Thursday, November 10, 2016

You need to know who voted for Trump

Loathe as I am to dive right back into that ill-fated American election, it continues to fascinate me, in the same kind of way as maggots chewing away at a carcass might fascinate me. As the media continues to dissect Tuesday's events in every conceivable way - from the geo-political to the personal, from how to explain it to your sensitive traumatized pre-adolescent child (I kid you not), to how Canadians might extract a grain of positivity from it, to why the polls got it all wrong yet again, to how thousands of fleeing Americans probably won't be beating down the doors to their northern neighbours - what still obsesses me is the demographics of who voted for the man, and why.
Some of the general demographics are no big surprise: maps of the voting patterns across the country show the usual blue (Democrat) support along both coasts, with wide swathes of red (Republican) support throughout the less populous, more rural interior.

Even within states, there is the usual marked dichotomy between urban areas (largely Democratic) and rural areas (largely Republican), although this is perhaps even more marked than usual. This trend is well exemplified by voting maps of Florida (which shows a completely red state punctuated by blue dots in the populous Miami, Orlando, Tampa and Tallahassee conurbations), Ohio (red but for blue pockets around Cincinnati, Columbus, Cleveland and Toledo) and Pennsylvania (all red except for Pittsburgh, Philadelphia, Harrisburg and Allentown).
But what kind of person votes for a person like Donald Trump? Data from exit polls, conducted by Edison Research and summarized in the Globe and Mail today, throw some light on the who, although much less on the why:
  • Race: overall, whites (still by far the largest group, even if declining in influence) favoured Trump by a substantial margin of 58% to 37%; African-Americans overwhelmingly supported Clinton with 88% compared to Trump's 8% (although this was still less than Obama's 95% support among black people); Clinton's 65% support among Latinos was double that of Trump (but again, still less than Obama's share 4 years ago).
  • Gender: Clinton's vote among all women was only 53% (about the same as Obama's was) compared to Trump's 41%; she was supported by over 90% of black women, but - crucially - more white women voted for Trump (53%) than for Clinton (43%).
  • Age: a graph of vote demographics by age shows a very clear downward trend in support for Clinton as the age increases, and the exact opposite for Trump's support, reinforcing the old axiom that young people lean left and old people lean right.
  • Income: contrary to many popular explanations for Trump's popularity, Clinton held a clear lead over Trump among the less wealthy (53% to 41% among those earning less than $30,000 a year), while Trump led at all levels of income above $50,000 a year (and especially among middle-earners from $50K to $100K a year).
  • Religion: as might be expected (even if it is not actually logical), Trump's support was highest among the more religious, particularly among those who attend religious services at least weekly, while Clinton's support is double that of Trump among those who never go to religious services.
  • Education: as has also been pointed out many times, the higher the voter's education level, the more likely they are to have voted for Clinton.
Maybe this stuff is not actually that interesting, but 1) I like graphs and maps, and 2) this cannot be allowed to ever happen again - for the sake of Canada and the rest of the world, as well as of America itself - so the Democrats need to know their enemy.

No comments: