The Drift Toward Pure Numbers Admissions

[by Bill Henderson, cross-posted to ELS Blog]

Law schools are part of a production function for entry levellawyers.  Therefore, if law schools alter their admissions practices,the character and complexion of the law school applicant pool can shiftin significant ways.  On the input side, the data are crystal clear: over the last 15 years, the rankings arms race has pushed U.S. law schoolstoward a pure numbers approach to admissions.   The more interestingquestion, however, is whether prestige-conscious law firms are now,inadvertently, experiencing any fallout.   First the data.

Law schools operate in an environment of supply and demand and arefamously counter-cyclical.  When Silicon Valley was booming in the late90s, law school applicants plummeted.  When the economy faltered in theearly 90s or after 9/11, applicants spiked.  Therefore, to examine howadmissions practices have changed over time, it is important to payattention to the underlying applicant pool.  Below are trend lines formedian LSAT scores by USNWR rank for 1994 and 2007, which reflectclasses that entered in the fall of 1993 and 2006 respectively.  Duringthose two admissions cycles, the number of applicants was virtuallyidentical:  89,600 (class entering fall 1993) and 88,700 (class entering fall 2006).

Obviously the blue line (2007) is higher than the orange line(1994).  In fact, despite slightly fewer law school applicants, theaverage median LSAT increased by 2.18 points (std. dev. of 1.99). Forthe record, only three schools fell out of Tier 1 between 1994 and2007.  And it cannot be explained by the ABA policy shift thatinstructs law schools to no longer average LSAT scores when reporting 25th, 50th, and 75thpercentile figures, thus slightly pumping up the volume of highLSAT scores.  That change was not enacted until the summer of 2006. 

Here is the same analysis for UGPA (1994 data came from the Princeton Review, 2007 from the ABA):

Although we might chock some of the higher UPGAs (avg. of +.17, std.dev. of +.12) on grade inflation between 1994 and 2007, it is likelythat schools were also trying to maximize this number. More after the jump … .

When admissions officers are under constant pressure to beat lastyear’s numbers, something has to give.  I suspect it is students whotook challenging majors but have LSAT scores slightly under thetarget.    Or applicants with impressive work experience, evidence ofleadership, or a history of overcoming major obstacles.  Although LSATand UGPA scores are strongly correlated within the applicant pool, theytend to be very weakly correlated (or sometimes negatively correlated)at individual law schools.  Why?  Because applicants who are above bothmedians tend to have admissions offers at higher ranked schools. After a school locks down its target LSAT and UGPA medians, the modestoverlap between the two groups means there are precious few spotsleft.  And often those spots are used to improve a school’s diversityprofile.

Over the years I have talked with many admissions officers.  Theycorroborate the sea change.  Further, many of the old hands argue thatthe current fixation on maximizing numbers is misguided–that, based ontheir experience, great candidates are being passed over fornondescript or unadventurous students with high numbers.   In otherwords, a large portion of candidates with compelling resumes andpersonal statements are being systematically pushed down to lowerranked law schools.

At a law firm level, there is a certain irony at work.  Manypartners could not get admitted to their alma mater; yet, between 2005and 2007, as NLJ 250 hiring increased rapidly, 53% of the new jobswent to students at USN Top 20 schools.  Rigid adherence to the elitelaw school model drove the starting salary cost structure from $125,000to $145,000 to $160,000–a legacy that is hard to swallow in a downmarket.  But were these intangibles–now less prevalent at most lawelite law schools–part of the firms’ secret sauce?   To my mind, thisis an interesting question.  Further, a recent Moneyball study by Kerma Partners suggests that the answer may be yes.