Donald Trump’s victory in Tuesday’s presidential election was a surprise to large swaths of American voters. But it shouldn’t have been. National polls were extremely close and deciding that only one outcome is possible in a close race is arrogance.
See the United Kingdom Brexit vote for a similar example.
Yes, many of those polls turned out to be wrong (more on that later). But let’s start with some fascinating research from Pew on what DID happen Tuesday.
• White non-Hispanic voters preferred Trump over Hillary Clinton by 21 percentage points – 58-37. That margin was almost identical to that by Mitt Romney in his failed challenge to Barack Obama in 2012.
• Clinton did not fare as well with minority voters as Obama. Clinton had an 80-point margin with African-American voters and a 36-point margin with Latinos. But Obama had an 88-point edge with African-Americans and a 44-point edge with Hispanics in 2012.
• Trump had a 67-28 edge among whites without a college degree, the largest such edge such Reagan-Carter in 1980. But Trump also led among whites WITH a college degree, 49-45.
• Gender-wise it was a push. Trump led by 12 points among men and Hillary led by 12 points among women. And it is worth noting that for all of the criticism Trump received regarding his treatment of women, he still received the backing of 42 percent of them. If Clinton had been able to widen the gap she might have won. Other pollsters found that Trump led Clinton 53-43 among white women.
• Among young voters (18-29) Clinton had a strong edge, 55-37. But the gap wasn’t as wide as it was for Obama against John McCain in 2008 (66-32) or against Romney in 2012 (60-36).
Simply put Clinton needed to put up Obama-like numbers with African-Americans, Latinos and young people to put her over the top. A bigger gender edge would have helped, too. But it didn’t happen.
Now, why were the polls wrong? Why did pre-election surveys show Clinton the likely winner … with some forecasts showing her having a 70 percent chance of winning … or more? If you looked at the Electoral College maps published by established groups such as CNN and Nate Silver it looked like Trump needed to win virtually all of the battleground states … and also pick off a few of the light blue states that were leaning to Clinton.
And he did just that, scoring wins in Pennsylvania and Wisconsin as well as taking swing states such as Ohio, North Carolina and Ohio.
How did this happen?
Pew and others note that polling has become more volatile since more Americans ditched their land lines. Those with cellphones only are harder to reach and it's more expensive to interview them.
Also, polling experts said, Trump’s support was underestimated and less-educated voters who were a key bloc of his backers are also hard to reach.
In addition, some pollsters noted that many voters just aren’t honest with survey workers either because they want to throw a wrench in the works or becomes it seems “socially desirable” to claim they do not support Trump.
I’ll be watching for more explanations of this as time goes on.
Oh, one other thing. Obama also did better in Benton County in 2012, with 65 percent of the vote. Clinton had just 60. I plan to publish precinct by precinct results of the presidential race and Measure 97 in an upcoming edition.