Professional Documents
Culture Documents
Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labor
Market Discrimination
Author(s): Marianne Bertrand and Sendhil Mullainathan
Reviewed work(s):
Source: The American Economic Review, Vol. 94, No. 4 (Sep., 2004), pp. 991-1013
Published by: American Economic Association
Stable URL: http://www.jstor.org/stable/3592802 .
Accessed: 31/10/2011 10:21
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to The
American Economic Review.
http://www.jstor.org
Are Emilyand Greg More EmployableThan Lakisha and
Jamal? A Field Experimenton Labor MarketDiscrimination
Every measure of economic success reveals dates, employers might favor the African-
significant racial inequality in the U.S. labor American one.' Data limitations make it
market. Compared to Whites, African-Ameri- difficult to empirically test these views. Since
cans are twice as likely to be unemployed and researcherspossess far less data than employers
earn nearly 25 percent less when they are em- do, White and African-Americanworkers that
ployed (Council of Economic Advisers, 1998). appear similar to researchers may look very
This inequality has sparked a debate as to differentto employers. So any racial difference
whether employers treat members of different in labormarketoutcomes could just as easily be
races differentially. When faced with observ- attributedto differences that are observable to
ably similar African-Americanand White ap- employers but unobservableto researchers.
plicants, do they favor the White one? Some To circumvent this difficulty, we conduct a
argue yes, citing either employer prejudice or field experiment that builds on the correspon-
employerperceptionthatrace signals lower pro- dence testing methodology that has been pri-
ductivity. Others argue that differential treat- marily used in the past to stud' minority
ment by race is a relic of the past, eliminatedby outcomes in the United Kingdom. We send
some combination of employer enlightenment, resumes in response to help-wantedads in Chi-
affirmative action programs and the profit- cago and Boston newspapersand measurecall-
maximizationmotive. In fact, many in this latter back for interview for each sent resume. We
camp even feel that stringent enforcement of
affirmative action programs has produced an
environment of reverse discrimination. They This camp often explains the poor performance of
would argue that faced with identical candi- African-Americansin terms of supply factors. If African-
Americanslack many basic skills enteringthe labor market,
then they will performworse, even with parityor favoritism
* Bertrand:GraduateSchool of Business, University of in hiring.
2 See
Chicago, 1101 E. 58th Street,RO229D, Chicago, IL 60637, Roger Jowell and PatriciaPrescott-Clarke(1970),
NBER, and CEPR (e-mail: marianne.bertrand@gsb. Jim Hubbuck and Simon Carter(1980), Colin Brown and
uchicago.edu); Mullainathan:Department of Economics, Pat Gay (1985), and Peter A. Riach and JudithRich (1991).
MassachusettsInstituteof Technology, 50 MemorialDrive, One caveat is that some of these studies fail to fully match
E52-380a, Cambridge, MA 02142, and NBER (e-mail: skills between minority and nonminorityresumes. For ex-
mullain@mit.edu).David Abrams, Victoria Bede, Simone ample some impose differential education backgroundby
Berkowitz,Hong Chung,AlmudenaFernandez,Mary Anne racial origin. Doris Weichselbaumer(2003, 2004) studies
Guediguian, Christine Jaw, Richa Maheswari, Beverley the impact of sex-stereotypes and sexual orientation.Rich-
Martis, Alison Tisza, GrantWhitehorn,and Christine Yee ardE. Nisbett and Dov Cohen (1996) performa relatedfield
providedexcellent researchassistance. We are also grateful experimentto study how employers' response to a criminal
to numerous colleagues and seminar participantsfor very past varies between the North and the South in the United
helpful comments. States.
991
992 THEAMERICANECONOMICREVIEW SEPTEMBER2004
race-specific names we have chosen might also Rouse (2000), for example, examine the effect
proxy for social class above andbeyond the race of blind auditioning on the hiring process of
of the applicant.Using birth certificatedata on orchestras. By observing the treatment of fe-
mother's educationfor the differentfirst names male candidatesbefore and after the introduc-
used in our sample, we find little relationship tion of blind auditions,they try to measure the
between social background and the name- amountof sex discrimination.When such pseu-
specific callbackrates. Second, we discuss how do-experiments can be found, the resulting
our results map back to the differentmodels of study can be very informative;but finding such
discriminationproposedin the economics liter- experiments has proven to be extremely
ature. In doing so, we focus on two important challenging.
results: the lower returns to credentials for A different set of studies, known as audit
African-Americansand the relative homogene- studies, attemptsto place comparableminority
ity of the racial gap across occupations and and White actors into actual social and eco-
industries.We conclude thatexisting models do nomic settings and measure how each group
a poor job of explaining the full set of findings. fares in these settings.9 Labor market audit
Section V concludes. studies send comparable minority (African-
American or Hispanic) and White auditors in
I. PreviousResearch for interviewsand measurewhetherone is more
likely to get the job than the other.'? While the
With conventionallabor force and household results vary somewhat across studies, minority
surveys, it is difficult to study whether differ- auditorstend to performworse on average:they
ential treatment occurs in the labor market.7 are less likely to get called back for a second
Armed only with survey data, researchersusu- interview and, conditional on getting called
ally measure differentialtreatmentby compar- back, less likely to get hired.
ing the labormarketperformanceof Whites and These audit studies provide some of the
African-Americans (or men and women) for cleanest nonlaboratoryevidence of differential
which they observe similar sets of skills. But treatmentby race. But they also have weak-
such comparisons can be quite misleading. nesses, most of which have been highlighted in
Standardlabor force surveys do not contain all Heckman and Siegelman (1992) and Heckman
the characteristicsthatemployers observe when (1998). First, these studies require that both
hiring, promoting,or setting wages. So one can members of the auditorpair are identical in all
never be sure that the minorityand nonminority dimensions that might affect productivity in
workers being comparedare truly similar from employers' eyes, except for race. To accomplish
the employers' perspective. As a consequence, this, researchers typically match auditors on
any measureddifferences in outcomes could be several characteristics(height, weight, age, di-
attributed to these unobserved (to the re- alect, dressing style, hairdo) and train them for
searcher)factors. several days to coordinate interviewing styles.
This difficulty with conventional data has Yet, critics note that this is unlikely to erase the
led some authors to instead rely on pseudo- numerousdifferences that exist between the au-
experiments.8 Claudia Goldin and Cecilia ditors in a pair.
Anotherweakness of the audit studies is that
they are not double-blind. Auditors know the
6
We also argue that a social class interpretationwould purpose of the study. As Turner et al. (1991)
find it hard to explain some of our findings, such as why
living in a betterneighborhooddoes not increasecallbackrates
more for African-American names thanfor White names. 9 Michael Fix and MarjeryA. Turner(1998) provide a
7
See Joseph G. Altonji and Rebecca M. Blank (1999) survey of many such audit studies.
for a detailed review of the existing literature on racial 10Earlier
hiring audit studies include JerryM. Newman
discriminationin the labor market. (1978) and Shelby J. McIntyre et al. (1980). Three more
8 William A. Darity, Jr. and Patrick L. Mason recent studies are HarryCross et al. (1990), FranklinJames
(1998)
describe an interestingnonexperimentalstudy. Prior to the and Steve W. DelCastillo (1991), and Turneret al. (1991).
