Cass Sunstein is back, this time with a new book claiming that the Internet is bad for democracy because it allows like minded people to hang out with other like minded people, and thus they all become partisan zombies and thats bad, because, umm, because people aren't supposed to have strong positions on things? The meta-argument kinda loses me there, as you can probably tell. I haven't read the book so I cannot speak to its in depth argument, but Mr. Sunstein is blogging at TMP Cafe this week and his first post is not a particularly impressive one.
Here is the core of his post:
As the experiment was designed, the groups consisted of predominantly liberal or conservative members—with the liberal groups coming from Boulder, and the conservative groups from Colorado Springs. (The groups were not mixed together.) It is widely known that Boulder tends to be liberal and that Colorado Springs tends to be conservative. The groups were screened to ensure that their members conformed to these stereotypes. (For example, if people in Boulder liked Vice President Cheney, they were cordially excused from the experiment.) People were asked to state their opinions anonymously both before and after a period of group discussion, and also to try to reach a public verdict before the final anonymous statement. What was the effect of discussion?
The results were simple. In almost every group, members ended up with more extreme positions after they spoke with one another. Discussion made civil unions more popular among liberals; discussion made civil unions less popular among conservatives. Liberals favored an international treaty to control global warming before discussion; they favored it more strongly after discussion. Conservatives were neutral on that treaty before discussion; they strongly opposed it after discussion. Mildly favorable toward affirmative action before discussion, liberals became strongly favorable toward affirmative action after discussion. Firmly negative about affirmative action before discussion, conservatives became even more negative about affirmative action after discussion.
A couple of things here. Even assuming that the degree of the effect was as intense as Sunstein implies, this is a poorly designed experiment, or, from Sunstein's description of it, this is a poorly designed experiment. He doesn't link to it or give us enough details to track down where it was published, so I cannot be one hundred percent certain. but there is no mention of what happens after the groups are shuffled and the rations are more 50-50. In other words, Sunstein assumes that the effect he describes is permanent without actually offering any evidence that it so. If your contention that the Internet makes partisanship worse, shouldn't and one of your supporting arguments is that groups in isolation become more partisan and have less internal differences, shouldn't you design an experiment that actually attempts to answer that question? Unless you assume that the groups always remain in isolation, the first thing you have to do is to determine whether or not that affect is lasting and to what degree it persists. This experiment as described does not do that.
And groups do not remain in isolation. People go to work, they go to Church, they go to movies and ball games and professional conferences. It is literally impossible for me to not be aware of the conservative view of the world, both because of the prevalence of conservative viewpoints in the mainstream media and because of the fact that I live a normal human life that brings into contact with other people all the time. Some of those people are conservatives, and some of those people talk about conservative issues. It's probably easier for conservatives, but even they would have a hard time going through life without exposure to liberals.
The second problem assumes that increased partisanship on the issues in questions is a bad thing. First, any argument that assumes "extremism" is a label that can be applied to civil unions is suspect in and of itself. Civil unions are the compromise position between the current level of discrimination and full equality before the law. So in Sunstein's experiment, the liberals became more strongly in favor of the compromise position. Describing that as an increase in extremism is ridiculous on its face. But its part and parcel of the problem with the underlying assumption. Sometimes, the compromise solution is not a viable one. Global warming, another area where "extremism" increased according to Sunstein is a good example. either global warming is a large problem that has ot be dealt with quickly and forthrightly or its an overblown ghost story that doesn't justify the material investment required to prevent its worst effects from occurring. Neither side is much served by a split the middle, compromise is always good attitude. Half measures may be better then nothing form the global warming perspective, for example, but they won't solve the problem and at the end of the day you still have to work to convince people to take the real steps required to prevent the worst form occurring. At best, compromise buys a little more time for such convincing -- the kind of convincing that went on in Sunstein's expiriement -- to continue. Affirmative action is less an all or nothing position than global warming, but it, too, can be argued that the compromise position is uselss for both sides in the argument: a little affirmative action is either an unnecessary violation of rights or a uselss tool that does not achieve the very real and needed societal changes.
It seems odd to argue that an increase in partisanship is bad without examining the actual merits of the questions involved; i.e. is one side already supporting a compromise position, as in the case of civil unions, or if a compromise position could have more than a minimal effect on the problem as in the case of global warming and affirmative action and is thus a rational choice for the players involved or not. Sunstein's experiment assumes that any two positions are equally extreme, that the group dynamic demonstrated is persistent, and that compromise is always the rational position. All three of the assumptions are dubious and the fact that they form the underpinning of his experiment leaves his grander thesis, to the degree that he is going to support it based on this example, on very shaky ground.