September 26, 2008

Justice Alito opts out of the cert. pool.

He's rejecting the efficiency of the system of shared law clerks in which clerk writes a memo relied on by Justices using the pool to decide whether to grant the petition to the Supreme Court to hear a case.
A petition accepted that must later be dismissed as “improvidently granted” is a significant embarrassment to the clerk in question. On the other hand, it is hard to get into trouble, [Pepperdine School of Law dean Kenneth] Starr said, by recommending a denial. “The prevailing spirit among the 25-year-old legal savants, whose life experience is necessarily limited in scope, is to seek out and destroy undeserving petitions,” he wrote.

The justices decided 67 cases last term, about half the number in an average year two decades ago. But Justice Alito has said the rise of the pool and the size of the docket are unrelated.
Starr's theory implies that they are related, and Alito's statement was made last year. Perhaps he's changed his mind.


wurly said...
This comment has been removed by the author.
Ann Althouse said...

He's rejecting something that is efficient, in pursuit of some higher value that we're left to wonder about.

wurly said...
This comment has been removed by the author.
Simon said...

Wurly, while I agree with your premise that the pool's efficiency would be tainted to some extent if causes meritorious petitions to be rejected, I do wonder how prevalent a problem it is and thus I doubt what I take to be the strong form of your thesis that "if there are [any] cases the Court should hear but doesn't, due to risk averse clerks, the pool system is not efficient."

To a very limited extent, this is a measurable problem. If the court is routinely turning down certworthy petitions, it should be no trouble at all to find examples of petitions that should have been granted but weren't. If we stipulate, arguendo, that the cert pool was the reason those meritorious petitions were turned down, that would give us a maximum threshold for the problem. If that number turns out to be minimal, we can conclude there is no problem, since the number of denials that actually attributable to the cert pool cannot exceed the total number of denials of meritorious petitions and is likely to be far less.

Bender said...

He's rejecting something that is efficient, in pursuit of some higher value that we're left to wonder about.

An even more efficient system would be to simply not read any of the cert petitions and reject them all. But then, that would mean that the justices weren't doing their jobs. But then again, delegating large amounts of the judicial power to unconfirmed clueless clerks is also a case of the justices not doing their jobs.

The fates of people's cases should not be determined to such a great degree by a group of kids a couple of months out of law school. It is bad enough that too many of the justices are brain dead ignoramuses, and we really do not need clerks being the de facto deciders of cases.

Simon said...

Bender, you don't seem to realize the volume of inputs to the court. There are in the region of eight to nine thousand cert petitions a year - that's between twenty and twenty four per day. Granted, some are plainly frivolous. Some are pro se. It doesn't seem outlandish, however, to assume that the average petition is going to involve a few hundred pages of reading - the opinion(s) below, the petition itself, the reply, less familiar cases. If the justices themselves have to do that, you've already given them several thousand pages to read every single day of the year before they've taken or decided a single case. That isn't a recipe for optimal performance of judicial duties.

Nor is it clear to me how having clerks write cert memos - an entirely separate question to the cert pool, by the way - is a delegation of power. You think that the clerks would abuse their position, or that they're not competent to summarize the issues in a petition? These aren't applicants who just walk in off the street. They are, ordinarily, the cream of the cream, highly competent young lawyers.

Your comment suffers from the same problem that afflicts all these criticisms of the docket, one I alluded to earlier: name a certworthy petition that was denied turned down where a reasonable person could conclude that there weren't good reasons why the justices might themselves have turned it down. Even if that class is overinclusive, it should be easy to find one, because the meritorious cases denied because of clerk bias must be a subset of all meritorious cases denied. Coming in with some of those cases should be a threshold requirement for these criticisms, even though it doesn't get you the whole nine yards.

Hey said...


We don't need to show a specific case, especially when your standard is so high. Evaluations of decision systems and processes are fairly standard, whether it's for terrorist screening, fingerprint identification, or granting cert.

The fewer improvidently granted petitions, the more improvidently denied - fewer false positives always equals more false negatives, when you have the same basic system. The pressure on clerks to not grant improvidently necessarily pushes more legitimate certs out, it's just the basic math (ok it's fairly serious applied stats, but it's junior year engineering and AI work, though maybe post-doc level for social science researchers).

For all decisions systems you have two bell curves representing actual positives and negatives, these curves overlap (the level of overlap depends on the design of the system but they ALWAYS overlap) and the tuning of the system is to decide where the cut-off line is. This cut-off gives you the %false positive and false negative. The design of the current system pushes that cut-off line far to the right in an attempt to eliminate false positives.

The big issue in systems design is deciding the "right" level of false positives and negatives. In very high stakes systems (like terrorist screening) you accept more false positives, in low-stakes high-volume systems you accept very many false negatives (6 sigma manufacturing and "five 9s" telecom reliability are predicated on large systems with low costs of false negatives - cheaper to reject a few good parts than to let bad parts get to clients).

With the importance of the court and the size of issues judged, it's easy to determine that the cert granting process should be biased towards recommending granting certs. This is true regardless of whether you want the court to have a small or large docket as it is the judges who should be making the decisions and they can better prune back cert grants but have a much harder time finding cases that should have been recommended but weren't.

Simon said...

I don't think I'm setting the bar particularly high! I do so by saying that if the court's processes are resulting in denials that should have been grants, there ought to be cases exemplary of the trend? How difficult can it be to point to a clearly certworthy case that was denied, if the problem is as endemic as some critics charge? Or, to put conversely, I think you do need to show an example - nay, examples - because if no examples of rejected certworthy cases can be found among the 8000 or so cases rejected every year, how widespread can the problem be?

In any event, I'm not defending the size of the docket or whatnot. The point that I'm getting at is that this issue can and ought to be tackled empirically in the first instance, and I don't see anyone doing that legwork. Ipse dixit and deduction have their place, but where it's substituted for obtainable data, it's pretty thin gruel. One might even say that it in fact serves to undermine your case: it suggests that the writer either doesn't want to do the work or is afraid of what the numbers might show, neither of which is a strong position to argue from.

Bender said...

Simon, you don't seem to realize how ignorant you are about what I know and don't know. I am well aware of how many certiorari petitions are filed each year, so your condescending comments are unnecessary. Indeed, I've gone through a couple of thousand of those very same petitions in my work. The fact remains, regardless of the case load, that our constitutional system did not envision people who are basically legal apprentices having such de facto power over the outcome of cases.