Civil Rights Act of 1964, employmentads would explicitly Heckman and Peter Siegelman (1992), Heckman (1998),
state racialbiases, providinga direct measureof differential and Altonji and Blank (1999) summarizethese studies. See
treatment.Of course, as Arrow (1998) mentions, discrimi- also David Neumark(1996) for a labor marketaudit study
nation was at that time "a fact too evident for detection." on gender discrimination.
994 THEAMERICANECONOMICREVIEW SEPTEMBER2004
note: "The first day of trainingalso included an examine the natureof the differentialtreatment
introduction to employment discrimination, from many more angles.
equal employmentopportunity,and a review of
project design and methodology." This may II. ExperimentalDesign
generate conscious or subconscious motives
among auditors to generate data consistent or A. Creating a Bank of Resumes
inconsistent with their beliefs about race issues
in America. As psychologists know very well, The first step of the experimentaldesign is to
these demand effects can be quite strong. It is generate templates for the resumes to be sent.
very difficult to insure that auditors will not The challenge is to producea set of realistic and
want to do "a good job." Since they know the representativeresumes without using resumes
goal of the experiment, they can alter their that belong to actual job seekers. To achieve
behaviorin front of employers to express (indi- this goal, we start with resumes of actual job
rectly) their own views. Even a small belief by searchers but alter them sufficiently to create
auditors that employers treat minorities differ- distinct resumes. The alterations maintain the
ently can result in measureddifferencesin treat- structureandrealismof the initial resumeswith-
ment. This effect is furthermagnifiedby the fact out compromisingtheir owners.
that auditorsare not in fact seekingjobs and are We begin with resumes posted on two job
thereforemore free to let their beliefs affect the search Web sites as the basis for our artificial
interview process. resumes.ll While the resumes posted on these
Finally, audit studies are extremely expen- Web sites may not be completely representative
sive, making it difficult to generate large of the averagejob seeker, they provide a prac-
enough samples to understandnuances and pos- tical approximation.'2We restrict ourselves to
sible mitigating factors. Also, these budgetary people seeking employmentin our experimental
constraintsworsen the problem of mismatched cities (Boston and Chicago). We also restrict
auditorpairs. Cost considerationsforce the use ourselves to four occupationalcategories:sales,
of a limited numberof pairs of auditors,mean- administrative support, clerical services, and
ing that any one mismatched pair can easily customer services. Finally, we further restrict
drive the results. In fact, these studies generally ourselves to resumes posted more than six
tend to find significantdifferences in outcomes months priorto the startof the experiment.We
across pairs. purgethe selected resumesof the person's name
Our study circumventsthese problems.First, and contact information.
because we only rely on resumes and not peo- During this process, we classify the resumes
ple, we can be sure to generate comparability within each detailed occupationalcategory into
across race. In fact, since race is randomly as- two groups: high and low quality. In judging
signed to each resume, the same resume will resume quality, we use criteria such as labor
sometimes be associated with an African- market experience, career profile, existence of
American name and sometimes with a White gaps in employment, and skills listed. Such a
name. This guaranteesthat any differences we classification is admittedly subjective but it is
find are caused solely by the race manipulation. made independentlyof any race assignment on
Second, the use of paper resumes insulates us the resumes (which occurs later in the experi-
from demand effects. While the researchassis- mental design). To furtherreinforce the quality
tants know the purpose of the study, our proto- gap between the two sets of resumes, we add to
col allows little room for conscious or each high-qualityresume a subset of the follow-
subconscious deviations from the set proce- ing features: summer or while-at-school em-
dures. Moreover, we can objectively measure ployment experience, volunteering experience,
whether the randomization occurred as ex- extra computerskills, certificationdegrees, for-
pected. This kind of objective measurementis eign language skills, honors, or some military
impossible in the case of the previous audit
studies. Finally, because of relatively low mar- 11The sites are www.careerbuilder.com and www.
ginal cost, we can send out a large number of americasjobbank.com.
resumes. Besides giving us more precise esti- 12
In practice, we found large variation in skill levels
mates, this larger sample size also allows us to among people posting their resumes on these sites.
VOL.94 NO. 4 BERTRANDAND MULLAINATHAN:
RACEIN THE LABORMARKET 995
experience. This resume quality manipulation which names are distinctively White and which
needs to be somewhat subtle to avoid making a are distinctively African-American.Distinctive
higher-qualityjob applicantoverqualifiedfor a names are those that have the highest ratio of
given job. We try to avoid this problem by frequency in one racial group to frequency in
making sure that the features listed above are the other racial group.
not all added at once to a given resume. This As a check of distinctiveness,we conducteda
leaves us with a high-qualityand a low-quality survey in variouspublic areas in Chicago. Each
pool of resumes.'3 respondent was asked to assess features of a
To minimize similarityto actualjob seekers, person with a particularname, one of which is
we use resumes from Boston job seekers to race. For each name, 30 respondentswere asked
form templatesfor the resumes to be sent out in to identifythe name as either"White,""African-
Chicago and use resumes from Chicago job American,""Other,"or "CannotTell." In gen-
seekers to form templatesfor the resumes to be eral, the names led respondents to readily
sent out in Boston. To implement this migra- attributethe expected race for the person but
tion, we alter the names of the schools and there were a few exceptions and these names
previous employers on the resumes. More spe- were disregarded.18
cifically, for each Boston resume, we use the The final list of firstnames used for this study
Chicago resumes to replace a Boston school is shown in Appendix Table Al. The table
with a Chicago school.'4 We also use the Chi- reportsthe relative likelihood of the names for
cago resumesto replace a Boston employerwith the Whites and African-Americansin the Mas-
a Chicago employer in the same industry. We sachusetts birth certificates data as well as
use a similar procedureto migrate Chicago re- the recognition rate in the field survey.19 As
sumes to Boston.15 This produces distinct but Appendix Table Al indicates, the African-
realistic looking resumes, similar in their edu- Americanfirstnames used in the experimentare
cation and career profiles to this subpopulation quite common in the population.This suggests
of job searchers.l that by using these names as an indicator of
race, we are actually covering a rather large
B. Identities of Fictitious Applicants segment of the African-Americanpopulation.20
Applicantsin each race/sex/city/resumequal-
The next step is to generateidentities for the ity cell are allocated the same phone number.
fictitiousjob applicants:names, telephone num- This guaranteesthat we can precisely trackem-
bers, postal addresses, and (possibly) e-mail ployer callbacks in each of these cells. The
addresses.The choice of names is crucialto our phone lines we use are virtualones with only a
experiment.17 To decide on which names are voice mailbox attachedto them. A similar out-
uniquely African-American and which are going message is recordedon each of the voice
uniquely White, we use name frequency data mailboxes but each message is recorded by
calculated from birth certificates of all babies someone of the appropriaterace and gender.
born in Massachusettsbetween 1974 and 1979.
