A Very Interesting Public Opinion Poll
I think I first saw this linked at SlateStarCodex, so HT2 him, but this recent NPR / PBS News Hour / Marist survey of 1,125 adults from across the nation has some interesting results.
First, on the question of what to do with Confederate statues, among African Americans 44% said they should “remain as a historical symbol,” while 40% said they “should be removed because they are offensive to some people.” (The margin in favor of “remain” was much higher among whites and Latinos.) Since everybody except Jim Buchanan is in favor of democracy, I guess this means…?
What’s really bizarre is that on the question asking if the person agrees with the beliefs of the “white supremacy movement”:
==> 3% of whites said they “Mostly agree” while 88% said they “Mostly disagree.”
==> 4% of African Americans said they “Mostly agree” while 90% said they “Mostly disagree.”
==> 7% of Latinos said they “Mostly agree” while 76% said they “Mostly disagree.”
The above results have made the rounds, with chortling bloggers understandably announcing, “More blacks and Hispanics agree with white supremacists than white people do.”
I’m not sure what to make of these results–I’m guessing there might be some trolling and misunderstanding of the question–but I think it’s safe to say that America is not on the verge of going skinhead. Even if you say, “Well it’s because the white white (sic) supremacists were afraid to admit it,” OK that’s part of the point. Everybody knows this is socially unacceptable in our culture.
A while back a polling company asked people whether they had ever been decapitated and 4% said yes. So there’s presumably 4% of people who will say yes to any question.
Frequently I see bloggers, especially left wing ones, go off on polls, particularly polls where some sizable fraction say something insane. They always take the number at face value. But insane answers usually correspond to a really stupid question. A lot of people will take the piss if they are asked a stupid question. One I recall was where does chocolate milk come from? Well about 10% picked the answer brown cows. So people were lamenting the public’s ignorance, and how this ignorance proved (surprise) that the blogger was right about everything. none seemed to imagine anyone might be joking.
The same survey lists KKK support as high amongst “Strong Democrats” and also “Latino” race (whatever that means) but these type of surveys have high statistical error and I suspect a lot of people deliberately say stupid stuff to phone surveys, because they are fed up with being called all the time.
Tel I haven’t looked back on this poll but I think someone else claimed it was an “in person” survey.
From the link:
‘Adults 18 years of age and older residing in
the contiguous United States were contacted on landline or mobile numbers and interviewed in English
by telephone using live interviewers. ‘
Ah OK maybe the person I read was stressing the “live interviewer” thing and I misunderstood. (Or maybe the person was wrong.) Thanks.
I saw the section on whether people thought race relations in the USA today were better or worse that a year ago. I noticed that 52% thought it was worse and thought this could be a damning indictment of Trump. However, looking at the figures from past surveys, 58% thought they were worse in Sept 2015.
This is interesting as people apparently think that race relations have been getting worse since 2014 and this appear not to have anything to do with Trump.
I think there were quite a lot if high profile stories about police actions against blacks in this period.
My hypothesis is that race relations have not generally been getting worse but awareness of how bad they are has grown.
A quick look down the results seems to show that Tea Party supporters are much more extreme than Trump supporters.
Anyone who understands how statistics works would note that there were twice as many Republicans in the poll who do NOT support the Tea Party as there were who DO suppose the Tea Party… thus we would always expect to see larger percentage variation on the smaller sample group (a.k.a. “Central Limit Theorem”).
The same happens with “Latinos”, who only composed 10% of the total sample, and therefore would have a statistical error approx 3x larger than the overall statistical error. Now, go and find a journalist who can comprehend these things and who might attempt to report it properly… good luck!
Any such reporter would be reassigned immediately.
The numbers are low and I would not draw firm conclusions, but for each if the right wing organisations Tea Party support is higher than among Trump supporter – Numbers below (Tea party/Trump).
White supremacists 6/4
White nationalists 11/6
KKK 4/1
Alt-right 17/7
If it was just scatter you would not expect all to be one way. However, if we look at the “mostly disagree” they are much more similar. Also there is the same pattern for Latinos – they support all the same organisations more than Trump supporters too, and support White Supremacists more than the tea party! It may be just scatter.
Perhaps this should not be surprising since Tea Party is specifically a “right wing” organisation in the colloquial sense, whereas Trump is something different, with some traditionally “left wing” ideas. However, if Trump is anything he is Nationalist.
Right Tel: very small numbers of Latinos and blacks in survey mean high sample std dev.
My overwhelming sense is that, for an efficient economist, it makes much more sense to vote than to waste time writing on white supremacists.
The story is told of a recent convention where there were 150 participants and 150 reporters (I’d look it up but see my opening sentence). The story has nothing to do with neo-nazis/white supremacists and everything to say about the “reporters”writing news rather than reporting the news.