When did pollsters become credible again? They failed to predict the Scottish Independence Referendum in 2014, the UK general election in early 2015, the Canadian federal election in late 2015, and the "Brexit" vote two months ago.
Not only did they fail to predict the outcome of any of these events but their assumptions, especially in the case of the Brexit vote, had a significant influence on the outcome. It is entirely plausible that Britons would not have been so casual about voting were the perception of a narrow "stay" victory not so pervasive.
Now that both parties in the US have concluded their respective national conventions, pollsters are attempting to ascertain whether either candidate now has any discernible lead over the other. Apparently both candidates enjoyed a noticeable lead following their own convention, as is generally expected, but have fallen back to more or less equal footing this week. As one poll observes: "The reasons behind the shift were unclear."
Right. So why are we listening to them?
• • •