1
EXPERT REPORT OF WILLIAM T. BIELBY, Ph.D.
Cremin et al. v. Merrill Lynch, Pierce, Fenner & Smith, Inc.
QUALIFICATIONS, ASSIGNMENT, AND MATERIALS REVIEWED
I have been retained by the law firms of Stowell & Friedman, Ltd. and Meites,
Mulder, Burger & Mollica, class counsel for plaintiffs in Cremin et al. v. Merrill Lynch,
to review statistical reports concerning gender disparities in the careers of Financial
Consultants (FCs) and to ascertain whether features of the firm's personnel policies and
practices were likely to have contributed to the differences in the career trajectories of
men and women. I have examined the expert reports of Dr. Jerry Goldman and Dr. Janice
Madden and have reviewed materials pertaining to the company's employment policies,
focusing primarily on the period from 1991 (when the Professional Development
Program for broker trainees was first introduced) to 1998 (when a settlement was reached
in the Cremin et al. litigation).
I have testified as an expert witness in both California Superior Court and Federal
Court on cases involving workplace discrimination. A list of cases in which I have been
identified as an expert or have given expert testimony can be found in my Curriculum
Vitae, which is attached as Exhibit A.
2
I received a Ph.D. in Sociology from the University of Wisconsin--Madison in
1976. I also have a B.S. in Electrical Engineering and a M.A. in Social Sciences from the
University of Illinois. I am currently Professor of Sociology at the University of
California, Santa Barbara, where I chaired my department from 1992 to 1998. I am also
affiliated with UCSB's Department of Statistics and Applied Probability. Among my
former positions are Visiting Professor of Management at UCLA and Fellow at the
Center for Advanced Study in the Behavioral Sciences at Stanford. I teach graduate and
undergraduate courses on organizational behavior, research methods for the social
sciences, labor markets, quantitative methods, and social inequality. I also specialize in
research in each of these areas. Over the past fifteen years, much of my research has
focused on issues of workplace discrimination, and on organizational policies and
practices more generally. My research on these topics has been supported by four grants
from the National Science Foundation, and it has been published in leading peer-reviewed
social science research journals.
I have received national awards from three different professional associations for
my research on gender, labor markets, social psychological aspects of work, and
organizational personnel practices. I have served on numerous panels, advisory
committees, and professional workshops on topics relating to workplace discrimination,
organizational personnel policies and practices, and research methodology. I have served
as an advisor, consultant, or reviewer for the following organizations and agencies: the
U.S. Bureau of the Census, the U.S. Department of Justice, the U.S. Department of
Labor, the National Science Foundation, the National Institute of Mental Health, the
National Academy of Sciences, the University of Michigan's Institute for Social Research,
3
Stanford University, and the Writers' Guild of America, West. I have also served on the
editorial boards of leading social science journals, and I regularly review manuscripts for
scientific journals on topics relating to organizational behavior, employment
discrimination, gender and work, and research methodology. I have been elected to
several offices in the American Sociological Association, including its presidency. I
served for three years on the ASA Council, the organization's governing body, and my
term as President of the Association begins in August, 2002.
In addition to the statistical reports of Dr. Goldman and Dr. Madden, I have
reviewed deposition testimony of Jerome A. Rosenthal, Robert Bowman, Deborah
Aldredge, Andrew Williams, and Linda J. Pittari and the exhibits accompanying their
depositions. Mr. Rosenthal, Director of Private Client Compensation from 1996 to 1999,
testified about compensation of Financial Consultants. Mr. Bowman, National Sales
Manger since May, 1999 and formerly a Regional Vice President, testified about branch
office policies and practices affecting the hiring, training, working conditions, and
compensation of Financial Consultants. Ms. Aldredge, Manager of the Learning and
Performance Group, has had responsibility for training of broker trainees since 1989 and
testified about policies and practices in that area. Mr. Williams, who testified about
training for advanced Financial Consultants, has had responsibility in that area since
1996. Ms. Pittari, who heads the Management Resource & Development group, testified
about management selection and training, based on her two decades of experience in that
area at Merrill Lynch.
Besides deposition exhibits, documents made available to me include: various
employee handbooks, brochures, and newsletters; documents describing EEO policies;
4
documents describing compensation for Financial Consultants, FC trainees, and
managers; documents relating to the litigation and settlement of O'Bannon, et al. and
EEOC v. Merrill Lynch; files of brokers who were evaluated for management positions at
the firm's Management Assessment Center, and files relating to Merrill Lynch special
incentive programs, local branch office documents dealing with policies and practice
related to office assignment, Professional Development Program training, account and
referral distribution and other areas under local management control. I have also
reviewed individual class members' complaints.
SUMMARY OF FINDINGS
To determine what factors explain disparities in the careers of men and women at
Merrill Lynch, I have reviewed the reports of the plaintiffs' statistical experts as well as
testimony and documents describing the firm's personnel policies and practices. I have
concluded that the gender segregation and compensation disparities documented in Dr.
Goldman's report cannot be explained by differences between men and women in the
skills, qualifications, and background required to be a successful Financial Consultant or
manager. I have concluded that the gender disparities in career trajectories at Merrill
Lynch are generated and sustained by a personnel system in which decisions that affect
the careers of Financial Consultants and Managers are based substantially on arbitrary
and subjective criteria, with little systematic oversight or accountability. I have also
concluded that the personnel policies and practices of the firm comprise a "successbreeds-success" or "cumulative advantage" system in which decisions that adversely
5
affect women at the start of their careers in the firm soon lead to substantial disparities
that persist over time. The basis for these conclusions is explained below.
OVERVIEW OF GENDER DISPARITIES IN THE CAREERS OF FINANCIAL
CONSULTANTS AT MERRILL LYNCH
Earnings Disparities
Dr. Goldman's report shows that gender disparities in compensation of Financial
Consultants emerge within two years of entering the Professional Development Program
(PDP) for broker trainees, widen during the early years of employment as an FC, and
persist throughout FCs' careers (see, e.g., Goldman report, Graphs 13 through 22, 24, 25,
28, and 31 and supporting tables). The gender disparity in median earnings is as wide as
15% to 20% within three years of entering the PDP, and regression results show that
experienced female financial consultants consistently earn approximately 13% to 18%
less than their male counterparts who have been with the firm for the same number of
years.
Gender Segregation in Career Paths
As of 1995, among the ranks of Financial Consultants at Merrill Lynch, men
outnumbered women by a ratio of more than six to one, and even among recently hired
trainees, the ratio exceeded five to one (Goldman report, Table 1). Women who enter the
PDP training program are less likely than their male counterparts to successfully complete
the program, and they are more likely than men to move instead into female-dominated
support positions such as Client Associate (Goldman report, Tables 3 through 7 and
Tables 11 and 12). Conversely, men in the female-dominated Registered Client
6
Associate position are more likely than women in that position to experience career
mobility into the position of Financial Consultant (Goldman reports, Tables 8, 9, and 10).
Furthermore, women FCs are much less likely than men to advance into branch-level and
higher management (Goldman report, Tables 1 and 2).
Institutionalized Patterns of Gender Segregation
Gender segregation in the career paths of FC trainees and Registered Client
Associates at Merrill Lynch takes place within an overall context of a highly segregated
workforce in the firm. As can be seen from Tables 1 and 2 of Dr. Goldman's report,
compared to the jobs dominated by males, those in which most of the firm's women are
employed have lower earnings and more limited opportunities for advancement. In effect,
the firm has distinct workforces. Women work mostly in lower-paid administrative and
support positions, while men are concentrated in highly-paid sales and management
positions.
Social scientists measure the overall degree to which men and women are
segregated into different job categories with a segregation index, also known as the
"index of dissimilarity."1 The segregation index ranges from 0 (when the ratio of men to
women is the same in every job category) to 100 (when every job category contains only
men or only women).2 The index computed for detailed Census occupational categories
for the entire U.S. labor force allows social scientists to assess the degree to which overall
sex segregation in the labor market has declined. For example, from 1970 to the early
1O. D. Duncan and B. Duncan, "A Methodological Analysis of Segregation Indices," American
Sociological Review, Vol. 20, 1955, p. 200-217.
2The segregation index can be interpreted as the percentage of individuals of either sex who would
have to change job classifications in order for the ratio of women to men to be equal in each category.
7
1990s, the segregation index for the entire U.S. workforce declined by 13 points (a 19%
decrease), from 68 to 55.3 In contrast, the segregation index for Merrill Lynch, computed
across the 8 job groupings listed in Table 1 of Dr. Goldman's report, is 68.7. In other
words, the Merrill Lynch labor force was about as segregated in 1995 as was the entire
U.S. workforce a quarter of a century earlier.4
Census data for the U.S. workforce as a whole show that in 1990, women
comprised 27.7% of those employed in the category "securities and financial services
sales," almost exactly double the representation of women in the FC category at Merrill
Lynch (13.8%) five years later in 1995 (U.S. Census EEO File for 1990 and Goldman
report, Table 1). In fact, the Merrill Lynch FC workforce was more heavily male
dominated in 1995 than was the nationwide security sales workforce fifteen years earlier
in 1980, when women comprised 18.6% of those employed in the securities and financial
services sales category (U.S. Census EEO File for 1980). In short, the segregated
workforce evident at Merrill Lynch in 1995 appears to reflect a longstanding, highly
entrenched pattern whereby women and men have rarely been employed alongside one
another in the same job titles, with equal access to career advancement.
