We are concerned that the American Association of University Professors’ general secretary, Roger W. Bowen, recently has made public statements that selectively evaluate academic research. Such public statements breach the AAUP’s established standards. Differentially criticizing the research of conservative scholars and withholding equivalent criticism from liberals suggest institutional bias. Such bias undermines the AAUP’s claim of evenhandedly defending the professoriat’s pursuit of truth.

We refer to statements that Mr. Bowen made in the April 8, 2005, issue of the Chronicle of Higher Education concerning the study “Politics and Professional Advancement among College Faculty” by an emeritus professor of government at Smith College, Stanley Rothman; a professor of communications at George Mason University, S. Robert Lichter, and a professor of political science at the University of Toronto, Neil Nevitte. Published in the online journal Forum, the study finds campus liberals to outnumber conservatives 5-to-1. It also concludes that conservatives tend to get worse jobs than liberals when research productivity and other personal characteristics are taken into account.

Mr. Bowen’s response to the study was to declare its methodology “suspect” because the sample size was “too small” and to assert, without empirical support, that in academia “the cream rises to the top.” But the sample size in the Rothman et al study, 1,643, is larger than samples in the majority of published social-science studies. It is larger, in fact, than the sample size that the Nielsen ratings use. Moreover, the Rothman et al methodology is as appropriate to its subject as are the methodologies of many studies that are published in scholarly social-science journals each month, and whose methodologies Mr. Bowen would no doubt not presume to evaluate.

Concerned about the inaccuracy of this response by an important academic leader, one of us e-mailed Mr. Bowen an inquiry to which he responded as follows: “Several studies have recently been done, but none of them approach the breadth or depth of those done over time by the Higher Education Research Institute, or HERI, since 1989-90. Its conclusions, based on a much larger sample, reveal a very different picture.”

Reviewing the most recent HERI survey, we discovered that it paints a picture not all that different from that in the Rothman et al report. To be exact, the 2002 HERI study finds that among four-year institutions, far left or liberal faculty outnumber conservatives by a ratio of roughly 3-to-1. Additionally, a low (below 15%) institutional response rate that is self-selected opens the HERI study to questions about methodology that are as serious as those that might be raised about the Rothman et al study. A balanced observer would raise questions about both studies, or be satisfied in the knowledge that social-science research is often imperfect.

When one of us then asked Mr. Bowen for further clarification of the basis for his remarks, he indicated that, rather than the HERI study, a professor at George Mason University, Jeremy Mayer, had provided him with information that was the basis for his remarks. When one of us questioned Mr. Mayer, he wrote that he had sent “Roger a preliminary e-mail slicing and dicing the methodology” in the Rothman et al study and that “I’d write a more sophisticated analysis if I’d known Roger was going to bandy my name about.” Mr. Mayer’s e-mail did not mention the HERI study.

Mr. Mayer’s e-mail did, however, raise a number of criticisms of the Rothman et al study that deserve scholarly debate. In particular, the e-mail mentions that the Rothman et al data defined too small a portion of their sample as moderate rather than liberal or conservative, omitted foreign policy and trade issues in its definition of liberal and conservative, used a less-than-objective self-appraisal for measuring academic achievement, and omitted controls for teaching at Christian and historically black colleges. While such points are deserving of debate, they are in fact debatable. Mr. Rothman has confirmed that Mr. Bowen never contacted him for his response to these points.

While mentioning that Mr. Rothman and his colleagues “simply won’t release their data,” Mr. Mayer also provided us with a slide presentation. The slide presentation notes that Mr. Rothman states that he intends to release his data when his planned publications are finished. In addition, the presentation points out that “discrimination in hiring is a bad thing” and that it is “possible to have an interesting conversation on the issue.” It goes on to ask “how [could] discrimination happen?” and in several slides speculates about potential reasons why there might be discrimination against conservatives. It then discusses econometric criticisms and alternative explanations for the preponderance of liberals over conservatives. Mr. Rothman has indicated to us that he intends to make his data available at the Roper Institute, as he has done with respect to six previous studies. He also indicates that he plans to respond to the various comments in future academic publications.

In short, like the HERI study, Mr. Mayer’s e-mail and the slide presentation do not support Mr. Bowen’s statements to the Chronicle, nor do they address the HERI study to which Mr. Bowen at first alluded as the basis for his published remarks. Based on correspondence with Messrs. Mayer and Rothman, Mr. Bowen’s statements about sample size and “cream rising to the top” were at best exaggerations.

Ms. de Russy is a trustee of the State University of New York and an adjunct fellow at the Hudson Institute. Mr. Langbert is an associate professor of business, management, and finance at Brooklyn College of the City University of New York.