That, of course, gets us back to the original issue that MoneyLaw suggested: that someone is worth more for what he's doing than for where he went to school and where he clerked. Not all law schools understand this concept in hiring, but they do in allocating annual raises. (I've never heard an executive committee of the faculty argue for higher raises for someone on the basis of school/clerkship.)
Should the analysts take those raises themselves into account? (From a personal point of view, I hope not--after I stepped down as dean and announced that I was going to Boyd School of Law at UNLV, I was told that I wasn't getting a raise for the rest of my time at Houston.) Dick Chait, of the Harvard Graduate School of Education and one of my personal gurus, has suggested that raises really don't affect what good professors do. We'd still want to teach well and continue our research whether or not the administration rewarded us for our past work.
What about raises that reflect a bidding war for a professor "in play"? There are good and bad reasons for an analyst to look at those types of raises. A large raise (or the granting of a professorship) indicates the current institution's valuation of the professor but could be limited by the current institution's funds, by a provost who doesn't approve the raise, or by a dean's unwillingness to retain that professor. Most retention packages fly slightly under the radar, making more internal news than external news. And the more that people have to resort to school-shopping to get large raises, the less institutional cohesiveness that school will have.
So what factors should MLPE analysts use?