"A remarkable pattern" of "almost uniformly conservative" dissents?

On Monday night, I was on a radio show with lawprof Cass Sunstein, talking about the nomination of Samuel Alito, which Bush had announced just that morning. Sunstein stressed a study of 41 Alito dissents, which he had -- amazingly -- completed that day. I see that he published an opinion piece about his study in the Washington Post yesterday:

As an appeals court judge, Samuel Alito has compiled a massive record that includes more than 240 opinions. Of these, the most illuminating may well be his 41 dissents -- opinions that he has written by himself, rejecting the views of his colleagues.

When they touch on issues that split people along political lines, Alito's dissents show a remarkable pattern: They are almost uniformly conservative. In the overwhelming majority of cases, he has urged a more conservative position than that of his colleagues. In his dissents, at least, he has been a conservative's conservative -- not always in his reasoning, which tends to be modest, but in his ultimate conclusions.

"Almost uniformly" -- what does that mean? There were 41 cases. In how many of the dissenting opinion did he take a position that Sunstein could code as "more conservative" than that taken by the rest of the panel (that is, the other two other judges from what, Sunstein concedes, is the "relatively liberal" Third Circuit)?

The WaPo piece doesn't give the number, but on the radio show -- which you can listen to here -- Sunstein says twice that "two dozen" of the dissents go in the conservative direction. 24 as compared to 17 is a "remarkable pattern"? 58% of the time is "almost uniformly"? I don't get it!

Or is it that there is some subcategory of the 41 cases that Sunstein viewed as containing "issues that split people along political lines"? So how many dissents in fifteen years are we talking about? Perhaps it's 27, because he did also mention that there were 3 cases in which he coded Alito's dissent as more liberal than the rest of the panel.

Can someone replicate the coding and counting of the Alito dissents? Apparently, it's a one-day job.

Also, when a judge dissents is he "rejecting the views of his colleagues"? Judges' conclusions upon analyzing legal issues aren't "views," and a disagreement among a panel of three judges isn't properly characterized as a "rejection" aimed at colleagues. They simply reached different conclusions and are saying as much.

UPDATE: Cass Sunstein emails:

In terms of counting: I looked over 41 dissents (not including the 14 or so concurring and dissenting opinions). Some of them are easy to code in ideological terms; some of them aren't. Somewhere between 13 and 20 are best treated as "neutral," that is, no ideological valence at all. Of those with an ideological valence, somewhere between 100% and 85% are to the majority's right. My best estimate is on the high end of that range. -- Reasonable people can differ, of course, about the precise calculation, but by any objective count, I think, the overwhelming majority (of those with an ideological valence) are to the right -- and more important, in a distinctive and interesting way, that is, they ask for deference to powerful institutions (and hence show little or no discernible libertarian streak). -- I confess that I wasn't looking (or hoping) for this pattern. It really surprised me. I really want to be in favor of Alito and haven't made up my mind -- this is a just a source of concern.

So, spread out over 15 years on the bench, we're talking about maybe one case a year. I'm not so ready to feel concerned about this. I'd like to see the actual cases. It could be that Alito is just turning out workmanlike analysis where the other two judges were stretching for a liberal result. Are the dissents on panels with especially liberal judges or in cases where it was tempting to the majority to overreach in a liberal direction? Are the dissents clustered in his early or later years or evenly spread over the years? Somehow these statistics don't speak to me other than to say: examine this more closely.