Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Ann Am Acad Pol Soc Sci. Author manuscript; available in PMC 2010 August 3.
Published in final edited form as:
Ann Am Acad Pol Soc Sci. 2009 January 1; 621(1): 70–93.
doi:  10.1177/0002716208324628
PMCID: PMC2914345

Bayesian Bigot? Statistical Discrimination, Stereotypes, and Employer Decision Making


Much of the debate over the underlying causes of discrimination centers on the rationality of employer decision making. Economic models of statistical discrimination emphasize the cognitive utility of group estimates as a means of dealing with the problems of uncertainty. Sociological and social-psychological models, by contrast, question the accuracy of group-level attributions. Although mean differences may exist between groups on productivity-related characteristics, these differences are often inflated in their application, leading to much larger differences in individual evaluations than would be warranted by actual group-level trait distributions. In this study, the authors examine the nature of employer attitudes about black and white workers and the extent to which these views are calibrated against their direct experiences with workers from each group. They use data from fifty-five in-depth interviews with hiring managers to explore employers’ group-level attributions and their direct observations to develop a model of attitude formation and employer learning.

Keywords: racial discrimination, employment, employer interviews, African Americans, stereotypes

The continuing significance of race in the minds of employers has been demonstrated in numerous contexts. Interviews with employers reveal the persistence of strong negative associations with minority workers, with particularly negative characteristics attributed to African American men (Kirschenman and Neckerman 1991; Moss and Tilly 2001; Wilson 1996). Studies of hiring behavior likewise suggest that employers strongly prefer white (and Latino) workers to otherwise similar African Americans (Pager 2003; Pager, Western, and Bonikowski 2007; Bertrand and Mullainathan 2004; Bendick, Brown, and Wall 1999; Fix and Struyk 1993). Where the continuing existence of discrimination is a matter of little controversy, however, the underlying causes remain widely contested.

Much of the debate over the causes of discrimination centers on the rationality of employer decision making. Economic models of statistical discrimination, for example, emphasize the cognitive utility of group estimates as a means of dealing with the problems of uncertainty (Phelps 1972; Arrow 1972). Group-level estimates of difficult-to-observe characteristics—such as productivity, reliability, or willingness to submit to authority—can provide useful information in the screening of individual applicants. If employers can accurately estimate differences in the skills or disposition of blacks and whites on average, this information can be helpful in guiding decisions about individual black and white candidates for whom these characteristics are more difficult to observe directly.

Sociological and social-psychological models, by contrast, question the degree to which group-level attributions reflect accurate assessments (Bielby and Baron 1986; Tomaskovic-Devey and Skaggs 1999). Although mean differences may exist between groups on some productivity-related characteristics, these differences may be inflated in their application, leading to much larger differences in individual evaluations than would be warranted by actual group-level trait distributions (Rothschild and Stiglitz 1982).1 Furthermore, estimates of group characteristics may reflect outdated associations, with factors such as occupational segregation, imperfect information flows, and negative feedback effects reducing awareness of changing distributions (Whatley and Wright 1994; Arrow 1998; Farmer and Terrell 1996). It thus remains unclear whether employers’ assessments of various racial groups represent accurate representations and to what extent these assessments are responsive to novel or competing sources of information.

In this study, we take one step toward investigating these questions by measuring the nature of employer attitudes about black and white workers and the extent to which these views are calibrated against their direct experiences with workers from each group. Using data from fifty-five in-depth interviews with hiring managers, we explore employers’ group-level attributions and their direct observations to develop a model of attitude formation and employer learning.

Prior Research on Discrimination

Direct observations of employer behavior suggest that race continues to shape employment opportunities in important ways. A recent field experiment of employment discrimination in New York City, for example, provided a rare glimpse into the pervasiveness of discrimination in low-wage labor markets (Pager, Western, and Bonikowski 2007). Across hundreds of applications for entry-level jobs, blacks were half as likely to receive a callback or job offer as equally qualified white applicants. Furthermore, blacks with clean records fared no better than a white man just released from prison. The results of this and earlier audit studies provide vivid illustration of the degree to which racial considerations continue to actively shape the employment opportunities available to young black men (Bendick, Brown, and Wall 1999; Fix and Struyk 1993; Bertrand and Mullainathan 2004; Pager 2003; see Pager [2007] for a review).

Although the foregoing studies provide an important measure of the demand side of the labor market, audit methods offer little insight into the motivations or attitudes that shape employer behavior. Does discrimination typically reflect racial animus? Have employers had negative experiences with African American employees in the past that have led them to shy away from hiring blacks? Or do other factors shape employer decision making?

Previous research leaves these questions only partially answered. The classic study by Kirschenman and Neckerman (1991) indicates that race is indeed salient to employers in their hiring decisions. Results from a large sample of in-depth interviews reveal employers’ blatant admission of their avoidance of young inner-city black men, attributing characteristics such as “lazy” and “unreliable” to this group (p. 213; see also Wilson 1996; Moss and Tilly 2001; Waldinger and Lichter 2003).2

At the same time, while this study has been widely cited as evidence of employers’ deep biases about African Americans, especially young black inner-city men, the findings themselves present a more complicated picture. While some employers spoke only in general terms about the assumed characteristics of black inner-city men, suggestive of the role of broad cultural stereotypes, others made specific reference to negative experiences with their own black employees, indicating that employer attitudes may be heavily shaped by direct observation of racial differences among their workers.

The underlying sources of employer attitudes thus remain somewhat unclear. Indeed, in framing their analysis with the concept of statistical discrimination, Kirschenman and Neckerman (1991) remain agnostic as to whether employers’ comments represent accurate depictions or exaggerated stereotypes. Employers are clearly using race as a proxy for employment-relevant characteristics, but the degree to which the use of this proxy is informed by empirical realities remains uncertain.

Moss and Tilly (2001) also find employers readily referring to negative characteristics among African American workers, with pervasive concerns about dependability, motivation, attitude, and skill. Many of these employers cite concrete experiences with their own black employees as the basis of their attitudes, though some also cite media representations and more general observations of African Americans as the source of their racial attributions (pp. 138-40). At the same time, Moss and Tilly note that a “silent majority” of employers claim not to notice racial differences among their employees. The authors speculate that these responses are due to some combination of social desirability bias, effective screening or training techniques (rendering a population-level racial skills gap irrelevant for the particular firm), or an honest experience of black and white workers as comparable. Among the largest group of employers, then, it is difficult to assess to what extent perceptions of racial differences are relevant for employers’ decision making.

