Thread by @magi_jay

Here is my new piece for @AlterNet. Some research came out recently and a lot of people used it to argue that Democrats should not mention race/racism in policy because doing so would be detrimental. The problem? That's not what the research said alternet.org/2021/05/demcor…

So many tweets and headlines carried the same message: mentioning race would hurt Democrats. But the research didn't find that. Mentioning race didn't hurt Democrats w/ all voters, it only hurt them with the subset of Republicans. Yet here were the headlines:

Even the NYTimes goofed this one up. They said mentioning race decreased support among independents for Democratic policies and neglected to include the fact that this was *not* statistically significant and was therefore not an actual effect.

Part of the problem here is w/ lack of statistical literacy in the media. If the effect wasn't significant, you can't say it was observed. Another issue is how the researchers reported the results, such that they stated they found effects when their results were not significant

Of course the media & twitter ate this up. It seemed to confirm priors that mentioning race was bad & focussing on class was good. But the study actually only showed small positive effects for mentioning class & **no** detrimental effect for including race alternet.org/2021/05/demcor…

The only time detrimental effects for mentioning race were observed was for Republicans.
Not Democrats
Not Independents
Not even the entire group of white people!
Only for Republicans

Also: Republicans didn't like the policies in the 1st place & were not moved by class framing

Why does this matter? Well, for one thing, because some people are already using this research to argue Democrats shouldn't talk about race or racism b/c it will harm Democratic causes.

So we need to say: it doesn't actually show that at all. alternet.org/2021/05/demcor…

I'm not just ranting about the misrepresentation of stats for the fun of it. I'm saying we have to confront these misrepresentations head-on because they will falsely inform the belief that we *must* minimize race/racism in order to win. There's no evidence of that in this study.

We should also view all of the evidence at hand. Democrats were told to shut up on social justice after 2016 and then we went on to win in 2018. We were told anti-racism protests would be our demise in 2020.

And then we won again.

In any case, the most important thing here is that if someone holds a media report of this study up to you and says, "We shouldn't talk about race b/c it harms Democrats," tell them they're wrong. The study didn't show that *at all*.

You don't have to know every result of the study, but you should know: headlines saying "Talking about race hurts Democrats" were false. They should've said: "Talking about race makes Republicans like Democratic policy less than they already did"
That's it alternet.org/2021/05/demcor…