Monday, April 24, 2017

Sure you want us to look at that poll, Donald?

Watchmen's Ozymandias, via Comic Vine

Popularity-obsessed Trump bragging on the results of the latest survey from ABC-Washington Post:

If they blew it so badly five months ago what makes him trust them now? As you know, they didn't blow it; their finding, that Hillary Clinton would have won a national election if the US had national elections, by a handy 2 or 3 million votes, was correct. The forecasters who were wrong (and not big but very narrowly), were those who studied the 50 separate state elections and missed the results in three, Wisconsin, Michigan, and Pennsylvania, where a hundred thousand or so votes gave Trump an unexpected 46 electoral votes and the presidency.

In any case, is he sure he wants us to look at that poll? Because 53% called him a strong leader? Sorry, all sorts of politicians are strong leaders, including Mussolini and, um...

Most watchers (including Boo) are taking more interest in the "how would you vote now?" question, where a startling 96% of the self-reported Trump voters say they'd vote for him again, compared to only 85% of the Clinton voters, suggesting that if they held another election today Trump would win an outright majority. I don't think it's especially important, because of two factors: one, that the poll is of all adults, not registered or likely voters, so its usefulness as a measure of voter attitudes is not very clear anyway, and two, that the number of self-reported Republicans is remarkably small, just 24%, where self-reported Democrats are at a more normal 31% (I couldn't find any other ABC-Wapo numbers to compare it to, but other polls have been finding GOP affiliation more often around 27 to 29 in recent months). We're not told how many Republicans voted for Trump, but just 84% of them approve of him, compared to 94% for Trump voters.

That low number of Republicans isn't likely a sign that voters are leaving the party, but more a registration of embarrassment (last time it hit 25% in the Gallup tracking was in the wind-down of the primary season, April 2016, as the inevitability of a Trump nomination became evident, and the party staggered in bewilderment, unable to defend itself against this bizarre assault). And you know what else a respondent might be embarrassed about: voting for Trump (an idiot), voting for Clinton (a loser), and not voting at all (extreme civic irresponsibility). In other words, pretty much all the alternatives.

In this way, I think, while the proportion of people who say they voted respectively for Trump and Clinton seems to align close enough to historical reality, 43% for him to 46% for her, the poll's identification of which particular adults voted for whom isn't very trustworthy. And those who are willing to say they voted for Trump are not the sort who ask themselves whether they've made a mistake, while those who are willing to to say they voted for Clinton are the opposite.

But the broad and deep contempt shown for Trump and most of his policy proposals is measured very well. There's just no telling whether a vote would reflect it or not.

No comments:

Post a Comment