Former Obama campaign manager Jim Messina has violent opinions about public pollsters – telling MSNBC’s Joe Scarborough “Joe, you know how I feel about public polls… I think all public pollsters should be shot.“
The former White House senior staffer who ran President Obama’s 2012 campaign was discussing the value of polling so far out from the election when he made the comment – telling Scarborough that he is more interested in how amped up voters are vs. generic polls.
What you care more about is passionate intensity. When I ran President Obama s campaign, the number I looked at everyday was intensity. Are my voters more motivated than Republican voters? Don t go generic. It s going to jump all over the place.
Pre-election polling from a variety of forecasters put Hillary Clinton’s chance of beating Donald Trump anywhere from 70% to as high as 99% – pegging her as the overwhelming favorite to win several states such as Wisconsin and Pennsylvania that ended up voting for Trump.
The fact that so many forecasts were off-target was particularly notable given the increasingly wide variety of methodologies being tested and reported via the mainstream media and other channels. The traditional telephone polls of recent decades are now joined by increasing numbers of high profile, online probability and nonprobability sample surveys, as well as prediction markets, all of which showed similar errors. –Pew
Some have suggested that the overly-optimistic (totally wrong) polls gave Democratic voters a false sense of security, leading to lower turnout.
The Pew Research center had some thoughts on what may have actually happened.
Nonresponse bias: Poor white Trump voters didn’t pick up the phone and participate in surveys. “The problem is if you get what pollsters call nonresponse bias, people are less likely to take your call or stay on the phone with you,” explained Claudia Deane, Vice President of research at Pew.
“Shy Trumper” effect: Fearing judgement from others, closeted Trump supporters told pollsters they were going to vote for Clinton, when in fact they voted for Trump.
Likely voter error: People who tell pollsters over the phone that they’ll turn out to vote, when in fact they don’t. “You might get 70 or 80% of the people on the phone telling you they’re going to vote because maybe that day they do plan to vote,” when in reality maybe 50 – 55% of people actually go to the polls.
Perhaps pollsters will figure it out before the next election. If not, they risk being shot by President Obama’s campaign manager – or perhaps someone inspired by him to take matters into their own hands.