BLM divides white Americans more than anyone else
Our polling shows that Americans aren't split along racial lines
Our exclusive US polling of over 4,000 respondents, in partnership with Focaldata, contains a striking finding about how different groups are responding to current events in America. One of a series of statements under the heading ‘To what extent do you agree or disagree with the following statements about rioting’ was the statement, ‘The ‘Black Lives Matter’ movement is dangerous.’
Respondents were asked whether they agreed strongly, mildly, or disagreed strongly or mildly. The inclusion of the preamble ‘about rioting’ in this survey experiment primes respondents to think about rioting, thus elevating agreement, so this shouldn’t be read as an assessment of BLM.
Like what you’re reading? Get the free UnHerd daily email
Already registered? Sign in
But its design helps us probe cultural fault lines in the American electorate. A majority — 56% — of respondents agreed with the statement. A third of African-Americans agreed, rising to 51% of Hispanics and Asians and 61% of whites, suggesting as you might expect that there is some racial variation in attitudes to BLM.
However, racial differences are dwarfed by the partisan divide: 86% of Republicans compared to just 30% of Democrats felt BLM to be dangerous, a gap of 56 points compared to the 29-point racial gap between whites and blacks.
This reinforces the point I made in a New York Times article last year in which I argued that the divide in America is not about race, but racial attitudes, which in large measure transcend race. That is, the partisan gap is underpinned in part by sharp differences in the position rival ideologies take on race-inflected issues.
Differences of opinion are less to do with racism than contrasting perceptions over why blacks are not as wealthy as whites, whether reparations should be paid, support for affirmative action, or perceptions of whether diversity makes American society stronger.
Breaking the responses down by race and partisanship in figure 1 reveals a similar pattern in our data. First look at the right-hand set of bars. White Trump voters are more likely to agree (87%) that BLM is dangerous than Hispanic and Asian Trump voters (82%) and black Trump voters (58%). This is what we would expect.
But then look at the left-hand set of bars. White Democrats (28%) are less likely than Hispanic and Asian Democrats (41%) and very close to black Democrats (26%) in their likelihood of agreeing that BLM is dangerous while. White Democrats are about 20 points less likely to criticise BLM than we would expect from a race-based interpretation. And we see the same anomalous pattern for white Democrats on other questions in this data, such as whether western civilisation is in decline.
It could be that age, education, region or income differences between the groups explain this curious anomaly. However, as figure 2 reveals, even when you control for these factors, the red line for white Americans bends sharply downwards due to white Democratic progressivism.
The explanation is that one group is acting in a manner that runs contrary to Shadi Hamid’s notion of racial self-interest, namely white Democrats. This also turns up significant in statistical models. It reflects Zach Goldberg’s finding that the only group to rate its own race lower than that of other races are white liberals. ‘Very liberal’ whites in America score whites 20 points lower on a 0-100 thermometer than other racial groups, whereas even minority conservatives feel more warmly toward their own race than others.
Political religion, which brooks largest among whites, especially white liberals, overrides ethnoracial attachment to a much greater extent than is true of other groups.
For full tables and methodology data please click HERE.
Eric Kaufmann is Professor of Politics at Birkbeck, University of London. His new book is Whiteshift: Immigration, Populism and the Future of White Majorities.
Note: Pseudo-R2=.174. Controls for nonvoter, education, age, gender, region, income. N=4,040.
Join the discussion
To join the discussion in the comments, become a paid subscriber.
Join like minded readers that support our journalism, read unlimited articles and enjoy other subscriber-only benefits.Subscribe