Four (more) things to consider when interpreting the election:

posted Nov 10, 2016, 9:12 AM by Daniel Laurison
  • If you’re trying to understand who voted how, remember what the vote counts actually are (with not-quite-all precincts reporting): Clinton won (just barely) the popular vote – 47.7% vs 47.5%.  Turnout is estimated at 56.3% of the voter-eligible population, which means that of the adult citizens over 18 in this country (who don’t live in Puerto Rico and haven’t been disenfranchised due to a felony conviction), about 27% voted for Clinton, 26% voted for Trump, 3.4% voted for someone else, and 43.7% stayed home.
  • Groups are not people – watch out for the ecological fallacy.  “Ecological fallacy” is the tendency to impute to individual members of a group the characteristics of the whole.  When you look at the election map, you see Pennsylvania in red. If you live in Pennsylvania, or followed the returns, you know that Pennsylvania was actually very close, and something like 48% of voters in PA voted for Clinton (it should be purple, and will be when this year’s purple maps come out). If you hear about the voting patterns of groups you don’t know or aren’t part of, try to remember that those groups are (almost) all mixed, too.  A few examples:
    • “Whites without a college degree supported Trump” – it looks like about 67% of those who voted did, which means 33% *didn’t*. 
    • The rust belt shifted to Trump – we know that many states that were “blue” last time are “red” this time; we don’t know (yet) whether that’s due to individuals who voted Obama last time voting Trump this time, or different people turning out vs staying home; probably some of both.
    • People who are harassing and assaulting visible minorities & Muslims seem to be vocal Trump supporters.  It’s awful that Trump’s victory seems to be emboldening these people, and many of us opposed Trump because we were worried this would happen. Most people who voted for Trump are not doing this, and a substantial portion of them would oppose it.
  • Beware single-factor explanations.  Just because if one thing *had* been different (and everything else had remained the same) the outcome would be different, does NOT mean that that factor is the cause of the result.  So, in a few states Clinton lost, the Jill Stein vote was larger than the Trump-Clinton difference.  Sure, if each and every one of those Stein voters had voted Clinton instead, Clinton would have won.  But Clinton *also* would have won if turnout among committed Democrats in that state had been % higher, or if a few more white women had voted for Clinton instead of Trump, or any number of other shifts of just a few percentage points. People will jump to their favorite explanation, and especially this early after the election, it’s pretty easy to find a single data point to back up most single-factor assertions.