We sort social groups into categories by our very nature.
Human beings are simplifiers. We are cognitive misers, exerting the least amount of mental effort that we can in making decisions. We rely on heuristics, or mental shortcuts, to take the fastest route from A to B. And we are categorizers.
The tendency to conceive of the world around us in categories is a strategy that is often adaptive, but has at least one unfortunate byproduct: the bias that results from associations we make with different categories. And while there is no getting rid of bias, we can design systems to correct for these errors.
Categorizing is critical to survival. It enables you to efficiently move through the world, managing the stream of stimuli that you encounter from millisecond to millisecond. It allows you to be comfortable handing a letter to the mail carrier and be assured that the supermarket clerk will give back your credit card after you’ve paid for your groceries.
All humans are predisposed to sort the world around them into categories. You quickly assign the empty, unlit alley as “risky” and the bustling, well-lit sidewalk as “likely safe.” You see a moldy apple and have no trouble labeling it “not good to eat.”
Categorizing can lead to bias
The innate tendency to categorize is essential for managing our lives, but grouping people into categories and making decisions based on those categories can bring with it pernicious side effects, particularly for systemically disadvantaged social groups. As behavioral scientist Jennifer L. Eberhardt puts it, “the very things that help us to see the world can also blind us to it.”
Research shows that these side effects are pervasive, and particularly pronounced in discrimination against people of color. For example, researchers who submit identical resumes to job ads receive more interview offers when the candidate has a white-sounding name than a Black-sounding name like “Lakisha” or “Jamal.” When a Black person is shown shoving a white person, 75% of study participants consider the action violent whereas only 17% rate the same shove as violent when the races are reversed.
The stakes of racial bias are critically high when police and guns are involved. When faced with an ambiguous situation, police may have to act quickly to decide whether a person is a risk to their safety, and react correspondingly. Police officers must routinely place people into categories: innocent bystander (don’t shoot!) or threatening criminal (take action). Unfortunately, this is where bias seeps in — often unknowingly, outside the officer’s awareness and in spite of their training.
In simulations, police are faster to shoot Black people with guns than white people with guns. Body camera footage shows that police are less respectful of Black drivers than white drivers that they pull over.
The spotlight has been shone on police bias due to innumerable high profile incidents, and much of the discourse has centered around police being good or bad characters, full of fault or free from blame. The debate gets bogged down in accusations of character flaws and racist cops. And while there are undoubtedly instances of bad cops, the much more nefarious problem is the more common form of racial bias that nearly everyone exhibits, including the most well-intentioned police officer. To accuse someone of bias may feel like an attack on their character, but the reality is that bias is a fundamental part of how our brains work.
Since humans are predisposed to recognize patterns and hang on to associations (erroneous as they may be), we cannot simply extract the bias from the person. Indeed, research in behavioral science shows that most implicit bias training does little to change the behaviors that count. Making people aware of their biases and attempting to correct them in a 2-day seminar is an ill-fated endeavor. Exacerbating this is the fact that we humans have a bias blind spot where we consider ourselves immune to the biases that we readily identify in others. So we excuse ourselves from our role in failing to stamp out institutional racism.
What does help counter racial bias?
What does work, however, is changing systems to correct for these errors. Rather than attempt to debias individuals, we must design systems built for equity. We can increase diversity and positive intergroup relations, and remove social reminders of inequity, such as the Confederate statues that continue to adorn cities and universities across America. We can redesign processes that reinforce systemic racism, and promote environments that reduce prejudice. Technology can be a powerful force in dismantling, or reinforcing, structures that discriminate against people of color, and can be strategically designed to reduce prejudice.
Professor Eberhardt’s research with neighborhood social network Nextdoor shows that significant strides can be taken through the simple redesign of forms. Eberhardt’s changes to the platform helped slow people down to avoid making snap judgments, reducing racial profiling by 75%. All she did was add friction to the process of reporting “suspicious activity,” requiring neighbors to specify what exactly the person they saw was doing that was suspicious, what the person looked like beyond their race and gender, and reminding users that racial profiling is prohibited on the platform.
What this suggests for police departments is the need for structural changes that promote calm, well-reasoned decision-making. It means that police should not be asked to serve in positions they aren’t trained for, such as dealing with mental health issues. It means that bias-inviting procedures should be eliminated, such as the routine traffic stop, foot pursuit, stop-and-frisk, and of course racial profiling. Timely reminders to stop and think can help officers, like all of us, exercise cognitive control over their automatic brains. Police forces can curate their environments to promote equitable decision-making that plays down biases and plays up the reason most people become police officers in the first place: to help communities and the people in them be safe places for all.
We must stop thinking of bias as something that people are or aren’t. We are all categorizers, and fighting bias itself is a Sisyphean endeavor. Once we accept the universality of bias, we can focus on designing systems to eliminate its negative effects, and unite in a more nuanced understanding of what it means to be human. By keeping in mind what works and relying on evidence-based strategies to bring out the best in people, offering real-time reminders and interventions that disrupt harmful automatic judgments, we can work toward Martin Luther King, Jr.’s dream of bending the long arc of the moral universe toward justice.
