Statement by Harry J. Holzer

The U.S. Equal Employment Opportunity Commission

Commission Meeting on Race and Color Discrimination of April 19, 2006, Washington D.C.

I would like to make some general comments this morning about racial discrimination, and its continuing presence in the U.S. labor market.

  1. Racial discrimination clearly persists in the labor market, though it is more powerful against some groups of minorities than others.

The evidence that discrimination in hiring persists is clearest in the many "audit" or "tester" studies of the past 10-15 years, in which matched pairs of white and minority job applicants with identical credentials on paper are sent to apply for jobs. Virtually all of these studies show statistically significant differences in the rates at which white and minority candidates receive "call backs" or are offered jobs (Fix and Struyk, 1994).1

Studies of self-reported employer data on job applicants and hiring show differences as well, with black applicants gaining fewer jobs than whites or Hispanics, and black men gaining fewer than black women (Holzer, 1996). The data show that smaller establishments, those with white customers, and those with white managers/owners make fewer offers to black applicants, all else equal (Holzer, 1998; Holzer and Ihlanfeldt, 1998; Holzer et al., 2000). Ethnographic evidence confirms that employers have negative stereotypes about blacks relative to other employees, and are more fearful of black men than women (Kirschenman and Neckerman, 1991; Moss and Tilly, 2001).

Employer discrimination also seems to occur more frequently at the hiring stage than at the job assignment or promotion stages, perhaps partly because employment discrimination lawsuits are more likely to focus on allegations of wrongful promotion or discharges rather than hiring (Donohue and Siegelman, 1991). Also, employers often prefer immigrants to native-born (minority) workers for low-wage, low-skill jobs; this reflect their perceptions of stronger work ethic and greater appreciation for the work among the former (Kirschenman and Neckerman, op. cit.).2

  1. While employer discrimination occurs less frequently than in earlier generations, and no doubt contributes less to persisting racial wage and employment gaps than was once true, its direct and indirect effects likely remain important.

Recent analyses of wage and employment gaps between whites and blacks suggest that existing gaps in wages are largely accounted for by racial gaps in schooling, cognitive achievement (as measured by test scores), and early work experience, while somewhat larger racial employment gaps remain even after accounting for these factors (Johnson and Neal, 1998). Wage gaps by race declined strongly in the decade or so following the passage of the Civil Rights Act, though they stabilized after that (and even widened somewhat among younger and more-educated groups, according to Bound and Freeman, 1992). Wage gaps among young Hispanics (particularly Mexican-Americans) and whites persist as well, though these seem to largely reflect education and language differences (Borjas and Katz, 2005). 

There is little doubt that persisting gaps in educational attainment and school achievement between whites and minorities do more damage to relative earnings than it did in the past. Also, a range of other factors – such as "spatial mismatch" between places of residence and work, weakening informal networks, as well as declining wages for less-educated men – continue to result in weaker employment outcomes among young blacks.

In fact, gaps in employment and labor force participation rates between young black and other men continued to widen in the 1980's and even the 1990's, while those among young women declined. Rising incarceration rates and child support policies have contributed to this development. As adolescents and teens, young minority men "disconnect" from school and work and become non-custodial fathers and offenders in large numbers (Holzer et al., 2005; Edelman et al., 2006; Mincy, 2006).        

However, it is likely that continuing employment discrimination plays some role in these trends. As Glenn Loury (2003) has pointed out, schooling and other human capital investment decisions by individuals will often reflect their perceived chances of success, which in turn can be affected by low expectations or widely-held racial stigmas. Indeed, the high rates of school disconnection and labor force withdrawal by young black men likely reflect a belief that few jobs outside of those paying low wages will be open to them over time (Edelman et al., 2006); and for those with limited schooling, these perceptions may not be inaccurate. The tendency of black men to be hired, conditional on applying for jobs, appears lower in the growing service sector than in declining blue-collar jobs in construction, manufacturing or wholesale trade (Holzer, 1996). Employers sometimes choose their recruitment methods, and their place of employment, at least partly to avoid black applicants and employees (Moss and Tilly, op. cit.). And the strength of the informal networks through which immigrants gain so many jobs at least partly reflects employer encouragement of immigrant referrals for their friends and relatives. 

Thus, the tendency of many young minority men (and perhaps women) to make personal choices at young ages that reduce their skills and labor market opportunities is likely reinforced by their perceptions of employer discrimination and hostility, among other factors. 

