By Erin Delmore
Correspondent
Donald Trump’s stunning win a shock to American voters, pundits and pollsters, even inside Trump Tower.
Monmouth University pollster Patrick Murray said, “Every single model — whether it was in the Trump campaign, the Clinton campaign — the vote casters who are trying to model the votes as they came in during the day and the polls themselves showed that Hillary Clinton was going to win enough states that she would win the electoral college.”
Trump broke through democratic candidate Hillary Clinton’s so-called “blue firewall” stealing swing states east, west and center flying in the face of national polls.
In the days leading up to the election, Hillary Clinton polled three, four percentage points ahead — with a clear path to victory, despite tight rac es predicted in bell-weathers like North Carolina and Florida. But both states went to Trump plus, Pennsylvania and Wisconsin the last a state where Trump didn’t win in a single poll this election season. He carried it handily Tuesday night.
“So right now people are scratching their heads saying ‘What happened? Who came out that we didn’t expect to be coming out,” said Murray.
In conversations with NJTV News, pollsters pointed to distrust in the polls themselves, fueled by the Trump campaign itself. The result: “non-response bias.” Hidden Trump support.
“Were we getting the rural Ameraican vote that seemed to turn out the, you know, working class lower educated. This was a notoriously hard group to study and given the nature of the rhetoric of rigged elections, and the system isn’t working, and the media is biased pollsters could be looped right along with that as far as Trump supporters may have been concerned,” said Ashley Koning, director of Eagleton Center for Public Interest Polling.
With so many new ways to reach voters and so many new voters to reach here’s what’s driving today’s Wednesday morning quarterbacking: Did pollsters talk to the respondents who turned out?
“When you’re doing a pre-election poll you’re not polling the general population, you are polling what are to be considered likely voters. And likely voters are a notoriously difficult group of people to identify for the simple reason that if you simply ask someone are you going to vote on Nov. 8 — or whenever election day is — most people will say yes. But we know from turnout data that 90 percent of the population does not vote,” said Krista Jenkins, executive director of FDU PublicMind Poll.
“Some voters in the electorate this past election cycle simply did not want to admit certain things to the interviewers that were interviewing them. They wanted to admit things that were socially desirable. Perhaps they were again shy Trumpers that didn’t see their responses or their candidate preference as socially desirable to the typical American citizen who may be interviewing them,” said Koeing.
Pollsters we spoke with were quick to note last nights results fell within many poll’s margin of error and they pointed out those polls are just a snapshot of a moment in an ever-changing horse race. They said, fewer than 10 percent of people respond to polls and they’re harder to reach now that people are cutting landlines and not picking up calls from unknown numbers. The American Association for Public Opinion Research sent an email today saying, “The polls clearly got it wrong.” As the organization’s done in year’s past, it’s convening experts to analyze the data and figure out what happened.