Sunstein: Lack of ideological diversity leads to extremism


Peer pressure, so often associated with the immaturities of adolescence, exerts substantial influence on the political decisions of adults and the jurisprudence of the federal judiciary, Felix Frankfurter Professor of Law Cass Sunstein ’78 stated before a packed audience in Griswold on Monday, November 3. Sunstein’s presentation was titled “Extremism: Politics and the Law,” and was the inaugural colloquium of the Harvard Legal Theory Forum.

Sunstein stated that extremism in multiple domains (labor unions, corporations, environmental protection, gay rights, and more) “is a product of a distinctive kind of crippled epistemology resulting from group polarization.” In other words, individuals tend to come to more extreme views if they deliberate a given issue with like-minded people.

Sunstein cited three empirical studies defending his position, First, he did a case study with two communities-liberals from Boulder, Colorado, and conservatives from Colorado Springs, Colorado. Sunstein took pains to note that there was group diversity on the issues prior to discussion (both conservatives supporting same-sex and liberals resisting climate change regulation, for example).

The two groups of people were asked their opinions on affirmative action, climate change, and same-sex civil unions. The participants were asked to write down their view privately at the start of the study. After that, the groups deliberated the issue, recording a public verdict, and then completed the study by recording their private views once more.

The results of the survey were in line with Sunstein’s hypothesis about group polarization. Before their discussion, people in Boulder liked civil unions. After their public conversation, the individuals loved civil unions. Likewise, some people in Colorado Springs were equivocal about the US signing an agreement about climate change prior to the discussion, but expressed uniform disdain for the prospect after discussing the issue. The internal diversity that had defined the study groups had been “squelched” by the forces of group polarization.

This polarization also takes place within the supposedly politically-insulated halls of the federal judiciary. In the federal courts, panels are polarized when the panels are unanimously one party. Thus, when comparing mixed panels (with 2 Republicans and 1 Democrat, or vise-versa) to panels with 3 judges of the same political party, the unanimous panel usually has 2-3 times the effect of a split panel. This pattern is found across fields of study, including gay rights (100% support from DDD, 86% opposed from RRR), environmental protection, labor disputes between unions and corporations, sex discrimination.

Three issues seem to be notable exceptions to polarization, the judges being impervious to panel influence: capital punishment, abortion, and national security post-9/11. And the Sixth Circuit also appears to be the statistical outlier, perhaps, Sunstein posited, because the judges on that court of differing political persuasions, “really do not like each other.”

A third study tested the theory of group polarization by focusing on punitive damage awards and considering what the effects of outrage are on such awards? The study, of 1000 diverse Americans, found that people across all demographics generally agree on the outrageousness of certain events (kids’ pajamas being flammable causing outrage where injury suffered by elderly on an exercise machine caused little). However, despite this agreement, the dollar outcome awarded by mock juries is not similar.

The cause of this phenomenon, Sunstein said, is that the scale used to calculate dollar awards has no benchmarks. Instead of measuring the brightness of a light on a scale of 1-8, it is akin to “measuring the brightness of a light on a scale of zero to infinity…which creates a recipe for unpredictability.”

Nevertheless, the data from the mock juries still showed the effects of group polarization For instance, there was a systematic shift toward more severe penalties where people began with antecedent outrage. Indeed, even if the median juror had an outrage level of 6 on the scale from 1-8, the jury would often end up with a cumulative outrage (manifested through the size of its award) of 7 or 8. The opposite effect occurred-a leniency shift-where jurors began with little to no outrage.

Shifting to a discussion of the historical studies, Sunstein stated that group polarization was first discovered at MIT regarding entrepreneurial decisions. Graduate students there had a risk-taking shift on all but two issues: boarding a plan when you are sick, and the question of whether you should get married if you are not sure. A subsequent study addressed the same questions with Taiwanese citizens and had the exact opposite results, perhaps leading one to conclude something about the nature of American/Taiwanese cultures.Ultimately, though, the best predictor of the end effect of deliberation is the pre-deliberation median. If the median is tacking toward caution, the end result will be cautious and vise-versa.

Sunstein offered a number of possible explanations about the theory of group polarization. First, he noted that deliberations produce an exchange of information that reinforced one’s own pre-existing beliefs about a given subject. “If you get a group of people who are inclined to think that climate change is a serious problem, the group will produce many arguments about why climate change is worth worrying over and fewer about why it is not.”

Indeed, as one of Sunstein’s friends told him, “When we animal rights people get together on a Friday of a three day meeting, we are sensible. By Sunday, we’ve lost our minds.” This same distortion, Sunstein says, has been manifested in the American Constitution Society and the Federalist Society as well.

A second reason for group polarization relates to how people want to portray themselves to others. If you consider yourself more favorably disposed to animal rights than others, you will trend toward extremism if the peer group to which you compare yourself is unanimously pro-animal rights. Sunstein said that this “preservation of self-presentation” grows out of people’s desire “to be different than other people in the right direction and to the right degree.”

Third, corroboration leads to confidence, which in turn leads to extremism. When people are initially uninformed of the opinions of others, they keep their option set open by remaining moderate (at least publically). Once people gain confidence in their initial inclination, they tend to become more extreme in their tendencies.

These findings, Sunstein stated, lead us to the great debate about the role of extremism in political discourse. On one hand, the Federalists, led by James Madison, believed that the solution to factions was the mixing of ideologies. Borrowing a phrase from noted historian Doris Kearns Goodwin, Sunstein stated that Madison (as well as Senator Barack Obama ’91, for whom Sunstein is an advisor) sought a, “team of rivals,” as the best safeguard against group polarization.

Others, including Justice Louis Brandeis (J.D. 1877), argue that second-order diversity (diversity across institutions, rather than within institutions) produces experiments of learning from which we all will benefit. Thus the diametrically opposed ideologies of the economics departments at Berkeley and the University of Chicago or the sharp contrasts in political decision-making between Utah and Massachusetts, produce choicesSunstein closed by stating how an understanding of group polarization can help us fight terrorism. Terrorists are not poor, insane, or without education. To the contrary, studies have shown that the common thread between terrorists is not poverty or insanity, but rather living in nations without civil liberties. Sunstein hypothesized that this is because citizens of those nations have access to fewer sources of information
and thus become increasingly more isolated. Sunstein concluded, “Closed social networks enable the like-minded to associate with one another and produce real outrage and action.”