Monday, May 16, 2016

How bigoted are Trump supporters?

Jason McDaniel and Sean McElwee have been doing great work analyzing the political movement behind Donald Trump. For example, they've showed pretty conclusively that Trump support is driven at least in part by what they call "racial resentment" - the notion that the government unfairly helps nonwhites.

But "racial resentment" is not the same thing as outright bigotry. Believing that the government unfairly helps black people doesn't necessarily mean you dislike black people. So McDaniel and McElwee did another survey asking about people's attitudes toward various groups. Here's a graph summarizing their basic findings:

So, from this graph, I gather:

1. Trump supporters, on average, say they like Blacks, Hispanics, Scientists, Whites, and Police. On average, they say they dislike Muslims, Transgenders, Gays, and Feminists.

2. Trump supporters, on average, say they like Whites a bit more than average, Muslims a lot less, and Transgenders a bit less. They also might say they like Hispanics, Gays, and Feminists somewhat less, though the statistical significance is borderline.

Now here's how Sean McElwee interpreted this same graph:

This interpretation doesn't appear to be supported by Sean's own data. In fact, his data appear to support the opposite of what he claims.

Now, the main caveat to all this is that surveys like this almost certainly don't do a good job of measuring people's real attitudes toward other groups. When someone calls you on the phone (or hands you a piece of paper) and asks you if you like Hispanics, whether you say "yes" or "no" is probably much more dependent on what you think you ought to say than what you really feel. So this survey is probably mainly just measuring differences in how Trump supporters feel they ought to answer surveys.

But even if this survey really did measure people's true attitudes, it still wouldn't tell us what Sean claims it does. Trump supporters, overall, say they like Blacks. And the degree to which they say they like Blacks is not statistically significant from the national average. Only when it comes to Muslims and Transgender folks do Trump supporters appear clearly more bigoted than the national average.

But going back to the main problem with surveys like this, it might be that Trump supporters are simply more willing to express their dislike of Muslims and Transgender people in a survey. This may just reflect their general lack of education. More educated people are plugged into the mass media culture, which generally discourages overt expressions of bigotry toward any group. Less educated folks are less likely to have gotten the message that you're not supposed to say bad things about Muslims and Trans people.

So in conclusion, this survey doesn't seem to support the narrative that Trump supporters are driven by bigotry. That narrative might still be true, of course - there are certainly some very loud and visible bigots within Trump's support base (and within his organization). But after looking at this data, my priors, which were pretty ambivalent about that narrative to begin with, haven't really been moved at all.


  1. What are you talking about? I love groups! Some of my best friends are in groups!

    1. I would like to allocate this comment 93.74 units of "like".

  2. On the matter of education, arguing that there is a bias because Trump supporters are less educated is probably not going to fly. They certainly have that image, but in fact some recent studies (sorry, no links) have suggested that they are above average in income, which probably suggests that they are not obviously below average in education.

    1. Anonymous9:57 AM

      Less educated still seems to fly. Might be that people with better education have been turning (even if reluctantly) to Trump, but his core has been among the white working class.

  3. Like less than average, even if not significant. Garbled but not difficult to interpret.

  4. Must be liars and bigots. I learned from PK that all Trump supporters hate THOSE people. And PK has ever lied right? Personally I think Trump is horrible...and I'm still going to vote for him over Hillary.

  5. Anonymous5:52 PM


  6. Anonymous8:02 PM

    I'm not sure if asking "do you like group x?" is the most effective way to determine bigotry (or lack thereof). Trump supporters might not dislike (or admit to disliking) African-Americans to an unusual degree, but they might dislike "thugs" or express disgust at rioting/looting in Ferguson and Baltimore to an extra normal degree. They might not dislike Hispanics as a group to an unusual degree, but dislike "illegals" and have negative perceptions of the impact of immigration on wages, employment and the public fisc to an unusual degree.

    And of course, they might not. But I think use of the nominal language of anti-racism has become so bipartisanly dominant as to reshape the expression of prejudice.

    Also, this is somewhat tangential to OP, but as a liberal I wish liberals were more capable of talking about empirical inter-group differences, and the origins thereof. I've often seen left-of-center publications like Vox include specific data about e.g. the higher rates at which African-Americans are arrested (4x IIRC) for drug possession despite equal rates of use in articles about criminal justice. But in the same article, when it comes to e.g. differences in homicide rates, no specific data is provided and there's immediate hand waving about vague "structural factors". It seems like its impossible to be a lefty in good standing and publicly state "African-Americans, on average, commit homicides at 7-8x the rate of whites, on average", even though that's a value-neutral empirical claim. It also seems like making the seemingly pretty milquetoast value claim "it should be a goal of public policy to reduce the rate at which African-Americans commit homicides to the age-adjusted level of other major ethnic groups" in public would earn you some suspicious glances. Additionally, I often hear "poverty" pronounced the main cause of inter-racial crime rate differences, but this overlooks the fact that poor white communities (often in Appalachia) and U.S. Hispanics as a whole have raw economic outcomes very comparable to poor African-American communities, but significantly (IIRC often 2-3x) lower rates of violent crime. To address the obvious elephant in the room, I don't think it's genetics so much as political institutions. (Like most people who have read the Better Angels of our Nature, I imagine.) But as long as the only people willing to write and think about race and crime, particularly in quantitative terms, are "race and IQ" alt-right nutjobs, there's a pretty significant hole in the liberal weltanschauung.