We tabulate these data by race to determine
18
For example, Maurice and Jerome are distinctively
African-Americannames in a frequency sense yet are not
perceived as such by many people.
13 19So
In Section III, subsection B, and Table 3, we provide many of names show a likelihood ratio of oc be-
a detailed summary of resume characteristicsby quality cause thereis censoringof the data at five births.If thereare
level. fewer than five babies in any race/namecell, it is censored
14
We try as much as possible to matchhigh schools and (and we do not know whether a cell has zero or was
colleges on quality and demographiccharacteristics. censored). This is primarilya problem for the computation
15Note that for
applicantswith schooling or work expe- of how many African-American babies have "White"
rience outside of the Boston or Chicago areas, we leave the names.
school or employer name unchanged. 20 We also tried to use more White-soundinglast names
16
We also generatea set of differentfonts, layouts, and for White applicantsand more African-American-sounding
cover letters to furtherdifferentiatethe resumes. These are last names for African-Americanapplicants.The last names
applied at the time the resumes are sent out. used for White applicants are: Baker, Kelly, McCarthy,
17 We chose name over other
potential manipulationsof Murphy,Murray,O'Brien, Ryan, Sullivan, and Walsh. The
race, such as affiliation with a minority group, because we last names used for African-Americanapplicantsare: Jack-
felt such affiliationsmay especially convey more than race. son, Jones, Robinson, Washington,and Williams.
996 THEAMERICANECONOMICREVIEW SEPTEMBER2004
Since we allocate the same phone number for sample four resumes (two high-qualityand two
applicantswith different names, we cannot use low-quality) that fit the job descriptionand re-
a person name in the outgoing message. quirements as closely as possible.23 In some
While we do not expect positive feedback cases, we slightly alter the resumes to improve
from an employer to take place via postal mail, the quality of the match, such as by adding the
resumes still need postal addresses. We there- knowledge of a specific software program.
fore constructfictitious addressesbased on real One of the high- and one of the low-quality
streets in Boston and Chicago using the White resumes selected are then drawn at random to
Pages. We select up to three addresses in each receive African-American names, the other
5-digit zip code in Boston and Chicago. Within high- and low-quality resumes receive White
cities, we randomly assign addresses across all names.24We use male and female names for
resumes. We also create eight e-mail addresses, sales jobs, whereas we use nearly exclusively
four for Chicago and four for Boston.21These female names for administrative and clerical
e-mail addressesare neutralwith respect to both jobs to increase callback rates.25Based on sex,
race and sex. Not all applicants are given an race, city, and resume quality, we assign a re-
e-mail address. The e-mail addresses are used sume the appropriatephone number. We also
almost exclusively for the higher-quality re- select at random a postal address. Finally, e-
sumes. This procedureleaves us with a bank of mail addresses are added to most of the high-
names, phone numbers, addresses, and e-mail quality resumes.26 The final resumes are
addresses that we can assign to the template formatted, with fonts, layout, and cover letter
resumes when responding to the employment style chosen at random. The resumes are then
ads. faxed (or in a few cases mailed) to the em-
ployer. All in all, we respond to more than
C. Responding to Ads 1,300 employment ads over the entire sample
period and send close to 5,000 resumes.
The experimentwas carriedout between July
2001 and January2002 in Boston and between D. Measuring Responses
July 2001 and May 2002 in Chicago.22 Over
that period, we surveyed all employmentads in We measurewhethera given resume elicits a
the Sunday editions of The Boston Globe and callback or e-mail back for an interview. For
The Chicago Tribune in the sales, administra- each phone or e-mail response, we use the con-
tive support,and clerical and customer services tent of the message left by the employer (name
sections. We eliminate any ad where applicants of the applicant, company name, telephone
were asked to call or appearin person. In fact, numberfor contact)to matchthe responseto the
most of the ads we surveyed in these job cate- corresponding resume-ad pair.27 Any attempt
gories ask for applicants to fax in or (more by employers to contact applicants via postal
rarely) mail in their resume. We log the name mail cannot be measured in our experiment
(when available) and contact information for since the addresses are fictitious. Several hu-
each employer, along with any informationon man resource managers confirmed to us that
the position advertised and specific require-
ments (such as education, experience, or com-
puter skills). We also record whetheror not the 23 In
some instances,our resumebankdoes not have four
ad explicitly states thatthe employeris an equal resumes that are appropriatematches for a given ad. In such
opportunityemployer. instances, we send only two resumes.
24
For each ad, we use the bank of resumes to Though the same names are repeatedly used in our
experiment,we guaranteethatno given ad receives multiple
resumes with the same name.
25
Male names were used for a few administrativejobs in
21The e-mail addresses are registered on Yahoo.com, the first month of the experiment.
26In the first month of the
Angelfire.com, or Hotmail.com. experiment, a few high-
22
This period spans tighterand slackerlabormarkets.In quality resumes were sent without e-mail addresses and a
our data, this is apparentas callback rates (and numberof few low-quality resumes were given e-mail addresses. See
new ads) dropped after September 11, 2001. Interestingly, Table 3 for details.
however, the racial gap we measureis the same across these 27 Very few employers used e-mail to contact an appli-
two periods. cant back.
VOL.94 NO. 4 BERTRANDAND MULLAINATHAN:
RACEIN THE LABORMARKET 997
Notes: The table reports,for the entire sample and differentsubsamplesof sent resumes, the callbackrates for applicantswith
a White-soundingname (column 1) an an African-American-soundingname (column 2), as well as the ratio (column 3) and
difference (column 4) of these callback rates. In bracketsin each cell is the numberof resumes sent in that cell. Column 4
also reportsthe p-value for a test of proportiontesting the null hypothesisthatthe callbackratesare equal across racialgroups.
employers rarely,if ever, contact applicantsvia name).28We returnto this issue in Section IV,
postal mail to set up interviews. subsection B.
Finally, and this is an issue pervasive in both
E. Weaknessesof the Experiment our study and the pair-matchingaudit studies,
newspaper ads representonly one channel for
We have alreadyhighlightedthe strengthsof job search. As is well known from previous
this experimentrelative to previous audit stud- work, social networks are another common
ies. We now discuss its weaknesses. First, our means throughwhich people find jobs and one
outcome measure is crude, even relative to the that clearly cannot be studied here. This omis-
previous audit studies. Ultimately, one cares sion could qualitatively affect our results if
about whether an applicant gets the job and African-Americansuse social networksmore or
about the wage offered conditional on getting if employers who rely more on networksdiffer-
the job. Our procedure,however, simply mea- entiate less by race.29
sures callbacksfor interviews.To the extent that
the search process has even moderatefrictions, III. Results
one would expect that reduced interview rates
would translate into reduced job offers. How- A. Is There a Racial Gap in Callback?
ever, we are not able to translateour results into
gaps in hiring rates or gaps in earnings. Table 1 tabulates average callback rates by
Anotherweakness is that the resumes do not racial soundingness of names. Included in
directly report race but instead suggest race brackets under each rate is the number of re-
through personal names. This leads to various sumes sent in that cell. Row 1 presents our
sources of concern. First, while the names are resultsfor the full data set. Resumes with White
chosen to make race salient, some employers
may simply not notice the names or not recog- 28
nize their racial content. On a related note, As Appendix Table Al indicates, the African-
because we are not assigning race but only American names we use are, however, quite common
among African-Americans,making this less of a concern.
race-specific names, our results are not repre- 29
In fact, thereis some evidence thatAfrican-Americans
sentative of the average African-American may rely less on social networksfor theirjob search (Harry
(who may not have such a racially distinct J. Holzer, 1987).
