HelpAbout LSAC

Crafting an Incoming Law School Class: Preliminary Results (RR-99-01)

Executive Summary

Each law school possesses unique characteristics that may reflect its goals, funding sources, location, etc., and these unique features translate into specialized admission practices. Clearly, no one admission model can be applied blindly to all law schools. However, there are some common practices shared by most law schools. Roughly speaking, most law schools attempt to attract and admit highly credentialed individuals, while at the same time seeking out individuals who possess other attractive characteristics. Special consideration may be given to applicants who might otherwise be foreclosed from a legal education or those who demonstrate an interest in pursuing a particular career after law school. The composition of incoming classes may also be driven in part by an aspiration to address implicit or explicitly stated mission goals of the law school.

More specifically: most law schools attempt to attract and admit those applicants who will be most likely to succeed at that law school. The measures most often used to predict applicants' success in law school are their Law School Admission Test (LSAT) scores and undergraduate grade-point averages (UGPAs), or some combination of these variables. Other measures may also be useful in predicting success at particular law schools.

As importantly, most law schools also attempt to attract and admit individuals who may bring other attributes (besides high LSAT scores and UGPAs, for example) to their classes. These additional attributes may enrich the law school experience for all or may further the goals of the law school. Applicant attributes that might enrich the law school experience could include strong leadership qualities or unique work experiences. Applicant attributes that might address law school goals could include a low-income background or a rural upbringing.

The approach proposed here formalizes the admission process by first requiring law schools to delineate the characteristics they want their incoming classes to possess (e.g., types of undergraduate majors, percent of in-state versus out-of-state residents, and levels of cultural diversity). These are then used as “constraints” on the selection of an incoming class or admit pool. A separate “optimizing” variable (e.g., average LSAT or UGPA) is used to choose among the subsets of applicants that satisfy the stated constraints. This is a common selection problem in the field of operations research and can be solved by a procedure known as constrained optimization.

Consider the following simple example: A law school wishes to select from a 1,000-person applicant pool a 200-person admit pool, who as a group have (1) an average age over 25; and (2) the highest average LSAT score. One way to accomplish this task would be to enumerate by hand all groups of 200 persons, chosen from the applicant pool, whose average age is at least 25. Then among these groups, the one group of 200 persons with the highest average LSAT score would constitute the optimal admit pool.

While the above simple task could be accomplished by hand, it would be exceedingly tedious and time consuming. A more complex problem, say one that involved 20 constraints, would be all but impossible to solve by hand. Fortunately, there are computer programs that can solve such problems, even extremely complex problems, in seconds. In practice, the actual numbers and types of constraints used will, of course, vary from law school to law school.

Note two very important features of this approach: (1) Applicants are not ranked in relationship to each other, and (2) applicant attributes are not weighted as being more or less important than other applicant attributes. Rather, an optimal subset of applicants is chosen to be admitted because they, as a group, satisfy certain constraints, such as an average age of at least 25, while simultaneously possessing certain maximal characteristics, such as the highest average LSAT score.

In this way, we propose the crafting of incoming law school classes according to a set of predetermined specifications, rather than the ranking of applicants. Empirical results presented in this paper demonstrate that this approach can select admit pools that are similar to those actually chosen by law schools. By using a variety of admit pool constraints, it is also shown that racially or ethnically diverse admit pools can be chosen without resorting to the use of race/ethnicity indicators.

Crafting an Incoming Law School Class: Preliminary Results (RR-99-01)

Research Report Index

All contents © Copyright 2004 Law School Admission Council, Inc.
All rights reserved. LSAT® & LSAC® are registered by Law School Admission Council, Inc.