… but on a different vector.  This started as a comment on my post about liberals playing the race card and neglecting WWCs (and thus having their hands dirty in the “where did Trump come from” converation).  It  got very long while at the same time seeming to unpack issues raised by CVFD and Barry.  So a post .. .

The Intercept ran a piece yesterday unpacking some polling data about WWC support for Trump.   The headline “The Great White Hype:  No One is Energizing the White Working Class, not Even Donald Trump” doesn’t reflect the data at the center of the piece.  The data shows

a reversal [of growing WWC support] from earlier in the summer, when Trump’s support among the group was in the 60s, higher than Romney’s, though not by leaps and bounds.

Trump’s support from WWC – defined in questionable ways (more on that below) – was in the high 60s before the convention, and has dropped to 49% in recent weeks.

These data only says that Trump’s support from the WWC is on the decline, and probably a significant one.  It doesn’t say that he isn’t energizing WWC or that he never did.

Still, it adds to our conversations elsewhere about what makes these people tick.  The data in the story offers a couple hypotheses:

1. that there is an overlap between Sanders’s WWC message and Trump’s and that voters are leaving Trump and going to Dems (per CVFD’s question);
2. primary behavior didn’t really tell us anything about the substance of the support for Trump (because, as the article notes, it was such a small percentage of voters in a serial fish bowl campaign mode)
3. Trump’s popularity was broader and deeper than WWC, and now even more mysterious

Those are hypotheses only.  The data only shows some correlations and we probably can’t prove anything for a while.

Or it may be hard to prove anything at all.  The article also highlights some tricky data/polling issues:

1. What is WWC – polls have fairly thin definitions of WWC – lower education levels, income levels, geographic overlay.  Sounds straightforward, but what about office workers, people with college degrees with working class incomes, people on the border and in fear of slipping.  This is classic Orwell territory, and worthy of more conversation and thinking, but I’ll just quote John Cleese playing Robin Hood in Time Bandits “The poor are going to be absolutely thrilled!  Have you met the poor?  … Oh you must meet them.  I’m sure you’ll like them.  Of course, they haven’t got two pennies to rub together but that’s because they’re poor.” Class words cover a lot of complexity.
2. the notion of base is tricky  – it’s never that only one segment makes a candidate, or that there is a finite set of attitudes that shape that support
3. the exceptional nature of 2016 – and for that matter all elections since and including 2008.  All of them are made exceptional, or at least discontinuous from previous elections and polling environments because of the financial crisis, and the various effects of the internet (microfragmentation of communities, disintermediation of experts, fund-raising for example).  The article goes back to 1992, reminding us that we’re 24 years into independent rumblings, starting with Perot, continuing with Nader, always in the air with Trump and Sanders and now with Gary Johnson (I can’t talk seriously about Jill Stein, I know that’s wrong, but I can’t.)
4. connections between what a candidates say and why people vote for them – there is a straight causal line drawn between candidate sound bites and votes.  In TV coverage, Trump talks about the wall, mocks people, rails against PC by being offensive.  When people vote for him, the only causal link we have is that Trump voters love those things.  That’s like saying I support flip-flopping on DOMA, the Iraq war, politicians who are super tight with Wall Street, resisting the minimum wage, resisting the New Glass Steagal, TPP because I’m voting for HRC.  James Fallows, Thomas Frank and others (I’ll try and add the links) continually point out that:  1) Trump talks a lot about jobs, trade, and economic issues for the rest of his speeches; 2) that Trump supporters don’t always gravitate to the xenophobia but will never vote for HRC, will never vote Dem, want the system messed with, are willing to roll the dice; and 3) he’s the Republican nominee, that’s who they vote for.  Polling and issue mapping don’t capture this complex process of political identity building, answers to pollsters, or final votes.  They hint at it and provide hypotheses, but that’s as far as we can go with certainty.
5. self-awareness about polling – fielding, following, reading polls has always had a box score feel to it, but now it feels national.  It seems like everyone talks about them and has enough of an understanding to manage a critique of them (see the revived “skewed poll” debate).  Trump refers to “nice polls” that show him doing well, questions the bias of those that don’t.  This is an edge argument, but with all this awareness of how polls play into politics, do we really understand what someone is doing when they answer questions?  We regularly entertain the notion that people might not be saying they support Trump because they don’t want to appear racist to the poll-taker – why can’t people answer questions in order to buck the trend, lodge a protest vote, which finally comes to
6. the notion of likely in polls – likely voters, likely to vote for a party – these are all concepts that have to standardized over time and which are based partly on self-identification and validated by historic behaviors.  As historic behaviors become less relevant (Trump’s strategy is to reach out to “low propensity voters” and people are aware of the impact of the surveys they’re participating in, we have a lot of threads to unpack.

Good thing there’s fivethirtyeight.

Apologies CVFD and Mister Jones – the only link I could be bothered to track down was the John Cleese clip.  I’m really dragging us down.