998 THEAMERICANECONOMICREVIEW SEPTEMBER2004
TABLE2-DISTRIBUTION OF CALLBACKS AD
BY EMPLOYMENT
Notes: This table documentsthe distributionof callbacks at the employment-adlevel. "No Callback"is the percentof ads for
which none of the fictitiousapplicantsreceived a callback. "1W + 1B" is the percentof ads for which exactly one White and
one African-Americanapplicantreceived a callback."2W + 2B" is the percentof ads for which exactly two White applicants
and two African-Americanapplicantsreceived a callback. "EqualTreatment"is defined as the sum of "No Callback,""1W +
1B," and "2W + 2B." "1W + OB"is the percent of ads for which exactly one White applicantand no African-American
applicant received a call back. "2W + OB" is the percent of ads for which excatly two White applicants and no
African-Americanapplicantreceived a callback."2W + 1B" is the percentof ads for which exactly two White applicantsand
one African-Americanapplicantreceived a callback. "WhitesFavored"is defined as the sum of "1W + OB,""2W + OB,"
and "2W + 1B." "1B + OW"is the percentof ads for which exactly one African-Americanapplicantand no White applicant
received a callback. "2B + OW"is the percent of ads for which exactly two African-Americanapplicants and no White
applicantreceived a callback. "2B + 1W" is the percent of ads for which exactly two African-Americanapplicantsand one
White applicantreceived a callback. "African-AmericansFavored"is defined as the sum of "1B + OW,""2B + OW,"and
"2B + 1W."In bracketsin each cell is the numberof employment ads in that cell. "Ho: WF = WB"reportsthe p-value for
a test of symmetrybetween the proportionof employers that favor White names and the proportionof employers that favor
African-Americannames.
no callbacks are recorded (83 percent of the employment ads we respond to, we send four
ads). Whites are favored by nearly 8.4 percent different resumes: two higher-qualityand two
of the employers, with a majorityof these em- lower-qualityones. Table 3 gives a bettersense
ployers contactingexactly one White applicant. of which factors enter into this subjective clas-
African-Americans,on the other hand, are fa- sification. Table 3 displays means and standard
vored by only about 3.5 percent of employers. deviations of the most relevant resume charac-
We formally test whether there is symmetry in teristics for the full sample (column 1), as well
the favoring of Whites over African-Americans as broken down by race (columns 2 and 3) and
and African-Americansover Whites. We find resume quality (columns 4 and 5). Since appli-
that the difference between the fraction of em- cants' names are randomized,there is no differ-
ployers favoring Whites and the fraction of ence in resume characteristicsby race. Columns
employers favoring African-Americansis sta- 4 and 5 documentthe objective differences be-
tistically very significant (p = 0.0000). tween resumes subjectively classified as high
and low quality. Higher-qualityapplicantshave
B. Do African-AmericansReceive Different on average close to an extra year of labor mar-
Returns to Resume Quality? ket experience,fewer employmentholes (where
an employmenthole is defined as a period of at
Our results so far demonstratea substantial least six months without a reportedjob), are
gap in callback based on applicants' names. more likely to have worked while at school,
Next, we would like to learn more about the and to report some military experience. Also,
factors that may influence this gap. More spe- higher-quality applicants are more likely to
cifically, we ask how employers respondto im- have an e-mail address, to have received some
provements in African-American applicants' honors, and to list some computer skills and
credentials.To answer this question, we exam- other special skills (such as a certification
ine how the racial gap in callback varies by degree or foreign language skills) on their re-
resume quality. sume. Note that the higher- and lower-quality
As we explainedin Section II, for most of the resumes do not differ on average with regardto
1000 THEAMERICANECONOMICREVIEW SEPTEMBER2004
TABLE3-RESUMECHARACTERISTICS: STATISTICS
SUMMARY
African-
Sample: All resumes White names American Higher quality Lower quality
Characteristic:
College degree 0.72 0.72 0.72 0.72 0.71
(Y= 1) (0.45) (0.45) (0.45) (0.45) (0.45)
Years of experience 7.84 7.86 7.83 8.29 7.39
(5.04) (5.07) (5.01) (5.29) (4.75)
Volunteeringexperience? 0.41 0.41 0.41 0.79 0.03
(Y = 1) (0.49) (0.49) (0.49) (0.41) (0.16)
Military experience? 0.10 0.09 0.10 0.19 0.00
(Y = 1) (0.30) (0.29) (0.30) (0.39) (0.06)
E-mail address? 0.48 0.48 0.48 0.92 0.03
(Y = 1) (0.50) (0.50) (0.50) (0.27) (0.17)
Employmentholes? 0.45 0.45 0.45 0.34 0.56
(Y = 1) (0.50) (0.50) (0.50) (0.47) (0.50)
Work in school? 0.56 0.56 0.56 0.72 0.40
(Y = 1) (0.50) (0.50) (0.50) (0.45) (0.49)
Honors? 0.05 0.05 0.05 0.07 0.03
(Y = 1) (0.22) (0.23) (0.22) (0.25) (0.18)
Computerskills? 0.82 0.81 0.83 0.91 0.73
(Y = 1) (0.38) (0.39) (0.37) (0.29) (0.44)
Special skills? 0.33 0.33 0.33 0.36 0.30
(Y = 1) (0.47) (0.47) (0.47) (0.48) (0.46)
Fractionhigh school dropoutsin 0.19 0.19 0.19 0.19 0.18
applicant's zip code (0.08) (0.08) (0.08) (0.08) (0.08)
Fractioncollege or more in 0.21 0.21 0.21 0.21 0.22
applicant's zip code (0.17) (0.17) (0.17) (0.17) (0.17)
FractionWhites in applicant's zip 0.54 0.55 0.54 0.53 0.55
code (0.33) (0.33) (0.33) (0.33) (0.33)
FractionAfrican-Americansin 0.31 0.31 0.31 0.32 0.31
applicant's zip code (0.33) (0.33) (0.33) (0.33) (0.33)
Log(medianper capital income) 9.55 9.55 9.55 9.54 9.56
in applicant's zip code (0.56) (0.56) (0.55) (0.54) (0.57)
Notes: The table reportsmeans and standarddeviations for the resume characteristicsas listed on the left. Column 1 refers
to all resumes sent; column 2 refers to resumeswith White names;column 3 refers to resumes with African-Americannames;
column 4 refers to higher-qualityresumes; column 5 refers to lower-qualityresumes. See text for details.