While the present study can move no closer to assessing the accuracy (or honesty) of employers’ racial assessments, it does attempt to provide a more explicit understanding of the connection between employers’ direct experiences and their more general racial attitudes. Following in the footsteps of the important employer-interview studies from the 1990s, this project seeks to better understand the degree to which employers’ extensive experience with black and white workers generates assessments of racial differences that reflect their own empirical observations. In exploring the link between experience and beliefs, we hope to provide some insight into the dynamic process by which racial attitudes are constructed and reinforced.

A Rational Actor Model of Hiring

Economic models of statistical discrimination suggest that race offers a useful proxy for difficult-to-observe characteristics. Because productivity is difficult to observe directly, particularly for new hires, employers rely on indirect information inferred from group membership. This model can represent rational action on the part of employers, provided that the information they have about groups is accurate, and provided that there is a mechanism for updating estimates of group characteristics over time. Where discussions of statistical discrimination typically focus on accounting for single-point hiring decisions, the theoretical propositions can be readily extended to a dynamic process. Figure 1 represents such a model schematically, in its simplest form.

A Rational Actor Model of Hiring Decisions

The model starts with information about known group characteristics: for example, employers may be aware of racial differences in graduation rates, test scores, incarceration rates, and other relevant disparities. Employers may also have direct prior experiences to which they refer in shaping their own beliefs. These antecedent sources of information then shape the employers’ general beliefs about blacks—about their productivity, reliability, intelligence, and other relevant characteristics. Those beliefs will then guide individual hiring decisions. Once hiring decisions are made, employers are exposed to a range of black employees, and to the extent that these experiences with black employees do not fit the assumptions of the employer about members of that group, employers will revise their beliefs in a sort of Bayesian updating process to reflect a more accurate set of expectations.3

This equilibrium model of statistical discrimination describes the mechanisms by which employers can incorporate their direct observations into more general assessments of group characteristics. This feedback loop provides an efficient means of calibrating expectations, observations, and behavior. Consistent with such a model, some research does indeed point to evidence of employer updating. Altonji and Pierret (2001), for example, show that as firms acquire more information about a worker (through posthire observation), their evaluations (as reflected in wage offers) rely less on general (noisy) characteristics, such as educational attainment, and more on individual-specific characteristics, such as cognitive skill (see also Oettinger 1996; Farber and Gibbons 1996).4 These studies provide compelling evidence that employers weigh their direct observations more heavily than inferences based on group proxies and that employer learning can improve on initial estimates. Nonetheless, this line of research applies specifically to learning about individual workers, whereby group-based estimates are replaced with the observed characteristics of individuals. It remains unclear, however, whether an employer’s learning about an individual employee affects the employer’s expectations about the broader group to which that individual employee belongs.

Indeed, Farmer and Terrell (1996) provide an elegant theoretical analysis of employer learning and statistical discrimination in which initial employer beliefs are revised through an updating process similar to that described above. According to the authors, however, the updating process might apply only to the specific employees under observation, rather than to members of the larger group. “Higher than expected output of one worker provides much information about individual ability, but only a single data point to estimate the average ability of a population of millions. In addition to observation of workers, an employer receives an abundance of information on average group ability from other sources. Observations of average output, or perhaps occupations, of other members of the group influence the assessment of group ability” (p. 206). The process by which employers generate estimates of group characteristics and update those estimates over time thus remains unknown. In the following analysis, we seek to make headway in understanding this dynamic process.

Source of Data

This research is based on in-depth, in-person interviews with 55 New York City employers. The employers in this study represent a subsample of firms advertising for entry-level positions in 2004.5 In selecting respondents, we aimed to capture the full range of entry-level employers according to industry, occupation, and other types of characteristics thought to be associated with discrimination. In all, we made 243 in-person contacts with 152 firms to solicit participation in our study, 55 of whom agreed, rendering a response rate of 36 percent. The majority of respondents were male (70 percent); white (59 percent); located in Manhattan (82 percent); and managing a firm in the retail industry (46 percent), the restaurant industry (31 percent), or the service industry (11 percent).6 Furthermore, 47 percent of respondents represented independent firms, 40 percent national chains, and 13 percent local New York City chains.

Interviews ranged in length from thirty minutes to two hours, with the average interview lasting fifty minutes. Content of the interviews reflected a wide range of topics, including recruitment strategies; screening procedures; concerns about entry-level workers; and criteria for selection, placement, and promotion decisions. The questions we focus on here come from a module focusing on employers’ racial attitudes. This core segment of the interview probed three primary sets of issues related to (1) employers’ general attitudes about the employment problems of black men, (2) their specific experiences with black applicants and employees, and (3) the relationship between employers’ concrete experiences and their general attitudes. In the following discussion, we examine the pattern of responses that emerged from our conversations with employers, focusing specifically on the link between employers’ general attitudes about blacks and their specific experiences.

The use of qualitative data is well suited to investigating complex processes given its ability to capture the nuance and depth of personal attitudes (Orbuch 1997). But this approach also has its limitations. The relatively small sample prevents us from drawing strong conclusions about the attitudes of employers more generally. More important, the validity of findings from interview data depends on respondents’ willingness to provide truthful answers to questions. Given the sensitive nature of this investigation, concerns over social desirability bias are highly relevant. As we discuss below, the candid answers we received from employers about their negative racial attitudes offer some reassurance that employers were not entirely self-censoring. At the same time, we must remain aware of the possibility that some responses may be affected by these concerns.

Despite these limitations, in-depth interviews provide a rare window into employers’ thought processes and offer some leverage in understanding the complex process of attitude formation among this group. While we cannot conclusively adjudicate among theories based on the results from this study, we hope that our findings will be generative of hypotheses for further testing.

How Do Employers Think about Young Black Men?

In talking about race, and in particular the employment problems of black men, employers’ responses represented a range of views.7 Some employers emphasized the structural barriers facing African Americans, including poverty, a lack of education, disadvantaged neighborhood contexts, and prejudice and discrimination. For example, one employer from a courier company emphasized the problems of residential segregation, discrimination, and incarceration as key barriers to black men’s employment opportunities.

Well, there are of course the obvious problems of racial profiling where people don’t want to hire them. They won’t tell you that and you don’t really think it happens as much in society. People think all of that segregation has ended, but it really hasn’t…. Another problem, as a good example in terms of hiring people that have criminal backgrounds, a lot of people are not looking to take the chance, and unfortunately the number of African American men that have been incarcerated has just, in the past few years, been phenomenal. And so people don’t stop to think that, let’s say someone wasn’t really involved in something. Maybe they just got caught up, so to speak. Some people may just think that, well you have been in trouble and so I don’t want to hire you. So that is another aspect that keeps people from hiring black men.

