Polarization in Small Groups and in Politics

April 13, 2014
Posted by Jay Livingston

In class last week, I tried replicating the “risky shift” experiments that date back to the 1950s. Groups discussed problems that pitted caution against risk. For example, down by three points on the last play of a football game, should you kick a field goal and settle for a tie, or try a play that might win but also risks a loss?* In the original studies, not only were group decisions riskier than individual decisions, but discussion persuaded more people towards risk than towards caution.

Later research showed that the risky shift was one instance of a more general effect – group polarization: When members of a group share a value, and they discuss something related to that value, group opinion will shift further out towards the pole on that dimension.

I hadn’t thought that the concept had much use outside of small groups, but now I wonder if something similar happens in politics.  “North Carolina Shows Strains with G.O.P.” says today’s Times (here) on page one.

 the divisions that are gripping the party nationally are playing out powerfully, expensively and often very messily.  And after haunting losses in 2012 in which far-right Senate candidates prevailed in primaries only to collapse in the general election, the Republican establishment is determined to stifle the more radical challengers.

Those divisions were always there. As someone pointed out even in the victorious Bush years, the party was an uneasy coalition of The Predators (pro big business), The Taliban ( religious and cultural conservatives), and NeoCons (foreign policy hawks). Now add the more populist, libertarian Tea Party, who accuse the others of being RINOs (Republicans in Name Only).

Republican primaries are basically group discussions among those who share conservative values. As in the small-group studies, participants are aware that others are evaluating them on their positions, so they move towards the valued end of that dimension. Those already further out provide an anchor – or perhaps a magnet – to pull the others further in that direction.**

Other things being equal, we might expect positions to get more extreme of the course of the primary season.  But of course, other things are not equal.  The difference between group discussion and politics is that in the small group experiments, all participants had an equal ability to voice their ideas to the group. In politics, thanks to the Supremes, the question is not just what someone wants to say; it’s who has the money to have his message heard most frequently. 

-----------------

* In those days, college football had not overtime. The game ended after the fourth quarter.

** The question in the experiments asked, “What is the lowest probabiblity that you would accept in order to go for the win rather than the tie?”  The person who went in choosing a 5-in-10 option might have thought himself reasonably risky. But when he got in the group, he found that others would be willing to take a 3-in-10 or even 1-in-10 chance.  His original position no longer seemed so in tune with the tacit value on risk, and he might shift to a riskier alternative.