Notes: PanelA reportsthe meancallbackpercentsfor applicantwith a White name (row 1) and African-American name (row 2)
dependingon whetherthe resume was subjectivelyqualifiedas a lower qualityor higherquality.In bracketsis the numberof
resumessent for each race/qualitygroup.The last columnreportsthep-valueof a test of proportiontestingthe null hypothesisthat
the callbackratesare equal acrossqualitygroupswithineach racialgroup.For PanelB, we use a thirdof the sampleto estimate
a probitregressionof the callbackdummyon the set of resumecharacteristics as displayedin Table 3. We furthercontrolfor a sex
dummy,a city dummy,six occupationdummies,anda vectorof dummyvariablesforjob requirementsas listedin the employment
ad (see SectionII, subsectionD, for details).We thenuse the estimatedcoefficientson the set of resumecharacteristics
to estimate
a predictedcallbackfor the remainingresumes(two-thirdsof the sample).We call "high-quality" resumesthe resumesthatrank
above the medianpredictedcallbackand "low-quality"resumesthe resumesthat rankbelow the medianpredictedcallback.In
bracketsis the numberof resumessentfor eachrace/qualitygroup.The last columnreportsthep-valueof a test of proportiontesting
the null hypothesisthatthe callbackpercentsare equal acrossqualitygroupswithineach racialgroup.
rate for similar improvementsin their creden- that have below-median-predicted callback. As
tials. African-Americanswith higher-qualityre- one can see from Panel B, qualitatively similar
sumes receive a callback6.7 percentof the time, results emerge from this analysis. While African-
comparedto 6.2 percentfor African-Americans Americans do appear to significantly benefit
with lower quality resumes. This is only a 0.51- from higher-quality resumes under this alterna-
percentage-point,or 8-percent, difference and tive classification, they benefit less than Whites.
this difference is not statistically significant The ratio of callback rates for high- versus
(p = 0.6084). low-quality resumes is 1.60 for African Amer-
Instead of relying on the subjective quality icans, compared to 1.89 for Whites.
classification, Panel B directly uses resume In Table 5, we directly report the results of
characteristics to classify the resumes. More race-specific probit regressions of the callback
specifically, we use a random subsample of dummy on resume characteristics. We, how-
one-third of the resumes to estimate a probit ever, start in column 1 with results for the full
regressionof the callbackdummyon the resume sample of sent resumes. As one can see, many
characteristicslisted in Table 3. We furthercon- of the resume characteristics have the expected
trol for a sex dummy, a city dummy, six occu- effect on the likelihood of a callback. The ad-
pation dummies, and a vector of job dition of an e-mail address, honors, and special
requirementsas listed in the employmentads.34 skills all have a positive and significant effect
We then use the estimated coefficients on the on the likelihood of a callback.35 Also, more
resume characteristics to rank the remaining experienced applicants are more likely to get
two-thirdsof the resumesby predictedcallback. called back: at the average number of years of
In Panel B, we classify as "high"those resumes experience in our sample (eight years), each
that have above-median-predicted callback;
similarly, we classify as "low" those resumes 35 Note that the e-mail address dummy, because it is
close to perfectly correlated with the subjective resume-
34 See Section qualityvariable,may in partcapturesome otherunmeasured
III, subsectionD, for more details on these resume characteristicsthat may have led us to categorize a
occupation categories and job requirements. given resume as higher quality.
1002 THEAMERICANECONOMICREVIEW SEPTEMBER2004
TABLE5-EFFECTOFRESUME OFCALLBACK
ONLIKELIHOOD
CHARACTERISTICS
Notes: Each column gives the results of a probit regression where the dependentvariable is the callback dummy. Reported
in the table are estimatedmarginalchanges in probabilityfor the continuousvariablesand estimateddiscrete changes for the
dummy variables. Also included in each regression are a city dummy, a sex dummy, six occupationdummies, and a vector
of dummy variablesfor job requirementsas listed in the employment ad (see Section III, subsection D, for details). Sample
in column 1 is the entire set of sent resumes; sample in column 2 is the set of resumes with White names; sample in column
3 is the set of resumes with African-Americannames. Standarderrorsare correctedfor clusteringof the observationsat the
employment-adlevel. Reported in the second to last row are the p-values for a x2 testing that the effects on the resume
characteristicsare all zero. Reportedin the second to last row is the standarddeviation of the predictedcallback rate.
Notes: Sample is all sent resumes (N = 4,870) unless otherwise specified in column 1. Column 2 reportsmeans and standard
deviations (in parentheses)for the job requirementor employer characteristic.For ads listing an experiencerequirement,50.1
percent listed "some," 24.0 percent listed "two years or less," and 25.9 percent listed "threeyears or more." For ads listing
an education requirement,8.8 percent listed a high school degree, 48.5 percent listed some college, and 42.7 percent listed
at least a four-yearcollege degree. Column 3 reportsthe marginaleffect of the job requirementor employer characteristic
listed in that row on differentialtreatment.Specifically, each cell in column 3 correspondsto a differentprobitregressionof
the callback dummy on an African-Americanname dummy, a dummy for the requirementor characteristiclisted in that row
and the interactionof the requirementor characteristicdummy with the African-Americanname dummy. Reportedin each
cell is the estimated change in probability for the interaction term. Standarderrors are corrected for clustering of the
observationsat the employment-adlevel.
ads mention some computerknowledge require- sources, we also try to identify the specific zip
ment, which can range from Excel or Word to code of the company (or company branch)that
more esoteric software programs. Good com- resumes are to be sent to. Finally, we use the
munication skills are explicitly required in Federal Procurementand Data Center Web site
about 12 percent of the ads. Organizationskills to find a list of companies that have federal
are mentioned 7 percent of the time. Finally, contracts.39The racial difference in callback
only about 11 percent of the ads list an explicit rates for the subsamples where employer char-
education requirement. Of these, 8.8 percent acteristicscould be determinedis very similarin
requirea high school degree, 48.5 percentsome magnitudeto that in the full sample.
college (such as an associate degree), and the Employer characteristicsdiffer significantly
rest at least a four-yearcollege degree.37 across ads. Twenty-nine percent of all employ-
Despite this variability,we find little system- ers explicitly state that they are "EqualOppor-
atic relationship between any of the require- tunity Employers." Eleven percent are federal
ments and the racial gap in callback. The point contractors and, therefore, might face greater
estimates in column 3 show no consistent eco- scrutinyunder affirmativeaction laws. The av-
nomic pattern and are all statistically weak. erage company size is around2,000 employees
Measures of job quality, such as experience or but there is a lot of variation across firms. Fi-
computerskills requirements,do not predictthe nally, 74 percentof the firms are privatelyheld,
extent of the racial gap. Communication or 15 percent are publicly traded, and 11 percent
other interpersonalskill requirementshave no are not-for-profitorganizations.