3D. A. Macpherson and B. T. Hirsch, "Wages and Gender Composition: Why Do Women's Jobs
Pay Less?" Journal of Labor Economics, Vol. 13, 1995, p. 426-471 (Table 1).
4 See Barbara F. Reskin and Heidi I. Hartmann, Women's Work, Men's Work: Sex Segregation on
the Job, National Academy Press, 1986, Table 2-4. The segregation index depends on the level of detail in
the categorization of jobs or occupations. More detailed job classifications yield larger segregation indices
(so, for example, an index computed across EEO-1 categories will generally be lower than one computed
across more detailed occupational titles). However, segregation at Merrill Lynch is consistent across job
categories, and thus aggregation has little impact on the index. For example, grouping together FCs and
trainees (rows labeled "Financial Consultant/FCIC" and "PDP/AFC" in Table 1 of Dr. Goldman's report)
and grouping together Client Associates and Registered Client Associates does not change the value of the
segregation index. Including the Associate Manager category (Goldman report, Table 2) in the calculation
increases the segregation index to 69.8.
8
Evidence of persistent segregation over time at Merrill Lynch appears in
documents related to the O'Bannon, et al. and EEOC v. Merrill Lynch litigation. Among
the issues involved in that litigation were barriers faced by women in moving from
Account Representative and Registered Sales Assistant Position into Account Executive
jobs (comparable to the Financial Consultant position). Under the terms of the original
consent decree covering the period from 1976 to 1980, hiring goals for Account
Executive trainee position were established starting at 10% for 1976 and increasing two
percent per year to 16% in 1980. Subsequent amendments to the consent decree
increased the goal to 25% for the years 1983 through 1995. Despite the 25% goal, the
firm's actual hiring over this period never exceeded 19%, and from 1990 (when
nationwide women represented nearly 28% of those in the occupation of securities and
financial services sales) to 1998 women accounted from approximately 14% to 17% of
the trainees hired annually by Merrill Lynch (see Figure 1).5 In Figure 1, I report the
gender composition of newly hired broker trainees at Merrill Lynch over the period from
1983 to 1998, and compare it with data on the gender composition of the occupation of
securities sales in the U.S. labor force in the 1980s and 1990s. As can be seen in Figure
1, the rate at which women were being hired as broker trainees at Merrill Lynch remained
virtually unchanged during an era when women's representation among securities
salespeople in the U.S. labor force increased dramatically, from about 20% in the early
1980s to over 30% in the early 1990s.6
5In contrast, at Smith Barney, women comprised 35% of broker trainees by 1999, approximately
double the level of women's representation among FC trainees at Merrill Lynch in the late 1990s.
6Data on Merrill Lynch broker trainees is from a document provided to me by plaintiff's counsel
titled "Employment Goals for Account Executive Trainees Pursuant to Consent Decree in EEOC v. Merrill,
9
Women's exclusion from management roles at Merrill Lynch is also part of a
longstanding pattern. Prior to 1976, no woman had ever attended the firm's Management
Assessment Center. From 1976 to June, 1999, only 33 women have successfully
completed the firm's training program for managers, and among that group, only eleven
had reached the level of Regional Vice President and one the level of District Director.7
Women comprised just eight percent of the individuals evaluated by the Management
Assessment Center from 1990 to 1999, and their representation in management remains
low, despite performing as well or better than men when they do have the opportunity to
be evaluated at the Center.8
In sum, the skewed gender ratio favoring males in the snapshot of Merrill Lynch's
Financial Consultant and management workforce as of 1995, documented in Dr.
Goldman's report, is part of an overall trend that has barely changed over two decades,
even though opportunities for women in securities sales were expanding overall in the
U.S. labor force.
Lynch, et al." Data for the U.S. Labor Force are from the Current Population Survey "merged outgoing
rotation group files" for the years 1979 through 1999, distributed by the National Bureau for Economic
Research. For years prior to 1983, data pertain to the 3-digit occupational classification "stock and bond
salesmen." For 1983 and later, data pertain to the 3-digit occupational classification "securities and
financial services sales occupations." Reported in Figure 1 is a three-year moving average of the percent
women in the relevant occupational category (i.e. the statistic for 1980 is an average of the relevant
percentages for 1979, 1980, and 1981). Sample sizes for individual years range from 241 to 726.
7Pittari, Exh. 16, "Merrill Lynch Assessment Center: Women at Assessment Centers," p. 2; Pittari
depo., p. 225.
8Merrill Lynch provided plaintiffs' counsel with Management Assessment Center files for 359
individuals assessed from 1990 to 1999, and 29 of those individuals were women. Women who proceed
through MAC assessment have a higher success rate than men. According to those files, the pass rate for
men over this period was 38%, compared to 48% for women. Because the rate for women is based on such
a small number of cases, the differential in the success rate is not statistically significant (also see Pittari
depo., p. 224.
10
DIFFERENCES IN SKILLS, QUALIFICATIONS, AND PRODUCTIVITY DO
NOT EXPLAIN GENDER DIFFERENCES IN CAREER TRAJECTORIES AT
MERRILL LYNCH
The Skills Required To Be a Successful Broker Do Not Differ by Gender
What explains the low representation of women among Financial Consultants at
Merrill Lynch and the earnings disparity by gender in the firm's FC workforce? One
possible explanation is that women are less likely than men to posses the skills and
qualifications required to be a successful broker. However, deposition testimony of the
firm's managers, written policies documenting the qualities sought in successful brokers
and managers, and the deposition testimony of an expert relied upon by Merrill Lynch for
developing selection instruments indicate that men and women are equally likely to have
the background, skills, and experience necessary to succeed as a Financial Consultant.
National Sales Manager Robert Bowman testified that there is no particular type
of education, degree, background, or prior job experience that he looks for in identifying
candidates who will be successful brokers.9 The firm's Professional Development
Program, introduced in 1991, is predicated on the assumption that regardless of
background, all new hires acquire necessary skills and knowledge through the twenty-four
month training program.10
9Bowman depo., p. 65-69.
10Aldredge depo., p. 24-25, 30-31.
11
Merrill Lynch initiated its Associate Financial Consultant Program to "develop
recent college graduates into successful Financial Consultants."11 Candidate
qualifications listed in AFC documents include: "work experience in finance or sales,
and/or interest in Financial Services;" "enthusiasm and entrepreneurial spirit;" "a balance
of academic and extracurricular achievement and employment-related accomplishments;"
and "performance-driven personality traits;" and "ability to overcome obstacles and
handle rejection."12 I am not aware of any research that suggests these qualities are
substantially more prevalent in men than in women (and certainly not in a ratio of five to
one).13
The Skills Required To Be a Successful Manager Do Not Differ by Gender
In its "Guide to Identifying & Developing Management Talent," the firm lists the
following as "dimensions critical to managerial success:" integrity, achievement
motivation, tolerance for stress, individual leadership, flexibility, sensitivity, visionary
leadership, coaching, initiative, applied learning, problem solving, external awareness,
planning and organizing, delegation, oral communication, and oral presentation.14
Again, these are not qualities that differ substantially between men and women.
11 "Associate Financial Consultant (AFC) Program," July, 1991, Bates D 0002543.
12"Associate F.C.", Bates D 0004138; "Consumer Markets Associate Financial Consultant (AFC)
Program," Bates D 0004195; "Associate Financial Consultant Program: Description for Management,"
Bates D 0002527.
13I have reviewed the social science literature on gender similarities and differences in work effort
and competitiveness and have concluded that the differences on these traits between men and women in
similar job situations are negligible. See W. T. Bielby and D. D. Bielby, "Telling Stories about Gender and
Effort: Social Science Narratives About Who Works Hard for the Money." Forthcoming in M. F. Guillen,
R. Collins, P. England, and M. Meyer (eds.), The New Economic Sociology: Developments in an Emerging
Field, Russell Sage Foundation, 2002.
14Bates D 0003082.
12
If men, on average, possess better management skills than women, we would
expect men to pass Management Assessment Center evaluations at a higher rate than do
women. However, this is not the case. Although women are largely absent in
management positions at the firm, those women who are able to attend the Management
Assessment Center have a higher first-time pass rate than men, 56% compared to 41%
overall, according to the deposition testimony of Ms. Pittari, and 48% compared to 38%
according to the Management Assessment Center files provided to plaintiffs' counsel.15
In short, it is implausible that gender differences in skills explains women's low
representation in management at Merrill Lynch.
Gender Differences in Interest in Management Careers Do Not Explain Women's Very
Low Representation Among Managers at Merrill Lynch
Another possible explanation for the low representation of women among
managers is that women FCs at Merrill Lynch are not interested in management careers in
the firm. However, prior to the settlement of the Cremin litigation, there was no objective
and systematic way for women to express their interest in management positions. But in
1999, when an invitation to the Opportunities in Leadership Seminar was sent to all FCs
who met minimum eligibility requirements, women responded at a rate that was roughly
double their representation in previous seminars.16 In short, when given an opportunity,
women expressed interest in management careers in numbers that greatly exceeded their
15Pittari depo., p. 224. Indeed the gender disparity in pass rates in favor of women may even be
greater than this. Pittari Exh. 3 indicates an overall pass rate of just 21% for individuals who attended the
Opportunities in Leadership Seminar.
