Preliminary Results on Empirical Study of Influence of Chief Justice:

Frank Cross reports some preliminary results on his empirical study of Chief Justices on case results:

I have completed a preliminary analysis of the 2004 term. There appears to be some result from Rehnquist's absence on outcomes. For example, in the cases over which Rehnquist presided, he had a 10% dissent rate. In the cases over which Stevens presided, Rehnquist had a 30% dissent rate. While this is a fairly small sample, and the changed probability of a Stevens dissent is much less, this is suggestive that Rehnquist's absence may have had an effect.

I went by their appearance at oral argument with help from some of your readers and found 40 cases for Stevens and 30 for Rehnquist, just using cases with full argument and full opinion.

He also adds a caveat (and invitation):

Posting it would be great, though please caveat that it's preliminary. It's only going to be a small part of the article, but getting this out there for discussion might uncover any mistakes I made. We'll be using a broader database to test the past effect of the Chief, i.e., what happened when Rehnquist was elevated? Did it affect his votes or the votes of others?

Sounds like a great research project.

Justice Fuller:
Isn't it possible that Rehnquist was tactical in picking the weeks in which he would not participate? He might have intentionally dropped out in weeks when he knew his side was likely to lose, as he knew he wouldn't get to assign the majority opinion in those key cases.
8.29.2005 12:12pm
Michael Last (mail):
If you make the assumption that the cases presided over by each chief justice is independent of how they feel about the cases (e.g. the opposite of what the first commenter suggests), and conduct a t-test comparing the estimated probability of dissent, the p-value is 3.3% (two-sided). This is a bit dicey since the proportions are so small, so I'd be hesitant to claim significance....
8.29.2005 12:26pm
Scott Scheule (mail) (www):
Fuller has a fine point.
8.29.2005 5:09pm