Editor's Note: The following is based on comments delivered by Robert Townsend, AHA's assistant director for research and publications, on April 15, 2002, to a meeting of the National Research Council's Committee to Examine the Methodology of Assessing Research Doctorate Programs, which is preparing for the decennial study of such programs (see News item on page 5). While delivered on behalf of the AHA, the opinions do not necessarily represent the opinions or policy of the AHA Council. We offer them here for further discussion and comment. Responses can be sent by e-mail to Perspectives or to the AHA headquarters offices. The committee's staff director, Charlotte Kuh, will provide a status report on the committee's work at the AHA's January 2003 annual meeting in Chicago.

We are grateful to the committee for requesting our input, and we certainly stand ready to provide any and all assistance you might need. Since the committee's initial concerns seem addressed to the practical utility of the study, I should state up front that as the person responsible for data analysis at the Association, we found the objective data in the last report exceptionally useful for doing cross-disciplinary analysis, and we certainly hope you will choose to conduct the survey again.

However, as someone that prospective graduate students call on for advice (and as someone who recently returned to school to complete a long-abandoned PhD), my impression of the survey is more ambivalent. The survey fails to rank the kinds of things that we consider truly important when helping candidates select a history PhD program-the quality of the faculty in specific subfields, the level of financial support and attached work requirements, and subsequent placements in the field. By that measure, the study will need to do a lot more before I could recommend its use to prospective students and advisers.

Moving beyond the practical value of the work, we have three specific recommendations to make. First, we would like to see history moved out of the social sciences and into the humanities. Second, we are troubled by the homogenization of the history PhD into a single category, and would recommend breaking them down at least into rankings comparable to the taxonomy used in the annual summary of doctorates. And last, we would like to see a wider range of statistical data in the survey, which would allow us to assess the entire life cycle of a history PhD.

Recommendation 1: Move History into the Humanities

We feel that including history in the social sciences no longer accords with our self-understanding as a discipline. With few exceptions, the search for "laws" and principles in human history is over, a point perhaps emphasized by the fact that quantitative history is increasingly rare and often relegated to the footnotes. Instead, there is currently an emphasis on exploring cultural forms and production. In fact, in a recent discussion of the AHA Council, members were quite adamant about the discipline's position in the humanities.

Perhaps equally important, categorizing history in the social sciences in the 1995 report introduced a couple of false measures into the data on history, specifically on patterns of publications and citations. By imposing social science criteria that privilege journal articles, the data presented in the NRC report indicated that historians were the least productive among social science scholars, averaging less than one publication between 1988 and 1992. These numbers contrast markedly with the NRC survey of humanities doctorates-L. Ingram, et al., Humanities Doctorates in the United States: 1993 Profile (Washington, D.C.: National Academy Press, 1995)-conducted the same year, which reported an average of almost 3.3 publications per history PhD between April 1992 and April 1993

As the survey of humanities PhDs suggests, the last report ignored many forms of publication-such as books, book reviews, book chapters, museum scripts, and other publications for a general audience-that are widely credited in the field. The patterns of citation show considerable unevenness between departments, and little correlation to any of the other measures in the study. When I reviewed the report back in 1996, it was my sense that this data simply elevated programs that did more quantitative history or history of science and medicine.

Recommendation 2: Include Subfields

We are equally concerned that the homogenization of the history PhD into a single category fails to supply the sort of evidence the report is intended to provide. After a couple of decades in which the discipline was fragmented into subfields, our surveys indicate that departments are now realigning and consolidating to specialize in particular subfields (especially for the training of PhDs).

This makes the rankings less useful to students and their advisers as they try to match students with the most appropriate program. In considering PhD programs, we advise students to think carefully about which program is best suited for their subfield-whether it be a particular aspect of U.S., European, or Latin American history, or a thematic field like diplomatic or cultural history.

Like many other disciplines, we are concerned that the reputational rankings especially privilege larger programs. As your respondents are considering their rankings, we fear they will only be able to rank based on who they know in their own subfield, and particularly the recent PhDs produced by the program in their field. So the number of bodies being pushed out the door can increase a program's recognition among a wider range of those being asked to rank the programs, and correspondingly reduce the rankings of a number of exceptionally fine programs with a narrower range. This problem has been exacerbated in recent years by the growth in one specific field-20th-century U.S. history-that currently accounts for almost one-third of the PhDs being produced, but only about 9 percent of the teaching faculty in the departments.

Recommendation 3: Provide More Data on the Life Cycle of the PhD

Finally, we feel the report could be significantly more useful for cross-disciplinary comparison, guidance for effecting real change in the departments, and practical advice for potential graduate students, if it provided additional data on what we think of as the life cycle of a PhD. The Association is currently confronting many of the same problems in a major survey (funded by the Carnegie Corporation) of history PhD programs, and we are working to develop data on some of the critical stages toward the PhD, including:

• Admissions (particularly of women and minorities, since we want to know more about why and where we are falling short in the production of new PhDs among these underrepresented groups).
• Attrition over time (a difficult task, we know, and an area where we have been unsuccessful in developing reliable information)
• Funding levels for graduate students and associated work requirements
• Debt levels upon completion
• Preparation for different job goals, such as research, teaching, and other employment outside the academy
• Outcomes (particularly in hiring) in relation to the department's goals

Our own recent experience leaves us with few illusions about the daunting nature of the task. As part of our study we distributed a detailed 42-page survey to every doctorate-granting history program in the United States (158 in all), and received completed surveys back from 105 programs. While the high response rate suggests a high level of interest, the responses also indicate that many departments fail to keep information on a number of key issues-such as accrued student debt and placement of their degree recipients-and simply do not know how to account for other key factors such as attrition. If the methodology committee could provide recommendations on the type of data that should be collected by departments, and a standard formula for measuring attrition, it would be doing an enormous service.

In sum, we are enormously grateful for your efforts and look forward to the next edition of the report.

—Robert Townsend is AHA's assistant director for research and publications.