On March 7th, Harry Windsor and Meghan Markle, former British royals, sat down for a tell-all interview with Oprah Winfrey. The interview raises a lot of questions. Why did Prince Charles stop taking Harry’s calls? Why is “working royal” a job that a person can have in 2021? Why did Oprah pick those glasses in particular?
One question we shouldn’t be asking, though, is whether Meghan really encountered racism at the palace. Irrespective of whether or how particular events occurred, it would have been almost impossible for her not to encounter racism. Even absent anything we know about the royals as individuals, or about the British monarchy as an institution, the social science literature is unequivocal that racism is incredibly common.[1]
Precisely measuring racism is difficult, due to something called social desirability bias: it is not polite to be racist, so if you ask someone outright if they dislike or disrespect people of other races, they might not tell you the truth. If fact, they might not know or admit their racism to themselves.
But social scientists have developed some ways to get around this problem. One approach is a list experiment. In this type of experiment, half of subjects are given a list of four or so innocuous statements (e.g. “I prefer to get my news from the radio”). The remaining subjects are given these same four statements plus an additional, more contentious statement (“I prefer to vote for someone of my own race.”) Subjects then report how many of the statements they agree with, but not which ones, meaning they can agree with the contentious statement without directly saying they agree with it. However, by calculating the average number of “agreements” in each group, and taking the difference, we can estimate how many people in the second group agreed with the contentious statement. This approach is how we learned that at least 42% of white people in the southern US were “angry” at thought of a Black family moving in next door and that only half of Irish people are comfortable with the idea of Black or Muslim people immigrating to Ireland.
Another option is an implicit association test (IAT), which captures evidence of bias before subjects can consciously notice and censor that bias. The test asks subjects to sort images and words to one side of the computer screen or the other, as quickly and accurately as they can. Subjects might be asked, for example, to sort images of Black people and words with positive meaning to the left side of the screen, and images of white people and negative words to the right. Then the words and images would switch, to match Black people with negative words and white with positive. The premise of the study is that if someone is biased against Black people, it will take them a split-second longer to remember to sort the images of Black people into a bin with positive words than to sort them into a bin with negative words. The computer captures this tiny delay and translates it into a measure of bias. The IAT shows that the average person has measurable pro-white bias in the US as well as in Europe
To be entirely clear, neither list experiments or IATs are perfectly reliable, and both can be misused or misinterpreted. But in truth we may not actually need them: a surprisingly high number of people are willing to admit their racism without much sneakiness at all. On the 2012 American Election National Survey, for example, 34% of respondents rated Black people’s average intelligence lower than they rated white people’s; 43% rated Black people as less hardworking. A similar number of Brits (44%) report that “some races” are born harder working than others.
So now what? If racism and bias are that common, does that mean we can stop worrying about them? Surely if everyone is racist, being racist isn’t really a problem, right?
Unfortunately, wrong. Racism does tangible harm, whether it’s intentional or not. Racism in employers means that applications with stereotypically Black names at the top are less likely to result in an interview than identical applications with white names, and both Black and Asian job-seekers are more likely to get an interview if they disguise their race on their resume. Racism in teachers means that children of color are more likely to be suspended and less likely to score well on standardized tests. Racism in medicine means that Black newborns in the US are twice as likely to die when they are cared for by white doctor than a Black one. And not many people in 2021 need to be told about racial disparities in use of force by police.
Besides, even if you have subconscious bias, you can become aware of your bias and prevent it from affecting your words or behavior: if people couldn’t control their racism, we wouldn’t need to worry about things like social desirability bias. So stop trying to figure out if you’re racist — you probably are! — and instead learn what harmful words and behaviors look like, sound like and feel like. And then stop doing those things too.
[1] There is some debate over what, exactly, racism is, and social scientists don’t always define it the way the general public does. For purposes of this post, racism is defined as a belief (conscious or otherwise) that people who are not white are different from, and in some way inferior to, white people.
Wait! We can explain! 