Another employer from a retail chain pointed to a range of factors, from racism to a lack of education, that disadvantage black men in the labor market:

Racism is still a huge issue in America…. Especially like black men have been repressed for so many years, like due to every, like due to racial issues, a lot of them are very like limited in their, their work experience, or in their education levels, because the education system in New York City is terrible, like in public schools especially, like in the inner city, and you know, in the poorer neighborhoods, so … there … is a lack of education … and so many people can’t, you know, afford to go to college…. And therefore there’s so many jobs that so many people are trying to compete for …. You know I think that that’s what a lot of the problem is for, for, for especially black men in, in this city.

Overall, more than 40 percent of employers commented on structural issues, about 15 percent of whom emphasized these factors as their primary explanation for the employment problems of black men.

By far, the most common explanations for black men’s employment problems, however, focused on the individual shortcomings of black men themselves. More than three-fourths of employers mentioned individual explanations at some point in their discussion of black men’s employment problems, with well over half (60 percent) emphasizing individual factors as their primary explanation. A wide range of perceived shortcomings were identified as primary explanations for black men’s employment troubles, including concerns over work ethic, attire, and attitude, which we discuss in detail below.

Of course, we know from other research that Americans tend to give individualistic explanations for inequality in general, whether racial or otherwise (Schuman and Krysan 1999; Kluegel and Smith 1985; Jackman 1994; Bobo 2004). For example, in his analysis of General Social Survey data collected between 2000 and 2004, Hunt (2007) finds that at least half of Americans believed contemporary racial inequality to be caused by a “lack of motivation,” relative to less than a third who cited discrimination as an important problem.8 To some extent, then, the distribution of responses we observe reflects a more general reluctance to view inequality in structural terms.

At the same time, however, the content in these interviews goes beyond the standard narrative of “pull yourself up by your bootstraps.” Certainly we also heard comments of this sort. One employer from a local restaurant chain, for example, insisted, “If you are persistent, something will eventually pan out for you. I am sure of it. If you really want a job, you will eventually find one.” The comments we focus on in this analysis, by contrast, go well beyond these generic beliefs in individualism to reveal far more specific attributions about black men.

In asking employers to reflect on their experiences with workers from different racial groups, we are not simply asking them about their general beliefs about inequality; we are asking them to draw from their expertise as employers to help us better understand why the economic outcomes for some groups are systematically better/worse than others. Given their unique vantage point, we might expect employers to express attitudes about the characteristics of black and white workers that diverge from mainstream American racial attitudes. Because of highly segregated social networks, many white Americans’ exposure to African Americans is limited to casual observation, brief encounters, and media representations. Most employers, by contrast, have had extensive contact with black workers and have had the opportunity to observe these workers perform specific tasks and responsibilities.

Under these conditions, then, we might expect employers’ direct experiences to play a larger role in shaping their racial attitudes relative to other white Americans. Previous research provides only partial insight into this question. Bobo, Johnson, and Suh (2002), for example, find that employers’ racial stereotypes are indistinguishable from those of the general public, suggesting that workplace power or experience does little to shift generalized racial associations.9 At the same time, close-ended survey measures capture only one dimension of attitudes, potentially overlooking some of the complexities and contradictions embedded in employers’ ideas about the characteristics of racial groups. The present study uses in-depth interviews to investigate employers’ attitudes about and experiences with black men. We begin with an investigation of employers’ general characterizations of black men, and then explore the degree to which these characterizations are rooted in direct experiences and observations.

Lack of a work ethic

One of the most common themes we heard from employers centered on the perceived lack of a work ethic among black men (fully 55 percent mentioned this issue). Some of these employers referred to a general lack of motivation to work among African American men. Others described a desire among black men to take advantage of the system instead of working. One employer at a retail store said simply, “I will tell you the truth. African Americans don’t want to work.” A manager of a retail store said similarly, “They don’t want to work—you can tell by the attitude, clothing, the general body language.” The owner of a dry-cleaning store commented, “They just don’t have any drive. No get up and go attitude.” Likewise, a young male employer working in a national clothing chain stated, “I think for a lot of them they are too lazy to work so they are not going to work.”

A human resources manager at a national retail chain, when asked what she thinks explains the employment problems of black men, explained,

Employer (E): They are not as motivated as other races to, you know, to get out and do more and to improve or make themselves better. … [T]hey are not as motivated or determined to move up or to even do anything to improve themselves.

Interviewer (I): What do you think this lack of motivation stems from?

E: I don’t know how to say this but probably laziness. Just being lazy and not wanting to work.

When asked the same question, the manager of a local restaurant chain in New York City first argued that more black men are employed than the statistics indicate because many work off the books. However, the employer then refined his initial explanation, stating, “Listen, I also think there is a degree of being lazy.” As with many of the other employers interviewed, this manager seemed to believe that if black men were motivated to search for employment in a responsible manner, their employment problems would be resolved.

In several cases, employers attributed the lack of work ethic to a complex history of racism and paternalism. For example, the long-term manager of an industrial supply store in the Bronx noted, “In America blacks believe the golden opportunity is to be taken care of.” Another employer who earlier had asserted that “African Americans don’t want to work” elaborated on his initial comment:

Maybe they think that this country owes them so much. Because of slavery and all of that. They, they tend to forget that was a couple of years ago. Instead of catching up with the world, they still keep that anger…. So African American men feel like they deserve something. That is basically why they don’t want to work. They feel that if they can get things for free, why not?

The manager of a national restaurant chain expressed a similar view, differentiating between black men and black women. “I think the mentality is changing. I think [black men] are more accepting of letting the female work and they stay home.” He went on to explain,

I think, you know, a percentage of it, probably has to do with them figuring out how to take advantage of the system. You know, with welfare or something like that. I do think that a lot of them have the mentality of: I’m black and the government doesn’t help. So, let me do what I can to get from the government. Though Martin Luther King has changed the world, a lot of them have the mentality that it is still a white world.

A different employer emphasized that programs exist to help those in need of work but that black men elect to ignore the programs as they want “things handed to them.” The main supervisor of a national food chain explained,

The key is, I think there are great programs out there, but people don’t take advantage of them. It is probably laziness. You know, there are programs. But people are lazy. A lot of people want things handed to them.

Another employer focused more specifically on negative consequences of welfare dependence:

Unfortunately we’ve bred generations of welfare—of people whose … careers are … welfare and social agencies. And, uh, and it’s unfortunate. A lot of these people just don’t have any work values.