effect on the racial gap either.38 Neither "EqualOpportunityEmployers"nor
We also study employer characteristics.Col- federal contractors appear to treat African-
lecting such informationis a more difficult task Americans more favorably. In fact, each of
since it is not readily available from the em- these employer characteristics is associated
ployment ads we respond to. The only piece of with a larger racial gap in callback (and this
employer information we can directly collect effect is marginally significant for federal con-
from the employment ad is whether or not the tractors). Differential treatmentdoes not vary
employer explicitly states being an "EqualOp- with employer size.40 Point estimates indicate
portunityEmployer."In several cases, the name less differential treatmentin the not-for-profit
of the employer is not even mentionedin the ad sector; however, this effect is very noisily
and the only piece of informationwe can rely on estimated.41
is the fax number which applications must be In an unpublishedAppendix (available from
submittedto. We thereforehave to turn to sup- the authors upon request), we also study how
plemental data sources. For employment ads the racial gap in callback varies by occupation
that do not list a specific employer, we first use and industry.Based on the employmentad list-
the fax number to try to identify the company ings, we classify the job openings into six oc-
name via Web reverse-lookup services. Based cupation categories: executives and managers;
on company names, we use three different data administrative supervisors; sales representa-
sources (Onesource Business Browser, Thomas tives; sales workers;secretariesand legal assis-
Register, and Dun and BradstreetMillion Dol- tants;clerical workers.We also, when possible,
lar Directory, 2001) to trackcompany informa-
tion such as total employment, industry, and
ownership status. Using this same set of data 39This Web site (www.fpdc.gov) is accurateup to and
including March 21, 2000.
40
Similar results hold when we measure employer size
37Other using a total sales measure rather than an employment
requirementssometimes mentionedinclude typ- measure.
ing skills for secretaries (with specific words-per-minute 41 Our measurementof the racial gap by firm or em-
minimum thresholds), and, more rarely, foreign language ployer type may not be a good indicatorof the fraction of
skills. African-Americansactually employed in these firms. For
38 Other
ways of estimatingthese effects produce a sim- example, "Equal OpportunityEmployers" may receive a
ilar nonresult.Among otherthings, we consideredincluding higher fraction of African-Americanresumes. Their actual
a city dummy or estimating the effects separatelyby city; hiring may thereforelook differentfrom that of non "Equal
we also estimatedone single probitregressionincluding all OpportunityEmployers"when one considers the full set of
requirementsat once. resumes they receive.
1006 THEAMERICANECONOMICREVIEW SEPTEMBER2004
classify employers into six industrycategories: Americans? Second, does our design only iso-
manufacturing;transportationand communica- late the effect of race or is the name
tion; wholesale and retail trade; finance, insur- manipulationconveying some otherfactorsthan
ance, and real estate; business and personal race? Third,how do our results relate to differ-
services;health,educational,and social services. ent models of racial discrimination?
We then compute occupation and industry-
specific racial gaps in callback and relate these A. InterpretingCallback Rates
gaps to 1990 Census-based measures of oc-
cupationandindustryearnings,as well as Census- Our results indicate that for two identical
based measures of the White/African-American individuals engaging in an identicaljob search,
wage gap in these occupationsand industries. the one with an African-Americanname would
We find a positive White/African-American receive fewer interviews. Does differential
gap in callbacks in all occupation and industry treatmentwithin our experimentimply that em-
categories (except for transportationand com- ployers are discriminating against African-
munication).While average earnings vary a lot Americans (whether it is rational, prejudice-
across the occupations covered in the experi- based, or otherform of discrimination)?In other
ment, we find no systematic relationship be- words, could the lower callback rate we record
tween occupationearningsand the racial gap in for African-Americanresumes within our ex-
callback.Similarly,the industry-specificgaps in periment be consistent with a racially neutral
callback do not relate well to a measure of review of the entire pool of resumes the sur-
inter-industrywage differentials.In fact, while veyed employers receive?
the racial gap in callback rates varies somewhat In a racially neutralreview process, employ-
across occupations and industries, we cannot ers would rank order resumes based on their
reject the null hypothesis that the gap is the quality and call back all applicants that are
same across all these categories. above a certain threshold. Because names are
The last row of Table 7 focuses on the mar- randomized, the White and African-American
ginal effect of employer location on the racial resumes we send should rank similarly on av-
gap in callback. We use as a measure of erage. So, irrespective of the skill and racial
employer location the zip code of the company composition of the applicantpool, a race-blind
(or company branch) resumes were to be sent selection rule would generateequal treatmentof
to. More specifically, we ask whether differen- Whites and African-Americans.So our results
tial treatmentby race varies with the fractionof must imply that employers use race as a factor
African-Americansin the employer's zip code. when reviewing resumes, which matches the
We find a marginallysignificantpositive effect legal definition of discrimination.
of employer location on African-Americancall- But even rules where employers are not try-
backs but this effect is extremely small. In re- ing to interview as few African-Americanap-
gressions not reportedhere (but available from plicants as possible may generate observed
the authors upon request), we reestimate this differential treatment in our experiment. One
effect separatelyby city. While the point esti- such hiring rule would be employers trying to
mates are positive for both cities, the effect is interview a target level of African-American
only statistically significantfor Chicago. candidates. For example, perhaps the average
firm in our experiment aims to produce an in-
IV. Interpretation terview pool that matches the population base
rate. This rule could produce the observed dif-
Three main sets of questions arise when in- ferentialtreatmentif the averagefirmreceives a
terpretingthe results above. First, does a higher higherproportionof African-Americanresumes
callback rate for White applicants imply that than the population base rate because African-
employers are discriminatingagainst African- Americans disproportionatelyapply to the jobs
and industriesin our sample.43
42
For previous work on the effect of employer location
on labor market discrimination,see, for example, Steven 43 Anothervariantof this argumentis thatthe (up to) two
Raphael et al. (2000). African-Americanresumes we sent are enough to signifi-
VOL.94 NO. 4 BERTRANDAND MULLAINATHAN:
RACEIN THE LABORMARKET 1007
Some of our other findings may be consistent is unlikely that the differential treatment we
with such a rule. For example, the fact that observeis generatedby hiringrulessuch as these.
