A small experiment using quotes with interchangeable races: which group is the most demonized?

This short experiment was published on Lee Jussim’s Substack site Unsafe Science (Lee has an addendum), but is by Michael Bernstein and April Bleske-Recheck, whose bona fides are below:

Michael Bernstein is an experimental psychologist and an Assistant Professor at Brown University. His research focuses on: cognitive biases, the placebo effect, pain, and substance use. He is an editor of the forthcoming book, The nocebo effect: When words make you sick.

April Bleske-Rechek is a differential and evolutionary psychologist. She is a full professor of psychology at the University of Wisconsin-Eau Claire, where she invests in mentoring undergraduate scholars and engaging students with viewpoints and data they are unlikely to be exposed to elsewhere. Her recent publications and presentations can be found on her personal website: bleske-rechek.com.

Click the screenshot to read the provocatively titled post, which in fact is appropriately titled. 

I could explain what they did, but it’s simple and they explain it themselves in the short post. The default hypothesis, I think, is that people would be more willing to accept quotes that denigrated white people than denigrated Jews or black people (whites are at the bottom of the “privilege” heap in today’s young people). Quotes from the article are indented:

Let’s play a game. It’s called “Who said it: Robin DiAngelo or Adolf Hitler?” DiAngelo, in case you don’t know, is author of the NYT best-seller book, White Fragility.  We’ll give you a couple quotes and you think about whether it’s from DiAngelo or Hitler. Ready, go.

1)    Did Hitler say: “Not having a group consciousness, Jews often respond defensively when grouped with other Jews.” Or did DiAngelo say: “Not having a group consciousness, Whites often respond defensively when grouped with other Whites.”

2)    Did Hitler say: “Jews… creep up on the workers in order to win their confidence, pretending to have compassion.” Or did DiAngelo say: “Whites… creep up on the workers in order to win their confidence, pretending to have compassion.”

For the record, the first quote is from DiAngelo and the second quote is from Hitler. Though whether you were right isn’t exactly the point, as an astute reader would probably know Hitler is likely to use the language of “workers” and DiAngelo is likely to use the language of “group consciousness”. The point is that DiAngelo and Hitler are both advocating an approach that reduces behavior to group membership. They describe the behavior of all Whites or all Jews in highly critical terms and conclude that this is the nature of Whiteness or Jewishness.

We know from decades of psychological research that people hold prejudices. But which groups in today’s society are more likely to be the target of expressions of prejudicial attitudes? And who is more likely to express them?

We decided to examine this empirically. Would agreement with the same statement, whether it be anti-White, anti-Black, or anti-Jew, vary depending on which group it referenced? And would political affiliation moderate attitudes

We took 3 real anti-Jew quotes from Adolf Hitler, 3 real anti-White quotes from Robin DiAngelo, and 3 real anti-Black quotes from Stephen Douglas. (Douglas was a 19^th century American politician who debated Abraham Lincoln). Then, we created anti-Jew, anti-White, and anti-Black variations of each quote, and showed it to 428 college graduates or college students (72% White). This means that 1/3 of participants saw the real quote verbatim, whereas the other 2/3 saw a version of the quote that was manipulated by changing the original (e.g., replace “Jew” with “White” or “Black”, or any other combination thereof). This is shown in the Table below. For each quote, participants were asked to imagine that an intellectual or political leader uttered the statement. They then indicated whether they agreed with the statement by selecting: “definitely no,” “probably no,” “probably yes,” or “definitely yes.” Participants answered this question for all nine quotes, and all were in the same frame (anti-Jew, anti-White, or anti-Black).

Note that whites were 72% of the sampled population of 428, which jibes with the 76% of whites we have in America if you count Hispanics as “white” (otherwise the country is 59% white.)

Here are the quotes, with the words targeted for replacement shown in red. So, for example, you could take the first Di-Angelo quote and substitute in “Jews” or “Blacks” for “whites”, and see if people agree or disagree with the quote. You’d also see whether they agreed with the quote as it stands.  (DiAngelo is, to my mind, off the rails.)

There are no statistics given, so I’m not sure which differences could be due to chance, but here are the results, which are as expected:

The results were surprising. For 7 of the 9 quotes, agreement differed according to target group. On each of these, agreement was highest in the anti-White condition versus the anti-Jew and anti-Black condition. The figures below show the percentage of college graduates (left) and college students (right) who either “probably” or “definitely” agreed with at least one statement, broken down by target group and the original author of the quote (Hitler vs. DiAngelo vs. Douglas). You can see that agreement with both Hitler and DiAngelo is much higher in the anti-White condition versus the other two conditions. Hardly anyone agreed with the Douglas quotes regardless of target group.

Douglas was a diehard racist, although in the Lincoln-Douglas debates Lincoln (to some extent) took out after him. At any rate, here are the data. College students (right) are much more likely to agree with antiwhite statements put into the mouths of Hitler and DiAngelo, while nobody agreed with Douglas no matter which group he was said to criticize.  Nearly 50% of college students agreed with the “antiwhite” statements of DiAngelo as well as when “white” was substituted for “Jews” in Hitler’s statements.  The percentages were lower for both when college grads were tested, but still about 35% agreed with Hitler’s spurious “antiwhite” statements.

There are other data (below) showing that antiwhite sentiments were higher among liberals than conservatives, but anti-black and anti-Jewish sentiments were higher among conservatives.

Clearly, though, antiwhite sentiments were the least offensive, though I suspect most of the students were white.  That is as expected, though Jews didn’t do as badly as I thought (most Americans still abhor antisemitic statements, though I predict that the “anti-Jew” numbers will rise in the future.

The authors conclude that their results aren’t really new, and then offer two caveats:

In a sense, our results are nothing new. We simply observed what has existed for millennia: People treat some groups preferentially to others. In the Bible, the Egyptian Pharoah enslaved the people of Israel. And interestingly, even God responded to this tribally by establishing Passover which “The whole community of Israel is to be included on the meal,” but “no foreigners are to eat it.”2

Sweeping claims about all members of certain demographic groups seem to be on the rise in some circles. But unless you’re tuned into a relatively small number of heterodox writers like Bari Weiss or Coleman Hughes, you will rarely hear someone speculate about the counter-factual (e.g. Bob said all Whites do X vs. Imagine if he instead said all Blacks do X).

The caveats (the first one is funny but, I suppose, necessary):

An incomplete list of disclaimers that should go without saying:

1)      Nothing in our essay is meant as an argument that DiAngelo is as evil a person as Hitler, or for that matter, evil at all. Hitler is responsible for the murder of 11 million people and the death toll from just the European theater of World War II was at least 40 million. DiAngelo is not responsible for the death of anyone. But we can recognize this fact while still pointing to similarities in their thinking.

2)      Just because a person agrees with a quote from Hitler does not mean that person agrees with Hitler’s genocide.

I should add another: this is a small experiment with no statistics (perhaps they’ll be in a subsequent paper). The differences are sometimes large (clearly Douglas’s quotes are despised no matter which group is substituted in them), but I’d like to see some error bars.