One of the most common questions being asked Wednesday is how pollsters and pundits managed to be so wrong about Donald Trump. Well, not everybody was wrong:
- The Los Angeles Times/University of South California tracking poll consistently had Trump in the lead through the final months, often to much derision. As USA Today notes, of 61 polls tracking a two-way race since October, only six put Trump in the lead—and all were from the LA Times/USC. (But even the LA Times' own electoral map predicted a landslide for Clinton on the eve of the election.)
- Back in September, American University history professor Allan Lichtman predicted a Trump victory based on a system that involves 13 true/false statements. One caveat: Lichtman's system predicts the winner of the popular vote, and Trump may yet lose that. Still, coming into this year, he's been perfect since 1984. In September, he spoke to the Washington Post.
- Stony Brook political science professor Helmut Norpoth predicted Trump would be the next president. He says his model, which assesses candidates' performances in the primaries, would have gotten every election right since 1912 except for 2000. See the New York Post.
- Two other tracking polls that consistently had Trump performing better than in most national surveys were from Rasmussen Reports and Investor's Business Daily, reports Politico.
- On Monday, an outfit called the Democracy Institute published a poll with Trump up 5 points and "Poised For Electoral College Win." The institute's press release that day boasted that its public poll was the only one to get Brexit correct.
- Stats guru Nate Silver managed to be both right and wrong in a sense. His final assessment at FiveThirtyEight gave Clinton a 71% chance of victory, but he also pushed back hard against polls calling her a shoo-in and observed Sunday that she was "one state away from potentially losing," notes the Washington Post. In other words, he covered his "backside," writes Dana Milbank.
(Read more Election 2016