"EqualOpportunityEmployers"or federal con-
tractorsdo not appearto discriminateany less B. Potential Confounds
may reflect the fact that such employersreceive
more applicationsfrom African-Americans.On While the names we have used in this exper-
the otherhand,otherkey findingsruncounterto iment strongly signal racial origin, they may
this rule. As we discuss above, we find no also signal some other personaltrait.More spe-
systematic difference in the racial gap in call- cifically, one might be concerned that employ-
back across occupationalor industrycategories, ers are inferring social background from the
despite the large variation in the fraction of personal name. When employers read a name
African-Americanslooking for work in those like "Tyrone"or "Latoya,"they may assume
categories. African-Americansare underrepre- that the person comes from a disadvantaged
sented in managerialoccupations, for example. background.44In the extremeform of this social
If employers matchedbase rates in the popula- background interpretation,employers do not
tion, the few African-Americanswho apply to care at all aboutrace but are discriminatingonly
these jobs should receive a higher callback rate against the social backgroundconveyed by the
than Whites. Yet, we find that the racial gap in names we have chosen.45
managerialoccupationsis the same as in all the While plausible, we feel that some of our
otherjob categories. This rule also runs counter earlier results are hard to reconcile with this
to our findingson returnsto skill. Suppose firms interpretation.For example, in Table 6, we
are strugglingto find White applicantsbut over- found that while employers value "better"ad-
whelmed with African-American ones. Then dresses, African-Americansare not helped more
they should be less sensitive to the quality of than Whites by living in Whiter or more edu-
White applicants (as they are trying to fill in cated neighborhoods.If the African-American
their hiring quota for Whites) and much more names we have chosen mainly signal negative
sensitive to the quality of Black applicants social background,one might have expected the
(when they have so many to pick from). Thus, it estimated name gap to be lower for better ad-
dresses. Also, if the names mainly signal social
background,one might have expected the name
gap to be higher for jobs that rely more on soft
cantly distort the racial composition of the entire applicant skills or requiremore interpersonalinteractions.
pool. This is unlikely for two reasons. First, anecdotal
evidence and the empirically low callback rates we record We found no such evidence in Table 7.
suggest that firms typically receive many hundredsof re- We, however, directlyaddressthis alternative
sumes in response to each ad they post. Hence, the (up to) interpretationby examining the average social
four resumes we send out are unlikely to influencethe racial
backgroundof babies born with the names used
composition of the pool. Second, the similar racial gap in in the experiment.We were able to obtain birth
callback we observe across the two cities goes counter to
this interpretationsince the racial composition base rates certificatedata on mother's education(less than
differ quite a lot across these two cities. Anothervariantof high school, high school or more) for babies
this argumentis that, for some reason, the average firm in born in Massachusetts between 1970 and
our sample receives a lot of high-quality resumes from
African-Americanapplicantsand much fewer high-quality
resumes from White applicants.Hypothetically,this might
occur if high-quality African-Americansare much more
likely to use help-wanted ads ratherthan other job search 44Roland Fryer and Steven Levitt (2003) provide a re-
channels. If employers perform within-race comparisons cent analysis of social backgroundand naming conventions
and again want to target a certain racial mix in their inter- amongst African-Americans.
viewing and hiring, our African-Americanresumes may 45 African-Americansas a whole come from more dis-
naturallyreceive lower callbacksas they are competingwith advantagedbackgroundsthan Whites. For this social class
many more high-qualityapplicants.This specific argument effect to be something of independent interest, one must
would be especially relevant in a case where the average assert that African-Americanswith the African-American
sampled employer is "known" to be good to African- names we have selected are from a lower social background
Americans.But our selection procedurefor the employment than the average African-Americanand/orthat Whites with
ads did not allow for such screening:we simply responded the White names we have selected are from a higher social
to as many ads as possible in the targeted occupational backgroundthan the average White. We come back to this
categories. point below.
1008 THEAMERICANECONOMICREVIEW SEPTEMBER2004
TABLE8-CALLBACKRATEANDMOTHER'S BYFIRSTNAME
EDUCATION
Notes: This table reports, for each first name used in the experiment,callback rate and average mother education. Mother
educationfor a given first name is defined as the percent of babies born with that name in Massachusettsbetween 1970 and
1986 whose motherhad at least completed a high school degree (see text for details). Within each sex/race group, firstnames
are rankedby increasing callback rate. "Average"reports,within each race-gendergroup, the average mother educationfor
all the babies born with one of the names used in the experiment."Overall"reports,within each race-gendergroup, average
mothereducationfor all babies bornin Massachusettsbetween 1970 and 1986 in thatrace-gendergroup."Correlation"reports
the Spearmanrankordercorrelationbetween callbackrate and mothereducationwithin each race-gendergroup as well as the
p-value for the test of independence.
1986.46For each first name in our experiment, name and, in thatgender-racecell, whose moth-
we compute the fraction of babies with that ers have at least completed a high school
degree.
46 This
In Table 8, we display the average callback
longer time span (comparedto thatused to assess rate for each firstname along with this proxy for
name frequencies) was imposed on us for confidentiality
reasons. When fewer than 10 births with education data social background. Within each race-gender
available are recorded in a particulareducation-namecell, group, the names are rankedby increasingcall-
the exact numberof birthsin thatcell is not reportedand we back rate. Interestingly, there is significant
impute five births. Our results are not sensitive to this
imputation.One African-Americanfemale name (Latonya)
and two male names (Rasheed and Hakim) were imputedin
this way. One African-Americanmale name (Tremayne) tatively similar when we use a larger data set of California
had too few births with available education data and was births for the years 1989 to 2000 (kindly provided to us by
thereforedroppedfrom this analysis. Our results are quali- Steven Levitt).
VOL.94 NO. 4 BERTRANDAND MULLAINATHAN:
RACEIN THE LABORMARKET 1009
variationin callback rates by name. Of course, exact opposite. The p-values indicate that we
chance alone could produce such variationbe- cannot reject independence at standardsignifi-
cause of the rather small number of observa- cance levels except in the case of African-
tions in each cell (about 200 for the female Americanmales where we can almost rejectit at
names and 70 for the male names).47 the 10-percentlevel (p = 0.120). In summary,
The row labeled "Average"reportsthe aver- this test suggests little evidence that social back-
age fraction of mothers that have at least com- ground drives the measuredrace gap.
pleted high school for the set of names listed in Names might also influence our results
that gender-racegroup.The row labeled "Over- through familiarity. One could argue that the
all" reportsthe average fractionof mothersthat African-American names used in the experi-
have at least completed high school for the full ment simply appear odd to human resource
sample of births in that gender-racegroup. For managers and that any odd name is discrimi-
example, 83.9 percent of White female babies nated against. But as noted earlier, the names
born between 1970 and 1986 have motherswith we have selected are not particularlyuncommon
at least a high school degree;91.7 percentof the among African-Americans(see AppendixTable
White female babies with one of the names used Al). We have also performeda similarexercise
in the experimenthave mothers with at least a to that of Table 8 and measuredthe rank-order
high school degree. correlation between name-specific callback
Consistentwith a social backgroundinterpre- rates and name frequency within each gender-
tation, the African-Americannames we have race group. We found no systematic positive
chosen fall below the African-Americanaver- correlation.
age. For African-Americanmale names, how- There is one final potential confound to our
ever, the gap between the experimentalnames results. Perhapswhat appearsas a bias against
and the population average is negligible. For African-Americansis actually the result of re-
White names, both the male and female names verse discrimination. If qualified African-
are above the populationaverage. Americans are thought to be in high demand,
But, more interestinglyto us, thereis substan- then employers with averagequalityjobs might
tial between-nameheterogeneityin social back- feel that an equally talented African-American
ground.African-Americanbabies namedKenya would never accept an offer from them and
or Jamal are affiliated with much higher moth- thereby never call her or him in for an inter-
ers' education than African-American babies view. Such an argumentmight also explain why
named Latonya or Leroy. Conversely, White African-Americansdo not receive as strong a
babies named Carrieor Neil have lower social returnas Whites to betterresumes, since higher
background than those named Emily or qualification only strengthens this argument.