16Women and minorities accounted for 21% of the 330 individuals who attended the 1999
seminar, whereas in previous years, there were typically two to twelve women and minorities in seminars
13
representation in the firm's management positions and that exceed their representation
among FCs.
Gender Differences in Skills, Productivity, and Office Location Do Not Explain the
Gender Gap In Compensation Among Financial Consultants
With a few exceptions, compensation for Financial Consultants is determined
largely on the commissions they receive on trades from their retail accounts, known as
their "production." The percentage payout on production is determined by a combination
of the broker's "length of service" (LOS) (up to five years) and production. The formulas
relating the commission percentage or payout to production is referred to in the firm as
"the grid."17 If production and LOS completely account for commissions, how can
gender influence FCs' compensation?18 As Mr. Rosenthal testified, "the grid is the
grid."19 Gender disparities in the earnings of FCs at the same LOS can be attributed to
factors that influence a broker's ability to generate production.
Thus, one possible explanation for the earnings disparity among FCs is that
women are less effective in managing their clients' assets than are men. However, this
seems unlikely, given the evidence summarized above regarding the traits that are
associated with being a successful Financial Consultant. Moreover, Dr. Madden's report
includes quantitative analyses that address the issue of whether women are less effective
that ranged in size from 50 to 80 candidates (Pittari depo., p. 109-114, p. 239-241). Also see "Merrill
broker Turnout Soars for Mngt Intro Session," Wall Street Letter, November 8, 1999.
17Rosenthal depo., p. 97-100, 106-127.
18Factors besides payout from the grid that influence a Financial Consultant's annual
compensation include, among others, discount sharing, Financial Foundation cash payout, the "small ticket"
policy, Asset Gathering awards, Wealth-Builder Program awards, and credit management compensation.
The full set of factors that comprise the compensation program for Financial Consultants is specified in the
Financial Consultant Compensation Plan, which is published annually.
14
than men in generating production from assets. Dr. Madden analyzed production
generated from assets that had been transferred from house accounts and from accounts of
brokers who had left the firm, and she compared the level of production from those assets
prior to the transfer (i.e. attributable to the efforts of a different broker) to the level of
production generated after the transfer. Dr. Madden found that when given equivalent
assets to manage, women FCs and FC trainees were as productive as their male
counterparts (Madden report, Tables 3, 6, and 8).
Another possible explanation for the lower compensation of female FCs is that
they are less likely to work in offices where there are substantial opportunities to generate
large commissions. Women may be disproportionately located in offices where neither
male nor female FCs are generating substantial production. Dr. Goldman tested this
hypothesis by comparing the distribution of FCs by gender between "high-earner" offices
that have at least one broker who generates more than a million dollars in production
credits, and those offices in which no broker generates production in excess of that
amount. Dr. Goldman found that women were not disproportionately located outside of
high-earner offices, but that women were less likely than men to be among the most
highly compensated brokers in the high-earner offices. In short, the gender disparity in
compensation cannot be attributed to differences in the distribution of men and women
across Merrill Lynch offices.
The materials I have reviewed indicate that differences in background, skills, and
productivity explain neither the pattern of gender segregation in the Merrill Lynch
workforce nor gender disparities in the compensation of Financial Consultants and FC
19Rosenthal depo., p. 144.
15
trainees. Below, I look instead at whether aspects of the firm's policies and practices may
create career barriers to women at Merrill Lynch. Key factors to examine are the exercise
of discretion and the reliance on arbitrary and subjective criteria for making personnel
decisions and allocating resources, the overall culture and environment of the firm, and
the effectiveness of the firm's policies and practices for monitoring equal employment
opportunity and for holding managers accountable for the criteria and procedures they use
in making decisions that affect their employees. Before examining the specific policies
and practices of Merrill Lynch I first summarize findings from social research on the
kinds of factors that create barriers to career advancement for women as well as those that
minimize bias.
FACTORS THAT CREATE AND MINIMIZE WORKPLACE GENDER BIAS:
FINDINGS FROM SOCIAL SCIENCE RESEARCH
External Validity of Social Science Research
Social research conducted across many decades has taught us much about what
generates and sustains workplace inequalities. That same research, either directly or by
implication, points to the kinds of workplace policies and practices that are likely to
minimize bias. The relevant research has applied multiple methodologies in a variety of
contexts, including experiments in controlled laboratory settings; ethnographies and case
studies in "real world" organizations both large and small, public and private, and in a
range of industries; surveys done with representative samples of workers and employers;
and historical studies based on archival materials from the United States and abroad.
Thus, the scientific evidence about gender bias, stereotypes, and the structure and
16
dynamics of gender inequality in organizations that I rely upon has substantial external
validity and provides a sound basis for analyzing the policies and practices of Merrill
Lynch. My method is to look at distinctive features of Merrill Lynch policies and
practices and to evaluate them against what social science research shows to be factors
that create and sustain bias and those that minimize bias. In litigation contexts, this
method of analysis is known as "social framework analysis."20
In this section of my report I summarize the scientific literature upon which my
opinions are based. In footnotes, I provide citations to sources in peer refereed journals,
in important books and edited volumes in relevant fields of social science research, and,
whenever possible, to review articles by leading experts who summarize the findings of
social science research on gender bias in organizations, stereotypes, and related topics.
Sources of Workplace Gender Bias
Depending on the job, organizational setting, and work environment, there are
many reasons why men and women can have different career trajectories. For example,
jobs may have job-related skill and experience requirements that differ, on average,
between men and women. Gender disparities arising from such factors would not be
considered discriminatory, so long as the employer is not responsible for differences in
men's and women's qualifications (e.g., by not providing equal access to training).
Conversely, employers create gender barriers when they make decisions about
individuals' suitability for jobs, training, and support or their compensation based on
beliefs about a person's gender rather than on his or her actual qualifications. Employers
20See J. Monahan and L. Walker, Social Science in Law: Cases and Materials, Fourth Edition,
Foundation Press, 1998, Chapter Five, "Social Science Used to Provide Context."
17
also create gender barriers when they ignore (or encourage) an organizational climate that
is hostile towards women and inhibits them from performing to their full potential.
Sometimes, practices that appear to be gender-neutral have the effect of denying to
women the same opportunities that are available to men. For example, using employee
referrals as a recruitment mechanism is likely to reinforce a workforce's existing gender
composition.21
One way gender bias affects career outcomes is when stereotypes are allowed to
affect personnel decisions. Gender stereotypes are beliefs about traits and behaviors that
differ between men and women.22 For example, men are believed to be competitive,
aggressive, assertive, strong, and independent, while women are thought to be nurturing,
cooperative, supportive, and understanding. Men are assumed to place a high priority on
their careers, while women are assumed to be more strongly oriented towards family,
even though research demonstrates that the commitments of men and women with similar
job opportunities and family situations are virtually identical.23 When women perform
successfully in male-dominated contexts, their accomplishments are more likely to be
21For a review of relevant research, see P. V. Marsden and E. Gorman, "Social Networks, Job
Changes, and Recruitment," p. 467-502 in Sourcebook on Labor Markets: Evolving Structures and
Processes, edited by I. Berg and A. L. Kalleberg, Plenum Publishing, 2001.
22K. Deaux and L. Lewis, "Components of Gender Stereotypes," Psychological Documents, Vol.
13, 1983, p. 25-34; K. Deaux and L. Lewis, "The Structure of Gender Stereotypes: Interrelationships
Among Components and Gender Label," Journal of Personality and Social Psychology, Vol. 46, 1984, p.
991-1004; J. E. Williams and D. L. Best, "Sex Stereotypes and Intergroup Relations," p. 244-259 in
Psychology of Intergroup Relations, edited by S. Worchel and W. G. Austin, Nelson-Hall, 1986; S. T.
Fiske, "Stereotyping, Prejudice, and Discrimination," p. 357-411 in The Handbook of Social Psychology,
Vol. 2 (Fourth Edition), edited by D. T. Gilbert, S. T. Fiske, and G. Lindzey, McGraw-Hill, 1998.
23For a review see D. D. Bielby, "Commitment to Work and Family," Annual Review of
Sociology, Vol. 18, 1992, p. 281-302. Also see P. V. Marsden, A. L. Kalleberg, and C. R. Cook, "Gender
Differences in Organizational Commitment: Influences of Work Positions and Family Roles," Work and
Occupations, Vol. 20, 1993, p. 367-390.
18
attributed to luck, help from others, or special circumstances than to their ability, whereas
comparable performance by men is more likely to be attributed to their superior skills.24
Moreover, stereotypical behaviors that are believed to be typical of men are often viewed
as inappropriate for women. For example, it is less acceptable for a married woman with
young children to place a high priority on her career than it is for a married man.
