2008 USN&WR Law School Rankings Under New Emp9 Formula
What impact will U.S. News & World Report's new formula for calculating the "employment at 9 months" variable have on next year's law school rankings? Most importantly, it will make the rankings more accurate. No longer will a law school be able to arbitrarily increase its rank by strategically classifying unemployed students studying for the Bar exam as "unemployed, not seeking employment." Even though flaws will continue to mar the Emp9 measure, as well as other aspects of USN&WR's methodology, we can look forward to a better and fairer assessment of law schools in next year's rankings.
The Emp9 formula that will kick in next year might not make everyone happy, however. It will doubtless cause some law schools to fare less well in the upcoming (the "2009") USN&WR rankings than they did in the most recent (the "2008") rankings. Which schools will drop in the rankings, and by how much? That is of course impossible to say. We can estimate the new Emp9 formula's future impact, however, by making it hypothetically effective a year earlier. I did just that, using my model of the 2008 rankings. Herewith the results:
In addition to those results, my experiment indicates that implementing the new Emp9 formula a year earlier would have caused:
How did I come up with these results? First, I downloaded the February 2006 employment data that the American Bar Association published for each law school it accredits. Second, I ran that data through the formula that USN&WR plans to use to calculate Emp9 scores in next year's rankings. Third, I plugged the resulting Emp9 scores into my model of the most recent rankings, substituting it for the Emp9 scores USN&WR calculated under its old formula. That gave me new scores for all the schools ranked by USN&WR. To best estimate the impact of the new formula, though, I didn't just use those scores. Rather, I added or subtracted to the scores USN&WR published for each of the law schools in the top two tiers of its 2008 rankings the difference between what my model had originally calculated for those schools and what it came up with under the new Emp9 formula. (For schools in other tiers, though, for which USN&WR does not publish scores, I simply compared the model's original and recalculated scores.) That generated the final result: an estimate of what effect the new Emp9 formula would have had had it been instituted this year.
Please allow me to emphasize that this experiment does not suffice to predict next year's rankings. My model does not perfectly mimic USN&WR's law school rankings, which at any rate rely on much more than only Emp9 measures. Also, as I observed earlier, it appears that for a few schools USN&WR used different Emp9 numbers than those published by the ABA. I of necessity used the latter numbers for this exercise. Speaking only as a rankings geek, though, I daresay that even these admittedly imperfect results prove interesting.
[Crossposted to Agoraphilia.]
Earlier posts about the 2008 USN&WR law school rankings:
Earlier posts about Emp9 measure:
The Emp9 formula that will kick in next year might not make everyone happy, however. It will doubtless cause some law schools to fare less well in the upcoming (the "2009") USN&WR rankings than they did in the most recent (the "2008") rankings. Which schools will drop in the rankings, and by how much? That is of course impossible to say. We can estimate the new Emp9 formula's future impact, however, by making it hypothetically effective a year earlier. I did just that, using my model of the 2008 rankings. Herewith the results:
In addition to those results, my experiment indicates that implementing the new Emp9 formula a year earlier would have caused:
- Florida International University School of Law to fall from the third tier into the fourth tier (to a point about 1/3rd down from the top of the fourth tier);
- Valparaiso University School of Law to (just barely) escape from the fourth tier to the bottom of the third; and
- Northern Illinois University School of Law to leap from the fourth tier to a point well up from the bottom of the third tier (about 1/3rd of the way up from the bottom of that tier).
How did I come up with these results? First, I downloaded the February 2006 employment data that the American Bar Association published for each law school it accredits. Second, I ran that data through the formula that USN&WR plans to use to calculate Emp9 scores in next year's rankings. Third, I plugged the resulting Emp9 scores into my model of the most recent rankings, substituting it for the Emp9 scores USN&WR calculated under its old formula. That gave me new scores for all the schools ranked by USN&WR. To best estimate the impact of the new formula, though, I didn't just use those scores. Rather, I added or subtracted to the scores USN&WR published for each of the law schools in the top two tiers of its 2008 rankings the difference between what my model had originally calculated for those schools and what it came up with under the new Emp9 formula. (For schools in other tiers, though, for which USN&WR does not publish scores, I simply compared the model's original and recalculated scores.) That generated the final result: an estimate of what effect the new Emp9 formula would have had had it been instituted this year.
Please allow me to emphasize that this experiment does not suffice to predict next year's rankings. My model does not perfectly mimic USN&WR's law school rankings, which at any rate rely on much more than only Emp9 measures. Also, as I observed earlier, it appears that for a few schools USN&WR used different Emp9 numbers than those published by the ABA. I of necessity used the latter numbers for this exercise. Speaking only as a rankings geek, though, I daresay that even these admittedly imperfect results prove interesting.
[Crossposted to Agoraphilia.]
Earlier posts about the 2008 USN&WR law school rankings:
Earlier posts about Emp9 measure:
posted by Tom W. Bell at 2:12 PM
0 Comments:
Post a Comment
<< Home