UW News

May 19, 2022

Q&A: Why discriminatory bias is a public health problem

UW News

Implicit bias leads to unintended discrimination in organizations, and leaders should take their cues from public health strategies in order to remedy the problem, explains Anthony Greenwald, emeritus professor of psychology at the University of Washington.


Over the past two years, the global pandemic and protests for racial and social justice have shone new light on disparities in health and other aspects of life for historically marginalized populations. Experts in a variety of fields called out racism as a public health crisis, leading in some cases to additional funding and research of the problem.

In parallel scientific work, Tony Greenwald, emeritus professor of psychology at the University of Washington, joined with five colleagues at other universities to understand effects of bias in all types of organizations, reviewing their own research along with that of many others. They determined that public health strategies can be a model for strategies that could effectively address unintended discrimination.

Their article, published May 19 in Psychological Science in the Public Interest, calls on decision-makers to reexamine and retool their organizations to remedy the profound effects of unintended bias.

Greenwald, who created the Implicit Association Test nearly 25 years ago, makes clear the distinction between bias and racism. Although both produce discrimination, he distinguishes racism, which is intentional bias motivated by hostility, from both implicit bias and systemic bias, which can produce discriminatory outcomes in the absence of hostility and intent.

“Damages due to implicit and systemic bias contribute significantly to contemporary social injustices without intention to discriminate and can do so without perpetrators’ awareness of the harms they are causing,” Greenwald said. “The personnel in workplaces and government institutions who carry out systemic biases may themselves have no intent to discriminate, and can also be unaware that their actions contribute to discriminatory outcomes.”

Greenwald addressed the issues raised in the new article for UW News.


Why do you recommend that discriminatory bias be treated as a public health problem?

Implicit and systemic biases are not illnesses. They are acquired as part of normal development. At the same time, implicit and systemic biases produce disparities in health status and in health care of those who are victimized by these biases. The main reason for considering it wise to treat implicit and systemic biases as public health problems is the likelihood that adaptations of well-known public health strategies can succeed in controlling them where widely used psychologically based efforts have not succeeded.


What about existing efforts to address unintended discrimination?

Two types of possible remedies have received the greatest attention in the past 20 years: mental debiasing, which has been a focus of many laboratory investigations by psychologists; and group-administered training, which has been adopted in many workplaces, where it may be called implicit bias training, anti-racism training, cultural competence training, or diversity training.  Unfortunately, neither mental debiasing nor group-administered training has been found reliably effective either in reducing biases or improving diversity within organizations.

However, it is premature to believe either that future efforts with these methods can succeed or that such future efforts are doomed to failure.


Which public health strategies could be successful here?

Many public health strategies are preventive. For discriminatory biases, effective prevention depends on requiring important decisions to be made without knowledge of demographic characteristics of those who are affected by those decisions. This is called decision blinding, and there are many presently unused opportunities to implement it.

Tony Greenwald

Implicit biases are acquired as part of normal development. Systemic biases arise in normal organizational and political process. Nevertheless, where psychologically-based methods have not succeeded, methods that have served effectively to manage public health problems can provide models for effective management of both implicit and systemic bias.

A widely usable method is a medical epidemiological approach called disparity finding. This requires analysis of existing data to identify demographic inequities. Unlike the presently unsuccessful methods of mental debiasing and group-administered training, both prevention and disparity finding depend on modifying decision-making procedures that otherwise can cause unintended discrimination.


Discuss the importance of the organizational “self-test,” which you and your co-authors offer at the end of your article. 

To reduce discrimination, business and political leaders must understand that it is in their power to implement effective fixes. We provide an organizational self-test that can be the starting point for their efforts. That test includes questions that aim to get at what an organization can and is prepared to do, such as obtaining data on how employees and others who receive their services are being treated, and whether there is someone in a leadership role who is responsible for diversity, equity and inclusion activities. Organizations that take these steps can start to repair unintended discrimination.

For more information, contact Greenwald at agg@uw.edu.