Read Stuff, You Should: More, Better Polling

Jonathan Bernstein is a Bloomberg View columnist. He taught political science at the University of Texas at San Antonio and DePauw University and wrote A Plain Blog About Politics.
Read More.
a | A

Happy Birthday to Jose Mijares, 30. Baseball is a brutal game; he was terrific for the Royals and Giants in 2012, and as far as I can tell he's out of baseball now.

Expect distracted blogging today. The Giants franchise doesn't exactly have a great history when it comes to Game 7 of the World Series. On the other hand, the Hunter Pence Giants don't lose elimination games. I'll try to focus on politics (other than the Game 6 post with my View colleague Kavitha Davidson, coming shortly), but it's not going to be easy. Even with the good stuff:

1. If you've read my posts about why the polls might be wrong, you should also read Eric McGhee at the Monkey Cage on why they might be better than ever.

2. Also at the Monkey Cage, Patrick Egan on going after North Carolina Democratic Senator Kay Hagan on immigration.

3. A campaign finance disclosure proposal worth considering from Heather K. Gerken at NPR's "On the Media."

4. Derek Willis at the Upshot looks at the controversy over a political science experiment in Montana.

5. As good as Mother Jones' Kevin Drum normal blogging is, his morphine-haze blogging is even better.

6. Steve Benen at MSNBC.com on Republican fearmongering.

7. And a good explanation from FiveThirtyEight's Nate Silver and Harry Enten about momentum (or lack thereof) in the polls. The bottom line is that there's no way to tell if a single poll, or even a few polls, indicate a new reality or just random noise. Look at the Gallup daily presidential approval line. It's been around 42 percent all year, but in late September it shot up to 45 percent, and it declined to 39 percent in early October. Because there are tons of presidential approval polls, we can be fairly sure that this was noise: President Barack Obama was at about 42 percent throughout the period, and Gallup happened to get a few different numbers. But in Senate elections, we have many fewer polls, and if we saw one showing a six-point swing, we'd be tempted to believe it was real, even though we can see (from the Gallup presidential numbers, if we don't believe it from basic statistics) that those sorts of swings can be meaningless. Except we also know that six-point swings can happen. So that's the problem for those who try to figure out what the polls are telling us: how to be sensitive to real changes without overreacting to noise. And the fewer polls there are, the harder it is to do.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Jonathan Bernstein at jbernstein62@bloomberg.net

To contact the editor on this story:
Max Berley at mberley@bloomberg.net