Between July 2001 and May 2002, Marianne Bertrand sent out over 1300 job applications. She wasn’t looking for work—since earning a PhD from Harvard in 1998, she’d been employed full-time as an economics professor—but instead she wanted to see if the name on the resumé affected the number of callbacks from potential employers. Specifically, half of the resumés had white-sounding names (Emily Walsh, Greg Baker) and half had black-sounding names (Lakisha Washington, Jamal Jones).

What she found was disheartening. White-sounding names received 50 percent more callbacks. About one in ten resumés with a white-sounding name received a callback, compared to one in fifteen with a black-sounding name.  The racial bias in responses cut across all industries, from fast food to accounting. Even companies that advertised themselves as equal opportunity employers discriminated at roughly the same rate.  

Bertrand’s finding raises questions: if most people do not want to appear racially biased—for example, in George W. Bush’s memoir, Decision Points, he considered being called a racist his most painful memory as president—why does it persist? Further, how can we get rid of it?  

Understanding two different systems of thinking helps explain where bias comes from and why it’s reluctant to go away.  One system performs fast, intuitive thinking and comes to automatic judgments all without any conscious control. The other computes effortful tasks, such as mentally calculating 18x29, is slower and requires more attention. The fast, automatic system is responsible for biases.

To get acquainted with how intuitive thinking works, consider this example from Daniel Kahneman’s book, Thinking, Fast and Slow:

Banana                              Vomit

When reading those two words, your heart rate quickened, your face contorted into an expression of disgust, and you may have imagined a cause and effect relationship between the two words, creating a scenario where bananas led to vomit. This all happened without any conscious influence. 

A 2000 study, led by Liz Phelps, a cognitive neuroscientist at New York University, suggests that the amydgala, a brain region involved in fear-processing and the fast, automatic thinking system, drives racial bias. Among other functions, the amygdala guides behavior by forming associations between experiences and unpleasant reactions, such as avoiding Taco Bell if you get sick after eating their chalupas. It also generates the emotional reaction to words, sights or ideas you find unpleasant. Your amygdala was likely active when you read the word “vomit.”   

To see if the amygdala was involved in racial bias, Phelps devised a simple experiment. She showed twelve white undergraduates pictures of faces of black and white male strangers while they were in an MRI scanner.  She also measured racial bias using a test designed to root out implicit, unspoken opinions that someone may suppress for fear of appearing politically incorrect or may simply be unaware of. In the test, subjects categorize words as either good (joy, love, peace) or bad (cancer, death, war) at the same time that they categorize the faces they saw in the MRI machine as black or white. In half of the trials, they use the same response for good and white (left button for good/white, right for bad/black). In the other half, the pairing switches and the they use the same response for good and black  (left for good/black, right for bad/white). Someone with greater racial bias finds the task harder when good is paired with black and it takes them longer to press the correct buttons during those trials. Bias is how much longer someone takes on good/black compared to good/white trials.

As expected, participants with stronger racial bias also had the most activation in their amygdala when viewing black faces compared to viewing white faces. The same people who were the slowest at responding when black was paired with good had the greatest amygdala activation when viewing black faces. Their slower response time suggests that they have to override the automatic response of the amygdala that tells them that black should be paired with bad, so it takes them longer to press the correct button when black is paired with good. Less amygdala reaction means fewer competing thoughts to prevent someone from pressing the left button for both black and good.

These results suggest that of the two systems of thinking, fast, automatic processing is more involved in racial bias. They also explain why bias stubbornly persists even if our cultural mores tell us it’s wrong. Negative thoughts crop up automatically, before we’re consciously aware of them, so tamping them down requires extra mental effort.

Phelps’s results help explain why Marianne Bertrand received fewer callbacks for the resumés she sent out with black-sounding names.  If fast, automatic signals from the amygdala—gut feelings—underlie racial bias, it can occur without conscious awareness. The amygdala guides us to avoid cues we associate with unpleasant experiences, even if we aren’t aware of the association. An employer may develop a gut feeling about a resumé after reading the name and subconsciously decide to move it to the discard pile before reading the applicant’s qualifications. That brings up the second question, how can bias be reduced?  

