Thursday, October 6, 2016

Harold Myerson thinks we have a white racist millennial Clinton-defector problem. He's wrong.

I anticipated this in my previous article, but since Harold Myerson is making the argument directly, I'd like to shoot it down directly:
Hillary Clinton is still having trouble winning the allegiance of the apt description of the millennials holding out for the third-party candidates: They’re all white...a hard core of young, white Bernie-or-Busters may yet believe that voting for Stein, or even Johnson, is an expression of their disdain for the system.
As noted, Clinton is not actually having trouble winning the allegiance of millennials - she's winning about 53% of them, compared with a minority of 38.9% among olds. Meanwhile, a plurality 39.3% of olds are voting for Trump, compared to just 24.7% of millennials. Clinton does not have a millennial problem - she has an old people problem, and a millennial solution. Analytically breaking down these age groups into conveniently gerrymandered sub-demographics (white millennials! third-party defector millennials! white third party defector millennials!) does nothing to contest the broader, obvious age trend; it just makes one's analysis increasingly narrow, and increasingly irrelevant.

Faced with these brute numbers, Myerson now only has one possible move: to compare today's young voters with those from 2012. But if we do that, the second premise of his argument collapses: Myerson wants to blame "white skin privilege" for the failure to support Clinton, but if the last election is our baseline, it's young people of color who are running from the Democratic party. He can, that is to say, only salvage his critique of millennials by abandoning his critique of white people, and vice versa.

A second problem with Myerson's race critique is that he repeatedly tries to make it into a specific critique of leftists. He opens with an anecdote about Stein and Sanders supporters; he quotes an organizer who singles out Sanders supporters; and he closes by once again brooding about the notorious "Bernie-or-Busters".

But to do this, he has to play fast-and-loose with the numbers, and in a way that strikes me as pretty deliberate. Consider, for example, the crux of his argument, which the article even highlights in a pull-quote:
Presumably, this 2% discrepancy demonstrates some kind of white privilege among leftist voters. That's why it's interesting that he omits a directly relevant fact: the same poll reports that Stein is also at 4% among Latinx and Asian Americans. An even more interesting point is that he includes those same numbers when he reports on Gary Johnson, who has the backing of "15 percent of whites...but just 8 percent of Latinos, 6 percent of Asian Americans, and 4 percent of African Americans." Comparatively, it seems clear that white supremacy is far and away the province of young libertarian voters, and that there is no detectable third-party voting trend that's unique to young white leftists; it also looks a lot like Myerson erased Stein's support among Asians and Latinxs precisely to obscure this point.

Finally, it's worth putting the millennial attrition issue into perspective. When Myerson notes that Stein and Johnson have 4% and 11% of the millennial vote, one's tempted to conclude that this amounts to 15% of the youth vote for Clinton. But in fact, as YouGov reports, only 35% of third party voters under 30 say that they prefer Clinton to Trump - meaning that she's really only losing about 5% of the youth vote to third parties. And since there's no significant evidence that this 5% is disproportionately white, it's hard to escape the impression that Myerson is accusing young leftists of racism for no good reason, while ignoring support for Trump in how own generation that's larger by several orders of magnitude.