Similarly, a woman who behaves in an assertive and independent manner often elicits
disapproval from those around her.25
Because of gender stereotypes, individuals tend to ascribe "masculine" traits to
men and "feminine" traits to women, and individuals tend to assume that the prevalence
of "masculine" traits among women and "feminine" traits among men is rare. A large
body of research demonstrates that the tendency to invoke gender stereotypes in making
24For a review of relevant research, see J. K. Swim and L. J. Sanna, "He's Skilled, She's Lucky: A
Meta-Analysis of Observers' Attributes for Women's and Men's Successes and Failures," Personality and
Social Psychology Bulletin, Vol. 22, 1996, p. 507-519. For studies of attribution bias in employment
contexts, see J. H. Greenhaus and S. Parasuraman, “Job Performance Attributions and Career Advancement
Prospects: An Examination of Gender and Race Effects,” Organizational Behavior and Human Decision
Processes, Vol. 55, 1991, p. 273-297; M. Igbaria and J. Baroudi, “The Impact of Job Performance
Evaluations on Career Advancement Prospects: An Examination of Gender Differences in the IS
Workplace, Management Information Systems Quarterly, Vol. 19, 1995, p. 107-123. Greenhaus and
Parasuraman studied managers in three companies in the communications, banking and electronics
industries, while Igbaria and Baroudi studied professionals in the field of information systems and data
processing. For reviews of social psychological research on attribution biases, see Fiske and S. E. Taylor,
Social Cognition, Second Edition, McGraw-Hill, 1991, Chapter 3; M. Hewstone, “The ‘Ultimate
Attribution Error’? A Review of the Literature on Intergroup Causal Attribution,” European Journal of
Social Psychology, Vol. 20, 1990, p. 311-335.
25T. W. Smith, "Working Wives and Women's Rights: The Connection Between the Employment
Status of Wives and the Feminist Attitudes of Husbands," Sex Roles, Vol. 12, 1985, p. 501-508; J. S.
Bridges and A. M. Orza, "The Effects of Employment Role and Motive for Employment on the Perceptions
of Mothers," Sex Roles, Vol. 27, 1992, p. 331-343; A. H. Eagly, M. G. Makhijani, and B. G. Klonsky,
"Gender and the Evaluation of Leaders: A Meta-Analysis," Psychological Bulletin, Vol. 111, 1992, p. 322; M. J. Martinkio and W. L. Gardner, "A Methodological Review of Sex-Related Access Discrimination
Problems," Sex Roles, Vol. 9, 1983, p. 825-839; C. L. Copeland, J. E. Driskell, and E. Salas, "Gender and
Reactions to Dominance," Journal of Social Behavior and Personality, Vol. 10, No. 6, 1995, p. 53-68; A.
M. Konrad and K. Cannings, "The Effects of Gender Role Congruence and Statistical Discrimination on
Managerial Advancement," Human Relations, Vol. 50, 1997, p. 1305-1328.
19
judgements about people is rapid and automatic.26 As a result, people are often unaware
of how stereotypes affect their perceptions and behavior, and individuals whose personal
beliefs are relatively free of prejudice or bias are susceptible to stereotypes in the same
ways as people who hold a personal animosity towards a social group.27
In the employment context, career barriers resulting from gender stereotypes and
gender bias are likely to be consequential for women working in a traditionally male
domains, such as the middle to upper managerial and professional ranks of large
corporations, engineering divisions of firms, in the military, and in historically maledominated industries such as skilled crafts and construction trades.28 A large body of
26See, for example, M. A. Zarate and Eliot R. Smith, "Person Categorization and Stereotyping,"
Social Cognition, Vol. 8, 1990, p. 161-185; M. R. Banaji, C. D. Hardin, and A. J. Rothman, "Implicit
Stereotyping in Person Judgment," Journal of Personality and Social Psychology, Vol. 65, 1993, p. 272281; M. A. Zarate and P. Sandoval, "The Effects of Contextual Cues on Making Occupational and Gender
Categorizations," British Journal of Social Psychology, Vol. 34, 1995, p. 353-362; M. R. Banaji and C. D.
Hardin, "Automatic Stereotyping," Psychological Science, Vol. 7, 1996, p. 136-141. For a review, see J. A.
Bargh and T. L. Chartrand, "The Unbearable Automaticity of Being," American Psychologist, Vol. 54,
1999, p. 462-479.
27S. T. Fiske and S. E. Taylor, op cit., p. 271-294; R. Brown, Prejudice, Blackwell, 1995; G. V.
Bodenhausen and C. N. Macrae, "The Self Regulation of Intergroup Perception: Mechanisms and
Consequences of Stereotype Suppression," p. 227-253 in Stereotypes and Stereotyping, edited by C. N.
Macrae, C. Stangor, and M. Hewstone, Guilford Press, 1996; S. T. Fiske, "Stereotyping, Prejudice and
Discrimination," p. 357-411 in Handbook of Social Psychology, edited by D. T. Gilbert, S.T. Fiske, and G.
Lindzey, McGraw-Hill, 1998; G. V. Bodenhausen, , C. N. Macrae, and J. Garst, “Stereotypes in Thought
and Deed: Social Cognitive Origins of Intergroup Discrimination,” p. 311-335 in Intergroup Cognition and
Intergroup Behaviors, edited by C. Sedikides, J. Schopler, and C. A. Insko, Erlbaum, 1998.
28W. T. Bielby and J. N. Baron, "Men and Women at Work: Sex Segregation and Statistical
Discrimination," American Journal of Sociology Vol. 91, 1986, p. 759-99; K. Deaux and J. C. Ullman,
Women of Steel: Female Blue-Collar Workers in the Basic Steel Industry, Praeger, 1983; R. M. Kanter,
Men and Women of the Corporation, Basic Books, 1977; Eagly et al., op cit.; M. J. Martinkio and W. L.
Gardner, "A Methodological Review of Sex-Related Access Discrimination Problems," Sex Roles, Vol. 9,
1983, p. 825-839; C. M. Jagacinski, "Engineering Careers: Women in a Male-Dominated Field,"
Psychology of Women Quarterly, Vol. 11, 1987, p. 97-110; P. Glick, C. Zion, and C. Nelson, "What
Mediates Sex Discrimination in Hiring Decisions?" Journal of Personality and Social Psychology, Vol. 55,
1988, p. 178-186; I. Padavic, "The Re-Creation of Gender in a Male Workplace," Symbolic Interaction,
Vol. 14, 1991, p. 279-294; J. S. McIlwee and J. Gregg Robinson, Women in Engineering. Gender, Power,
and Workplace Culture, State University of New York Press, 1992; J. Mettrick and G. Cowan, "Gender
Stereotypes and Predictions of Performance: Women in Air Combat," Journal of Social Behavior and
Personality, Vol. 11, 1996, p. 105-120; J. H. Stiehm, Arms and the Enlisted Woman, Temple University
20
social science research demonstrates that stereotypes are especially likely to influence
personnel decisions when they are based on informal, arbitrary, and subjective factors.29
In such settings, stereotypes can bias assessments of a woman's qualifications,
contributions, and advancement potential, because perceptions are shaped by stereotypical
beliefs about women generally, not by the actual skills and accomplishments of the
person as an individual.30 In decision-making contexts characterized by arbitrary and
Press, 1989; R. Hertz, "Guarding Against Women?: Responses of Military Men and Their Wives to Gender
Integration," Journal of Contemporary Ethnography, Vol. 25, 1996, p. 251-284. A job category is
considered male-dominated when men comprise at least eighty to eighty-five percent of the incumbents.
For additional research on the bias women face in such settings, conducted in a variety of contexts, see: E.
Spangler, M. A. Gordon, and R. Pipkin, "Token Women: An Empirical Test of the Kanter Hypothesis,"
American Journal of Sociology, Vol. 84, 1978, p. 160-170; M. E. Heilman, "The Impact of Situational
Factors on Personnel Decisions Concerning Women: Varying the Sex Composition of the Applicant Pool,"
Organizational Behavior and Human Performance, Vol. 26, 1980, p. 386-395; J. Crocker and K. M.
McGraw, "What's Good for the Goose is not Good for the Gander: Solo Status as an Obstacle to
Occupational Achievement for Males and Females," American Behavioral Scientist, Vol. 27, 1982, p. 357369; P. R. Sackett, C. L. DuBois, and A. W. Noe, "Tokenism in Performance Evaluations: The Effects of
Work Group Representation on Male-Female and Black-White Differences in Performance Evaluations,"
Journal of Applied Psychology, Vol. 76, 1991, p. 263-267; J. Rosenberg, H. Perlstadt, and W. R. Phillips,
"Now That We Are Here: Discrimination, Disparagement, and Harassment at Work and the Experiences of
Women Lawyers," Gender & Society, Vol. 7, 1993, p. 415-433; and J. D. Yoder, "Looking Beyond
Numbers: The Effects of Gender Status, Job Prestige, and Occupational Gender-Typing on Tokenism
Processes," Social Psychology Quarterly, Vol. 57, 1994, p. 150-159.
29For a review, see American Psychological Association, "In the Supreme Court of the United
States: Price Waterhouse v. Ann B. Hopkins: Amicus Curiae Brief for The American Psychological
Association," American Psychologist, Vol. 46, 1991, p. 1061-1070. Also see V. F. Nieva and B. A. Gutek,
"Sex Effects on Evaluation," Academy of Management Review, Vol. 5, 1980, p. 267-275 (especially pages
270-274); D. N. Bersoff, "In the Supreme Court of the United States: Clara Watson v. Fort Worth Bank &
Trust, Amicus Curiae Brief for the American Psychological Association" (reprinted in American
Psychologist, Vol. 43, 1988, p. 1019-1028); Messick and Mackie, "Intergroup Relations," Annual Review
of Psychology, Vol. 40, 1989, p. 49-50; Fiske and Taylor, op cit. (especially Chapter 9, "Social Inference");
L. H. Krieger, "The Contents of our Categories: A Cognitive Bias Approach to Discrimination and Equal
Employment Opportunity," Stanford Law Review, Vol. 47, 1995, p. 1161-248; B. F. Reskin, The Realities
of Affirmative Action in Employment, American Sociological Association, 1998; W. T. Bielby, "Minimizing
Workplace Gender and Racial Bias," Contemporary Sociology, Vol. 29, 2000, p. 120-129.