Applying a similar line of reasoning, the young manager of a local telemarketing firm expressed frustration with black men playing the “race card” as a way of escaping responsibility:

I mean, I do understand that sometimes the black man is racially profiled. So it is something they are always going to deal with. This is unfortunate for them, but I think they shouldn’t even go there with playing the race card. I mean, a lot of them are like, “You treat me this way because I am black,” and “I can’t do this and I can’t do that.” I just think for a lot of them, it is that they just don’t want to do things. They want to smoke weed and be a rapper.

Although this employer does see racial discrimination as a reality, he views it as more often used as a convenient excuse for failure than a direct cause. Underlying the employment problems of black men, according to this employer, is more often a simple lack of motivation and effort.

According to each of these employers, a lack of work ethic, motivation, and personal responsibility is pervasive among African American men, and together, these factors represent the primary causes of this group’s employment problems. Historical social policies (e.g., slavery and welfare) are seen less as explanations for these patterns and more as convenient excuses available for African Americans in the abdication of responsibility.


A second persistent theme in the interviews was the issue of black men’s self-presentation (45 percent of employers mentioned this issue). In particular, employers highlighted problems of unsuitable appearance, negative attitude, and inappropriate conduct among black men. Particularly for positions involving customer service, employers expressed concern about the image projected by black men in their attire and attitude. An employer representing a local New York City clothing chain remarked,

Sometimes these people looking for a job, why would they have a do-rag on and jeans? Why would they? You know, just the way you present yourself, it’s how you are…. The black male, yes. It’s just the way they carry themselves.

Another employer, the manager of a children’s clothing store, when explaining why black men have difficulty securing employment, stated,

The way they present themselves in the store. If they come in, and excuse the word, but they are all thugged out, it is not somebody I want on my sales floor.

Also placing significant weight on the role of appearance in securing employment, an employer representing a popular local New York City retail chain commented,

If you go out looking for a job with caps, baggy pants and triple x t-shirts or whatever, you can’t expect someone to hire you like that. Why don’t you put on some decent clothes and go look for a job? That is the mistake minorities, black and Hispanics tend to do. They look for a job, and when they don’t get hired, they automatically say it is discrimination.

The manager of a retail clothing chain complained, “[Black men are] usually, like very urban looking, baggy pants, you know, just like baggy clothes, hat.” Likewise, an employer for a moving company said, “I think people who come in wearing baggy pants or something like that just make a bad impression. You can be green, orange, purple, or whatever. It doesn’t matter, it isn’t good.”10 For these employers, the problems of attire—and specifically the issue of baggy pants—signaled a lack of professionalism or an ignorance about appropriate work attire, something they often associated with black men.

In addition to matters of appearance, employers’ comments about the self-presentation of black men also emphasized issues of attitude and conduct. Black men were often perceived as having a “bad attitude” about work or relations of authority, or at times behaved in ways considered inappropriate for mainstream work environments. When considering the employment problems of black men, an employer from a national clothing chain explained, “It’s just the way they carry themselves.” Another described black men’s tendency to present “language and the attitude, like gangsta or street.”

The manager of a large national retail chain noted that differences in the self-presentation and conduct of black and white men affected the way they are perceived by employers and customers:

I have heard people say that it is easier to deal with a white person more than a black person. I guess because they feel black people are always loud and hyper. Which is true. White people may know how to carry themselves a little better than black people. Black people always want to make a scene and always want to be heard.

A manager of a national retail store emphasized the distinction between qualifications and self-presentation, with the latter undermining the former in the case of black men:

You know, I know a lot of black males and I know how they react to things and I know why they don’t get jobs. I kind of know why. Some of them are actually qualified for jobs but they go in with that attitude. It is all in how they present themselves.

Several employers commented on the attitude problems they perceived among young black men as reflecting a sense of hostility or resistance to authority. An employer from a small, independent retail store stated, “[Black men have] this kind of attitude that is, is umm, resentful. It is hard. It is not an attitude that is favorable for business.” An employer representing a large national clothing chain emphasized that black men don’t present themselves to employers appropriately because they have a chip on their shoulder:

I: Why do you think they are not presenting themselves in what you think is the best way? Where does this stem from?

E: Um, ego and insecurity. The insecurity part is that they don’t feel that they are right for the job. They lack that confidence but yet their ego won’t let them admit it so they have a chip on their shoulder.

The owner of a local garment factory echoed some of the same concerns:

[Black men] act a little more belligerent than others. There is the attitude and pattern of animosity with this group. They have a chip on their shoulders. They think, man, you are white and that is why you don’t give me a job. That kind of thing…. The black male will come and say I am better than this and better than that and so on. So there is more of a macho type of attitude with the black male. When you go for a job you have to be, besides appearing decent and trustworthy, you can’t look macho or act like you are better than other people.

Another employer at a retail store found it difficult to advise young black men about appropriate attire because of their resistance to authority.

I think that it is hard for these men because they are too proud. Especially the younger guys, the eighteen- and nineteen-year-old black guys that come in here, are like, “Who are you to tell me no?” They have a real problem with authority.

These employers perceive black men to exhibit styles of dress or demeanor that present barriers to their mainstream involvement in the labor market. Concerns about the self-presentation of black men, ranging from baggy pants to bad attitudes, were viewed as a key obstacle to employment for this group.

A threatening or criminal demeanor

A final major theme in employers’ general comments about black men focused on a perceived threatening or criminal demeanor. Extending comments about the “hardness” or “animosity” of black men discussed above, roughly a third of employers mentioned concerns of this sort. According to an employer from a local garment factory,

I find that the great majority of this minority group that you are talking about either doesn’t qualify for certain jobs because they look a little bit more, they come on as if, well, they are threatening.

An employer from a popular national clothing chain dismisses the racial frame initially but ultimately reached a similar conclusion:

I don’t know if I consider it on a race level. I just consider it more on approachability. And if someone seems intimidating, you know, and which, stereotypically might be, you might consider like a Hispanic person or a black person more intimidating than like a white person.11

Other employers focused more specifically on concerns about criminality. According to one employer from an ice cream chain, for example,

I notice here working in the store, sometimes, a group of young black males will come in. And sort of, a red flag goes up. Everything stops, and you wonder, what are they going to do? Are they here to buy something? There is a general belief that, because of the way they dress and how they carry themselves, that they are trouble.

An employer from a retail store explained, “I mean, black males are not expected to go out and work, because they are doing other illegal things in the neighborhood.” Likewise, an employer for a moving company pointed to the lure of criminal activity among African Americans:

They see the drug dealer who is driving around in a Lexus, and then they see me in a Chevy, and they say I don’t need to be a mover. I don’t need to work sixty hours a week. I can go hustle stuff on the corner and drive a Lexus.

