Summary

Corporations often start their diversity journey by providing their managers or all workers with diversity training. These trainings were first offered as race-relations sessions in the 1960s and are among the most popular tools for diversity managers. Diversity training programs have changed their content during the decades, but they usually include live or online explanations about unlawful discrimination and bias, often supplemented by discussions of cultural differences and business needs for diversity. Despite their popularity and often high costs, a large body of research conducted over decades shows that most diversity training programs do not lead to long-term improvements in participants’ bias, attitudes, behavior, or workforce diversity. Some studies also show that training has negative effects on bias and diversity.

Factors that impede the success of diversity training or make them backfire include the hardwiring of cognitive biases and people’s complex reactions to direct attempts to change their biases, as well as the broader systemic biases rooted in everyday organizational routines. These suggest that common diversity training simply may not be the right tool for reducing bias and generating the changes needed for increasing workplace diversity. Some studies suggest that trainings’ effects could be improved by carefully designing them. This includes: avoiding training features that increase participants’ alienation, such as mandatory attendance, quizzes, and legalistic content; designing long-term training such that meaningful learning can be achieved; calibrating training to specific organizational challenges rather than using off-the-shelf content; and including ongoing collaborative contact with members of underrepresented groups and integrating training as part of broader diversity and accountability efforts. More research is needed to determine whether these types of training indeed produce sustained improvements in bias and diversity.

Alternative bias reduction mechanisms can be found in popular management models that increase collaboration between workers, such as cross-functional teams and cross-training. Such collaborative teams and training improve corporate performance and, as a byproduct, also reduce bias. Cognitive biases are affected by the work contexts in which individuals operate. Highly segregated workplaces, where White people and men meet women and people of color (or other underrepresented groups) primarily in marginalized jobs, deepen group boundaries and strengthen stereotypes. When organizations create cross-functional collaborations using self-directed teams and cross-training, workers from different groups have more opportunities to collaborate and, as studies show, biases and group boundaries are reduced, and leadership diversity increases.