Thursday, November 10, 2016

How Donald Trump Won the 2016 Presidential Election, According To NPR

How Donald Trump Won 2016 the Presidential Election, According To NPR

Key stat:

Whites without college degrees preferred Trump by 39 points.

Whites with college degrees were split almost evenly (Trump up by four, which was probably within the MOE).

In the 1990s, whites with and without degrees voted almost identically. The white blue-collar union guys were once a solid bloc for the Democratic party over the years, and that sub-group kept the overall tally for the white non-degreed group split. Well, union households still prefer Democrats, but only slightly (Hillary got 51% of them), so the Donkey stranglehold has been broken. With that sub-group split, the non-degreed group has moved ever more toward the right. Exit polls in the previous three elections in this century showed a gap of 12-14 points between degreed and non-degreed whites, and Trump just about tripled the gap to 35.

The CNN exit polls showed the education split in more detail - further split between men and women:

White women with degrees: Clinton by 6
White women without degrees: Trump by 28
White men with degrees: Trump by 15
White men without degrees: Trump by 49

Consider this: if white men without degrees are splitting 72-23 for Trump, but the ones in unions are splitting about in half, then non-degreed white males who are not in unions must be voting very close to 100% for Trump.

There were some other important facts and some surprises within the CNN polls:

* Trump won 13% of black men and 33% of latino men. My guesses would have been lower.

* Trump won the independent vote 48-42. That is a critical fact, because those voters who consider themselves independents are nearly a third of the electorate. Since self-professed Democrats outnumber Republicans by four points, a big margin in the independent vote is essential for Republican victory. That margin (approximately a third of six points - therefore, roughly two points) was not enough to shift the country's popular vote to Trump, but it seems to have been enough to swing a lot of important states. Trump won Wisconsin by a point, Pennsylvania by a point, Florida by a point - that's 59 electoral votes swung by a single point in each state.


  1. Even with all those shifts, Clinton got more total votes.

    1. Which is meaningless.

  2. The real driver was deep and deserved personal hatred of Hillary. Bernie would have won without breaking a sweat, the polls were there for the reading. DLC, Debbie Wasserman Schultz: go die. Burn the party to the ground, start over. Or don't start over.

  3. One detail often overlooked: when they say people with college degrees vote Democratic, it implies that they have the college-educated crowd sewn up. But all that really means is that they have the ones who finished a four-year degree. The voters described as non-college graduates aren't just people with only a high school education or less, that's also people who have anything less than a full four years of college. Since far more people attend some college than finish it, it's a big difference in total voters.

  4. I would be cautious before believing what these preliminary exit polls are reporting. There will be a handful of later polls that will come out.

    The reason for this is not based on some general distrust of polls but because the reporting on the Latino vote simply makes no sense to me. It's possible, of course, that it's the reporting on the Latino vote in 2012 that was wrong and not the reporting this time around.

    Hillary Clinton according to the exit polls, received around 65% of the Latino vote whereas Barack Obama received around 72%.

    I simply don't see how this can be correct given that Barack Obama lost Texas by around 15% whereas Hillary Clinton lost it by around 9% and Obama lost Arizona by around 9% while Hillary Clinton lost it by 4%.

    In the other states with large Latino populations, Florida, Colorado and Nevada, Hillary Clinton received a lower percent of the vote than Obama, but not by anywhere near as much as she gained in Texas and Arizona.

    I think, at most, the results in Florida, Colorado and Nevada cancel out the results in Texas and Arizona.

    If this is the case, that would suggest Hillary Clinton received roughly the same percentage vote from Latinos that Obama received, not taking into account the roughly 5% that the third party candidates received.

    If that is the case, I would put that down to the same thing that in every Presidential election around 25% of homosexuals vote for the Republican candidate.

    Whether they are right or wrong, obviously some Latinos and homosexuals vote on reasons other than this status.

  5. I think it is unwise to combine the Latino vote into a single monolithic bloc, as pollsters often do. I doubt if Trump pulled 33% of MEXICAN males, but Cubans, Puerto Ricans, and South Americans may view the issues from entirely different perspectives. (Puerto Ricans are unaffected by immigration restrictions, and I think Cuban-Americans still tend to vote Republican, although the younger generations are trending away from the GOP.)

    The Miami Herald predicted that the Cuban-American vote gave Trump a good chance in Florida, even though many pundits were placing the state in the blue corner.

    I would also like to know the reasons why those latino males voted Trump. Is it because his macho posturing conforms to their concept of proper masculinity? Is it because the least educated latino males can be culturally unprepared to have a woman in charge? I don't know, but those issues may provide a part of the answer key.

    1. This comment has been removed by the author.

  6. This comment has been removed by the author.

  7. Fair enough on the first point, I think.

    Two other points:
    1.I would be interested in knowing why Trump's voters supported him as well. This should be something those polls can be believed on because it would only be answered by Trump voters.

    I would guess it breaks down into five categories:
    1.Resentment of the 'cultural elites.'/Resentment of the 'SJW'/Resentment of illegal aliens and build the wall!
    2.Belief Trump will bring back manufacturing jobs
    3.Trump is a businessman/outsider
    On a Venn diagram I'm sure these three would intersect somewhat

    4.Anti Hillary Clinton
    5.Did not vote for Trump per se but always vote for the Republican Presidential Nominee.

    2.In regards to some of the Latino men, it had been observed in prior polling that Trump had some of the same appeal to Latino working class men as he did to white men, and, as well to Latino working class women.

    I've already written why I think these exit polls on the number of total Latinos who voted for Trump are wrong.

    But, there is no question in regards to why Hillary Clinton lost in the electoral college that there was, in fact, a group of shy voters for Trump: the college educated whites, and that the Clinton campaign mistakenly believed that many more white working class women would vote for her than ended up doing so, especially in the 'rust belt' states.

  8. How are they so sure who voted for whom? Polls? I never answer polls. I would think only the most strident people would. That doesn't represent most people.