Get the latest on the midterms & more in the NPR AppAhead of Election Day, there are a lot of questions about what might happen, what the implications of different outcomes might be and even what it's like to cover elections.In collaboration with the NPR Politics Podcast, listeners sent in their questions about the election and politics for our political correspondents to answer.This is a common question every election cycle, but it's a little like saying, "I don't like the media because the media is biased." There are good polls and bad polls just like there are good reporters and bad reporters.There have no doubt been some high-profile misses, particularly at the state level.
National polls have actually been pretty good when it comes to the horse race, with a slight underestimation of Republican turnout.But remember, these polls have margins of error.
That means a poll with +/- 3 or 4 percentage points is actually a range of 6 to 8 points.That's important because in a country with very close elections, the kind of precision that Americans expect sets up the polls for failure because they're not intended to be that specific.
They set more of a fuzzy range.State surveys have seen a lot of junk polls, even a couple fake ones this time around.So it's really important to pay attention to the quality of the pollster.
The reason for this is because of the decline of local journalism.Local newspapers and TV stations just don't have the kind of money that national outlets do to pay for good, rigorous and frequent polling.Horse-race polling isn't even the best use of polling.
I talk a lot about polls, but you rarely hear me talking about specific margins between candidates, unless there's been a trend moving the needle in a certain direction.The races are close, within margin of error in many places, and that's all that really needs to be conveyed.
Unfortunately, some outlets emphasize the horse race, but the best stuff is underneath.Are there trends over time...