Well, frustrated as I am with Medium turning into a paid lefty echo chamber, I’ve recently taken to Twitter to see if there’s any intelligent debate to be had.
I think everyone knows the answer to that question. :-)
That aside, I have been recently engaged with an individual, opposed to me politically it appears, regarding racism, and vis a vis his claim that the GOP is, in its essence, a “racist party”. (Oddly, he took issue with my view that such rhetoric is demagogic, while I view painting with such a broad brush to be irresponsible, prima facie.)
At any rate, while we were both attempting to convince the other that the South was either horribly racist (him) or not nearly as racist as the Dems make it out to be (moi’) this factoid got thrown out by him:
More than half of republicans think AA’s less evolved from a Kellogg study.
More. Than. Half. Urk. Because that datapoint is so out of whack with the General Social Survey and other professional studies, statistician red flags started going off in my mind. Somebody is really screwed up someplace, and it’s either the GSS, Kellogg, or the poster.
So, I did some research, and this quickly became a poster child for how a serious study can quickly turn into “fake news”. So, I thought I’d share. To wit:
- My Twitter adversary quoted an article from The Root, which is an african-american-centric news and blog site, which of course leans left because african americans, in general, lean left. The quote from The Root article was this:
52 percent of voters who supported Donald Trump in the 2016 presidential election believed blacks are “less evolved” than whites, according to researchers at the Kellog School of Management.
- So, you see the first issue: Kellogg’s study was on voters who “supported Donald Trump”, while my Twitter “pal” changed that to “all Republicans”. Bad, bad, boy. :-)
- So, the next question that the statistician in me wants to ask is “how did Kellogg segment the voting population to get “Trump supporters?” Was that ALL REPUBLICANS, or did they use another metric to separate those of us that supported someone else originally (be that Cruz, Rubio, Fiorina, Carson, whoever) from those who supported Trump originally? The Root article links to a Slate article (yes, this is usually not a good thing, if you’re looking for facts) that does, however, answers the question:
Here we compare Trump’s strongest opponents (defined here as those who rate Trump at a 25 or below) to Trump’s strongest supporters (those who rate Trump higher than 75). Twenty-eight percent of white Trump opponents rate blacks as less evolved than they rate whites. In contrast, a majority of Trump supporters — 52 percent — rate blacks as less evolved than whites.
- So, what Kellogg did was use a scientific method to separate the strong Trump supporters from the rest of the GOP. Ergo, that 52% number is not “All Republicans”, but the much smaller number of Republicans who are the strongest Trump supporters.
- OK. Now that we’ve discredited the idea that study proves that the GOP is a racist party, let’s look for additional context to understand exactly what that study showed. And it’s interesting. :-) So, there’s this tidbit:
In this nationally representative sample, a full 38 percent of white subjects rated blacks as less evolved than they rated whites.
Now, that’s a stunning number…..does that mean that 38% of us are racist? Well, studies to determine racism are very difficult, because one has to find a question(s) which excludes anything OTHER than racism as a reason for answering in a certain way. And to put it bluntly, it’s very difficult to formulate a study which eliminates all the “not racism” reasons why a nonracist person would answer the same way as a racist.
So, you can see from the Slate article how Kellogg attempted to solve that problem. And whatever you think of how well they did that, their answer completely runs counter to the General Social Survey question which asks “Would you ever vote for a black president”, where 97% of respondents from the South answered “yes”.
- So finally (and this has gone on too long, already) does the context also dispute the “GOP as racists” meme? Well, yes it does (the GSS questions already dispute that, by finding that there is some — — but not a huge amount— — difference between White Republican racism and White Democrat Racism):
But, at any rate, the Kellogg study did not find much difference between the parties when it comes to racism, either. To quote:
These results are not limited to groups of whites that are often stereotyped as racially prejudiced. For instance, 33 percent of white Democrats and 34 percent of high-income whites rated black people as less evolved than white people, compared with 39 percent of white Republicans and 41 percent of low-income whites. Dehumanizing views are pervasive across white social groups.
Critically, surveys we conducted earlier in the Republican primary reveal that dehumanization of blacks is not strongly associated with support for other Republican candidates such as Sens. Ted Cruz and Marco Rubio.
So, what can we conclude from all this?
Whether you accept the higher numbers of the Kellogg study or the lower numbers of the General Social Survey, it’s clear that both parties have a racist minority hiding within it. The GOP’s is somewhat higher than the Dems, but trying to claim that the GOP is a “racist” party while the Dems are not is sheer demogogery.
The math doesn’t work.