A Bonus (!) Social Science Roundup (Feb. 11, 2022)
Brand spanking new social science research to tickle your fancy
Welcome to the first bonus post of the Holbein Social Science Roundup! Don’t you feel lucky? :)
This week’s roundup focuses on racial bias—first, how we measure it and second, what its implications are in all aspects of society.
Subscribe if you’re new here!
And if you have a study you think I should feature here, feel free to email me: holbein@virginia.edu.
How do you measure racial bias? This may seem like a question with an obvious answer, but there are reasons to think that’s not true. It may also seem unnecessary to measure racial bias when we can look out of our (proverbial) windows and see bias in many aspects of modern society. However, there are many good reasons for wanting to systematically quantify and document racial bias and to do so well; measuring bias—and understanding where it is/isn’t, who exhibits it, and why they do so—is the first step on the path to addressing bias.
So, how do social scientists measure racial bias? It’s hard! But, there are two dominant approaches, broadly conceived, to measuring bias. The first uses survey questions to try and get respondents to self-report bias. The second uses observations of individual behavior, often—but not always, in a controlled environment like a lab.
One survey-based approach to measuring racial bias is the Racial Resentment Scale. This scale was developed by social scientists Donald Kinder and Lynn M. Sanders for the American National Election Studies in the 1980s. They who wanted to understand the levels of bias in our society and its causes and consequences.
Here’s what goes into the racial resentment scale. In a survey, respondents are asked to answer to what extent they agree/disagree with the following statements. From good ol’ Wikipedia:
The racial resentment scale is widely used in the political and social sciences, partly because collected survey data is so cheap and representative survey samples are so ubiquitous. See…
However, three recent papers pose challenges to what the Racial Resentment Scale actually measures.
In this newly published piece at The Journal of Politics, by Matthew T. Pietryka and Randall C. MacIntosh, the authors argue that the Racial Resentment Scale (along with other scales measured in the ANES) doesn’t measure what we think it does. Succinctly, they argue that “the Racial Resentment Scale conflates policy views with racial animus.”
Another paper also recently published at The Journal of Politics, by Kyle Peyton and Gregory A. Huber, takes a slightly different tact. The authors examine whether the Racial Resentment Scale predicts discriminatory behavior in a controlled environment. They find that “resentment and prejudice are distinct constructs.” That is academic for “racial resentment isn’t as strong of a predictor of behavior in a controlled environment as you might think.” Fascinatingly, they also find that “explicit prejudice is widespread among white Americans and significantly less partisan than resentment.” Wow!
This working paper from Riley Carney and Ryan Enos lofts a similar critique: “Modern racism scales, which are the most common way to measure anti-Black prejudice in political science, were created in response to a shift in the attitudes of white Americans toward African Americans, and reflect a mix of social conservatism and anti-Black affect…Modern racism scales measure attitudes toward any group, rather than African Americans alone.” They then call for “new instruments … to measure group-specific prejudice.”
So, where does that leave us? I have two thoughts. (I’d love to hear if you have other perspectives!)
First, I think many of these criticisms of the racial resentment scale do, in fact, land. I believe that the scale isn’t perfect. I buy that is capturing many different attitudes. And I subscribe to the belief that social desirability may play a big role in self-reports of bias. That said, I read the original scholars who collected these scales as acknowledging with that fact in their early work. That said, I would love to see a counter-perspective on the issues that are brought up from those who develop/use resentment scales. I agree with Carney and Enos that we need more work in this space!
Second, the issues brought up above are why I—in my own research on bias—focus on behavioral measures. If you’re interested in learning more about behavioral measures of racial bias, you might check out this approach that I and my colleagues have developed…
Or for an overview/meta-analysis of the work that other scholars have done, see here:
Here are a few other recent cool studies of behavioral measures of racial bias…
Racial bias in soccer referee calls: “Darker-skinned [soccer] players receive more foul calls and more cards than lighter-skinned players, controlling for a range of confounders… Fans may play a key role in inducing poor calls against darker-skinned players.”
Confederate memorials can have direct effects on asset markets: “Houses on streets with names that honor the Confederacy sell for 3% less than other similar nearby houses…Confederate listings are also more likely to experience long durations and sell at large discounts relative to listing prices.”
That’s all! See you all next week!
Update #1: Yusaku Horiuchi (and Jonathan Robinson) pointed me to his article (with colleagues) on racial resentment. They find that “the Racial Resentment Scale captures favoring of Blacks substantially more than disfavoring.”