The Press, Race, and Sloppy Surveys

Days before the 2012 election I read an article citing results from a survey The Impact of Anti-Black Racism on Approval of Barack Obama’s Job Performance and on Voting in the 2012 Presidential Election. When a study’s title is longer than a run on sentence, you know you’re in for a joy ride.

I read through the study and found it modestly interesting, but not nearly as interesting as how it was presented and interpreted in the press.

Reading online comments at the time, it was clear many readers took away a simple message that Republicans were racists. Even though the survey indicated “In 2012, the proportion of people expressing anti-Black attitudes was 32% among Democrats, 48% among independents, and 79% among Republicans” apparently Democratic readers seized on the high Republican number to reinforce their existing beliefs. Never mind that the survey indicated one in three Democrats is explicitly racist.

Well, that kind of selective understanding is common enough, but then I wondered how the numbers would look for non-black Democrats and Republicans. Blacks make up 22% of Democrats but only 2% of Republicans(*)  The survey results are actually measuring two things in one variable: they’re combining a) a measure of racism against a single group with b) differences in party racial composition. Since the published survey results did not reveal the more granular results (**) I had to do some simple number crunching…

Assuming blacks are not racist against blacks (***) we find that 41% of non-black Democrats were explicitly racist against blacks vs. 81% for Republicans.

But wait, there’s more. The study’s “implicit anti-Black attitudes” metric indicates “the proportion of people expressing anti-Black attitudes was 55% among Democrats, 49% among independents, and 64% among Republicans”. So we find that 71% of non-black Democrats were implicitly racist against blacks vs. 65% for Republicans.

That’s right. Study finds Democrats more racist than Republicans.

So if you’re a Democrat and you liked the data that reinforced your notions about Republican racism, are you prepared to embrace the other findings? Will you proudly proclaim that “only” 40% of non-black Democrats explicitly display racist attitudes against blacks and that deep down more non-black Democrats are racist than Republicans? If you don’t believe the survey about that, will you reject all of the findings?

And before Republicans get all frothy about how the study shows Democratic racism, remember it shows four out of five Republicans are explicit racists.

Do you really believe these results are accurate after examining them in a bit more detail? Perhaps the survey wording and study definitions lead to unreliable conclusions. It may be better to totally ignore these kinds of widely circulated results about hot button issues unless you’ve got the time to look in detail or find an unbiased, competent person who will do it for you.

Aside:  Some journalists emphasized the top level results that about 50% of Americans harbored these racist attitudes. But would they also then agree that non-black Democrats are implicitly more anti-Black than Republicans? Well the same survey questions and analysis lead to both results. If you reject one, you can’t believe the other. Logically anyway.

* Unfortunately the Hispanic designation makes these data imperfect. These numbers are non-Hispanic blacks. Hispanic blacks are included in the Hispanic category in the referenced data.

** I’ve had a very high success rate contacting and getting helpful responses from various publishers – journalists, government officials, and academics – when I’ve had a substantive question or objection to their work. It is heartening and helpful. I emailed Mr. Krosnick and received no reply. Given the amount of detail in the study results and survey questions I was puzzled by a couple of things and asked, among other things “Is there a reason the study does not present the results excluding blacks for some segments? For example, is there a reason it does not reveal the results for, say, “Non-Hispanic White Democrats”? (Since the attitudes towards blacks are expected to heavily depend on the respondent’s race and since the racial make-up of the Democrats differs meaningfully from Republicans, wouldn’t aggregation at the party level obscure rather than reveal the nature of these attitudes between the two parties?)” Crickets.

*** I’m more comfortable about this assumption for explicit racial attitudes.  But I’ve got to make an assumption. See **.

**** In case you missed it the study is showing that 79% of Republicans are explicitly racist against blacks but only 64% are implicitly so. OK. So I guess there are a bunch of Republicans boasting about their racism, but they’re fooling themselves. They’re so stupid they think they’re racist but they’re really not. Something like that.

This entry was posted in Partisanship and tagged , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s