  1. The very high rates of incarceration among young black men clearly feed discriminatory behavior by employers. 

Employers are much more reluctant to hire men with criminal records than any other stigmatized group in the labor market (Holzer et al., 2004). Partly this reflects a range of state laws which forbid their employment in a variety of sectors; it also reflects employer fears of legal liability, financial loss or possible personal injury if they hire a potentially dishonest or violent employee. The employer aversion to hiring varies with the nature of the jobs and tasks required; in particular, jobs that involve reading or writing skills, the handling of cash, or direct contact with customers are less likely to be filled by offenders than more traditional blue-collar or service-sector jobs that do not involve these traits.

Audit studies in Milwaukee by Pager (2003) and in New York City by Pager and Western (2005) confirm that, among both white and black men, employers have a strong aversion to hiring those with criminal records. But, in these studies, black men without criminal records receive no more offers than white men with records, and with comparable educational and employment histories. The exact interpretation of this finding is somewhat unclear, since employers might well fear the presence of criminal records among the black men who do not actively report it.

Indeed, "statistical discrimination" on the part of employers, who know that black men are more likely on average to have criminal records or some other impediments to employment, implies that the provision of accurate information to employers might be a way to reduce such discrimination. Holzer et al. (2004) report that employer willingness to hire black men actually rises when the employers do criminal background checks, and can thus determine with greater certainty which men have criminal records and which do not.3

  1. A range of approaches might be useful in reducing hiring discrimination, though the cost-effectiveness of each is not really known right now.

Can Equal Employment Opportunity (EEO) laws be enforced with greater effectiveness than they are to date? Perhaps, though any answers to this question are clearly speculative. Their efficacy, especially in smaller establishments with relatively fewer minority applicants, might be enhanced by: 1) Greater use of tester studies to establish discriminatory behavior; 2) More systematic review of EEO-1 data from employers who are not government contractors; 3) The systematic collection and review of wage as well as employment data by race and gender from employers; and 4) Achieving somewhat greater balance between the numbers of hiring and promotion/discharge cases in employment discrimination lawsuits. Of course, a variety of practical, legal and administrative issues would need to be settled before these approaches could be advanced.4

Other activities and circumstances matter as well. Discrimination became less pervasive in the tight labor markets of the late 1990's, as employers could no longer find alternative job applicants to the groups in question (Holzer et al., 2006); and after welfare reform, during which employer and public attitudes towards welfare recipients as workers became more positive.5 Also, there is some evidence that when labor market "intermediaries" provide employers with more information about the personal qualities and recent work experiences of stigmatized employees - such as those with criminal records - employer opposition to them weakens somewhat (Holzer et al., 2003; Giloth, 2003).6 The use of such intermediaries – whether public or private, for-profit or not-for-profit – in labor market placements more broadly to combat discrimination might be considered as well, though the difficulties of achieving scale with such approaches would need to be considered. Intermediaries can also work with both employees and employers to improve their "soft skills" and "cultural competence" in dealing with members of other racial groups.

And, since better information on individual skills and attributes can help reduce labor market discrimination, developing better certifications of such skills (for example, for ex-offenders who have successfully completed a "transitional job" assignment) might be helpful as well. On the other hand, the importance of information about people also implies that ensuring some degree of accuracy in such information is critical. For example, it has become much easier for employers to do criminal background checks over the internet, but there is little regulation of private provision of such information to ensure that it is accurate.

Overall, there is a need for continued experimentation with and evaluation of these various approaches, to determine which can be cost-effective and used on a large enough scale to reduce the incidence of racial discrimination. And, as Hispanics and Asians account for more of the minority population over time, we need more research on these particular groups and the extent of labor market discrimination that they might experience.


Autor, David and David Scarborough. "Will Job Testing Harm Minority Workers? Evidence from the Retail Sector." National Bureau of Economic Research (NBER) Working Paper, 2003.

Borjas, George and Lawrence Katz. 2005. "The Evolution of the Mexican-Born Workforce in the U.S." NBER Working Paper.

Bertrand, Marianne and Senil Mullinathan. 2004. "Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment in Labor Market Discirmination." American Economic Review.

Bound, John and Richard Freeman. 1992. "What Went Wrong?" Quarterly Journal of Econonics. Vol 107, No. 1. February.

Darity, William and Patrick Mason. 1998. "Evidence on Discrimination in Employment: Codes of Color, Codes of Gender."  Journal of Economic Perspectives. Vol. 12, No. 2.

Donohue, John and Peter Siegelman. 1991. "The Changing Nature of Employment Discrimination Litigation." Stanford Law Review, Vol. 43.

Edelman, Peter; Harry J. Holzer and Paul Offner. 2006. Reconnecting Disadvantaged Young Men. Washington DC: Urban Institute Press.

Fix, Michael and Raymond Struyk. 1994. Clear and Convincing Evidence. Washington DC: Urban Institute Press.

Giloth, Robert ed. 2003. Workforce Intermediaries for the 21st Century. Philadelphia: Temple University Press.