While it’s troubling that even people who embrace equality have subconscious biases, knowing that their slant is related to amygdala activation offers a target for correcting the problem. A study out this year in Psychopharmacology, led by graduate student Sylvia Terbeck at the University of Oxford, set out to see if reducing negative gut feelings could curb bias.   

Terbeck used a drug called propranolol, a beta-blocker that has proven successful at treating post-traumatic stress disorder, or PTSD. PTSD usually develops after a terrifying experience and it involves vivid recall of the experience along with crippling anxiety. The beauty and beast of the brain lies in its web of connected neurons, which allows us to perform complex thoughts but also means that recalling a traumatic event also reignites everything associated with it, including terror. Propranolol reduces anxiety by disrupting the actions of epinephrine and stress hormones by blocking beta-receptors (hence beta blocker); many musicians take it before performances to ease stage fright. For PTSD treatment, taking propranolol before therapy sessions helps patients feel relaxed while recalling a traumatic event, linking the memory of the event to calmness rather than fright.      

In the study, Terbeck gave the same test of subliminal racial bias that Phelps used to 36 white male undergraduates. Terbeck also tested open prejudice, the kind someone is consciously aware of. Using a thermometer that ranged from 0-100 degrees, she asked how warm they felt about white and black people. The difference between their temperature for white and black people was their level of prejudice. Half of the participants were given a capsule filled with 40mg of propranolol before the tests, and the other half were given placebo.

As expected, propranolol didn’t affect open prejudice. The groups that took placebo and propranolol showed the same difference in warmth between white and black people. Reducing anxiety had no effect on conscious opinions of race.  

However, the males who took propranolol showed significantly less subconscious bias. On the test, the group that took placebo took over a second longer on average to press the button for black/good compared to white/good. The group that took propranolol took just a quarter of a second longer to press the button for the black/good on average, an improvement of three-quarters of a second. Further, over a third of the participants who took propranolol hit the button faster for black/good compared to white/good. No one in the placebo group was faster in the black/good condition. By reducing anxiety, propranolol also wiped out subconscious racial bias.

These results show that propranolol can diminish bias by curtailing negative thoughts. Like in the treatment of PTSD, where pairing calmness with traumatic memories numbs the pain associated with the event, racial bias may be shelved by reining in the fearful amygdala response associated with other races.

Although it may be impractical to require employers to take propranolol before looking through job applications, this study offers hope for fairness. Knowing how bias forms and how its knots can be untied, we stand a better chance at overcoming it. Like appreciating other cultures, the neuroscience of racial bias has humble beginnings: it starts with understanding.

image credit:


Bertrand, M and Mullainathan, S (2003) Are Emily and Brendan More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination. (Working Paper). Chicago, IL: Graduate School of Business, University of Chicago

Phelps, E.A., O'Connor, K.J., Cunningham, W.A., Funayma, E.S., Gatenby, J.C., Gore, J.C., Banaji, M.R. (2000). Performance on indirect measures of race evaluation predicts amygdala activity, Journal of Cognitive Neuroscience, 12, 1-10.

Kahneman, D (2011) Thinking, Fast and Slow. Farrar, Straus and Giroux. New York, NY

Terbeck S, Kahane G, McTavish S, Savulescu J, Cowen PJ, Hewstone M. (2012) Propranolol reduces implicit negative racial bias. Psychopharmacology (Berl). Aug;222(3):419-24.

Recent Posts in You, Illuminated

How the Brain Processes Keeping up With the Joneses

The brain may use two systems for social comparisons

Long Term Effects of Marijuana on the Brain

Part 3 exploring how marijuana alters brain function: long-term effects

7 Short-Term Effects of Marijuana on the Brain

Part 2 exploring how marijuana alters brain function: short-term effects

What Does Marijuana Really Do to the Brain?

Part 1 exploring how marijuana alters brain function: the mechanism.

Making Learning Fun

An interview with Sooinn Lee about teaching math the fun way

Have We Discovered a Hidden Danger of Pornography?

A new study gets a lot of attention, but what does it mean?