Another explained simply, “Half of them are in jail.” These employers associate black men with danger, crime, and the criminal justice system, factors that appear incompatible with legitimate work.

Overall, then, we find fairly pervasive negative attitudes about black men as a group. While some employers did place more emphasis on lack of opportunity, prejudice, or other more structural factors, the majority of employers pointed to specific deficiencies among black men that led them to have difficulty in the job market. Our next set of questions seeks to investigate to what extent these negative attitudes are a function of employers’ direct experiences.

Are Employers’ Perceptions Informed by Experience?

In discussing their general attitudes about black men, many employers inevitably referred to specific observations or experiences in their own workplace. These anecdotes point to one potential source of employers’ general attitudes, though they tell us little about the overall distribution of experiences. Our next set of interview questions, then, attempted to elicit more systematic information about employers’ experiences with applicants and employees from various racial groups.12 In particular, we asked employers to consider the extent to which the characteristics they had described to us were reflected in the black applicants or employees they had encountered over the past year.

In some cases, there was clearly a link between employers’ general attitudes and their direct experiences. For example, one employer who had expressed negative views of blacks generally (“basically these people are lazy”) went on to describe the differences he has observed between black and white applicants.

Black people, mostly, yeah I can say that they come in, and “Are you hiring?” That’s their question, and then that’s it. They don’t have, they are not prepared at all. I give them an application and they don’t show up back with it. It’s like, “I’m taking it, but let’s see what happens….” White people … they mostly come with resume already done, the paper with them.

Another employer who had commented earlier on dysfunctional culture affecting “people that come from two or three generations that are on welfare” described the differences he observed:

Well for my business I am looking for somebody that comes dressed for the interview. If you walk in with baggy jeans it is not even worth talking to you. This is something I have observed with black men. I mean, yeah, I definitely have noticed that a lot of black male applicants typically don’t know how to properly talk to me, or they leave messages on the phone that aren’t really what I am looking for. I mean, they just don’t seem energetic or like they really want this job.

An employer from a national service organization commented on her experiences with black applicants who seemed to take job opportunities for granted, not demonstrating suitable interest or seriousness of purpose.

People will call here and you will try to schedule them for an interview. So you will offer them an interview, and they call up asking for one, and they say things like, “Yeah, I don’t know if that time works for me.” You know what I mean? It is just not proper etiquette.

While we were conducting an interview with the manager of a local telemarketing firm, a black male job applicant entered the room, dressed in baggy clothing, and asked to schedule an interview. The manager replied that the individual would have to come back later in the week, to which the applicant stated, “Gotcha. See you then.” After the applicant left, the manager stated, “Not gonna hire him. See what I mean? Where did he learn that ‘gotcha’ was the right thing to say to a potential boss? And see what I mean about attire?”

A number of the employers we spoke to reported at least some negative experiences with black applicants and/or employees, ranging from comments about individual workers to descriptions of more general patterns. Overall, among our sample of white respondents, employers were more likely to note racial difference among their applicant pool (46 percent) than among their employees (34 percent).13 This suggests that employers are doing an effective job at weeding out good employees from a more heterogeneous applicant pool. At the same time, note that more than half of employers claim not to notice racial differences in the quality of their applicants, and fully two-thirds of employers notice no racial differences among their own employees. In these cases, there is often little relationship between the impressions they report about African Americans generally and their own direct observations of black employees.

For example, one employer earlier emphasized the lack of work ethic among blacks (“just being lazy and not wanting to work”) as the major factor for their employment problems; but when asked if she had observed these problems in her workplace, she replied, “Well no…. Of course once in a while they goof off, but that is across the board. I don’t see any differences between groups.” A video store employer acknowledged that stereotypes were often quite different from reality: “There are people that say there are differences in work ethics of black men, but I have no specific thing like that that I have noticed. Absolutely not. The worst employees have been the non people of color. They have been the worst.”

An employer who earlier had alluded to the problems black men have with “presentation” and “attitude,” went on to describe his employees with the following comments:

I: Among your employees, have you noticed differences in the work performance of blacks and whites?

E: Um, do you mean in terms of work ethic?

I: Sure.

E: They really have a nice work ethic.

I: What about comparing whites to blacks?

E: In my experience blacks will outdo them.

I: Blacks will outdo whites?

E: Yeah, once you get the right person. Sometimes with an entry-level, they don’t seem as committed.

I: You mean your white employees?

E: Yeah. It is like they think they are above this. I don’t find this with the black employees.

An employer for a retail clothing store (who had referred to blacks as being more “intimidating” than whites) came to a similar assessment about her own employees: “It’s hard to compare because it’s different types of work that they’re doing, but I would say that the people who are not of Caucasian descent work a little harder than the white kids.”

In several cases, employers did acknowledge differences among workers but accounted for them along nonracial lines. We asked an employer who had earlier emphasized the importance of attire for her workers:

I: Do you observe that black male applicants are more likely than white male applicants to present themselves to you inappropriately?

E: No not really. I think it is a problem for all of them. I think just men in general.

In another case, the employer started out describing the problematic applicants she deals with. When we pressed her to consider whether these characteristics were more prevalent among black men, she thought about the question and then reframed her comments in terms of age.

E: They don’t come in dressed, they don’t come in and speak to you in a correct way. They speak like they are hanging out and not looking for a job.

I: Okay, so do you notice that black males come in not prepared or dressed appropriately more often than white males that are looking for jobs?

E: I think it is about the same honestly. It is more about the age. The high school kids are the ones that don’t come in dressed like they are looking for a job. As they get older you can see the difference because they are coming in in slacks and they are coming in in a shirt and they are speaking to you. They aren’t just like, “Here is an application,” and then they walk out the door.

These employers appear to have identified a cluster of behaviors or styles of presentation that signal poor performance, some of which are often associated with African American men, but when pressed to make sense of their observations, they focus on a different set of categories. Of course, it may be the case that employers feel uncomfortable making racial attributions and thus retreat to a language of gender or age out of social desirability concerns. We have no way of conclusively ruling out this possibility. At the same time, given the extremely candid racial remarks made by these employers just moments before, we feel some reassurance in taking these responses at face value. These comments lead us to believe that employers felt sufficiently at ease to speak in racial terms, and thus, we see little reason that they would suddenly retreat into a more politically correct style of discourse. In fact, we were concerned with the opposite effect, that employers would feel some pressure to come up with specific examples of poor performance among their black employees, if only to maintain consistency with their earlier comments. Quite the contrary, we found employers repeatedly emphasizing the lack of clear distinctions between black and white applicants and employees, even as these comments appear to contradict earlier statements about the generalized characteristics of black men.