30For review articles on gender bias in evaluation, see Nieva and Gutek, op cit. and R. Kalin and
D. C. Hodgins, "Sex Bias and Occupational Suitability," Canadian Journal of Behavioral Science, Vol. 16,
1984, p. 311-325. Examples of studies of gender bias in assessments of qualifications, performance and
promotion potential include: G. L. Gerber, "The More Positive Evaluation of Men Than Women on the
Gender-Stereotyped Traits," Psychological Reports, Vol. 65, 1989, p. 275-286; G. H. Dobbins, R. L.
Cardy, and D. M. Truxillo, "The Effects of Purpose of Appraisal and Individual Differences in Stereotypes
of Women on Sex Differences in Performance Ratings: A Laboratory and Field Study," Journal of Applied
Psychology, Vol. 71, 1988, p. 551-558; T. H. Shore, "Subtle Gender Bias in the Assessment of Managerial
21
subjective criteria and substantial decision-maker discretion, individuals tend to seek out
and retain stereotyping-confirming information and ignore or minimize information that
defies stereotypes.31
Social research establishes clearly that the historical representation of women in a
job has a tremendous impact on compensation and other job rewards, mobility prospects,
and workplace culture.32 Experimental studies on stereotyping show that male and
female job applicants with identical personal traits are matched according to their gender
to jobs that are considered predominantly-male and predominantly-female.33 And
studies done in both experimental and natural settings demonstrate the impact of "sex role
spillover," whereby gender-linked traits associated with male-dominated occupations can
Potential," Sex Roles, Vol. 27, 1992, p. 499-515; M. Foschi, L. Lai, and K. Sigerson, "Gender and Double
Standards in the Assessment of Job Applicants," Social Psychology Quarterly, Vol. 57, 1994, 326-339; J.
Landau, "The Relationship of Race and Gender to Managers' Ratings of Promotion Potential," Journal of
Organizational Behavior, Vol. 16, 1995, p. 391-400.
31This kind of biased information-processing has been examined and replicated in numerous
experimental studies. See, for example, Banaji, Hardin, and Rothman, op cit.; Banaji and Hardin, op cit.; J.
Crocker, D. B. Hannah, and R. Weber, "Person Memory and Causal Attributions," Journal of Personality
and Social Psychology, Vol. 44, 1983, p. 55-66; S. M. Belmore, "Determinants of Attention During
Impression Formation," Journal of Experimental Psychology: Learning, Memory, and Cognition, Vol. 13,
1987, p. 480-489; M. Schaller, "Social Categorization and the Formation of Group Stereotypes: Further
Evidence for Biased Information Processing in the Perception of Group-Behavior Correlations," European
Journal of Social Psychology, Vol. 21, 1991, p. 25-35; T. E. Ford and C. Stangor, "The Role of
Diagnosticity in Stereotype Formation: Perceiving Group means and Variances," Journal of Personality
and Social Psychology, Vol. 63, 1992, p. 356-367; C. N. Macrae, A. B. Milne, and G. V. Bodenhausen,
"Stereotypes as Energy-Saving Devices: A Peek Inside the Cognitive Toolbox," Journal of Personality and
Social Psychology, Vol. 66, 1994, p. 921-935; C. N. Macrae, C. Stangor, and A. B. Milne, "Activating
Social Stereotypes: A Functional Analysis," Journal of Experimental Social Psychology, Vol. 23, 1994, p.
370-389; L. C. Johnston and C. N. Macrae, "Changing Social Stereotypes: The Case of the Information
Seeker," European Journal of Social Psychology, Vol. 24, 1994, p. 356-367; M. Zuckerman, C. R. Knee,
K. Miyake, and H. S. Hodgins, "Hypothesis Confirmation: The Joint Effect of Positive Test Strategy and
Acquiescence Response Set," Journal of Personality and Social Psychology, Vol. 68, 1995, p. 52-60. For
reviews, see Brown, op cit. and Fiske, op cit.
32P. England, Comparable Worth: Theories and Evidence, Aldine de Gruyter, 1992; B. F.
Reskin, D. B. McBrier, and J. A. Kmec, "The Determinants and Consequences of Workplace Race and Sex
Composition," Annual Review of Sociology, Vol. 25, 1999, p. 335-361.
33Glick et al., op cit.
22
profoundly affect the working climate for women.34 Women employed in traditionally
male-dominated jobs often attract more attention, are evaluated more extremely, are
perceived as different, receive less support, compared to male co-workers. As a result,
women in such settings often receive inadequate mentoring from their male colleagues,
are less likely to have full access to training and developmental work assignments, and
are less likely to receive useful and accurate feedback about their performance than are
similarly situated males in such settings.35 A large body of research in industrial
sociology, dating back to the 1950s, shows that individuals who find their opportunities
for advancement blocked respond by lowering their goals and aspirations, and by
lowering their commitment to their work compared to others with more promising career
prospects.36 More recent research documents that sexual harassment and a hostile work
34Gutek and B. Morasch, "Sex Ratios, Sex-Role Spillover, and Sexual Harassment of Women at
Work," Journal of Social Issues, Vol. 38, 1982, p. 55-74; B. Gutek, Sex and the Workplace, Jossey-Bass,
1985; D. Burgess and E. Borgida, "Sexual Harassment: An Experimental Test of Sex-Role Spillover
Theory," Personality and Social Psychology Bulletin, Vol. 23, 1997, p. 63-75.
35Kanter, op. cit.; D. F. Izraeli, "Sex Effects or Structural Effects" An Empirical Test of Kanter's
Theory of Proportions," Social Forces, Vol. 62, 1983, p. 153-165; M. B. Brewer and R. M. Kramer, "The
Psychology of Intergroup Attitudes and Behavior," Annual Review of Psychology, Vol. 36, 1985; D. M.
Messick and D. Mackie, op cit.; R, A. Noe, "Women and Mentoring: A Review and Research Agenda,"
Academy of Management Review, Vol. 13, 1988, p. 65-78; R. Ragins, "Barriers to Mentoring: The Female
Manager's Dilemma," Human Relations, Vol. 42, 1989, p. 1-22; B. R. Ragins and J. L. Cotton, "Easier Said
than Done: Gender Differences in Perceived Barriers to Gaining a Mentor," Academy of Management
Journal, Vol. 14, 1991, p. 939-951. Also see T. F. Pettigrew and J. Martin, "Shaping the Organizational
Context for Black American Inclusion," Journal of Social Issues, Vol. 43, 1987, p. 41-78.
36R. Guest, "Work Careers and Aspiration of Automobile Workers," American Sociological
Review, Vol. 19, 1954, p. 155-63; E. Chinoy, Automobile Workers and The American Dream, Doubleday
1955; T. V. Purcell, Blue Collar Man: Patterns of Dual Allegiance in Industry, Harvard University Press,
1960; R. Blauner, Alienation and Freedom, University of Chicago Press, 1964; O. Grusky, "Career
Mobility and Organizational Commitment," Administrative Science Quarterly, Vol. 10, 1966, p. 489-502;
Kanter, op cit.; J. A. Jacobs, Revolving Doors: Sex Segregation and Women's Careers, Stanford University
Press, 1989; and K. Loscocco, "Reactions to Blue-Collar Work: A Comparison of Men and Women,"
Work and Occupations, Vol. 17, 1990, p. 152-177; L. M. Shore and S. J. Wayne, "Commitment and
Employee Behavior: Comparison of Affective Commitment and Continuance Commitment with Perceived
Organizational Support," Journal of Applied Psychology, Vol. 78, 1993, p. 774-780; J. E. Wallace,
"Organizational and Professional Commitment in Professional and Nonprofessional Organizations,"
23
environment affects women's productivity on the job, physical health, and emotional will
being.37
Minimizing Workplace Gender Bias
Organizational policies and practices that create barriers to career advancement
for women and minorities, once in place, become institutionalized and rarely change in
the absence of any substantial change in a firm's business, technical, or legal
environment.38 This is especially true of personnel practices and policies that are
reinforced by the firm's culture.39 However, gender bias in the workplace is by no means
Administrative Science Quarterly, Vol. 40, 1995, p. 228-255; R. P. Settoon, N. Bennett, and R. C. Liden,
"Social Exchange in Organizations: Perceived Organizational Support, Leader-Member Exchange, and
Employee Reciprocity," Journal of Applied Psychology, Vol. 81, 1996, p. 219-227. For reviews, see W. T.
Markham, S. L. Harlan, and E. J. Hackett, "Promotion Opportunity in Organizations: Causes and
Consequences," Research in Personnel and Human Resource Management, Vol. 5, 1987, p. 223-87 and D.
