A briefing paper prepared by Alison T. Wynn and Shelley J. Correll, Stanford University, for the Council on Contemporary Families’ Gender Matters Online Symposium (.pdf).

Research consistently shows that unconscious or implicit gender biases systematically hinder women’s advancement in the workplace. Such biases operate outside of conscious awareness, which makes them particularly difficult to detect and combat. Even people who are not explicitly sexist or racist are susceptible to subtle, unconscious biases, such as weighing a man’s opinion as more credible than a woman’s, which can unconsciously affect our judgments and, ultimately, the rewards men and women earn in settings like workplaces.

In recent years, organizations have become interested in reducing these biases by training their employees. For example, in the wake of an incident where employees called the police on two Black customers for actions that were ignored when engaged in by white customers, Starbucks recently closed its 8,000 U.S. stores to provide unconscious bias trainings to its 175,000 employees.

While unconscious bias trainings are an important first step, research finds that organizations must do more if they want to produce sustainable change. Unconscious bias trainings, while helpful, can wear off over time, or can even risk exacerbating bias by painting it as something normal and unavoidable. Specifically, organizations must alter the conditions that are known to enable and exacerbate bias.

The Small Wins Model.

At the Clayman Institute for Gender Research at Stanford University, researchers are collaborating with companies to engage in such change efforts. Using a “small wins model” of organizational change, we first educate employees about unconscious bias and then work with them to develop and assess new processes and tools to get beyond bias.

For example, at one large technology company, we collaborated with managers to improve gender equality. When we began our work, the company had a less-than-stellar reputation for gender equality and no consistent performance management process in place. Through a targeted intervention, we worked with managers to reduce the ambiguity in their performance assessment processes, since ambiguity is known to exacerbate bias. Research has found that when the criteria for evaluation are not clearly defined or spelled out, they leave room for unconscious biases to have a particularly robust impact on people’s judgments. In our intervention, managers developed new clear, measurable criteria to assess employees; ensured that the same criteria were being applied to all employees; and allotted equal amounts of time for discussing each employee during their calibration meetings. Prior to these changes, women were more likely than men to receive criticisms about their personality, and they were more likely to have their performance ratings downgraded in calibration meetings. After the intervention, these differences were no longer significant.

These small wins inspired other changes at the company, including reworking their job ads to be more appealing to women and other groups. Today, half the entry-level engineers hired are women, and the company has since been named one of the top workplaces for women in tech.


Alison T. Wynn, Stanford University, atp5@stanford.edu. Shelley J. Correll, Stanford University, scorrell@stanford.edu. They are authors of “Combating Gender Bias in Modern Workplaces” in the Handbook of the Sociology of Gender.