Roger Bowen and the Banality of Academic Ethics

The American Council of Trustees and Alumni (ACTA) held its annual meeting this past Friday, and one of the speakers was Roger Bowen, president of the American Association of University Professors. Candace de Russy and I have previously taken Bowen to task in the New York Sun (also see the Hudson Institute website) for unfairly casting aspersions upon Rothman, Lichter and Nevitte , who produced a study finding that liberals outnumber conservatives in academe, a claim controversial only to academics themselves, who seem to lack self-awareness. Roger Bowen is to be commended for speaking to a conservative audience--yet why the head of a mainstream academic organization needs to have an ideological affiliation is difficult to fathom (why should Roger feel like a "masochist", as he put it, when speaking to a conservative group any more than to a liberal group?).

Roger Bowen called Rothman, Lichter and Nevitte “unethical” because they have not “shared their data”. This claim is unfair because Rothman has repeatedly stated that he would share his data as soon as he and his coauthors are through analyzing it and he has said so on a panel at the Manhattan Institute last summer where Roger Bowen was also a panelist. Roger Bowen’s insistence on name calling is troubling because Bowen ought to represent all university faculty in his role as head of the American Association of University Professors.

Bowen selectively defames Rothman while avoiding comment on powerful liberal academics’ long standing refusal to share their own data. I’m speaking particularly of William Bowen (no relation to Roger, I believe) and Derek Bok, who have for years refused to share their data on affirmative action that was in their book Shape of the River. William Bowen is former president of Princeton and currently head of the Mellon foundation and Derek Bok is former president of Harvard. They share control of the Shape of the River data with the Spencer Foundation, headed by Mike McPherson.

The leading American scholar on the subject of affirmative action is Thomas Sowell, and Dr. Sowell has complained about McPherson’s and Bowen’s refusal to share their data in his book Affirmative Action Around the World, which I reviewed for the Journal of Interdisciplinary Studies. Concerned about Sowell’s complaints, I contacted William Bowen and Derek Bok, who directed me to Mike McPherson. All three give reasonable explanations for not sharing the data, but their reasons are much less convincing than Rothman’s. Yet, Bowen insists on calling Rothman “unethical.” I am waiting for Roger to publicly call William Bowen, Mike McPherson and Derek Bok “unethical.” This double standard is revealing about ethics in academia.

Copies of E-mails from William Bowen and Mike McPherson follow:

Mike McPherson of the Spencer Foundation writes:

This is in reply to your inquiry of April 22. Please allow me first to clarify my role in relation to the College and Beyond database. The College and Beyond data continue to be held by the Andrew Mellon Foundation. Some years ago, the Foundation created a committee to advise the Foundation on proposals by researchers from outside the Foundation to use the College and Beyond data. I was asked to chair that advisory committee, as I continue to do. That responsibility is unrelated to my work for the Spencer Foundation and indeed it substantially predates my coming to Spencer. The task of the committee is simply to appraise researchers' requests for use of the C&B database in terms of the quality of the research design, the appropriateness of the C&B data for the proposed purpose, and the potential significance of the findings.

Your email raises a concern that there has been political bias in the Foundation's decisions about who will have access to the database. You cite the case of Robert Lerner, whose request for use of the data was turned down by the Foundation. But this decision was purely procedural. As you know from reading the correspondence, Lerner wanted access to an adjunct to the C&B database consisting of files of application and admissions data for five institutions. Access to these data was granted to the Foundation on the explicit condition that they would never be made available to outside researchers -- a condition that does not pertain to other parts of the database. The institutions who provided the data had what I am sure they considered good reasons for this restriction. They regarded these data as particularly sensitive, and with a dataset involving only five institutions, the risk that a data release might inadvertently allow identification of information about individual institutions (or individual students) is heightened. You might wish nevertheless that the institutions had not imposed this restriction but it seems entirely clear that the Foundation, having obtained the data on the condition of this restriction, would never proceed to violate it. Lerner's application was thus properly denied on procedural grounds, and indeed was never brought to our committee for substantive consideration of his proposed research.

This is the only case you cite. Your email seems to suggest that either Professor Sowell or Professor Thernstrom applied and was turned down, but if I read correctly the commentary you have provided, neither of them ever did apply, and we have no record of such an application. Apparently Sowell and/or Thernstrom may have concluded that if they applied they would be turned down, but I see no basis for such a conclusion -- provided of course that they intended to present a well-thought out research proposal drawing on aspects of the database other than the one described above to which a special restriction applies. It seems, then, that you have not offered any evidence at all of a political bias in access to the data.

As to your request for the names of applicants for use of the data, I would strongly recommend against the Mellon Foundation providing any such information to you. The Foundation has never given applicants reason to believe that their names would be shared in this manner, and I see no good reason why they should be shared. For that matter, knowing the names and political affiliations of applicants would be of little use, since the key question is the quality of the research proposals. And I would consider it highly inappropriate for the Foundation to share these proposals for purposes unrelated to the Foundation's decision about whether to make the data available to the applicant.

My committee stands ready to recommend approval of any well-reasoned and analytically sound proposal to use data from College and Beyond to address a scientifically significant question, provided the proposal is offered in good faith by a qualified researcher who is willing to abide by the confidentiality restrictions that apply to these data, and provided that the question cannot be addressed well through the use of other, less sensitive, data. These requirements, and our judgments, are independent of the researcher's political views or his or her views on affirmative action.

Sincerely,

Mike McPherson
William Bowen writes on 4/22/05:

If you have questions about the fairness of access to the data, you could contact Dr. McPherson, who chairs the committee that makes these determinations. Our criteria are entirely non-ideological -- and in at least one important case resulted in giving access to the data to scholars who produced findings very different from our assumptions. (I have absolutely no recollection of the Lerner application, and there is no reason why I would have.) Also, in many cases other uses of the data could not be said either to support or criticize race-sensitive admissions -- there are, after all, many other uses for these data. For instance, one study was done of in-state vs out-of-state tuition using these data. Another study examined differences between men and women in the choice of major, SAT scores held constant. And then of course there are the two big books on college sports. As I said before, we stress hypothesis-testing and rigorous methodology -- as I think we should do -- and as we promised those who provided the data that we would do.

Bill Bowen

William Bowen writes on 4/18:

Mr. Langbert,

If Prof. Sowell wishes to use the data in the database, he will have to submit his own request, and it will need to contain -- as all requests do -- a carefully worked out research proposal demonstrating what his proposed research would contribute, what statistical, econometric techniques he would employ, and so on. He would also have to conform to all established confidentiality agreements, as I'm sure he would want to do. Finally, I should be clear that proposals that are submitted are reviewed on their research merits -- as, again, I'm sure you would expect them to be.

Bill Bowen