D. Bielby, "Commitment to Work and Family," Annual Review of Sociology, Vol. 18, 1992, p. 281-302.
37 L. F. Fitzgerald, F. Drasgow, Fritz, C. L. Hulin, M. J. Gelfand, and others, "Antecedents and
Consequences of Sexual Harassment in Organizations: A Test of an Integrated Model," Journal of Applied
Psychology, Vol. 82, 1997, p. 578-589; K. T. Schneider, S. Swan, and L. F. Fitzgerald, "Job-related and
Psychological Effects of Sexual Harassment in the Workplace: Empirical Evidence from Two
Organizations," Journal of Applied Psychology, Vol. 82, 1997 p. :401-415; L. F. Fitzgerald "Sexual
Harassment: Violence Against Women in the Workplace, American Psychologist, vol. 48, 1993, p. 10701076.
38L. Stinchcombe, "Social Structure and Organizations," p. 142-93 in Handbook of
Organizations, edited by J. G. March, Rand McNally, 1965; M. T. Hannan and J. H. Freeman, "Structural
Inertia and Organizational Change," American Sociological Review, Vol. 43, 1984, p. 143-164; J. N. Baron,
"Organizational Evidence of Ascription in Labor Markets," in New Approaches to Economic and Social
Analyses of Discrimination, edited by R. Cornwall and P. Wunnava, Praeger, 1991. The concept of
organizational inertia has been applied in scientific studies conducted in a wide range of industrial settings.
See, for example, J. Roggema and M. H. Smith, "Organizational Change in the Shipping Industry: Issues in
the Transformation of Basic Assumptions," Human Relations, Vol. 36, 1983, p. 765-790; E. Abrahamson
and C. J. Fombrun, "Macrocultures: Determinants and Consequences," Academy of Management Review,
Vol. 19, 1994, p. 728-755; L. Gardenswartz and A. Rowe, "Diversity Management: Practical Application
in a Health Care Organization," Frontiers of Health Services Management, Vol. 11, 1994, p. 36-40; G. T.
Fairhurst, S. Green, and J. Courtright, "Inertial Forces and the Implementation of a Socio-technical Systems
Approach: A Communication Study," Organization Science, Vol. 6, 1995, p. 168-185; C. Doucouliagos,
"Conformity, Replication of Design and Business Niches," Journal of Economic Behavior & Organization,
Vol. 30, 1996, p. 45-62; and M.. Ruef, "Assessing Organizational Fitness on a Dynamic Landscape: An
Empirical Test of the Relative Inertia Thesis," Strategic Management Journal, Vol. 18, 1997, p. 837-853.
39P. Doeringer and M. Piore, Internal Labor Markets and Manpower Analysis, D. C. Heath, 1971
24
inevitable, and social science research shows what kinds of policies and practices
effectively minimize bias.
Through deliberate efforts, the effects of stereotypes can be controlled.40
Research studies show that the effects of stereotypes and outgroup bias on evaluative
judgments such as those involved in recruitment, hiring, job assignment, promotion, and
assessments of skills and qualifications can be minimized when decision-makers know
that they will be held accountable for the criteria used to make decisions, for the accuracy
of the information upon which the decisions are based, and for the consequences their
actions have for equal employment opportunity.41
Formal written policies alone are not sufficient to minimize bias in personnel
decisions. A written EEO policy that is simply reactive and lacks effective accountability
40P. G. Devine, "Stereotypes and Prejudice: Their Automatic and Controlled Components,"
Journal of Personality and Social Psychology, Vol. 56, 1989; S. T. Fiske, M. Lin, and S. L. Neuberg, "The
Continuum Model: Ten Years Later," p. 231-54 in Dual Process Theories in Social Psychology, edited by
S. Chaiken and Y. Trope, Guilford Press, 1999.
41T. E. Nelson, M. Acker and M. Manis, "Irrepressible Stereotypes," Journal of Experimental
Social Psychology, Vol. 32, 1996, p. 13-38; J. L. Eberhardt and S. T. Fiske, "Motivating Individuals to
Change: What Is a Target to Do?" p. 369-415 in Stereotypes and Stereotyping, edited by C. N. Macrae, C.
Stangor, and M. Hewstone, Guilford Press, 1996; A. M. Konrad and F. Linnehan, "Formalized HRM
Structures: Coordinating Equal Employment Opportunity or Concealing Organizational Practices?"
Academy of Management Journal, Vol. 38, 1995, p. 787-829; T. F. Pettigrew and J. Martin, "Shaping the
Organizational Context for Black American Inclusion," Journal of Social Issues, Vol. 43, 1987, p. 41-78;
G. R. Salancik and J. Pfeffer, "Uncertainty, Secrecy, and the Choice of Similar Others," Social Psychology,
Vol. 41, 1978, p. 246-55; C. T. Schreiber, K. F. Price, and A. Morrison, "Workforce Diversity and the
Glass Ceiling: Practices, Barriers, Possibilities," Human Resource Planning, Vol. 16, 1993, p. 51-69; P. E.
Tetlock, "Accountability: The Neglected Social Context of Judgment and Choice," p. 297-332 in Research
in Organizational Behavior, Vol. 7, edited by L. L. Cummings and B. M. Staw, Jai Press, 1985; P. E.
Tetlock and J. I. Kim, "Accountability and Judgment Processes in a Personality Prediction Task," Journal
of Personality and Social Psychology, Vol. 52, 1987, p. 700-709; P. E. Tetlock, "The Impact of
Accountability on Judgment and Choice: Toward a Social Contingency Model," Advances in Experimental
Social Psychology, Vol. 25, 1992, p. 331-376; Reskin, op cit.; P. M. Tetlock and M. Lerner, "The Social
Contingency Model: Identifying Empirical and Normative Boundary Conditions on the Error-and-Bias
Portrait of Human Nature," p. 571-585 in Dual Process Theories in Social Psychology, edited by S.
Chaiken and Y. Trope, Guilford Press, 1999; W. Bielby, 2000, op cit. To see how human resource
professionals apply these principles in the design of personnel systems, see R. D. Gatewood and H. S. Field,
25
is vulnerable to bias against women and minorities. Often, such a system is simply a
symbolic exercise in "going through the motions," with little substantive impact on
creating a work environment that is free of bias.42 True "EEO accountability" has three
key elements.43 One is the regular monitoring and analysis of patterns of segregation and
differences by gender and race in pay and career advancement as a routine part of an
organization's personnel system. Such monitoring assess whether disparities are greater
than what plausibly might be expected based on differences in job-related knowledge,
skills, abilities, and interests and other job-related factors that influence an employee's
contributions to the organization.44
A second component of EEO accountability is systematic analysis of feedback
from employees about perceptions of barriers to and opportunities for career
advancement. Systematic monitoring of trends in employees’ perceptions of barriers to
Human Resource Selection, Third Edition, Dryden Press, 1994; and H. G. Heneman, III, R. L. Heneman,
and T. A. Judge Staffing Organizations, Second Edition, Mendota House and Richard D. Irwin, 1997.
42L. B. Edelman, "Legal Ambiguity and Symbolic Structures: Organizational Mediation of Civil
Rights Law," American Journal of Sociology, Vol. 97, 1992, p. 1531-1576; L. B. Edelman, S. Patterson, E.
Chambliss, and H. S. Erlanger, "Legal Ambiguity and the Politics of Compliance: Affirmative Action
Officers' Dilemma," Law and Policy, Vol. 13, 1991, p. 73-97; L. B. Edelman, H. S. Erlanger, and J. Lande,
"Employers' Handling of Discrimination Complaints: The Transformation of Rights in the Workplace," Law
& Society Review, Vol. 27, 1993, p. 497-534; L. B. Edelman, Lauren B. and S. Petterson, "Symbols and
Substance in Organizational Response to Civil Rights Law," in Research in Social Stratification and
Mobility, 1999; M. E. Heilman, " Sex Stereotypes and their Effects in the Workplace: What We Know and
What we Don't Know," Journal of Social Behavior and Personality, Vol. 10, 1995, p. 3-26; J. S. Leonard,
"Women and Affirmative Action," Journal of Economic Perspectives, Vol. 3 (No. 1), 1989, p. 61-75. Also
see J. S. Leonard, Use of Enforcement Techniques in Eliminating Glass Ceiling Barriers, Report prepared
for the U. S. Department of Labor, Glass Ceiling Commission, April 1994.
43W. Bielby, 2000, op cit.
44Organizations with Affirmative Action Plans usually do something like this under the rubric of
"availability and utilization analyses," but often such analyses are generic reports generated by off-the-shelf
programs with little real connection to a company's overall personnel system. Effective monitoring is not
based on the generic formulae and broad occupational categories typically used in Affirmative Action
Plans, but instead relies on actual job transitions and is based on the same information used by those who
make decisions about hiring, job assignment, training, performance evaluation, promotion, compensation,
and the like.
26
career advancement and of top management's commitment to EEO can be used to identify
subtle forms of bias and related problems not immediately apparent from analyses of
more objective workforce data.
The third component of EEO accountability is explicit evaluation of managers and
supervisors on their contributions to an organization's EEO goals. Nearly all medium- to
large-scale organizations have a written antidiscrimination policy, and many have a
written policy stating that implementing the objectives of the Affirmative Action Plan is
the responsibility of every employee. However, such policies are merely symbolic unless
they also delineate explicit duties and responsibilities relating to equal employment
opportunity in each manager's or supervisor's job description, which can then be related to
specific evaluative dimensions in the performance reviews of those employees.
