A year ago today, Donald Trump became the 45th President of the United States of America. The left has been alternating between rioting, crying, and declaring war ever since. The right is alternating between celebrating and...declaring war.
So already there is some common ground between the two sides who usually disagree on everything. However, there is lots of common ground available, isn't there? There are lots of lessons we had learned on November 9th, and we all agree what they are. It's heartwarming, it's comforting, and it's inspirational.
#6 - Undecideds won Trump the election: This comes up in every election in every country, and it never holds any water. Please stop, take a deep breath. Cosmopolitan very very very thinly disguised their contempt for non-voters, essentially accusing them of putting Trump in the White House. It doesn't matter, frankly, what the percentage of eligible voters who didn't vote was: they didn't cast a vote, and you can't assume that if you strong-armed all of them into a polling booth that they would pull the level you wanted them to pull. Maybe every one of them would have chosen Hillary and put her in the White House. Or maybe every one of them would have picked The Donald and he would have swept the entire country. After all, Trump did well with undecided voters in the third debate, but poorly in the first debate. The ultimate answer is that you don't know how they would have voted. In fact, they may not know either. They don't count, they never count. Never count voters who didn't vote.
#5 - The results are a major cultural shift: Remember what I said above? Ignore it. Now we need to count the undecided voters who didn't vote. As a percentage of the general population, in 2016 the Republican candidate won 26.07% of the potential votes. In 2012 the Republican candidate won 27.83% of the popular vote, and in 2008 the Republican candidate won 26.58% of the popular vote. Conversely, the Democratic candidate in 2008 won 30.82% of the popular vote, in 2012 the Democratic candidate won 30.1% of the popular vote, and in 2016 won 26.36% of the popular vote. Now look at this on a graph:
If you're looking at this image on a small enough screen and/or squinting, you might not even realize that these bars aren't all the same size. The distribution of votes looks pretty much identical. Any real discrepancies are probably from the estimates of eligible voters anyways: all of these figures are from Wikipedia, however estimates vary widely. StatisticsBrain.com repored 231,556,622 eligible voters in 2016, Wikipedia claims 251,107,000. In 2012 the Bipartison Policy Center says there were 218,959,000 eligible voters while Wikipedia says 235,248,000. That's a 9% gap in both cases, meaning the margin of error for the size of the bars far exceeds the actual gap between the parties (though the number of votes, unlike the number of eligible voters, is a firm well-agreed number).
If anything, looking all the way back to 1992, it appears the Democrats have a pretty solid line of support (a slight blip in 2008) with Republicans getting their voters out with a much less consistent rate of production. But there's no real cultural shifts going on here: the tiniest change in the number of vote % can make a wild difference. Look at the Electoral College totals which I included for each party above the graph. The difference between the Democrats in 2004 and 2008 is striking: their seats in the Electoral College increased 45% (from 251 to 365)...by increasing their share of the national vote by 12% (and in real terms only by 3.3%). For those objecting to the Electoral College, that isn't uncommon. The Conservative Party of Canada won 99 seats with 32% of the ballots cast in 2015 while the Liberal Party won 184 seats with 39% of the popular vote...meaning each 1% of vote gained one party 4.71 seats versus 3.09 seats for another. That's just how it goes. -ed]
#4 - Whites voted for Trump: Whites tend to vote Republican. In 2012, 59% of white voters supported Romney. In 2008, 55% of white voters supported McCain. In 2004, 58% of white voters supported Bush. In 2000, 55% of white voters supported Bush. So how many white voters did Trump get? Drumroll please...
...
...
...
...hey, did you ever watch that "House of Cards" show? Doesn't it open up with that faggot Kevin Spacey choking a dog to death? Isn't he the hero of the show, and also a Democrat? Isn't that worse than 2/3rds of what Trump is alleged to have done and probably worse than anything he actually has done? I have no point here (there is a point to the actions of real-Spacey vs Trump), just an open question...
...
...are you ready?...
58%.
That's right. Donald Trump's share of the white vote is slightly less than Mitt Romney in 2012. Some of that is the Utah Mormons supporting Evan McMullin, and some will be going to Gary Johnson tripling his vote share from four years ago. Still, whites voted for the chap with (R) next to his name at a pretty consistent rate from election to election.
