Do White Liberals Talk Down to African Americans? New Research Hints At ‘Yes’

A new study out of the Yale University School of Management has found that white liberals use less competent-sounding words when speaking with blacks than they do when talking with whites. White conservatives, in contrast, use similar words with both groups.

Let that sink in for a moment. White liberals tend to avoid words that might make them look very  competent when talking with blacks. White conservatives don’t. What does that suggest about how white liberals view blacks?

Two Possibilities

The full report hasn’t been published yet, but lead research Cydnee Dupree wrote up a summary for the web. It’s titled “White Liberals Present Themselves as Less Competent in Interactions with African-Americans.”

If findings hold up in further research, there are at least two ways to interpret them.

I’ll begin with the less obvious one: The difference could come from liberals’ desire to appear likable to blacks. There’s tons of research on “likability,” and its results are consistent. People come across as more likable by showing their weaknesses than their strengths.

So that’s one possibility. The other possibility was “kind of an unpleasant surprise,” says Dupree. “Even if it’s ultimately well-intentioned, it could be seen as patronizing.” That’s putting it gently. More bluntly, it could be that white liberals think they have to talk down to African-Americans.

Unconscious Bias

Of course it’s unintentional. Liberals don’t actually intend to use less competent-sounding words when speaking with blacks. They do it unconsciously instead — which lands this study in line with another famous (or infamous) test of unconscious bias, Harvard University’s Implicit Association Test.

On a first view at least, this new study appears much more robust than the IAT. The IAT has been online since 1998, and millions of people have taken it. It claims to detect bias people are completely unaware of, based on how quickly they relate black and white faces to positive and negative words. It’s all very science-y — it comes down to differences on the order of just a few tenths of a second. It almost feels magic: “You mean they can scope out my biases that way?”

It’s fueled a movement of “implicit bias training” in colleges and corporations. For all its Harvard origins and its highly influential two decades of history, though, the IAT has been pretty much debunked.

Debunking the IAT

It’s unreliable, for starters. Reliability, in this context, is a lot like repeatability. IAT scores aren’t stable; persons often get significantly different scores when they re-take the test. It’s hard to believe underlying biases would alter as fast as as these changing scores seem to indicate.

It’s also no help at all in predicting how people will act in real-world settings. Liberals and conservatives score the same on it. More or less educated people score the same. Older and younger people score the same. Real differences only show up when blacks are compared with whites, or when people admit outright they prefer one race over the other. It’s more than just a little bit silly, though, to call those biases “unconscious,” and there’s hardly any need to go to all the trouble of the IAT to find them out. 

And to top it all off, there are a whole list of other possible causes to explain what’s really going on when people take this test.

Yale Research Appears Much Stronger

The new Yale research is completely different, however. Completely.

It comes in two flavors, for starters. In one study, researchers asked volunteers to compose emails to people with stereotypically white and black names, like “Emily” and “Lakisha.” Participants tended to less competent-sounding words in emails to blacks than to whites.

I don’t have the report in hand to see what researchers did to assure that test would be reliable. At least it’s got good face validity, meaning it actually looks like it’s measuring what it claims to be measuring. That by itself sets it apart from the IAT’s science-y hocus-pocus.

And there’s no question that the other side of this new research relates to how people behave in the real world. Especially presidential candidates.

Researchers examined actual campaign speeches over a 25-year period, and found a very interesting difference. Democrats’ speeches for mostly-minority audiences used fewer high-competence words than their speeches for mostly-white crowds. Republicans’ speeches treated both audiences the same.

What’s Going On?

So what’s really going on here? Do Democratic candidates speak this way to African-Americans so they can seem more likable? Or are they talking down to them? Whichever is true, does it also go for liberals in other interactions with minorities? How strong is this white-liberal unconscious bias? And again, what does it say for white liberals’ true attitudes toward blacks?

These questions all come out of a preliminary report; we need to see more details from the current research. Follow-up studies will be needed, too, to confirm the results and sort out the different possible interpretations. So it’s too early to say for sure. But it’s not too early to wonder.