IN MAKING PERSONNEL DECISIONS AND DISTRIBUTING RESOURCES
THAT AFFECT FINANCIAL CONSULTANTS' CAREERS, MERRILL LYNCH
MANAGERS RELY ON ARBITRARY AND SUBJECTIVE PROCEDURES
WITH LITTLE OVERSIGHT OR ACCOUNTABILITY
As noted above, reliance on arbitrary and subjective criteria with inadequate
oversight and accountability makes personnel decisions vulnerable to gender bias. This is
especially true when decisions are made in the context with a male-dominated work
culture. In this section, I describe the system and criteria used for making decisions about
hiring new financial consultants, making promotions into management, and allocating
resources that affect the ability of FC trainees and FCs to generate commissions. The
features of Merrill Lynch's policies and practices I describe in this section are comparable
27
to those identified in social science research as factors that create career barriers to
women. After describing these features, in the section that follows I describe how these
aspects of the firm's personnel system create and sustain gender differences in the careers
of Financial Consultants.
Hiring Into the Financial Consultant Position
Decision-making regarding the process and criteria used to hire trainees into the
Professional Development Program is delegated to local branch management. There are
no company-wide written guidelines regarding qualifications, and there is no systematic
monitoring of the hiring process, criteria used, or the impact on equal employment
opportunity.45 In explaining the criteria he relied upon, National Sales Manager
Bowman testified that "there is no boiler plate for success," that he looks for no specific
background or education, and that instead it is "a mismash of everything."46
According to the testimony of Merrill Lynch executives, in hiring new trainees, managers
rely substantially on referrals from current FCs and on word-of-mouth, recruitment
techniques that social science research shows tend to reproduce the gender composition
of the existing workforce.47
45Aldredge depo., p. 61-67; Bowman depo., p. 51-54, 60-69.
46Bowman depo., p. 65.
47Bowman depo., p. 70-71, 75-6; Aldredge depo., p. 63. Research on recruitment practices shows
that firms that rely upon employee referrals and personal connections in recruiting and screening are less
likely to hire minorities and women than are firms who do not rely heavily on such mechanisms. J. H.
Braddock and J. M. McPartland, "How Minorities Continue to Be Excluded from Equal Employment
Opportunities: Research on Labor Market and Institutional Barriers," Journal of Social Issues, Vol. 43,
1987, p. 5-39; K. M. Neckerman and J. Kirschenman, "Hiring Strategies, Racial Bias, and Inner-City
Workers," Social Problems, Vol. 38, 1991, p. 433-447; H. Holzer, What Employers Want: Job Prospects
for Less-Educated Workers, Russell Sage Foundation, 1996; R. Waldinger, "Black/immigrant Competition
Re-assessed: New Evidence from Los Angeles, Sociological Perspectives Vol. 40, 1997, p. 365-386; B. F.
Reskin and D. B. McBrier, "Why Not Ascription? Organizations' Employment of Male and Female
28
After hiring a new trainee, the branch manager has discretion regarding whether
to assign the new employee to the PDP program or to put that person into the AFC
program, which provides additional training and support before the trainee is required to
compete with other new hires.48 In addition, as described below, local branch
management has discretion over the allocation of transferred accounts, leads, referrals,
walk-ins, and other resources that can affect the ability of an FC trainee to meet his or her
production goals. The discretionary allocation of those resources has a continuing impact
on the ability of an FC to build a book during the early stages of her or his career.
Promotion Into Management
To prepare oneself to become a candidate for the Management Assessment Center
and for a Resident Vice President position at Merrill Lynch, it is important to gain
experience in a range of positions that build management skills, such as Resident
(Associate) Manager, Financial Consultant in Charge (FCIC), Producing Sales Manager,
or a Product Coordinator.49 Assignment to those positions is left to the discretion of
local branch management, with no written firm-wide guidelines. In addition, to
participate in the firm's Opportunity in Leadership Program (which is preparatory to the
Management Assessment Center) requires the sponsorship of the FC's Regional Vice
President and District Director.50 I have seen no evidence that the allocation of
development assignments is systematically monitored, either for the process and criteria
Managers," American Sociological Review, Vol. 65, 2000, p. 210-233; P. V. Marsden and E. Gorman, op
cit.
48Aldredge depo., p. 64-65, 70-73.
49Aldredge depo., p. 35-36, 44=45.
50Pittari depo., p. 73-77, 129.
29
used or for its impact on equal employment opportunity. While Ms. Pittari, who heads
the Management Resource & Development group, monitors statistics on the success rate
of women once they are admitted to the Management Assessment Center, she does not
monitor gender disparities in the process whereby candidates are selected for the MAC.51
Allocation of Career-Enhancing Resources
Local branch managers have a tremendous amount of discretion in the allocation
of resources that affect the ability of an FC trainee or Financial Consultant to generate
commissions, including, most importantly, accounts of brokers who have left the firm or
gave up their book, but also leads, referrals, walk-ins, "broker-of-the-day" positions,
office space, and administrative support.52 During the period covered by the Cremin
litigation, there was virtually no systematic monitoring or oversight by the firm regarding
process used in local offices for allocating these resources or the potential impact on
equal employment opportunity.53
Cumulative Advantage
The management philosophy at Merrill Lynch is based on a principle of "successbreeds-success." The resources described above are typically concentrated on those
Financial Consultants and FC trainees who have been the most successful in generating
51Pittari depo., p. 74, 215.
52On discretion in distributing accounts of departing brokers, see Bowman depo., p. 117-119;
Rosenthal depo., p. 57. On the impact of such distributions on an FC's ability to earn commissions, see
Rosenthal depo., p. 57-61. On discretion in the distribution of walk-ins, leads, referrals, call-ins, and
administrative support, and the potential impact on compensations, see Rosenthal depo., p. 59-62; Bowman
depo., p. 113-120, 134-138. On office space, see Bowman depo., p. 122-126, 144-146.
53Bowman depo., p. 211-216, 236-239. Also see Rosenthal depo., p. 232-235, 405-441 on
management compensation. Mr. Rosenthal's testimony shows that compensation of Resident (Associate)
Managers, FCICs, Regional Vice Presidents, and District Directors is not based in any way on performance
relative to equal opportunity objectives.
30
production and are believed by firm management to be the ones who will be most
productive in the future. For example, the PDP program for FC trainees, which provides
the foundation for successful performance as a broker, is structured such that only those
trainees who surpass specific performance hurdles are invited to attend national training
sessions. At those sessions, trainees learn important practical skills for selling and are
able to establish relationships and learn from role models who are some of the most
successful FCs in the firm.54 Similarly, access to many forms of advanced training at the
national level is typically available only to the most successful FCs.55 According to Mr.
Bowman, the overall philosophy is to spotlight the most successful FCs, so that others
will emulate them. As a result, brokers who generate the most production tend to receive
the most desirable office space, more and higher quality administrative support, subsidies
for business expenses, prestigious job titles, and more and better quality accounts,
referrals, leads, and walk-ins. Each of these resources enhances the successful FC's
capacity to generate future production.56
Social scientists describe these kinds of reward systems as systems of cumulative
advantage. In a cumulative advantage system, factors that result in small disparities
between individuals early in their careers have a cumulative impact, and generate growing
disparities over time.57 At Merrill Lynch, the discretionary allocation of career-
54Aldredge depo., p. 50-53, 129-135.
55Bowman depo., p. 93, 122; Williams depo., p. 65-66.
56Bowman depo., p. 123-128, 139-141, 144-155.
57The first cumulative advantage models were developed to explain differences in the career
trajectories of scientists, including gender disparities in scientists' careers. See, for example, P. D. Allison,
J. S. Long, and T. K. Krauze, "Cumulative Advantage and Inequality in Science," American Sociological
Review, Vol. 47, 1982, p. 596-606; S. M. Clark, and M. Corcoran, " Perspectives in the Professional
31
enhancing resources gives some FCs a "head start" over others. The "success-breedssuccess" principle launches FCs who are favored by this system into a career trajectory
that typically sustains itself at a level that those left behind have little chance of matching.
A Male-Dominated Organizational Culture
The exercise of discretion and the "success-breeds-success" system
described above is carried out in the context of a distinctive firm-wide organizational
culture. An organizational culture is a shared set of values and beliefs about how things
are done in the organization.58 A strong and widely-shared organizational culture
promotes uniformity of practices throughout an organization. Merrill Lynch's practices of
emphasizing the company's founders and unique history in its publications and webpage,
it articulation of distinctive principles and slogans in employee handbooks, newsletters,
and other documents, the ubiquitous display of the corporate symbol (the "bull" logo),
and the reliance on intensive training and socialization of new employees are widely
recognized mechanisms for sustaining a strong culture.59
Socialization of Women Faculty: A Case of Accumulative Disadvantage?" Journal of Higher Education,
Vol. 57, 1986, p. 20-43; R. K. Merton, "The Matthew Effect in Science, II: Cumulative Advantage and the
Symbolism of Intellectual Property," Isis Vol. 79, 1988, p. 606-623; and W. T. Bielby, "Sex Differences in
Careers: Is Science a Special Case?" p. 171-187 in The Outer Circle: Women in the Scientific Community,
edited by H. Zuckerman, J. R. Cole, and J. T. Bruer, Norton, 1991. For its application more broadly, see
A. M. O'Rand, "The Cumulative Stratification of the Life Course," p. 188-207 in Handbook of Aging and
the Social Sciences (4th ed.), edited by R. H. Binstock, L. K. George, et al. Academic Press, 1996.
