Law professor says data backs up mismatch theory

UCLA School of Law Professor Richard Sander said data he reviewed from two California law schools supports his so-called “mismatch hypothesis” that giving applicants’ admissions preferences, such as for race, leads to them not performing as well on the bar exam.

The draft paper Sander co-authored and released last week said it “analyzes law school mismatch with more precise data than hitherto has been available.”

For years, Sander has unsuccessfully sought to obtain data from the State Bar about California bar exam applicants to study the mismatch hypothesis.

The two schools from which Sander said he obtained information are not named in the paper posted on the Social Science Research Network, but the document indicates their median LSAT scores were 165 and 161, respectively.

The information from the higher-ranked school, referred to as “School A,” covers graduates from 2001, 2005 and 2006. Data from School B covers graduates from 1998-2012.

Data on the graduates’ racial background was obtained from both American Bar Association-accredited schools, he wrote.

First-time California bar takers with LSAT scores of 162 and above from School A performed better than takers from School B with the same credentials.

First-time takers with LSAT scores of 158 or lower at School B outperformed their counterparts at School A, in some cases by large margins. For example, 64 percent of School B bar takers with LSAT scores in the 153-155 range passed on their first attempt, while only 47 percent of their peers from School A did.

The data from the two California schools was compared to 2005-2011 information about graduates from a public law school in Arkansas. After controlling for differences in the exams, a similar trend in bar passage rates was demonstrated, Sander’s paper stated.

“It should now be obvious to everyone that mismatch is real,” Sander said in an interview.

Sander’s paper, co-authored with University of Arkansas law professor Robert Steinbuch, states, “When we control for mismatch effects and entering credentials, racial deficits in bar passage rates either disappear or become statistically insignificant.”

While Sander’s focus on mismatch has previously generated criticism, at least a half-dozen legal education experts declined to comment on his draft paper or did not respond to requests for comment.

Some proponents of lowering the passing score on the state’s bar exam argued that doing so would provide the biggest boost to racial minorities.

Bar staff acknowledged racial disparities in bar passage in a report provided to the agency’s Board of Trustees last month.

“While the root causes of disproportionate rates of passage are beyond the scope of this report, it is clear that applicants of color pass the bar exam at rates that are disproportionate to those of their white counterparts,” the document said.

For example, 51.2 percent of Caucasians passed the July 2016 California exam compared to 20.6 percent of African-Americans, according to the bar.

The state Supreme Court decided last Wednesday to maintain the passing score on the bar exam, which is the nation’s second highest.

“Given the size and seriousness of racial gaps in bar passage, it is of great importance that available data sources for studying mismatch more generally — in particular, the California bar’s admissions dataset — should be made available for objective research,” Sander’s paper said.

A San Francisco County Superior Court judge ruled last year that the bar did not have to provide decades of admissions data sought by Sander and the First Amendment Coalition. Sander v. State Bar of California, CPF08508880 (S.F. Super. Ct., Nov. 7, 2016).

The decision is on appeal.

The bar announced plans earlier this year to seek student data from law schools, such as LSAT scores and grade point averages, to assess what role student credentials played in the pass rate on the July exam dropping from 62 percent in 2008 to 43 percent in 2016.

Dr. Roger Bolus, the Committee of Bar Examiners’ psychometrician, will be the principal investigator. The study he will lead will examine multiple variables contributing to changes in bar exam performance, he said, but there are no plans to examine the mismatch theory.

“Once you start going down that path, it is kind of a dark hole,” Bolus said.

Law schools have been hesitant to provide the bar with data about their students to conduct the study because of concerns about violating student privacy laws, Bolus said.

He said the study ultimately should assist the state Supreme Court in future deliberations about potential changes to the bar exam.

The court has called on the bar, law schools and others to evaluate whether improvements in law school admission, education and graduation standards could raise bar exam pass rates and “boost the availability of competent and effective attorneys across all demographics and for all Californians.”

The court called for the study of other potential factors that have contributed to declining pass rates on the licensing test.

“I applaud their call for the bar to set up a consortium of internal and external scholars to study a host of questions relative to bar passage and lawyer competency,” Sander said.