What about subdividing whites up? Here there are some notable discrepancies: Trump received the support of 67% of non-college educated whites, versus only 61% for Romney. It's balanced by Romney getting 56% of college educated whites while Trump only secured 49%. Trump also lost Romney's 30% share of (religious, not ethnic!) Jewish voters, going from 30% to 24%. But this is still better than both 2000 and 2008 and comparable to 2004.
#3 - Women voted for Hillary: Similar to the bit about whites, Hillary did well with women, garnishing 54% of the female vote. She certainly can lord that over her predecessor who won in 2012 with...55% of the female vote. Which itself was a dropoff from 2008, where he had support from 56% of women voters. You can't even attribute any of this to Jill Stein: like Gary Johnson, she tripled her share of the national vote from 2012...however the sheer numbers are too small to make a difference. She went from 0.36% to 0.98%.
#2 - Blacks and other minorities voted against Trump: I put all of these as separate issues just to pad the numbers, frankly. It's all the exact same phenomenon. "Hateful racist" Trump received 8% of the black vote. Loving and inclusive Mitt Romney received less of the black vote share than Trump did, getting only 6%. In 2008, McCain received only 4%. In other words, Trump received the highest proportion of black votes for a Republican Presidential nominee since 2004, and almost the same share as in 2000. Trump received 29% of the asian vote, lower than McCain received in 2008 but higher than Romney received in 2012. Trump received 37% of the "other" racial votes, Romney received 38% and McCain only received 31%.
Finally, we come to Hispanics. Trump was infamously going to collapse with Hispanic voters and that alone would cost him the election, pollsters assured us. His comments on Mexican rapists and an end to illegal immigration across the southern border poisoned any possible relationship with American's hispanic population and doomed his candidacy. Now first you have to remember that 2000 & 2004 were huge outliers for the Republicans. George W. Bush (pbuh) was legendary in his relationship with Hispanic voters. In 2004, 44% of them supported him for President and was a major factor in securing his dominant win over John Kerry. In 1996 Dole only received 21% of the Hispanic vote, and in 2008 John McCain received 31%. 2012 saw Mitt Romney's share fall to 27%, shedding three-quarters of the 2004 gains over 1996. Trump again saw an increase of Hispanic voters versus Romney. Not 'yuge' by any means, but 29% of Hispanics (including 33% of Latino men) put a vote down for the guy with (R) next to his name, even though with the exception of "really large wall" his immigration policies were basically identical to the guy who got 27%. In other words, Hispanics voted for Trump at about the same level as they have been since spillover from a President highly regarded by Hispanics.
#1 - Polls are important and the pollsters got it all wrong: Those lying biased pollsters have upset literally everybody. Before the election, Trump supporters took to the internet to claim that organizations running the polls and the media reporting the both are both so biased towards Hillary that you couldn't take any of their claims at face value. Trump, they said, was going to win, and win big, and if the pollsters didn't agree they were lying to you. After the election, it's the left irate at these pollsters. Hillary supporters were under the impression that she had a strong lead going into the polls, and therefore weren't as motivated to get out and vote, thinking their girl had it all sewn up. The Guardian reports that the Democrats didn't devote much money to Wisconsin and Michigan because polls told them their lead was secure. However, the polls said Hillary would win and instead Trump won. I mean, the numbers don't lie, do they? You promise X and when X doesn't happen, you were wrong. What could be more obvious than that?
I suppose the better question is why are you believing polls in the first place? The old adage that the only poll that matters is held on election day carries a lot of weight, frankly. Much like how we never cared about movie box office receipts back in the day, we never used to rely so heavily on polls before either. It's mostly a function of an endless media cycle and our willingness to boil anything complicated down to a simple "who's winning" question. Polls are almost like watching a game, and keeping your eye on the score. It's important, I suppose: at the end of the game, the score is the most important thing. As it goes on, the score can help you understand a lot of what's going on...the question of "why are they bunting" or "why are they playing a neutral zone trap" is often answered by looking at the score. Similarly, polls could have helped you understand why Clinton had been going on about Muslims for the last three days of the campaign (a new poll had showed Trump's support among Muslims was increasing)†.
