When Latanya Sweeney Googled herself, the African-American professor at Harvard University came across a surprising search result indicating she might have an arrest record.

After determining she didn’t, Sweeney decided to conduct a study of Google AdSense to see if it had a racial bias towards people with “black sounding” names.

Her conclusion: “black-identifying names” (such as Aaliyah, Lakisha and DeShawn) were much more likely than “white-identifying names” (Jill, Emma and Geoffrey) to generate ads that included the word “arrest” (60% vs. 48%).

According to Sweeney, the odds are less than 1 in 1,000 that the disparity was a result of random chance. She adds that “the ‘Google algorithm’ learns over time which ad text gets the most clicks from viewers of the ad. It does this by assigning weights (or probabilities) based on the click history of each ad copy.”

Google responded to research by saying the company “does not conduct any racial profiling.”

-Noel Brinkerhoff

To Learn More:

Discrimination in Online Ad Delivery (by Latanya Sweeney, Harvard University) (pdf)

Is This the New Face of Racism? (by James Taranto, Wall Street Journal)

Google Searches Expose Racial Bias, Says Study Of Names (BBC News)