Defining America’s division

A look at the exit polls

The sharpest divisions are, in order of the distance between them: race, religion, age, gender, education. (Age and gender are actually about even in terms of the splits.)

In general, minorities voted 76% blue, compared to 44% for whites.

In general, Protestants and Evangelicals voted 42% blue, compared to 60% for all others. This is a key division because the two groups are not only sharply divided, but of nearly equal size.

In general, people under 45 voted 61% blue, compared to 49% for older Americans. (Surprisingly, people under 45 represent only about 1/3 of the voters.)

In general, women voted 59% blue, compared to 47% for men.

In general, the college educated voted 59% blue, compared to 49% for others.

Here’s an interesting fact I never noticed before: there are no additional significant splits within the minority votes. Younger minorities vote like their elders. Female minorities vote like their male counterparts. College-educated minorities vote like those without educations. This is completely opposite to the white group, which is deeply divided in all respects: males are more conservative; older people are more conservative; and the poorly educated are more conservative.

On the basis of the aggregated data, education doesn’t seem to divide Americans as deeply as other factors, but that is misleading because of a point I just noted: educated minorities vote exactly the same as uneducated minorities. When you look at the splits among whites alone, they become more sharply defined. That is Trump’s base: the poorly educated white people. And that’s the them-vs-us that defines America at the moment. It’s the uneducated white people against everyone else, and they are more than 40% of the electorate, so they are pouring red all over the ballot boxes. (See the chart below)

If you consolidate the info in that chart, it looks like this:

Voted Democratic Voted Republican
whites with no college degree 37% 61%
everyone else 64% 34%


There is no specific exit poll data for older white men with no degree, but you have to figure that is the truly major red territory. It may split 4-1 for the GOP. President Trump and some of his surrogates know exactly how to play to that audience. (Sean Hannity is a good example of someone who not only plays to them, but is actually one of them. Trump himself is not an older, uneducated white man, but you would certainly assume he was if you came out of a coma not knowing who he is, then met him socially.)

7 thoughts on “Defining America’s division

  1. This is always a faulty way of presenting voter information. It is a liberal presentation highlighting the majority of college educated voters leaning blue. The reality is that key division is by gender in which white men across the Baird support red. Trumps base are men. Period.

    1. Not so. One of the greatest divides of all is between white men with college degrees and those without. The split is 15 points.

      To underline the point about education being more important than gender, ask this question: “Which was more likely to vote red: a college-educated white man or a non-college white woman?” The answer is the non-college woman. Education trumps gender. (Although it’s close).

      Education has become a powerful separation. (Race, of course, is #1 by several lengths. It’s the Secretariat of demographics.)

      1. My only point is that white college educated women overwhelmingly condemn trump which skews the “white college educated” vote. White college educated men prefer trump, just not by as wide a margin as women.

  2. Election updates:

    Arizona: with 100% of the precincts reporting (but I don’t think every vote has been counted) Democrat Krysten Sinema has taken a nearly 10,000 vote lead over Republican Martha McSally. This lead change appears to be due to absentee ballots (and probably provisional ballots as well.)

    Florida: Governor Rick Scott’s lead over U.S Senator Bill Nelson has dwindled to just 15,000 votes (with some counties not having even started counting provisional ballots) and is now threatening to sue counties in South Florida if they continue counting, or don’t allow Republicans to scrutinize the count, or something like that.

    If Nelson takes the lead and wins in Florida, and if Sinema wins in Arizona, that would leave the Republicans at just +1 on the night in the U.S Senate with a 52-48 majority.

    This is clearly something that I think needs to be addressed: that in tight state wide races anyway, Republicans appear to have won with nearly all the precincts reporting and this is reported on, but overlooked is that in many Republican governed states many of the remaining votes to be counted are either in the big cities or are provisional ballots, both of which tend to heavily favor the Democrats. So, this creates a false narrative of a ‘Republican win’ followed by false claims that the late lead change is due to ‘Democratic Party cheating.’

  3. Good stuff. I had not been following up on this.

    The very narrative you describe happened here in Cheeseland. Scott Walker was ahead by a few hundred votes, then all of a sudden Evers was up by 30,000 and the conservatives were calling foul – especially since recounts can be requested at 1 point of differential or less, and the extra ballots put Evers barely over that hurdle.

    It was a bunch of uncounted ballots, some 50 thousand, in inner city Milwaukee. They went 80% for Evers, which is nothing so unusual. African-American voters usually go blue at 75% or more (sometimes way more).

  4. It also happened here, NM-2. The Republican candidate declared victory on Tuesday night, but 8000 mail-in ballots still hadn’t been counted, mostly from the two liberal enclaves in the district. End result: the Democrat won.

Comments are closed.