How Did the Media Really Do in 2016? Check the Evidence
I hope people pay attention to Nate Silver's essay on the media, probability and the 2016 election. He's ... I guess "annoyed" is the right word ... that too many people in the news media don't know some of what they should know about quantitative evidence. "The media keeps misinterpreting data -- and then blaming the data," he says.
It's not that he's wrong, but overall I'm much more of an optimist about this than Silver is. As I see it, political coverage over the last several years, particularly when it comes to quantitative evidence, is vastly better than it was 20 or 30 years ago, in part of course because of FiveThirtyEight's excellent work. I can't blame anyone for impatience when people still get it wrong, as they do, but there's much more good stuff than there ever was before. And while there are still plenty of people who treat quantitative analysts as if they are either wizards or charlatans, I think that too is fading a little.
Silver's main substantive point, and one that I think is exactly correct, is that the polls (and Silver's polling forecast) were reasonably good in the 2016 general election. Yes, that's right -- reasonably good. There's some reasonable disagreement among analysts about the extent to which some state polling wasn't very good, but the national polling was fine, and a lot of the state polling was fine, too. When you flip an honest coin 10 times and it comes up tails seven or eight times, the most likely explanation is going to be that somewhat improbable things happen, and we should expect them to happen sometimes.
I do think Silver is slightly unfair to the media in that the FiveThirtyEight forecast was quite a bit better for Donald Trump than most of its competitors'; it's possible the media would still have treated Silver's 70 percent chance as close to certain anyway, but that was presumably reinforced by other forecasters who actually did portray the election as a lock for Hillary Clinton.
My other disagreement is that I don't think it was "liberal media bias" that produced the media assumption Clinton would win; I think standard (poor) interpretation of previous elections was far more important -- indeed, the mistaken Electoral College "lock" assumptions in 2016 were exactly the same as the Republican-favoring bias after the 1984 election, which was upended in 1992. I suspect it was also to an extent a case of some in the media following political insiders, both Democrats and Republicans, as well.
I think the better criticism of the media in the 2016 presidential general election wasn't so much that they were too certain that Clinton would win, but that some made very poor decisions about what to cover and how based on that certainty.
1. Sheri Berman at the Monkey Cage on the German election.
2. Lisa Blaydes and Martha Crenshaw at the Monkey Cage on the consequences of the Iraq War.
3. Robert Farley asks whether the U.S. could have won in Vietnam.
4. My Bloomberg View colleague Timothy L. O'Brien on Trump becoming presidential. Well, maybe not.
5. See also Mike Allen on Trump as "the great divider."
Get Early Returns every morning in your inbox. Click here to subscribe.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
To contact the editor responsible for this story:
Brooke Sample at firstname.lastname@example.org