  7. what they call "racial resentment" - the notion that the government unfairly helps nonwhites. <<
    Let's see -- isn't there this policy called "Affirmative Action"? And isn't the purpose of this policy to help nonwhites? -- or at least, blacks, maybe also Hispanics, tho not Asian Americans.

    So, why not ask the question: Is Affirmative Action a racist policy?
    Well, what about it Noah, is Affirmative Action a racist policy?
    If you can't answer this question honestly, you're being ... intellectually dishonest.

    Note that I mostly like your blog and usual willingness to look at issues: >>
    it might be that Trump supporters are simply more willing to express their dislike of Muslims and Transgender people in a survey. <<
    This idea that Trump supporters might be more honest in expressing themselves seems very possible, even likely (60%), and most Rep-haters wouldn't even include this idea.
    So please keep up the good work you do, even if you can't be honest about Affirmative Action.

    1. I'm just curious. Traditional racism might be compared with a vital running race, in which blacks are burdened with cubic-foot concrete blocks on their feet, and are horsewhipped every time they go around a curve.

      1. Is affirmative action that bad towards whites?

      2. Does affirmative action come anywhere close to compensating for that treatment? Let alone stop the current hidden forms?

      In other words, does affirmative action amount to anything more than removing a minuscule fraction of the unfair advantage accruing to white persons?

    2. Anonymous6:49 PM

      1. Affirmative Action mostly damages Asian-Americans, at least in elite university admission, not whites.

      2. The black beneficiaries of Affirmative Action in college recruitment (where it's most widely used) are disproportionately Afro-Caribbean and African immigrants whose ancestors weren't disadvantaged by U.S. slavery. The non-immigrant black recipients are disproportionately from wealthy, well-educated families.

      3. Even if descendants of U.S. slaves were the primary black beneficiaries of college Affirmative Action, it has been expanded to benefit Hispanics. It's not really clear to me on what distributive justice grounds the Hispanics who benefit from Affirmative Action deserve special treatment as opposed to the Asians who lose out from it.

      4. The existence of greater injustice somewhere else doesn't negate a particular injustice. Funnily enough, arguing otherwise is usually a conservative rhetorical technique.

      5. Obviously, the real (at least de facto) purpose of Affirmative Action isn't to make America a more racially just society. It's to provide a thin veneer of "diversity" (i.e. over representation of low performing non-white ethnic groups, not literal diversity) to colleges dependent on large federal grants and implicit subsidies to serve as a shield against discrimination lawsuits. Noah has previously stated that he doesn't think Affirmative Action is a problem because "private" universities do it, but this drastically misunderstands how American higher education works. The vast majority of U.S. colleges are legally run as non-profits; they receive charitable contributions to perform a quasi-public mission. As Thomas Sowell has noted, this is why they were both able to discriminate against blacks and Jews in the past and able to discriminate against Asians and to a lesser extent whites in the present. Universities aren't subject to market discipline (if they were, as the number of qualified applicants to Ivies increased the number of seats offered would have increased), so they can practice racial discrimination with impunity.

  8. The problem with racial resentment is that it results in the same outcomes as racial hatred. Even in the good old days down in Dixie it wasn't about the hatred of Negros - they had Negros back then - at least not as long as they knew their place. Even now you'll run into paeons to the good n----r.

    (This reminds me of the old joke about the Southern slave owner who said, "I have nothing against Negros, I own hundreds of them." You'll recognize the mentality.)

  9. This comment has been removed by the author.

  10. You might interpret this data in another way, one which seems to confirm something I've become more convinced of over time. The variety of bigotry that we see more often targets beliefs, or what the bigots perceive to be a choice. This, rather than the more overt form of bigotry which is racial discrimination. In other words, bigotry on account of belonging to a race is unjustified (in the mind of the modern bigot) because one cannot choose their race. On the other hand, a person chooses to be a Muslim and (again, in the mind of the bigot) they choose to be gay or trans-gender.

    And I think this is an important distinction in the mind of the bigot. The line as drawn also traces individual choice, and this plays well within the network of ideas that are espoused as conservative values. Of course, I don't believe, morally, that this is any better. The chief moral problem we face today is how to live amicably with people who hold beliefs that diverge fundamentally from our own.

    1. Anonymous6:26 PM

      Nice one. All too often the intolerant use the word bigot to try to marginalize someone who doesn't share their view. It's funny to see someone who claims to be progressive and tolerant shout out "bigot" to anyone who shows doubt about their belief. If they were really tolerant, and not just being an echo chamber for the democrat party one liners, they would actually listen to evidence supporting the idea that they aren't infallible.

  11. Anonymous3:12 PM

    There's no need to study this in the first place. Whether Trump is garnering a larger percentage of the bigots is just totally irrelevant. If Trump were not running, these bigots would still be bigots and they would vote for somebody else.

    Bigotry and economic unfairness are not the same thing. They seem correlated, but it is a disservice to call someone desiring economic fairness a bigot because the unfair things seem associated with other ethnicities.

    It would be ridiculous to expect people to feel good about giving up wages to an ethnic minority just because it is PC. This is much ado about nothing.