Gender Bias against Women of Color in Science
An interview with Joan Williams
By Lisa Levey

Research Spotlight Series:
A Project of the Committee to Connect Research, Policy and Practice
May 2015

Image result for joan williams uc hastings

Over several decades, social scientists have identified clear patterns of gender bias that women
encounter at work. Yet little is understood about the nuances of how these patterns manifest for
women of color. Joan Williams, Distinguished Professor of Law at UC Hastings, wanted to explore how
gender plays out in the everyday interactions of women scientists and how they differ by race and
ethnicity.

Professor Williams partnered with colleagues Katherine Phillips at the Columbia Business School and
Erika Hall at Emory University (formerly a graduate student at Northwestern) to conduct 60 in‐depth
interviews with women of color scientists. In addition, they surveyed more than 500 women to quantify
the experiences of White, Black, Asian‐American and Latina women in STEM. The researchers are
indebted to the Association of Women in Science (AWIS) for their assistance in recruiting study
participants.

Key Findings

Women encounter rampant gender bias in the scientific community with important distinctions in how
bias differs by race and ethnicity.

  •  Approximately 2/3rd’s of the women ‐ 66.7% of those interviewed and 63.9% of those surveyed –
    reported Prove‐It‐Again bias or the need to provide more evidence of competence than their
    male colleagues in order to be seen as equally competent. Black women were sharply more
    likely to report this phenomenon, which they attributed to race, while other women attributed
    Prove‐It‐Again bias to gender.
  •  More than 3/4th’s of the women interviewed reported being seen as either ‘too feminine to be
    competent’ or ‘too masculine to be likeable’, known as Tightrope bias. For Asian women, the
    acceptable band of behavior was even narrower with greater backlash for stereotypically
    masculine behaviors such as self‐promotion. Latinas reported being labeled as emotional or
    angry when being assertive and faced greater pressure to manage ‘office housework’ such as
    scheduling meetings. Black women are afforded more space to act assertively as long as this
    behavior is not construed as anger.
  • Gender bias against women often fuels conflict among women. Over half of the scientists
    interviewed and approximately one in four women surveyed reported the Tug of War among
    women. An example is when women who’ve faced discrimination early in their careers distance
    themselves from other women, a pattern documented in several studies. Among 3/4th’s of the
    women surveyed reported that women generally supported each other in their work
    environments—but only a bit over half (56%) of Black women did.
  •  Once triggered, the strongest form of gender bias is the Maternal Wall reported by nearly 2/3rd’s
    of women scientists interviewed. After having children, women face stereotypes about a
    reduced level of commitment to ‐ and interest in ‐ their careers. The Maternal Wall affects
    women of all races and ethnicities but was heightened for white and Asian women more often
    advised by colleagues to work fewer hours after becoming mothers.

Implications of the Research for Organizations Seeking to Drive Change

The research underscores that women scientists of all races and ethnicities face gender bias—but that
their experience of gender bias differs by race. A clear implication of the research for diversity
practitioners is the need to recognize and address differing patterns of bias in organizational women’s
initiatives. Too often women’s initiatives are seen as primarily for white women without adequately
reflecting the needs and experiences of women of color. Efforts to proactively understand how gender
biases manifest for women of color ‐ and to reflect back those differences in both messaging and
targeted interventions – is a powerful means to help women of color feel included. Just as diversity
initiatives focused on race alone are by default male and inadequately address the differing experiences
of women, gender initiatives must also seek to reflect the unique needs and issues for women of color.
A second key learning from the research, that is the continuing prevalence of gender bias for women
scientists, demands a new and more results‐based approach. Of the total sample of 127 women
interviewed for Williams’ co‐authored book (with Rachel Dempsey), 96% reported gender bias. Many
gender initiatives are focused on enriching the women – with targeted trainings, mentoring programs,
and network opportunities among other efforts – and while these certainly have merit, this approach is
incomplete. Because implicit and subtle bias is the real culprit in most workplaces today, rather than the
overt discrimination of the past, it is much more challenging to see and thus to address. In addition,
implicit bias is powerfully self‐perpetuating and gets reinforced in countless ways day after day.
Remember that ALL of the women interviewed for the research reported experiencing one or more of
the major patterns of gender bias.

The chief weapon used by organizations to combat gender bias is typically a single implicit bias training.
According to Williams, “While this is a good start, doing anything just once will not change a culture. It
just doesn’t work like that.” Eliminating gender bias requires addressing the structural and behavioral patterns that under grid how organizations operate such as women being penalized for utilizing the
expanded tenure clock to manage the transition to parenthood or men naturally gravitating to sponsor
other men for key projects and positions, in advertently disadvantaging their women colleagues. A
better solution is for organizations to identify and disrupt the bias through a four step process:

  1. Identify the patterns of bias through conversations with women at the organization.
  2. Develop a hypothesis and an objective metric to measure progress. For example, if
    women are doing large loads of office housework—taking notes and playing backroom
    roles—while the glamour work is more likely done by men, a metric might be to
    measure who’s doing the glamour work and who’s doing the office housework.
  3. Use a bias interrupter to interrupt the bias. A gentle interrupter might be a training that
    describes the office housework problem and how it relates to gender bias. (Keep in
    mind that gender bias trainings typically are more effective if they are built into a
    training that’s already expected and in place.)
  4. Review the metric to see whether the bias interrupter has been effective. If it hasn’t,
    ratchet up to a stronger interrupter. An example would be to create policies identifying
    that the responsibilities for taking notes should be rotated and/or that a more formal
    assignment system for allocating office housework is necessary.

Combatting deeply entrenched patterns of gender bias requires an iterative, evidence‐based process.
Progress results when organizations try different approaches through time and apply what they learn.
Gender equity needs to be seen as an ongoing management issue, like research and development or
marketing, and treated with the same rigor anchored by metrics, consequences, and ongoing innovative
approaches. The UC Hastings College of the Law has founded the Working Group on Bias Interrupters,
which brings corporate partners together with leading researchers to identify and pilot bias interrupters.
There is space for a few additional corporate partners. Please inquire if you are interested or would like
to learn more.

Potential Research Questions to Explore

Two promising avenues to explore through additional research include:

  •  Creating a comprehensive repository of evidence‐based Bias Interrupters – concrete changes to
    organizational structures and practices ‐ that are making a quantifiable difference for women
    scientists.
  •  Exploring the enduring work‐life imbalance many employees experience, despite efforts such as
    offering part‐time professional options which too often are marginalized by the always‐on work
    culture, by focusing on the deep link between masculinity norms and cultures of pervasive
    overwork.

Research Links

Below we have highlighted researchers focused on key issues highlighted in this Research Spotlight.
Research focused on differing expressions of gender bias by race and ethnicity:

  • Erika Hall, the Goizueta Business School at Emory University
  • Robert Livingston – University of Sussex
  • Katherine Phillips, Columbia Business School

Research focused on systemic changes – or disrupters – to address the ongoing crisis of work overload
and imbalance which so profoundly impacts women, and increasingly men professionals:

  • Lotte Bailyn, Sloan School of Management
  • Jennifer Berdahl, University of British Columbia
  • Erin Kelly, University of Minnesota
  • Phyllis Moen, University of Minnesota
  • Leslie Perlow, Harvard Business School