Extraordinarily Popular Candidates and the Madness of Crowds
Categories: misc.
In the two days after the Iowa caucuses, Barack Obama zigged so far ahead, in the polls and in media opinion, that he suddenly looked unbeatable. Hillary Clinton, for so long the front-runner, was reduced to tears of despair and frustration.
Then Obama zagged back, more quickly than he had zigged, and Clinton won New Hampshire – and after that Nevada. In the New York Times, political scientist Daron Shaw wondered if the bandwagon effect – a.k.a. “momentum” – had somehow gone away:
Dr. Shaw expected Mr. Obama in New Hampshire and Mr. McCain in Michigan to benefit from voters’ long-demonstrated propensity to flock toward expected winners.
“If Giuliani wins Florida,” he said, “you’ve got a real case for saying, ‘Is momentum dead?’ ”
But Giuliani has already faded in the polls, in Florida and elsewhere. And “momentum,” a.k.a. the “bandwagon effect,” hasn’t died. It just has competition, in our ultra-wired society, from other factors that can swing the electorate at short notice.
A huge percentage of “independent” New Hampshirites admitted in exit polls that their minds had been made up only days — or less — before the vote. Flung towards Obama by the post-Iowa bandwagon effect, these people then apparently swung back towards Hillary, perhaps influenced by her show of vulnerability, or by Obama’s supercilious treatment of her in a day-before-the-vote debate — both events having been widely televised.
The same mercurial voters feature, by the way, in political science accounts of the bandwagon effect. A study from 1994, by researchers from the University of Southwestern Louisiana and the University of Kentucky, found the effect to be particularly strong among “independent” voters not affiliated with any party. A study in 1996 blamed “particularly unsophisticated independent voters” for the effect in a German election. A UCLA study published in the Journal of Applied Social Psychology in 1998 hinted that the effect was rooted in traits similar to suggestibility:
Bandwagon effects were stronger for women compared with men, and for 2 of 3 PAD (Pleasantness, Arousability, Dominance) basic temperament factors; that is, for individuals with more arousable and less dominant temperaments.
Just how irrational the bandwagon effect can be was shown by the famous Asch Social-Conformity experiments, published in the early 1950s by psychologist Solomon Asch. In these experiments, subjects would frequently give the wrong answer to simple, seemingly obvious perception tests merely because people around them (i.e., the experimenter’s lying confederates) expressed certainty about the same wrong answer. A much more recent study using functional-MRI scans indicated that the subjects in experiments like these really do perceive things differently under the influence of other (wrong) opinions, and that contrarian thinking – though correct – lights up subjects’ emotional centers, suggesting an inner struggle.
One of Asch’s students, Stanley Milgram, would later go on to do similar experiments showing that the same instinct towards social conformity can also severely distort moral perception. Ditto for romantic perception, which psychologists recently have found to be skewed in favor of prospective mates whom the crowd seems to find attractive. Ditto again for economic perception, and the bubbles and bursts to which it can lead when groupthink sets in.
Some people like to think of all this as a kind of “social contagion,” which in a sense it is. But it is a broadcast contagion. Remember Marshal McLuhan, and his worries about a wired, “global-village” with its “panic terrors, exactly befitting a small world of tribal drums, total interdependence, and superimposed co-existence”? Even today, with all the different ways we have to plug into the media, we still tend to get very similar, very centralized coverage for big events like a presidential race.
The immediate effects of mass media are obvious: We are inoculated by the same influencing events — like Hillary’s tears — more or less all at once. To the extent that these events are unpredictable, their broadcast injects unpredictability into our opinions. But when the event in question is an expression of our own opinions, for example a poll or primary vote, there is potentially even more volatility — because our opinion now can feed off itself.
Broadcast and bandwagon effects will sometimes conflict, and sometimes align. But on the whole they will turn what arguably should be a sober debate over policies, experience and character into something like a sporting contest, with a strong emphasis on the flow of events and a strong element of unpredictability. No wonder we (and the rest of the world) find the process so fascinating to watch.
What we really ought to do is to consider it dangerous, because to the extent that our individual opinions move towards McLuhan’s “total interdependence” they are prone to be unwise as well as unstable.
Remember the concept of the “wisdom of the crowd”? As described by James Surowiecki in his book of the same name, it requires that opinions be relatively independent of each other. In principle, random errors of judgement cancel out this way, effectively amplifying whatever valid information remains.
If, on the other hand, opinion biases take hold via broadcast effects or bandwagon effects, whatever wisdom a crowd had is apt to swing towards collective madness.
And by the way: The more tightly our attention is wired to the media, the more rapidly these volatile, irrational effects will be able to move us, and the further we’ll be liable to drift from the anchoring influences of family, community and tradition.
Can we at least make our elections less susceptible to these effects? Probably not in any structural way. To do so would mean limiting our connection to the media, coalescing the primaries — perhaps into a single nationwide election — and of course censoring poll results. None of these is likely to happen soon.
The one handy tool we have is awareness. If people learn how to recognize when broadcast and bandwagon effects can sway them, they have a chance to combat those effects with more analytical thinking.
Unfortunately, the concept that human social behavior is largely media-driven and crowd-driven doesn’t seem to get much traction in the popular imagination. It seems to be one of those things we’d just rather not think about. Apparently we prefer to see not chaos but order in the world, to find patterns and to navigate by them. Thus when a candidate surges in the polls we are tempted to rationalize it by reference to his or her attributes – electability, freshness, plainspoken honesty – just as we rationalize the to-ings and fro-ings of the stock market, the fates of sports teams, and all the other things in life that are never fully rationalizable.
Note: This is an update/rewrite of a piece — “The Bandwagon Effect” — I posted in early December.