Geoffrey. This allows for a direct test of the But this interpretation would suggest that
social backgroundhypothesis within our sam- among the betterjobs, we ought to see evidence
ple: are names associated with a worse social of reverse discrimination,or at least a smaller
backgrounddiscriminatedagainst more? In the racial gap. However, as we discussed in Section
last row in each gender-racegroup, we report III, subsection D, we do not find any such
the rank-order correlation between callback evidence. The racial gap does not vary across
rates and mother's education. The social back- jobs with different skill requirements,nor does
ground hypothesis predicts a positive correla- it vary across occupation categories. Even
tion. Yet, for all four categories, we find the among the better jobs in our sample, we find
that employers significantly favor applicants
47 We
with White names.48
formally tested whetherthis variationwas signif-
icant by estimating a probit regression of the callback
dummy on all the personal first names, allowing for clus-
tering of the observationsat the employment-adlevel. For
all but African-Americanfemales, we cannot reject the null 48 One
might argue that employers who reverse-discrim-
hypothesis that all the first name effects in the same race- inate hire through less formal channels than help-wanted
gender group are the same. Of course, a lack of a rejection ads. But this would imply that African-Americansare less
does not mean thereis no underlyingpatternin the between- likely to findjobs throughformalchannels.The evidence on
name variationin callbacks that might have been detectable exit out of unemploymentdoes not paint a clear picture in
with larger sample sizes. this direction (Holzer, 1987).
1010 THEAMERICANECONOMICREVIEW SEPTEMBER2004
Notes: This table tabulatesthe differentfirst names used in the experimentand their identifiability.The first column reports
the likelihood that a baby born with that name (in Massachusettsbetween 1974 and 1979) is White (or African-American)
relative to the likelihood that it is African-American(White). The second column reportsthe probabilitythat the name was
picked as White (or African-American)in an independentfield survey of people. The last row for each group of names shows
the proportionof all births in that race group that these names account for.
Darity, William A., Jr. and Mason, Patrick L. ers in Britain."Race, November 1970, 11(4),
"Evidence on Discrimination in Employ- pp. 397-417.
ment: Codes of Color, Codes of Gender." Lundberg, Shelly J. and Starz, Richard. "Private
Journal of Economic Perspectives, Spring Discrimination and Social Intervention in
1998, 12(2), pp. 63-90. Competitive Labor Market."American Eco-
Fix, Michael and Turner, Margery A., eds. A nomic Review, June 1983, 73(3), pp. 340-47.
national report card on discrimination in McIntyre, Shelby J.; Moberg, Dennis J. and Pos-
America: The role of testing. Washington, ner, BarryZ. "Discriminationin Recruitment:
DC: Urban InstitutePress, 1998. An Empirical Analysis: Comment." Indus-
Fryer, Roland and Levitt, Steven. "The Causes trial and Labor Relations Review, July 1980,
and Consequences of Distinctively Black 33(4), pp. 543-47.
Names." Mimeo, University of Chicago, Mullainathan,Sendhil."ThinkingThroughCat-
2003. egories." Mimeo, MassachusettsInstitute of
Goldin, Claudia and Rouse, Cecilia. "Orchestrat- Technology, 2003.
ing Impartiality:The Impact of Blind Audi- Neumark,David. "Sex Discrimination in Res-
tions on Female Musicians." American taurantHiring: An Audit Study." Quarterly
Economic Review, September 2000, 90(4), Journal of Economics, August 1996, 111(3),
pp. 715-41. pp. 915-42.
Heckman,James J. "DetectingDiscrimination." Newman,Jerry M. "Discriminationin Recruit-
Journal of Economic Perspectives, Spring ment:An EmpiricalAnalysis."Industrialand
1998, 12(2), pp. 101-16. Labor Relations Review, October 1978,
Heckman, James J.; Lochner, Lance J., and Todd, 32(1), pp. 15-23.
PetraE. "FiftyYears of Mincer EarningsRe- Nisbett, Richard E. and Cohen, Dov. The culture
gressions." Mimeo, University of Chicago, of honor: The psychology of violence in the
2001. South. Boulder, CO: Westview Press, 1996.
Heckman, James J. and Siegelman, Peter. "The Phelps, Edmund S. "The Statistical Theory of
UrbanInstituteAudit Studies:TheirMethods Racism and Sexism." American Economic
and Findings,"in Michael Fix and Raymond Review, September1972, 62(4), pp. 659-61.
J. Struyk, eds., Clear and convincing evi- Raphael, Steven; Stoll, Michael A. and Holzer,
dence: Measurement of discrimination in Harry J. "Are SuburbanFirms More Likely
America. Lanham, MD: Urban Institute to Discriminateagainst African Americans?"
Press, 1992, pp. 187-258. Journal of Urban Economics, November
Holzer,HarryJ. "InformalJob SearchandBlack 2000, 48(3), pp. 485-508.
Youth Unemployment."AmericanEconomic Riach, Peter A. and Rich, Judity. "Testing for
Review, June 1987, 77(3), pp. 446-52. Racial Discriminationin the LabourMarket."
Hubbuck, Jim and Carter, Simon. Half a CambridgeJournal of Economics, September
chance? A report on job discrimination 1991, 15(3), pp. 239-56.
against young blacks in Nottingham.London: Turner, Margery A.; Fix, Michael and Struyk,
Commission for Racial Equality, 1980. RaymondJ. Opportunitiesdenied, opportuni-
James,Franklinand DelCastillo,SteveW. "Mea- ties diminished:Racial discriminationin hir-
suring Job Discrimination by Private Em- ing. Washington, DC: Urban InstitutePress,
ployers Against Young Black and Hispanic 1991.
Seeking Entry Level Work in the Denver Weichselbaumer, Doris. "Sexual Orientation
Metropolitan Area." Mimeo, University of Discriminationin Hiring." Labour Econom-
Colorado-Denver,1991. ics, December 2003, 10(6), pp. 629-42.
Jones, Melinda.Social psychology of prejudice. . "Is it Sex or Personality?The Impact
Saddle River, NJ: Pearson Education,2002. of Sex-Stereotypeson Discriminationin Ap-
Jowell, Roger and Prescott-Clark, Patricia. "Ra- plicant Selection." Eastern Economic Jour-
cial Discriminationand White-CollarWork- nal, Spring 2004, 30(2), pp. 159-86.