In the wake of the presidential election, many people blamed the polls for being wrong about who would win. But were the polls really that far off? Nate Silver makes a very good case that they were not.
Silver’s case is based on two facts. First, that the election changed direction near the end, when James Comey sent his October surprise letter to Congress. Polls respond slowly to changes. Even worse, people, and especially the media, misinterpreted Clinton’s national lead in the popular vote as implying a win in the Electoral College. According to Silver, it wasn’t the polls that were at fault.
Why is this important? Because every time a poll says that people are unhappy with Trump, he is going to claim that the polls are rigged, “just like they were before the election”. He has already done this for polls that said that the majority of people were not happy with the way his transition was going.
Silver points out that polls about the popularity of Trump are easier to get right than polls about whether he would win the election, because (for example) election polls have to take into account who is going to actually show up to vote, which is notoriously difficult to predict.
I remember some posts/interviews with Nate Silver in the weeks before the election where he was quite vehement that Trump had a good shot of winning. It was easy to see that he was seething at how people were misinterpreting the polls.
Blame some of the liberal networks for the misunderstanding about the polls. Huffpost was a bit culprit.