When Trying To Interpret Polls Now, Beware The Bandwagon Effect

A frustrated Obama supporter e-mails:

Obama is going backwards in polls which makes claims that people don't know him ring hollow. He shouldn't be going backward, especially in the early states. But it is pretty obvious why he's going backwards in NH and SC and nationally, it isn't because people don't know him, it is because they think HRC is gonna win so they join the bandwagon. The problem is that unless Obama actually starts caring about polls this will continue."

I think there's something to be said for this interpretation. There is no real way to separate the aggregate influence of what people perceive to be common knowledge from the snapshots of public opinion that polls provide.

Pollsters and psychologists know that, when then it's too early to pay close attention, many voters assimilate and then give voice to the opinions they know that others hold. We don't like to feel as if they are outliers, so we jump on the bandwagon and temper down our cognitive dissonance.

So part of Clinton's rise can be attributed to a cognitive bias. And the Clinton campaign is smart to exploit its self-perpetuating cycle.

The good news for everyone is that there has to be some reason for Clinton to rise in the first place. The Bandwagon Effect is, at most, a multiplier of a real phenomenon.

It is also not determinative. It can be reversed. And the more voters take time to pay attention, the more confident they'll feel in their own judgments. The more confident they feel, and the more information they have, and the more politics saturates their lives, thay may well feel less pressure to identify with the frontrunner for the solidary benefits.

Iowa and New Hampshire voters have been conditioned to reject the bandwagon effect; voters in states after Iowa and New Hampshire... not so much.

Here's from the master, Mark Blumenthal.