In 1970, under the leadership of the late Clark Kerr, the Carnegie Commission on Higher Education created a classification of colleges and universities to address a specific problem: existing classifications were not well-suited to the commission’s analytic needs. Three years later, the commission published “A Classification of Institutions of Higher Education” because, as Kerr put it in the foreword, “we felt that it would be helpful to many individuals and organizations that are engaged in research on higher education” (emphasis added). The purpose of the Carnegie Classification was, and continues to be, to support the study of higher education.
But in American higher education, curious things happen with such lists -- lists that say, “these colleges are similar, and they are different from those colleges.” When some groupings are populated by institutions of national prominence (let's admit it, they have much in common), and those groupings are numbered “I,” it doesn’t take long for a research-oriented classification system to look like a ranking. Add to the mix the periodic revision of the classification, and the genie is out of the bottle: Senior administrators searching for metrics of progress and success, as well as ways to win more state funding and alumni support, recruit the best talent, and perhaps advance their own careers along the way, scrutinize the architecture of the lists to ask, “How can we move up?” And the higher education press can tell a story of institutions in ascendance and falling behind, of winners and losers.
The 1973 edition of the Carnegie Classification had four sets of institutions that were subdivided into “I” and “II,” and only two of the four “I” lists contained high-status institutions. So not all joined the frenzy to move up. But enough did that the classification was frequently identified as an object of institutional strategy -- the research tool had become an unintended policy lever. Nowhere was this more evident than among the group that was most finely differentiated in the original classification, doctorate-granting universities.
As this process unfolded over the course of four classification revisions, some truly ridiculous claims were made as colleges and universities sought to spin their new “Carnegie rankings.” They boasted that Carnegie had placed them among the top 4 or 6 percent of colleges and universities nationwide, asserting by implication that all research universities are superior to all comprehensive universities, liberal arts colleges, and community colleges. They proclaimed that overnight, the degrees of their graduates had become more valuable. When U.S. News repackaged the classification in its rankings, using terms like “national universities" and "national liberal arts colleges,” and a college's classification determined not only its U.S. News label but the colleges and universities it would be compared with, it focused the mind at many institutions whose classification had not previously been a subject of discussion or concern.
Over the years, many voiced concerns over the classification's impact, calling for various changes. Some charged that it rewarded institutions for prioritizing research over teaching. Some said it failed to adequately address the teaching and service functions. Some wanted to use the classification strategically to shape institutional behavior. In addition to the tournament mentality that had grown up around parts of the Carnegie Classification, higher education itself became more complex. Between 1970 and 2000, the number of accredited, degree-granting institutions grew from about 2,800 to nearly 4,000. The student population and patterns of participation changed, as well. And yet the analytic framework chosen by the Carnegie Commission in 1970 remained the predominant way to characterize similarity and difference among colleges and universities (with periodic changes to criteria and labels).
So starting in the late 1990s, the Carnegie Foundation began thinking about overhauling the classification system. We wanted to identify and remedy blind spots in the classification framework, while providing more nuanced and flexible ways to represent the diversity of higher education. An interim update to the basic classification framework was issued in 2000 because the then-current 1994 edition was out of date, and we began a serious effort to develop new tools for examining institutional similarity and difference along several dimensions.
This week, the Carnegie Foundation will release five new classification schemes, each of which offers a different perspective on how institutions resemble or differ from one another (institutions will be classified on each of the new schemes). In December, we will also release a substantially revised version of the basic classification framework. A preliminary version of the five new classifications has been available online for about a month, but preview's purpose is to identify and correct possible errors in individual classifications; the preview site does not provide category listings. As part of the official release, we will provide a Web-based facility for generating standard and customized classification listings and downloads.
The new classifications are organized around three central questions: What is taught? To whom? and In what setting? A system of several independent and parallel classifications provides different lenses through which to view institutions. By expanding the system in this way, we can take selected attributes that are considered in a limited way in the basic framework and delve more deeply. For example, the basic classification says nothing about undergraduate education for institutions awarding more than a threshold number of master's or doctoral degrees (and it says fairly little about it for the others). With multiple classifications, we can address undergraduate and graduate education independently. Thus we can classify with respect to the undergraduate program regardless of the extent of graduate education, and look at graduate education in a separate classification. In addition to these two classifications focused on the instructional program (what is taught), two characterize the student population (to whom) and one groups institutions according to size and residential character (in what setting).
Viewing an institution from these multiple perspectives offers a fuller, more textured portrait than is possible with a single classification. Colleges and universities will no longer be characterized on the basis of a single view of what they do. For example, a research university’s "portrait" will reflect not only its commitment to graduate education, but also the nature of its undergraduate program, the characteristics of its undergraduates, the relative size of undergraduate and graduate populations, and the absolute size and residential character of the campus. These are all important dimensions of a complex institution, but all but the first are rendered invisible in the conventional classification.
By addressing selected dimensions of similarity and difference in independent classifications, we can also examine the points of intersection among the different classifications. Which institutions emphasize professional fields at the doctoral level yet emphasize the arts and sciences in their undergraduate education? Which ones emphasize business programs at the master's level and blend the arts and sciences with occupational/professional training at the undergraduate level? Which ones serve a primarily nonresidential undergraduate student body with an appreciable share who attend part-time? Which exclusively undergraduate institutions serve a highly residential student body but admit an appreciable number of students as transfers?
Starting with the new classifications, users will be able to build customized classifications aligned with specific analytic purposes. This new flexibility also has potential to assist college and university personnel in identifying peer institutions (Which institutions share our classification with respect to undergraduate program, undergraduate profile and size and setting? Which ones have similar undergraduate and graduate program classifications?). A Web-based listing facility will allow users to aggregate within and intersect across classifications to suit their needs.
The new classifications will not solve all the problems of the old classification, and they will make some things harder. The Carnegie Classification has always been based on secondary analysis of empirical data on what institutions do, and this is at once a strength and a limitation. It is a strength because institutions are classified not on the basis of rhetorical assertions about mission, but on actual behavior as revealed in the data. It is a limitation in other respects: not everything that counts can be counted, and empirically based classifications offer a retrospective account of what institutions do. Colleges and universities are always changing, but the classifications provide a time-specific snapshot. Classification using national data is undertaken at a distance, without interviews, site visits, or close reading of institutional documents. The national data can never fully represent an institution's character or identity, and it's simply unrealistic to expect it to.
There are also important things that colleges and universities do that are not reflected in national data collections. Two new "elective" classifications -- based on voluntary participation -- will attempt to incorporate new information on two areas where institutions have special commitments. One of these is outreach and community engagement. This year a diverse group of 14 institutions participated in a pilot project to develop a framework for documenting the range of ways that they engage with community. The results of that work will inform a wider effort for institutions to participate in a new classification for community engagement. The second elective classification, to be developed in 2006, will focus on how institutions seek to analyze, understand, and improve undergraduate education.
The new classifications will complicate how we talk about similarities and differences among institutions as the simple, mutually exclusive terminology of the traditional classification gives way to a more complicated multidimensional framework. But this framework will do more justice to the multifaceted nature of the institutions it seeks to describe. Still more classifications may emerge as well, because the Carnegie Foundation does not hold an exclusive franchise on classifying colleges and universities.
Alexander C. McCormick is a senior scholar at the Carnegie Foundation for the Advancement of Teaching, where he directs the Carnegie Classification project.