As a final example of the disconnect that often appeared between employers’ expectations and outcomes, one employer from a small retail store describes her recent experience with a black man she had hired for a stockperson position in her retail store:

Like, last year. I had this guy come in, with a big hat and a big jacket. I don’t know if he had a criminal record, but he looked like it. But I was so desperate that I hired him. [Laughs] He was OK. But the way he dressed. Sometimes the way they dress. It is bad. I mean, a big hat. You try to be nice, but at the same time, I don’t need a guy with a big hat. [Laughs] I was surprised that he worked out well. He finished the season very nicely.”

In this case, despite the fact that the employer viewed this man as poorly dressed and potentially criminal, he ultimately wound up being a reliable employee. Indeed, employers frequently acknowledged that their first impressions of applicants were often completely off base. Here, the employer recognized her misjudgment, and yet, there is little sign that this experience caused her to rethink her more general, negative impressions of black men. Earlier in the interview, this employer had characterized black men as having attitude problems. “Socially there is a difference. In the neighborhood [black men] have a kind of attitude, that is, is um, resentful. It is hard. It is not an attitude that is favorable for business.” Although here (and in other comments), she insists that her own experiences with black employees have been overwhelmingly positive, she retains strong negative impressions about black men “in the neighborhood,” the source of which remains unknown.

How to Explain the Gap between Employers’ Perceptions and Experience?

Whereas most of the employers in our sample expressed consistent negative attitudes about black men, far fewer could identify specific patterns or experiences among their own applicants or employees to fit these characterizations. How can we account for this surprising disconnect? While the results of this study cannot provide a definitive answer, we consider several plausible explanations.

The salience of negative events

First, it is possible that while employers’ experiences with black applicants or employees on average may be similar to those with whites, a small number of negative experiences may hold especially strong weight in shaping attitudes (Fiske 1998). Indeed, several of the employers in our sample referred to singular experiences that, while not representative of their experiences more generally, seemed to shape their associations of specific groups. According to one employer from a national retail chain,

You know, everyone has a couple of bad hires. And you remember those very vividly. And who that person is can really impact. [He describes a negative experience with an African American female employee.] That person just stuck in my head. That was the first time I had done hiring during a holiday season, which is pretty stressful. And that person just stuck in my head. And I could see her. It was hard to not see her in other people that you meet.

According to this employer’s account, one particularly negative experience with a black employee colored the employers’ expectations of blacks in the future. Where this employer also noted a number of positive experiences he had had with African American employees, this negative experience appeared to carry especially strong weight.

In another case, an employer for a retail clothing store spoke about his negative experiences with a few black employees at a previous company.

One of them in particular was threatening me and telling me after I fired him that he was going to wait for me outside and that he was going to get me. So that kind of thing sometimes gives you a vengeance. So, you know, you become biased a little bit.

Again, this employer acknowledges the ways in which a particularly unpleasant encounter contributed to his biases against black men more generally. It may be the case, then, that where employers may have had only a small number of unusually negative experiences with African Americans, these incidents may be the driving force behind generalized negative attributions. Benign or positive experiences create less salient memories, even if more prevalent overall.

At the same time, only a few of the employers in the sample reported extreme negative experiences, and more than half reported that their experiences with black and white applicants and employees were roughly comparable.14 Without salient negative experiences coloring attitudes, what then might explain employers’ persistent negative racial attitudes?

Selection and screening

A second possible explanation for the disconnect between employers’ characterizations of black men in general and their direct experiences has to do with the various selection and screening procedures that weed out the most undesirable members of the group. Indeed, recall that employers were significantly more likely to report noticing racial differences in the characteristics of their job applicants than among their employees (46 versus 34 percent), suggesting that the hiring process leads to a more even distribution of productivity characteristics among black and white employees than exist in the general population. As one employer mentioned, “Yeah, once you get the right person …,” indicating that the selection process can yield high-quality black employees, even if there is a great deal of variation within the general population. Recognizing that employees are not picked at random, we asked employers to comment separately on their perceptions of the applicant pool, expecting that racial differences may be more pronounced before the employers’ active screening. The fact that more than half of employers claim not to notice racial differences even at this stage suggests that more remains to be explained than employer sorting.

Of course, the hiring process is not the only selection process at work. Showing up to apply for a job itself involves a process of selection, with the search process requiring some degree of organization, motivation, and a commitment to formal work. Particularly in recent years, as we have seen increasing numbers of young black men exit the labor force altogether (Holzer, Offner, and Sorensen 2005), selection into labor force participation may differentiate the types of black men employers encounter among their applicant pool from those in the general population.

This research does not make claims about employers’ accuracy in detecting population-level characteristics relative to those observed in the workplace. Rather, we simply note that for many employers, whatever sources of information they use to infer general characteristics of black men, direct experiences with black applicants and employees do not appear dominant.

Updating versus subtyping

A third potential explanation for the disconnect between employers’ general attitudes and their specific experiences draws upon the social-psychological concept of subtyping. Where Bayesian models of updating assume that individuals incorporate new information by refining their expectations in ways consistent with their experiences, social psychological models emphasize the cognitive resistance to information that is disconfirming of expectations (Fiske 1998). A wealth of experimental evidence illustrates ways in which individuals are prone to view those who do not conform to stereotypes as exceptions, unrepresentative of the group as a whole, rather than as impetus for stereotype refinement (Taylor 1981; Weber and Crocker 1983). The creation of a subtype allows group stereotypes to remain intact in the face of disconfirming information.

Figure 2 applies the theory of subtyping to the schematic model proposed earlier to reveal a potentially different set of processes shaping racial attitudes. Here, instead of (or in addition to) known group characteristics and prior experiences being the predominant sources of racial attitudes, cultural stereotypes and media imagery play a strong role in shaping group expectations.15 Those beliefs then translate into hiring decisions, which in turn provide a range of direct experiences for employers with black workers. But instead of positive experiences with black employees—as many of these employers report—leading to an updating of beliefs about blacks in general, we see a process of subtyping. As the saying might go, “My blacks aren’t like blacks in general.” Employers view their own black workers as a special subtype whose characteristics have little bearing on their evaluations of the larger group. Correspondingly, we see no pathway linking direct experiences with general beliefs.

