Cass Sunstein explains the root cause of Rathergate

I know he did not intend to. In fact, this article was written three years before CBS ran their TANG story and Sunstein is most worried about right-wingers on the internet, not the honest liberals who dominate broadcast journalism. Still, this passage explains how good people could run such an embarrassingly bad story:

We can sharpen our understanding of this problem if we attend to the phenomenon of group polarization. The idea is that after deliberating with one another, people are likely to move toward a more extreme point in the direction to which they were previously inclined, as indicated by the median of their predeliberation judgments. With respect to the Internet, the implication is that groups of people, especially if they are like-minded, will end up thinking the same thing that they thought before-but in more extreme form.

Consider some examples of this basic phenomenon, which has been found in over a dozen nations.3 (a) After discussion, citizens of France become more critical of the United States and its intentions with respect to economic aid. (b) After discussion, whites predisposed to show racial prejudice offer more negative responses to questions about whether white racism is responsible for conditions faced by African Americans in American cities. (c) After discussion, whites predisposed not to show racial prejudice offer more positive responses to the same question. (d) A group of moderately profeminist women will become more strongly profeminist after discussion. It follows that, for example, after discussion with one another, those inclined to think that President Clinton was a crook will be quite convinced of this point; that those inclined to favor more aggressive affirmative action programs will become more extreme on the issue if they talk among one another; that those who believe that tax rates are too high will, after talking together, come to think that large, immediate tax reductions are an extremely good idea.

The phenomenon of group polarization has conspicuous importance to the current communications market, where groups with distinctive identities increasingly engage in within-group discussion. If the public is balkanized, and if different groups design their own preferred communications packages, the consequence will be further balkanization, as group members move one another toward more extreme points in line with their initial tendencies. At the same time, different deliberating groups, each consisting of like-minded people, will be driven increasingly far apart, simply because most of their discussions are with one another.

Why does group polarization occur? There have been two main explanations, both of which have been extensively investigated and are strongly supported by the data.

The first explanation emphasizes the role of persuasive arguments, and of what is and is not heard within a group of like-minded people. It is based on a common sense intuition: any individual's position on any issue is (fortunately!) a function, at least in part, of which arguments seem convincing. If your position is going to move as a result of group discussion, it is likely to move in the direction of the most persuasive position defended within the group, taken as a collectivity. Of course-and this is the key point-a group whose members are already inclined in a certain direction will offer a disproportionately large number of arguments supporting that same direction, and a disproportionately small number of arguments going the other way. The result of discussion will therefore be to move the group, taken as a collectivity, further in the direction of their initial inclinations. To be sure, individuals with the most extreme views will sometimes move toward a more moderate position. But the group as a whole moves, as a statistical regularity, to a more extreme position consistent with its predeliberation leanings.

The second mechanism, which involves social comparison, begins with the claim that people want to be perceived favorably by other group members (and to perceive themselves favorably). Once they hear what others believe, they adjust their positions in the direction of the dominant position. People may wish, for example, not to seem too enthusiastic, or too restrained, in their enthusiasm for affirmative action, feminism, or an increase in national defense. Hence their views may shift when they see what other people and in particular what other group members think.
Group polarization is a human regularity, but social context can decrease, increase, or even eliminate it. For present purposes, the most important point is that group polarization will significantly increase if people think of themselves, antecedently or otherwise, as part of a group having a shared identity and a degree of solidarity. If, for example, a group of people in an Internet discussion group think of themselves as opponents of high taxes, or advocates of animal rights, their discussions are likely to move toward extreme positions. As this happens to many different groups, polarization is both more likely and more extreme. Hence significant movements should be expected for those who listen to a radio show known to be conservative, or a television program dedicated to traditional religious values or to exposing white racism.

Eureka. A non-right-winger explains why ideologically uniform newsrooms create distorted news. All the conditions he points to exist inside CBS news or The New York Times. The people who work there will tend to move to a more "liberal" (or more "urbane") viewpoint. This, in turn, makes it easy to caricature non-liberals and hard to rein in an ideologue on a mission.

What a shame the Thornburg panel did not know about this phenomena.