernandez et al., 2019). This links biases to employers’ ability to exercise monopsony power by nurturing stereotypes and shows that workers’ information frictions are not the main driver of monopsony exploitation experienced by nonwhite workers.

Moreover, employers continue utilizing practices that bolster their monopsony power well beyond the hiring process. Employers pay for general training because part of its returns will accrue to employers—better trained and higher paid employees are less likely to leave. Current employer’s information about worker’s ability relative to other firms gives the employer an ex-post monopsony power over the worker that encourages the firm to provide training (Acemoglu & Pischke, 1998).

Democratic nations propelled by labor and popular interests tend to actively regulate labor markets even when the government efforts are challenged by elites and employers. Most of the government policies are aimed at reducing monopsony exploitation, even though they aren’t billed that way. For example, the pro-labor legislation in many if not all countries is consistent with attempts to reduce monopsony power. Section 1 of the U.S. National Labor Relations Act of 1935 says “the inequality of bargaining power between employees who do not possess freedom of association or actual liberty of contract, and employers who are organized in the corporate and other forms of ownership association substantially burdens and affects the flow of commerce” (National Labor Relation Board, 1935, Section 1.[§151.]). This pattern of labor regulation is an assumption that employers set wages—which is key to the monopsony model. That regulation does not result in unemployment, which the knife-edge neoclassical model assumes is consistent with monopsony. Equal pay legislation substantially raises the pay of women and does not appear to harm their employment; this is readily explained by a monopsonistic perspective but a serious problem if one believes the labor market is perfectly competitive. And, similarly, finding that the minimum wage does not harm employment prospects in some situations is no particular mystery if one believes in monopsony.

With the rise of alternative work arrangements, the number of employees receiving subminimum wages has increased. Lawmakers need to reconsider regulations governing employment contracts, including non-compete clauses, notice periods, and zero-hours contracts where employers are not obligated to provide a minimum number of working hours to the employee. Lack of proper regulation means workers will enter employment contracts that are plagued with exploitation. In the context of gig work, lack of regulations curtailing employers’ outsized bargaining position will disproportionately impact nonwhite workers because workers of color are more likely to be in nontraditional arrangements that are lower paid (U.S. Bureau of Labor Statistics, 2017).

Well-meaning policies can also create monopsony exploitation. Take, for example, policies to reduce search friction for workers. As Stelzner and Bahn (2021, p. 23) argue:

At least a portion of the search frictions experienced by women are a result of increased caregiving responsibility within families and the gendering of paid caregiving work as feminine. A suite of family economic security policies, including universal access to paid family and medical leave, paid sick leave, affordable childcare, improving transportation, would all have a disproportionate effect on women workers who bear the brunt of caregiving responsibilities within families.

But these policies could inadvertently increase workers’ attachment to a firm, unless as the authors note, the policies are universal. Take, for instance, on-site, employer specific childcare benefits. On-site childcare creates a monopsony situation relevant to female parents because patriarchal gender division of labor means women are more responsible to arrange for care. Findings indicate childcare initiatives are associated with lower female collective turnover rates in following years (after establishing on-site childcare) and that higher female collective turnover rates are also associated with a higher likelihood of adopting a childcare initiative in subsequent years (Piszczek, 2020). These results provide support for the argument that childcare initiatives can increase mobility barriers for women causing their labor supply elasticities to be smaller.

That the earned income tax credit is applauded by employers and taken up primarily by women can also have the unintended consequence of lowering wages if a higher binding minimum wage does not exist. The economic incidence of the earned income tax credit is diluted because the influx of new workers floods the labor market, and wages fall as employers discover they can fill their jobs at lower cost than they paid previously (Rothstein & Zipperer, 2020). Moreover, this subsidization of low-paid work will incentivize employers to expand low-paying jobs.

The general omission of gender and race from the monopsony literature may be remedied as more researchers recognize that marginalized groups remain at the bottom of income distribution (Akee et al., 2019). The demise of unions and the erosion of worker rights have widened the Black-white wage gap (Rosenfeld & Kleykamp, 2012) that from a monopsonistic perspective, could be increasing the level of employers’ monopsony power and also their ability to exercise this exploitative capability.

### Conclusion: Employers Actively Nurture Monopsony Power

Monopsony power is created not given. A dynamic system of intergenerational privileges and resources, legal rights, and social norms make employers dominant when bargaining for wages, hours, security, and working conditions. Most monopsony models presume that monopsony conditions are given to and received by the employer in the form of worker immobility, worker lack of information, and worker idiosyncratic and pre-market tastes and decisions. In contrast, emerging new and rediscovered evidence finds support for the view that employers actively and persistently create monopsony power by fighting unions, resisting minimum wage hikes, imposing non-compete clauses, keeping salaries secret, promoting race and sex occupational segregation, colluding with other employers, and, perhaps by being an employer unfriendly to woman and nonwhites. Future work on gender and race in the framework of monopsonistic labor markets explored here should help economists better understand ways employers use personnel tactics and promote government policies to nurture monopsony power by utilizing and maintaining structural racism and sexism.