Resilience of Stereotypes

Of course, this simple model does not capture the many nuances of attitude formation and change, which certainly contain a more diverse set of influences and pathways than are represented here. Rather, this basic schematic serves to represent one important alternative to the model proposed earlier, in which information flowed from the general to the specific and back again in ways that enhanced accuracy over time. Here, by contrast, an updating of expectations may occur with respect to specific members of the group (the subtype), but little revision of employers’ more general attitudes is expected.

The analysis likewise does not imply that employers are necessarily wrong in their assessments of various group characteristics. It may be the case that employers’ information about African Americans as a group is accurate and that the various selection processes at work (with individuals selecting into labor market participation and employers selecting workers) simply yield a more advantaged subgroup to which employers are exposed. This research makes no claims about the relationship between employers’ attitudes and the “true” characteristics of African Americans. Rather, the findings suggest that, whatever sources of information they may be drawing from in forming racial attitudes, employers do not seem to draw heavily from their own personal experiences.

Conclusion and Discussion

The findings of this research suggest that, while most employers expressed strong negative views about the characteristics of African American men, fewer than half of these employers reported observations of their own applicants or employees consistent with these general perceptions. Where employers may update their expectations regarding individual workers, these experiences do not seem to have noticeable effects on their attitudes about the group as a whole. Rather, employer attitudes appear more consistent with a model of subtyping, in which individuals who do not conform to a stereotype are viewed as exceptions whose characteristics have little bearing on the larger group.

Employers surely receive relevant information about various groups from sources other than direct workplace experience. They observe and interact with people in public spaces, they read newspaper coverage and watch TV news, and they are aware of racial inequality and have their own ideas about how and why this inequality is generated and maintained. These sources of information surely provide valuable complements to direct workplace experience. At the same time, it is surprising that the experiences employers report from their own direct observations do not carry greater weight in their general attitude formation. These results suggest that simple contact and exposure are themselves insufficient to revise deeply embedded racial attributions.

This analysis holds potentially troubling implications for hiring behavior. We know from the results of field experiments that employers consistently avoid black workers, hiring them at roughly half the rate of equally qualified whites. Where models of statistical discrimination might interpret this behavior as the rational response to observed differences in the productivity of black and white workers, the present research questions this conclusion. The majority of employers who report positive experiences with black workers (or no differences between black and white workers) nevertheless maintain strong negative attitudes about black men generally. To the extent that these attitudes shape hiring decisions, even in the scenario of equal productivity among black and white workers, we would expect the problems of hiring discrimination to persist well into the future.


We thank Bruce Western for his contributions to the original interview schedule and for valuable comments on the article. Thanks also to Susan Fiske and Jeremy Freese for helpful comments and suggestions. Support for this project came from grants from NSF (SES-0547810), NIH (K01-HD053694), and a William T. Grant Foundation Scholar’s Award.



Devah Pager is an associate professor of sociology and a faculty associate of the Office of Population Research at Princeton University. Her research focuses on institutions affecting racial stratification, including education, labor markets, and the criminal justice system. Her current research has involved a series of field experiments studying discrimination against minorities and ex-offenders in the low-wage labor market. Recent publications include Marked: Race, Crime, and Finding Work in an Era of Mass Incarceration (University of Chicago Press 2007) and “Walking the Talk: What Employers Say versus What They Do” (with Lincoln Quillian), in American Sociological Review 70, no. 3 (2005): 355-80.


Diana Karafin is a PhD candidate at Ohio State University whose research focuses on consequences of neighborhood integration, racial democracy and crime, and discrimination in the housing and labor markets.



1See Armour (1997) for an extensive discussion of the logical, legal, and moral dilemmas of “reasonable racism.”

2Although the results of this study are indeed striking, it is important to keep in mind that more than 50 percent of Kirshenman and Neckerman’s (1991) sample “either saw no difference [in the work ethic of whites, blacks, and Hispanics] or refused to categorize in a straightforward way” (p. 210).

3A similar process of updating is described in the social psychological literature on stereotype change, referred to as a “book keeping model,” according to which new information is incrementally incorporated into existing beliefs or attitudes about a group (Weber and Crocker 1983).

4In this analysis, cognitive ability is observed to the researcher (by the respondent’s score on an Armed Forces Qualification Test [AFQT] test measured earlier) but not by the employer at the point of hire. The assumption is that the worker’s cognitive ability becomes observable to the employer with time on the job.

5Roughly 80 percent of the employers in this sample were drawn from a random sample of employers advertising for an entry-level position in 2004 (see Pager, Western, and Bonikowski 2007). The remaining 20 percent were drawn through a purposive sampling technique enabling us to better represent large employers and industries underrepresented by the primary sample.

6Note that within these industries are represented a wide range of job titles, including stockers, sales assistants, busboys, kitchen staff, waiters, couriers, and customer service positions.

7The vast majority of employers offered multiple explanations within a single response series. In categorizing employers’ sentiments, we distinguish between individual and structural explanations using several coding schemes. The first coding scheme takes into account the “first-mention,” or whatever explanation was first proposed by the employer; the second takes into account any factor mentioned by an employer, with most employers being coded into several categories; the third uses a “holistic” approach in which we coded the comments according to what appeared to be the employers’ main point. In many interviews, we specifically probed employers who reported multiple explanations with the question, “What do you think is the most important factor?” In other cases, this coding is based on our interpretation of the transcript. The main substantive conclusions are consistent across coding schemes, and where relevant, multiple measures are reported here.

8An additional 10 percent cited “less inborn ability” as a major explanation, and 43 percent cited a “lack of chance for education.” Respondents were allowed to choose more than one explanation, and thus, the categories do not sum to 100 percent (Hunt 2007, 400, Table 2).

9Bobo, Johnson, and Suh’s (2002) study draws data from the Los Angeles Study of Urban Inequality in which respondents were asked to rate members of a series of racial groups according to a range of characteristics (unintelligent, prefer welfare, hard to get along with, poor English).

10Note that here (and in several other interviews), the job in question requires primarily manual work and few customer interactions. Jobs at a cleaning company or a moving company do not typically require professional dress, and yet, for these employers, a worker’s attire sends an important signal about his reliability and commitment to work.

11A number of other employers emphasized the perceptions of others about blacks as threatening or intimidating. For example, one employer said, “I think a lot of white people are scared of black people for some reason. I think they are scared of them, intimidated by them, they don’t feel comfortable around them.” Similarly, the manager of a small restaurant expressed concern over the aggressive demeanor of black men, stating that employers may be hesitant to call black applicants back because “immediately a black male is perceived as being aggressive.” Although these perceptions may also be highly relevant for hiring decisions, we do not include these responses here as this analysis focuses on employers’ own views about African Americans.

