First, we dispute that Sander is conducting responsible research. As we state in our commentary, there’s been quite a few critiques of Sander’s methodology and an absence of concurring published research.

The privacy in this case is mandated by law and the confidentiality the State Bar agreed to when exam takers provided their information.

Disclosing this data is against federal and state law. Period.

CFAC continues to frame this issue as one of malicious refusal to provide the data requested. Again, by law, this data cannot be released.

Your hypothetical about what EJS’s position would be on a student assignment case does not support your claims.

On your final point, our stand on this issue is primarily about the protection of privacy and adherence to the law. Our disagreements with Sander’s hypothesis pale in comparison to those two paramount issues.

Keith,

I respect your concerns about the privacy of test results, but they are misplaced. Bar data produced according to the specifications of our record request would not allow identification of the test-takers–not by Sander, not by researchers who disagree with him, and not by anyone else.

When data are “anonymized” to the extent that individuals are not identifiable, privacy is not compromised. That is true not just for this data, but in many other situations where, I’m quite confident, your organization would be among the first to argue that privacy considerations cannot stand in the way of disclosure.

Consider, for example, a lawsuit against a public school district alleging a pattern of student class assignments that effectively denies qualified minority students the opportunity to take advanced placement courses. This claim can only be litigated by analysis of confidential student records. Surely you would not support a school district’s refusal, on privacy grounds, to produce the records.

It is your right to disagree with Sander–and we would stand with you to defend that right. But the way to disagree is to dispute his interpretation of the data, not to block his (or any other scholar’s) academic research on an issue of clear public importance.

Best Regards,

Peter Scheer
CFAC exec director

As a practical matter, if the position taken by CFAC is correct — that they seek to make the data broadly available, not just to Sander — that only heightens the danger to privacy interests. If the records are truly public as they claim, then anyone can get them and will not be bound by any protocols or standards regarding what they do with them.

Mr. Chadwick argues that if the data is not made available, an important policy question cannot be resolved. As many important research questions as there are in the world, we as a society have decided that you have to follow certain rules and ethical guidelines in investigating them. This is particularly so the investigation involves personal information about human beings and can impact their lives and well being. The problem here is that the rules require the consent of the subjects. And in this case, the subjects haven’t given it.

I respectfully disagree. The premises of the Equal Justice Society’s objections to the litigation are mistaken.

One premise is that the data is being sought to support Prof. Sander’s claims.

That is not true. The records are being sought so that all interested researchers can have access to them, in order to conduct research that the data they contain are uniquely capable of supporting. If the records are made public in response to CFAC’s petition, they will be available to all researchers, including those who disagree with Prof. Sander. Moreover, your premise implies that the data will be manipulated to support Prof. Sander’s hypothesis, regardless of what it actually demonstrates. That is not accurate. The records will be used to evaluate whether there is or is not support for that hypothesis, and the results of that analysis will be made public regardless of what the records show.

Your assertion that state and federal law prohibit the disclosure of these records without individual consent is also incorrect.

As the Equal Justice Society is aware, the requests by Prof. Sander, Mr. Hicks, and CFAC specifically call for the elimination of any information that would directly identify bar examinees. Moreover, they propose the adoption of redaction procedures that will ensure that no individual examinee can be indirectly identified. Thus, for example, disclosure pursuant to the requested procedures would ensure that no combination of characteristics that might conceivably make an examinee indirectly identifiable will ever apply to a group smaller than five. Thus, no individual examinee will be identifiable. Experts in data anonymization have confirmed this.

Although EJS does not identify the basis for its assertion that disclosure would violate state and federal law, we have thoroughly researched and considered the applicable provisions of state and federal law, and it is clear that they do not prohibit disclosure of the records in the manner that CFAC and the other petitioners have requested. The Family Educational Rights and Privacy Act (FERPA) governs the privacy and disclosure of educational records. Under FERPA, anonymized educational records are routinely made public, by education institutions and by the federal government itself. Every case we have found addressing a request such as that made by CFAC held that disclosure was required, and does not violate FERPA. Notably, in its opposition to our petition in the California Supreme Court, the State Bar does not assert that disclosure would violate FERPA. Nor would disclosure of anonymized records violate the California Constitution’s protection for privacy, because California law does not preclude the public disclosure of records from which identifying information has been removed.

To the extent that EJS nonetheless believes that disclosure of the records in the manner requested by CFAC is improper, it should examine and consider the reports and studies that have been publicly disseminated by the State Bar for over thirty years. For example, the State Bar’s reports on each of the semi-annual bar examinations routinely combine indirect identifiers such as law school, race, and bar passage in a way that describes groups of examinees that may be as small as a single individual. The State Bar has determined that the disclosure of this information is not contrary to federal or state law. It cannot coherently assert that its own reports and studies comply with the law, but that disclosure pursuant to the more protective procedures proposed by CFAC in this case would not.

Finally, EJS suggests that the proposition that Prof. Sander’s mismatch hypothesis is “uncorroborated” means that CFAC should not seek records relevant to proving or disproving that hypothesis, and that the Supreme Court should not release such records.

Assuming, for the moment, the validity of the proposition that Prof. Sander’s hypothesis has not been corroborated, I am surprised that EJS would take the position that a state agency should not be required to release records sought to determine whether such a hypothesis can be corroborated. First, where, as here, government records will cast light on an important social and governmental issue, the need to test such hypotheses is among the most compelling reasons for making records public. Second, the logic of this objection is perfectly circular: the hypothesis has not been corroborated, so the records that might corroborate it should not be made public. By that reasoning, disclosure would be appropriate only when it was unnecessary, i.e., if the hypothesis had already been proven.

I encourage EJS to reconsider its opposition to disclosure.

If these records cannot be released in order to provide a basis for research that might support Prof. Sander’s thesis, then they cannot be released in order to support research that might demonstrate, for example, that Proposition 209 has negatively impacted the diversity of the California Bar, that the California bar examination is racially biased, or that law schools are engaging in practices that systematically impair the performance of racial minorities. EJS should carefully consider whether it is content to rely upon the conclusions of the agency charged with administering the bar examination and the bar admissions process that there are no problems with the examination or the admissions process that might contribute to the reduction of diversity in the California Bar.

The objective of the litigation is to provide a sound basis for evaluating issues regarding bar examination performance and admissions–regardless of the hypothesis sought to be tested. Our hope is that sound objective data will provide the foundation for sound objective decisions regarding how best to enhance and ensure the diversity of the California bar. Consistent with CFAC’s mission, we do not believe that suppressing information about public issues that can be provided without compromising personal privacy is consistent with anyone’s interests.
===
James Chadwick, an attorney in Palo Alto, represents the petitioners in the suit against the State Bar.