By Robert Folsom | October 16, 2012
We’re close enough to the presidential election for the latest poll numbers to be a fixture in the news cycle, especially when those numbers suggest that “the race is tightening…”
… And it’s no stretch to say that poll results are taken more at face value than is most other reporting. Polls are supposed to be data collections of what voters say they will do. Data is neutral.
While that’s the perception, the reality is more nuanced. And by nuanced I do not mean “margin of error.”
From a socionomic perspective, one has to keep in mind the thing which polls attempt to measure — specifically, a future expression of public mood.
At the same time, the volume of polling today is such that no one person can keep up with it all: Multiple polling groups and organizations that operate at state, local and national levels, which employ numerous models and methods, and collect a variety of sample sizes in an assortment of ways.
There are even people and groups who aggregate poll results, into a poll of the polls.
All of which is to say: Political polling today amounts to one collective trying to measure a much larger collective.
And where you have collective behavior(s), herding is bound to happen. And apparently pollsters do indeed herd.
Now, it would be one thing for me to make such an observation, but quite another when it comes from one of the most highly respected statisticians in the business, namely Nate Silver in his FiveThirtyEight blog in the New York Times (Oct. 11):
Here’s a dirty little secret: pollsters herd. Or to put it less politely: it’s probable that some polling firms, especially those that use less rigorous methodologies, cheat off the stronger ones — seeing what the consensus results are before weighing in on their own.
I won’t suggest Silver intended to make a profoundly relevant socionomic observation, yet that’s exactly what this is. Furthermore, the comment implies that the herding happens exactly where you would expect to see it — namely, in a context of uncertainty.
It’s always satisfying to come across someone who understands that a collective behavior has a deeply psychological component, though I need to point out that “psychological component” is a synonym for social mood.
More importantly, social mood is bigger than polls and presidential elections: It also drives financial markets, to the point that markets can actually help forecast the outcome of elections.
Sound far fetched? Read the paper co-authored by Bob Prechter, which has recently been mentioned in the “A” list of news outlets from the Washington Post to the Huffington Post to NBC News.
Also feel free to speak up in the comments section below or on our Twitter feed.
Andrea Dibben contributes research.