12We asked separately about applicants and employees, given that effective employer screening would lead to a different distribution of characteristics among those seeking employment relative to those who become employed.

13African American employers appeared slightly more inclined to notice racial differences among applicants but substantially less likely to report racial differences among employees. Immigrant employers noted more racial differences at both stages. Note, however, that our sample of African American and immigrant employers is small and that these indications must be taken as tentative.

14Even respondents who described just one negative event were coded among those who had observed racial differences.

15While we cannot measure the influence of cultural stereotypes directly, several employers did explicitly comment on the ways in which cultural representations of black men in the media affected their perceptions of black men. According to one employer, “I’d probably say 90 percent of the crimes you see on TV are African Americans, female or male, and that’s something that’s in the back of your head, you know.” Similarly, another employer commented, “We have the media sending all this negative information about the young black male. All this negative information constantly. … We are getting the wrong image of what they look like.”


  • Altonji Joseph G, Pierret Charles R. Employer learning and statistical discrimination. Quarterly Journal of Economics. 2001;116:313–350.
  • Armour Jody David. Negrophobia and reasonable racism: The hidden costs of being black in America. New York: New York University Press; 1997.
  • Arrow Kenneth J. Models of job discrimination. In: Pascal Anthony H., editor. Racial discrimination in economic life. Lexington, MA: D.C. Heath; 1972. pp. 83–102.
  • Arrow Kenneth J. What has economics to say about racial discrimination? Journal of Economic Perspectives. 1998;12:91–100.
  • Bendick Marc, Jr, Brown Lauren, Wall Kennington. No foot in the door: An experimental study of employment discrimination. Journal of Aging and Social Policy. 1999;10:5–23. [PubMed]
  • Bertrand Marianne, Mullainathan Sendhil. Are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination. American Economic Review. 2004;94:991–1013.
  • Bielby William, Baron James N. Men and women at work: Sex segregation, and statistical discrimination. American Journal of Sociology. 1986;91:759–799.
  • Bobo Lawrence D. Inequalities that endure? Racial ideology, American politics, and the peculiar role of the social sciences. In: Krysan Maria, Lewis Amanda E., editors. The changing terrain of race and ethnicity. New York: Russell Sage Foundation; 2004. pp. 13–42.
  • Bobo Lawrence D, Johnson Devon, Suh Susan A. Racial attitudes and power in the workplace: Do the haves differ from the have-nots? In: Bobo Lawrence D, Oliver Melvin, Johnson James H, Jr, Valenzuela Abel., Jr, editors. Prismatic metropolis: Inequality in Los Angeles. New York: Russell Sage Foundation; 2002. pp. 491–523.
  • Farber Henry S, Gibbons Robert. Learning and wage dynamics. Quarterly Journal of Economics. 1996;111:1007–1047.
  • Farmer Amy, Terrell Dek. Discrimination, Bayesian updating of employer beliefs and human capital accumulation. Economic Inquiry. 1996;34:204–219.
  • Fiske Susan. Stereotyping, prejudice, and discrimination. In: Gilbert Daniel, Fiske Susan, Lindzey Gardner., editors. The handbook of social psychology. New York: McGraw Hill; 1998.
  • Fix Michael, Struyk Raymond J., editors. Clear and convincing evidence: Measurement of discrimination in America. Washington, DC: Urban Institute Press; 1993.
  • Holzer Harry J, Offner Paul, Sorensen Elaine. Declining employment among young black less-educated men: The role of incarceration and child support. Journal of Policy Analysis and Management. 2005;24:330–333.
  • Hunt Matthew O. African American, Hispanic, and white beliefs about black/white inequality, 1977–2004. American Sociological Review. 2007;72:390–415.
  • Jackman Man R. The velvet glove: Paternalism and conflict in gender, class, and race relations. Berkeley: University of California Press; 1994.
  • Joleen Kirschenman, Katherine Neckerman. We’d love to hire them, but …: The meaning of race for employers. In: Jencks Christopher, Peterson Paul E., editors. The urban underclass. Washington, DC: Brookings Institution; 1991. pp. 203–234.
  • Kluegel James R, Smith Eliot R. Beliefs about inequality: Americans’ views of what is and what ought to be. New York: Aldine de Gruyter; 1985.
  • Moss Philip I, Tilly Christopher. Stories employers tell: Race, skill, and hiring in America. New York: Russell Sage Foundation; 2001.
  • Oettinger Gerald S. Statistical discrimination and the early career evolution of the black-white wage gap. Journal of Labor Economics. 1996;14:52–78.
  • Orbuch Terri L. People’s accounts count: The sociology of accounts. Annual Review of Sociology. 1997;23:455–478.
  • Pager Devah. The mark of a criminal record. American Journal of Sociology. 2003;108:937–975.
  • Pager Devah. Marked: Race, crime, and finding work in an era of mass incarceration. Chicago: University of Chicago Press; 2007.
  • Pager Devah, Western Bruce, Bonikowski Bart. Working Paper, Office of Population Research. Princeton, NJ: Princeton University; 2007. Discrimination in low wage labor markets.
  • Phelps Edmund. The statistical theory of racism and sexism. American Economic Review. 1972;62:659–661.
  • Micahel Rothschild, Stiglitz Joseph E. A model of employment outcomes illustrating the effect of the structure of information on the level and distribution of income. Economic Letters. 1982;10:231–236.
  • Howard Schuman, Krysan Maria. A historical note on whites’ beliefs about racial inequality. American Sociological Review. 1999;64:847–855.
  • Taylor Shelley E. A categorization approach to stereotyping. In: Hamilton DL, editor. Cognitive processes in stereotyping and intergroup behavior. Hillsdale, NJ: Lawrence Erlbaum; 1981. pp. 83–114.
  • Tomaskovic-Devey Donald, Skaggs Sheryl. An establishment-level test of the statistical discrimination hypothesis. Work and Occupations. 1999;26:420–443.
  • Waldinger Roger, Lichter Michael. How the other half works: Immigration and the social organization of labor. Berkeley: University of California Press; 2003.
  • Weber Renee, Crocker Jennifer. Cognitive processes in the revision of stereotypic beliefs. Journal of Personality and Social Psychology. 1983;45(5):961–977.
  • Whatley Warren, Wright Gavin. Race, human capital, and labour markets in American history. In: Grantham George, MacKinnon Mary., editors. Labour market evolution: The economic history of market integration, wage flexibility, and the employment relation. London: Routledge; 1994. pp. 270–291.
  • Wilson William Julius. When work disappears: The world of the new urban poor. New York: Vintage Books; 1996.