Heckman, James. 1998. "Detecting Discrimination." Journal of Economic Perspectives. Vol. 12, No. 2.

Holzer, Harry J. 1996. What Employers Want: Job Prospects for Less-Educated Workers. New York: Russell Sage Foundation.

Holzer, Harry J. 1998. "Why Do Smaller Establishments Hire Fewer Blacks than Larger Ones?" Journal of Human Resources. Fall.

Holzer, Harry J. and Keith R. Ihlanfeldt. 1998. "Customer Discrimination and Employment Outcomes of Minority Workers." Quarterly Journal of Economics. August.

Holzer, Harry J. and David Neumark. 2006. "Affirmative Action: What Do We Know?" Journal of Policy Analysis and Management. Spring.

Holzer, Harry J.; Paul Offner and Elaine Sorensen. 2005. "Declining Employment Among Young Black Men: The Role of Incarceration and Child Support." Journal of Policy Analysis and Management. Spring.

Holzer, Harry J.; Steven Raphael and Michael Stoll. 2003. "Barriers to Employment for Ex-Offenders." Paper presented at the Reentry Roundtable, Sponsored by the Urban Institute, New York University, May 21-23.

Holzer, Harry J.; Steven Raphael and Michael Stoll. 2006. "Employers in the Boom: The Hiring of Less-Skilled Workers in the 1990's." Review of Economics and Statistics. Forthcoming. 

Holzer, Harry J.; Steven Raphael and Michael Stoll. 2000. "Are Suburban Firms More Likely to Discriminate Against African Americans?" Journal of Urban Economics.

Holzer, Harry J.; Steven Raphael and Michael Stoll. 2004. "Will Employers Hire Former Offenders? Employer Preferences, Background Checks and their Determinants." In M. Pattillo, D. Weiman and B. Western eds. Imprisoning America: The Racial Effects of Mass Incarceration. New York: Russell Sage Foundation.

Holzer, Harry J. and Michael Stoll. 2001. Employers and Welfare Recipients: The Effects of Welfare Reform in the Workplace. San Francisco: Public Policy Institute of California.

Johnson, William and Derek Neal. 1998. "" In C. Jencks and M. Phillips eds. The Black-White Achievement Gap. Washington DC: The Brookings Institution.

Kenney, Genevieve and Douglas Wissoker. 1994. "An Analysis of the Correlates of Discrimination Facing Young Hispanic Jobseekers." American Economic Review. Vol. 84.

Kirschenman, Joleen and Kathryn Neckerman. 1991. "We'd Love to Hire Them But...." In C. Jencks and P. Peterson eds. The Urban Underclass. Washington DC: The Brookings Institution.  

Loury, Glenn. 2003. The Anatomy of Racial Inequality. Harvard University Press.

Mincy, Ronald ed. 2006. Young Black Males Left Behind. Washington DC: Urban Institute Press.

Moss, Philip and Chris Tilly. 2001. Stories Employers Tell. New York: Russell Sage Foundation.

Pager, Devah. 2003. "The Mark of a Criminal Record." American Journal of Sociology. Vol. 108.

Pager, Devah and Bruce Western. 2005. "Discrimination in Low Wage Labor Markets." Unpublished, Princeton University.


(1) See reviews and critiques of these studies in Darity and Mason (1998) and Heckman (1998). For evidence on Hispanics see Kenney and Wissoker (1994). Also, Bertrand and Mullinathan (2004) show that employers are less likely to call back individuals with African-American names than those with more conventional names.

(2) This is true even when employers consider legal immigrants, who are generally paid the same wages for those jobs as are paid to other employees. For those who are illegal (or undocumented), wages paid might well be lower.

(3) See also Autor and Scarborough (2003 ) for evidence that employers who provide a variety of tests engage in less discrimination against African Americans than those who do not.

(4) For instance, the usefulness of testers as part of a direct enforcement strategy is limited not only by legal considerations, but also by the practical fact that we would need many data points per employer in order to infer bias at the individual level, which would make large scale enforcement efforts using this technique very difficult. But additional tester studies might help us identify more clearly which types of employers, in which business sectors and geographic areas, are most likely to discriminate, and against whom.

(5) See Holzer and Stoll (2001). The change in attitudes towards welfare recipients likely occurred partly because they were now being portrayed as struggling working parents rather than non-working welfare recipients, and also partly because employers reported relatively positive experiences when hiring them.

(6) For example, the employer survey data reported in Holzer et al. suggests that employer opposition to hiring offenders softens as they here more about the (non-violent) nature of the offenses committed or any positive work experience that the offender might have had since being released.

This page was last modified on June 12, 2006.