But that's in a case where all you care about is whether your side is winning or losing. In sports, that's most of us. But what if you're a professional scout looking at a junior or minor league team? Should our team sign that hot young prospect? If you're an NBA scout who gets asked that question, and your answer is "his team won the game 102-98" you are probably going to be fired within twelve minutes. Who cares? Did this hot young prospect even play in the game? If he did, what was his PER? And how did he look? Did you see things in his game that made you confident when he went up to the next higher levels that he would succeed? Does he have the specific qualities that the team is looking for? Does his nature and personality mean he would be a good fit in your locker room? You can see how for a person wanting to evaluate two candidates these analogies are far more applicable than score-watching. I suppose you could say it's a factor in how polarized we've all become...but then remember that graph about how few voting age Americans actually move their vote from election to election. The undecideds tend to decide all of these races by relatively narrow margins, and therefore how the media covers a campaign and what information the electorate is exposed to are major factors into how that election plays. As a result, poll watching is completely useless to the small sliver of the voting-aged public who will ultimately need to decide what the candidates are all about and decide the actual election.
† This is just an example.
Now understanding that the polls shouldn't be important, they were still wrong. Here's The Guardian again:
I spent almost two years working for Nate Silver’s website FiveThirtyEight, where I hoped to learn the secrets of political forecasting. I walked away totally disillusioned. It sometimes seemed as though their interpretation of the math wasn’t free from subjective bias. There was also a certain arrogance that comes from being part of an elite that “gets the numbers”, and an entrenched hierarchy meant that predictions weren’t properly scrutinised.There's a clue in there, for one. As Mona Chalabi writes, humans aren't subjected to Newtonian mechanics, so you can't always treat information gleaned from them reliably. But polls have been highly accurate before, haven't they? Nate Silver didn't come out of nowhere: FiveThirtyEight.com rose to prominence after correctly predicting all 50 states during the 2012 Presidential election. Yet the day before the election Nate Silver assured his liberal audience that Donald Trump was going to slink away on November 9th with his tail between his legs.
But analysts such as Silver, a man dubbed an oracle, a soothsayer and a savant have an interest in continuing to share these predictions. Where would the man’s career be if he simply replied “don’t know” when asked what Americans would do? It’s not just FiveThirtyEight. The New York Times also got it wrong, along with Reuters, NBC news and countless others. Just about everyone did – because they couldn’t resist the temptation to try to guess human behaviour.
Our forecast has Clinton winning the national popular vote by 3.6 percentage points, which is similar to her lead in recent national polls. Her chances of winning the popular vote are 81 percent, according to our forecast.A 3.6 lead in percentage points looks impressive...until you ask what the error is. 538.com admitted that Hillary's lead wasn't that much more than the margin of error.‡
‡ Whenever somebody wants to show you an average, you really need to look at two pieces of information. The average itself, and the standard deviation. In a way, that's what the above analysis of white/female/minority voters was. Sure Trump only got 8% of the black vote, but when you consider his range was basically 4-11%, it puts the numbers into better perspective. If you really want to be a pro at this, when you receive an average ask what kind of average it is as well.
To be fair, the polls were consistently biased towards Clinton. It's not like any major polls were predicting a Trump win...but the "shy Tory" concept wasn't new for this election. Pollsters gave a cold hard reading of the results, such as they were. It was when spokesmen for polling firms and their media contacts started interpreting the results that things got hairy. Subconsciously or not, because none of them knew any "shy Tories" they clearly must not exist...and it didn't help that numerous polling firms self-selected their sample size to begin with. Silver himself has to note that statewide polling errors can easily compound for a particular candidate, and no poll can be accurate when the biased left-wing media essentially forces people to lie to pollsters.
People can change their mind, too. In June of 2015 Tom Mulcair was leading in the national polls...by October he the leader of the third place party and by April he was fired. Likewise, particularly in Wisconsin and Pennsylvania, Trump scored big in people deciding in the last week of the campaign (similar to how Trudeau became Prime Minister), and in many of the swing states rigorous polls weren't even being conducted. Polls can be right or wrong, but what they should never be is stale.
Finally, swinging back to the Shy Tory question, perhaps if you know about the effect you might break with tradition and look at candidate internal polling. Trump's pollsters called Ohio, Michigan, Iowa, and Wisconsin in the days before the election.
So "the pollsters" didn't necessarily get it wrong. They didn't fail to predict that whites would "vote for Trump in droves" because they didn't, anymore than Hispanics rallied the vote to "stop Trump" anymore than you could have stopped Trump by forcing undecideds into a polling station. They're all myths about the 2016 Presidential Election.
Oh, and here's a bonus truth about the election both sides believe: one of the two candidates colluded with Russians to try and influence the election.
