Major media pollsters and pundit forecasters wrongly called the presidential election for Hillary Clinton, had her winning states that Donald Trump actually captured and wrongly predicted Democratic victories in the Senate.
ABC News analyst Nate Silver, 38, the national polling guru who runs the website FiveThirtyEight.com and who confidently predicted a Clinton victory, said, “It’s the most shocking political development of my lifetime.”
Big media outlets such as The New York Times, The Washington Post and NBC News — which put much effort into aggressively vetting Mr. Trump’s campaign — released final polls and assessments that said the Manhattan real estate developer would lose the popular vote by 3 to 5 percentage points as well as the decisive Electoral College.
The countrywide vote on Wednesday evening stood at a near-tie — 47.7 percent for Mr. Trump, 47.5 percent for Mrs. Clinton.
NBC had said she would win by 5 percentage points nationwide, a spread that likely would have won her enough states to be president-elect. Reuters also said she would win by 5 points. The ABC/Washington Post poll put it at 3 percentage points. The most accurate appears to be TIPP/Business Investors Daily, which projected a 1-point Clinton national margin.
It was not that the national polling was way off. The mistake was that, based on those numbers and individual state polls, newspapers predicted a Clinton win. The Washington Post did not just predict a victory for her; it said flat-out that it was mathematically impossible for Mr. Trump to gain the needed 270 electoral votes.
“Trump’s path to an electoral college victory isn’t narrow. It’s nonexistent,” read The Washington Post’s Oct. 18 headline over a story by voting analyst Stuart Rothenberg.
The New York Times gave Mrs. Clinton an 80 percent chance of winning as the polls were closing.
Mr. Trump is projected to have won as many as 306 electoral votes.
It was in the state battlegrounds, where elections are decided, that pollsters goofed. They had Mr. Trump losing in Wisconsin, Pennsylvania and Michigan. He won the first two and was leading in the third. The surveys also missed key Senate races by undercounting Republican turnout.
The poor prognostications on Election Day were perhaps signaled two years ago. A Washington Times analysis showed that in 2014, as on Tuesday, major pollsters misjudged the Republican turnout in key races. Maryland Gov. Larry Hogan, for example, was down double digits in late polling but won by 4 percentage points.
Also Tuesday, Edison Research delivered exit polls to TV networks predicting that four Republican Senate incumbents would lose. They instead won. The Edison handouts influenced the chatter by cable TV election analysts, who rely on the numbers to guide their commentary while waiting for the actual data. On live TV, some intimated that Hillary Clinton and Senate Democrats had won.
“They say it’s a Hillary landslide,” one cable network pundit told The Washington Times on Tuesday night.
That pro-Clinton tone faded as real returns rolled in and Mr. Trump picked off must-win states of Ohio, North Carolina and Florida, and then led in Wisconsin, Pennsylvania and Michigan.
Larry Sabato, who runs the University of Virginia Center for Politics, appeared on “Fox and Friends” to humbly acknowledge that his ballyhooed “Sabato’s Crystal Ball” failed him and the public. He had predicted a decisive Clinton win, with more than 300 electoral votes.
“That ball is shattered in a thousand pieces,” Mr. Sabato said. “I’ve got to order a new one.”
He said he fed thousands of polls into the computer and 90 percent of them turned out wrong.
“We were wrong. The entire punditry industry, the entire polling industry, the entire analysts industry. And I want to take this opportunity to take my share of the blame. We were wrong. I apologize,” he said.
Here are some of the big misses:
• Virtually every national poll had former Democratic Sen. Evan Bayh defeating Rep. Todd Young, a former Marine Corps officer. Mr. Young won by 10 percentage points.
• All but one of 25 polls had former Democratic Sen. Russ Feingold beating incumbent Ron Johnson, who won by 3 points.
• Of 12 late polls, only one had Republican Sen. Patrick J. Toomey of Pennsylvania ahead of Democrat Katie McGinty — by 1 percentage point. He was behind in the rest by as many as 12 points but won by 2 points.
• For Wisconsin, more than 30 national polls said Mrs. Clinton would win the state by 4 to 15 points. Mr. Trump won by 1 point.
• For Michigan, every national poll had Mrs. Clinton winning by 4 to 10 points. She was trailing Mr. Trump by 0.3 percentage point with all precincts reported.
• For Pennsylvania, all national polls said Mr. Trump would lose. He won by 1.2 percentage points.
A small, somewhat quirky polling firm in Atlanta went against the grain. It predicted Trump victories in Michigan and Pennsylvania on the eve of voting and turned out to be right when all of the national firms were wrong.
Trafalgar Group and its seven employees discovered during comparison polling that some Trump voters would not disclose how they planned to fill out their ballots. After all, the liberal news media and Mrs. Clinton had labeled Trump followers as deplorable people.
So the company used robotic calls for which Trump voters seemed more comfortable. They also added a “neighbor” question, figuring that a respondent would be more willing to answer truthfully if a neighbor was voting for Trump.
The also created a demographic of people who had not voted in a half-dozen years or so but planned to vote for Mr. Trump. To capture these voters, Trafalgar created a large sample, about 1,000 respondents, compared with a usual poll of 400 or so.
The result: A day after the election, when searchers looked at the Real Clear Politics webpage for Michigan polls, there was a long list of predictions, all of which were in blue because they had Mrs. Clinton ahead — except one posted on Nov. 6. It had Mr. Trump winning by 2 percentage points. He was leading by 0.3 percent with all precincts reporting. That outlier poll was by Trafalgar Group.
The Real Clear Politics Pennsylvania page tells the same story. A sea of 30 “blue” poll results up to the top of the list, where Trafalgar predicts, in red, a 1-point Trump win, which actually materialized at 1.2 percentage points.
“We knew Monday night that Trump was going to win,” said Robert Cahaly, the group’s senior strategist.
Trafalgar also nailed results in North Carolina and Georgia and predicted a Trump win in Florida, though by 5 points rather than the actual 1.3 points.
It also said Mr. Trump would receive 306 electoral votes. He is likely to end up with that exact number.
Trafalgar’s bet that a “hidden” Trump vote existed was dismissed by the mainstream press.
But Matt Briggs, a teacher of statistics at Cornell University, issued a study on Wednesday concluding that indeed there was a “shy” Trump effect that the major pollsters did not try to identify and thus was left out of their turnout models. Trafalgar did use such voters.
“Everybody around elites — namely other elites — were saying Hillary was going to take it,” Mr. Briggs wrote. “It’s difficult to discount the opinion of all your colleagues. Wishcasting, a type of confirmation bias well known to meteorologists, also played a role. Wishcasting is when the evidence that supports a strong desire is given more weight than it deserves. Is there any doubt the whole of the elite media were desperate for Hillary to win?”