58See T. E. Deal and A. A. Kennedy, Corporate Cultures. Addison-Wesley, 1982; E. H. Schein,
Organizational Culture and Leadership, Second Edition, Jossey-Bass, 1992; and J. Kerr and J. W. Slocum,
"Managing Corporate Culture Through Reward Systems," Academy of Management Executive, Vol. 1,
1987, p. 99-108.
59Deal and Kennedy, op cit.; Schein, op cit., W. Ouchi, "Markets, Bureaucracies, and Clans,"
Administrative Science Quarterly, Vol. 25, 1980, p. 129-141. Merrill Lynch articulates its official values in
a publication titled "Merrill Lynch Principles" (Bates D 0000464-0000474). These principles, along with
the founder's slogan, "our business is other people's money," appears on the first page of it's employee
handbook, "Welcome to Merrill Lynch: An Employee's Guide to Policy and Practices" (Bates D 0000484).
The company's founders, slogans, symbols, and claims to uniqueness are featured prominently at the
32
A company's informal culture can be especially consequential for personnel
practices, especially when that culture embraces the values and contributions of one
group over another.60 As noted above, in organizations that are highly segregated by sex,
with men dominating positions of power and authority and far outnumbering women in
jobs with the most favorable career prospects, the informal culture of the organization is
likely to be male-dominated. Merrill Lynch is no exception.
The male-dominated values of Merrill Lynch are evident in qualities that are
emphasized in the Management Assessment Center files for 1990 through 1999, produced
as part of this litigation. As noted above, the actual skills required for management
positions at Merrill Lynch do not differ by gender. Yet there is a remarkable emphasis on
gender-linked activities and background in the MAC assessment files. For example, a
majority of the files make reference to the candidate's experience or accomplishments in
either team sports or golf, and many make reference to fraternity membership and
activities. Reference to military service is common, and military imagery is often
invoked in describing a candidate's contributions to the firm, through terms such as
"military bearing," "good soldier," "corporate soldier," "a real soldier's soldier," "would
fall on a sword for...," and "good to be in a foxhole with."
Explicitly gender-linked references such as "good old boy," "all-American boy,"
"one cool guy," and "appeals to senior guys" appear frequently. Even outright sexist
language is not uncommon in descriptions of a candidate's background. For example, one
corporate history page of the company's website. On managers' role in transmitting the company culture,
see Pittari depo., p. 40-41.
60S. A. Sackmann, "Culture and Subcultures: An Analysis of Organizational Knowledge,"
Administrative Science Quarterly, Vol. 37, 1992, p. 140-161; W. B. Stevenson and J. M. Bartunek, "Power,
33
candidate is described as having attended Harvard where he "found the girls less than
glamorous," whereas another is described as going to a specific college because of the
"beautiful women," and a female candidate is described as "a little bossy." In their
complaints in this litigation, class members report numerous instances of sexist and
offensive language.
Inadequate Oversight and Accountability
As noted above, prior to the settlement of the Cremin litigation, Merrill Lynch did
virtually no systematic monitoring or oversight of the process used in local offices for
allocating resources that influence an FCs productivity. Ms. Pittari testified that the firm
did not attempt to monitor women's representation in the Management Assessment Center
relative to the gender composition of the firm's workforce.61 In the testimony and
documents I have reviewed there is no evidence of ongoing, systematic monitoring
regarding disparities by gender in the numbers of individuals entering into the PDP,
employed as Financial Consultants, entering the Management Assessment Center, or
employed in the various management positions, prior to the Cremin settlement. Nor does
it appear that there was any systematic evaluation of Resident Managers, FCIC's,
Regional Vice Presidents, or higher-level managers on their performance relative to the
firm's Equal Employment Opportunity objectives.
Finally, because of the industry-mandated arbitration procedure, the firm was
insulated from public scrutiny and faced no effective external legal accountability to
comply with antidiscrimination laws. A recent article in the Employee Relations Law
Interaction, Position, and the Generation of Cultural Agreement in Organizations," Human Relations, Vol.
49, 1996, p. 75-104.
34
Journal identified several features of arbitration systems that can increase fairness and
reduce bias. These include provisions for joint selection of the arbitrator, discovery, a
written award, a clear and specific waiver clause, limited judicial review, and a
requirement that arbitrators be trained in employment law.62 None of these features were
present in the industry-run arbitration system in use during the period covered by the
Cremin litigation.63
THE FEATURES OF THE FIRM'S POLICIES AND PRACTICES DESCRIBED
ABOVE CREATE AND SUSTAIN BARRIERS TO WOMEN'S CAREER
ADVANCEMENT AT MERRILL LYNCH
The personnel system described above has features that a large body of research
has shown create barriers to women's career advancement. When important resources are
allocated in an arbitrary and consistent manner, with an absence of systematic monitoring
and oversight, it is often difficult to get quantitative measures of disparate treatment of
women. While it is not possible to quantify the outcome of decisions regarding office
space, leads, referrals, walk-ins, access to developmental assignments, mentoring, and
administrative support, it is possible to quantify the value and type of accounts transferred
to men and women from the accounts of brokers who have left the firm and from house
accounts. Dr. Madden conducted an analysis of gender disparities in the distribution of
61Pittari depo., p. 214-215.
62D. R. Casey and B. A. Lee, "Mandatory Arbitration Clauses in Individual Employment
Contracts: Enhancing Fairness and Enforceability." Employee Relations Law Journal, Vol. 25, 1999 (No.
3, Winter), p. 57-75.
35
accounts and found that women FCs receive lower quality accounts (as indexed by prior
year's commissions, the asset value of prime accounts, and the asset value of accounts
that stayed with the firm) than their male counterparts. Dr. Madden also found that
women FC trainees were less likely than men to received high quality accounts. In
addition, she found that compared to their male counterparts, women received accounts
with lower asset value and production credits in the year that they moved from the broker
training program to broker status.
Given the organizational context in which decisions about distributing accounts
are made, the gender disparities detected by Dr. Madden come as no surprise. Within this
context, it is likely that there are significant gender disparities in the distribution of the
many resources that are difficult to quantify, including leads, referrals, walk-ins, office
space, administrative support, mentoring, etc. These disparities, especially early in an
FCs career, are likely to be quite consequential in the "success-breeds-success" system
fostered by Merrill Lynch. For example, being advantaged in the distribution of accounts
has an immediate payoff in terms of production and, therefore, compensation. But it also
has an indirect impact in several ways. Those distributed accounts can generate leads
directly, and the production generated off those accounts enhance an FC's likelihood of
receiving better administrative support, office space, mentoring, leads, walk-ins, referrals,
and other resources that in turn build assets and generate still more production.
In sum, key to the way cumulative advantage affects gender disparities in earnings
of Financial Consultants are the following factors: (1) managers' discretion in
63Baker, Debra, "Tangled Up in Ticker Tape," ABA Journal, Vol. 85, 1999 (No. 44, December),
p. 44.
36
distributing resources and support, based substantially on informal, arbitrary, and
subjective criteria; (2) an absence of oversight over the process used by managers to
allocate resources and support; (3) an absence of monitoring of gender disparities in
compensation and in the factors that influence compensation; (4) a history of sex
segregation in which broker and sales management positions have historically been
"men's work" at the firm; and (5) an organizational climate that has been hostile and
unsupportive of women working in the male-dominated position of Financial Consultant.
The patterns of gender segregation documented here and in Dr. Goldman's report and the
experiences and perceptions of class members reflected in their complaints are consistent
with the phenomenon of "sex role spillover" described in social science literature and
summarized above. This is precisely the kind of setting in which managers' discretion
will be influenced by stereotypes and gender bias. The materials I have reviewed strongly
support the conclusion that female FCs and FC trainees were systematically
disadvantaged in this way. They also support the conclusion that such disparities in the
allocation of resources and support have had a cumulative impact, generating the large
and significant gender gap in compensation that emerges early and persists over the
course of brokers' careers at Merrill Lynch.
CONCLUSION
In sum, there is no reason to believe that women and men are differentially
qualified to perform as Financial Consultants or in management positions at Merrill
Lynch. Nevertheless, among Financial Consultants and FC trainees, men continue to
outnumber women by more than five to one, and women's representation in sales
37
management positions has been negligible. Women who work alongside men in the
male-dominated FC positions earn substantially less than their male counterparts. Given
the historically male-dominated culture, both in the industry overall and at Merrill Lynch
in particular, and the absence of any substantive evaluation or accountability of managers
with respect to equal employment opportunity, the most plausible explanation for these
disparities is the unequal treatment of men and women in the application of the informal,
arbitrary, and subjective features of the firm's personnel system. Resources that affect an
FC trainee's or Financial Consultant's ability to generate commissions are allocated in a
manner that is often arbitrary and subjective, with virtually no oversight over the process
and criteria used or on the impact on equal employment opportunity. As a result, women
Financial Consultants have not received the same level of resources and support as their
male counterparts, which has created gender disparities in earnings that emerge early in
brokers' careers and persist over time. Features that are known to minimize bias in
organizational settings have been largely absent from the firm's policies and procedures,
and in their absence the inequities that were first identified in the mid-1970s have
persisted for a quarter of a century.
William T. Bielby, Ph.D.
January 3, 2002
38
EXHIBIT A
Curriculum Vitae