Carnegie Classifications FAQs
[Jump to Institution-specific Questions]
The Carnegie Foundation does not rank colleges and universities. Our classifications identify meaningful similarities and differences among institutions, but they do not imply quality differences.
The 2010 edition continues the multiple-classification approach used in 2005, but updated with the most recent national data available, as of 2010. To ensure continuity of the classifications and allow comparison across years, the 2010 edition represents minimal changes from the 2005 edition; that said, there are a couple of exceptions due to changes in reporting practices of critical national datasets:
- In the Basic Classification, we used the most recent data from the National Science Foundation (NSF) Survey of Research and Development Expenditures at Universities and Colleges (fiscal year 2008). Due to recent changes in NSF reporting, the data contained disaggregated R&D expenditures in psychology and social sciences, which had been previously grouped within the Science and Engineering (S&E) expenditures (therefore inseparable from the S&E expenditures). Due to this new data regrouping possibility, in the 2010 update of the Basic Classification, R&D expenditures in the fields of psychology and social sciences are subtracted from the Science and Engineering (S&E) expenditures and added to the R&D expenditures in non-S&E fields.
- In 2008, the Integrated Postsecondary Education Data System (IPEDS) Completions survey revised the way it classifies doctoral degrees. As a result, the 2008-2009 completions data (IPEDS data name “c2009,” which corresponds to degree conferrals from 2008 to 2009) contains both the “old” doctoral degree categories (doctoral degree and first-professional degree) and the new doctoral degree categories (doctoral degree–research/scholarship, doctoral degree–professional practice, and doctoral degree–other). The first-professional degree category was eliminated. Since over 50 percent of institutions still reported the old doctoral categories, we converted the new doctoral degree categories into the old categories in the 2010 classification update, that is treating doctoral degree–research/scholarship as doctoral degrees, and doctoral degree–professional practice and doctoral degree–other as first-professional degrees.
The Research I & II and Doctoral I & II categories of doctorate-granting institutions last appeared in the 1994 edition. The use of Roman numerals was discontinued to avoid the inference that the categories signify quality differences. The traditional classification framework was updated in 2005 and since identified as the Basic Classification
. Many of the category definitions and labels changed with this revision.
Starting with the 2000 edition of the Carnegie Classification, we eliminated the use of federal funding to differentiate doctorate-granting institutions for two reasons. First, federal support is at best a rough proxy for an institution’s research activity that suffers from several weaknesses. Not all research is federally funded, and institutions differ in the proportion of all research that is funded from federal sources. Similarly, academic fields differ in their reliance on federal research funding and also in the costs associated with research. Thus a focus on federal dollars pays greater attention to fields that are heavily dependent on federal funding and also to fields where research requires substantial investments. The federal obligations data are also blind to the pass-through of funds from one institution to another, as happens in the case of large projects involving research teams at different institutions. All of these factors compromise the accuracy of federal obligations as a gauge of overall research activity.
From its inception, the Carnegie Classification’s purpose has been to assist those conducting research on higher education. Researchers need a way to reference the great diversity of colleges and universities in the United States, and classifications enable them to identify groups of roughly comparable institutions. The primary audience is the research community, including academic researchers and institutional research staff as well as other education analysts. By providing a set of distinct classifications as well as a set of online tools for creating custom listings (combining categories within classifications, identifying institutions in similar categories across classifications, or filtering listings by selected criteria), researchers now have much greater analytic flexibility, allowing them to match classification tools to their analytic needs.
All accredited, degree-granting colleges and universities in the United States represented in the National Center for Education Statistics IPEDS system (as of the year the classification was issued, and subject to the availability of required data)are eligible for inclusion in the Carnegie Classifications. Branch campuses that are separately identified in IPEDS are classified separately.
A single classification cannot do justice to the complex nature of higher education today. When the Carnegie Classification was created in 1970, there were about 2,800 U.S. colleges and universities. Today there are more than 4,500.
Colleges and universities are complex organizations, and a single classification masks the range of ways they can resemble or differ from one another. As valuable as it has been, the basic framework has blind spots. For example, it says nothing about undergraduate education for institutions that award more than a minimum number of graduate degrees. Yet most of these institutions enroll more undergraduates than graduate or professional students.
Another motivation for these changes has to do with the persistent confusion of classification and ranking. For years, both the Carnegie Foundation and others in the higher education community have been concerned about the extent to which the Carnegie Classification dominates considerations of institutional differences, and especially the extent to which it is misinterpreted as an assessment of quality, thereby establishing aspirational targets. This phenomenon has been most pronounced among doctorate-granting institutions, where it is not uncommon to find explicit strategic ambitions to “move up” the perceived hierarchy. By introducing a new set of classifications we hope to call attention to the range of ways that institutions resemble and differ from one another and also to de-emphasize the improper use of the classification as informal quality touchstone.
In 2005, for the first time, we adopted a multiple-classification approach; that is, a set of new classifications that illustrate a range of ways to think about how colleges and universities resemble or differ from one another. These classifications offer a set of different lenses through which to view higher education. Each institution appears in each of the new classifications.
The classifications are organized around three key questions: What is taught? To whom? In what setting? Two of the classifications focus on the instructional program (one on the undergraduate program, and one on the graduate program). Two describe the profile of enrolled students (one describes the mix of undergraduate and graduate/professional students, while the other focuses on the undergraduate population). Finally, a fifth differentiates institutions with respect to size and residential character.
In addition to these all-inclusive classifications, work continues on a set of “elective” classifications in which institutions voluntarily participate and document aspects of institutional activity that are not adequately reflected in the national data. The first such classification, released in 2006 and updated in 2008 and 2010, focuses on Community Engagement.
The all-inclusive classifications are conducted based on nationally available datasets. Institutions do NOT need to apply for all-inclusive classifications. All accredited, degree-granting colleges and universities in the United States represented in the National Center for Education Statistics IPEDS system are classified in the all-inclusive classifications, which include:
An elective classification is based on voluntary participation of individual institutions. Institutions need to apply and supply documentations for review to be included in an elective classification. Currently, the only elective classification is the Carnegie Community Engagement Classification.
The new classifications were intentionally designed so users can combine categories to suit their analytic needs. The Custom Listings
tool provides an easy way to do this.
We used the most current data available to us for all institutions. It is very important to recognize that empirically-based classifications are time-specific and inherently retrospective: they portray a snapshot of individual institutions and of the U.S. higher education system at a particular point in time. Because colleges and universities are constantly changing, many might be classified differently if we repeated our analyses with more recent data—but those classifications would still be retrospective.
The update cycle has not yet been decided.
The Research University categories are defined according to a statistical combination of several correlates of research activity. There are no cutoff values for the individual measures. (See Basic Classification Methodology.)
In 2008, the reporting of professional and doctoral degrees to IPEDS changed. Before that year, graduate degrees above the master’s level were divided into “first professional” degrees (J.D., M.D., etc.) and “doctorates” (Ph.D., Ed.D., etc.). Under the new system, the "first professional" category has been eliminated and doctoral degrees have been divided into three categories: "professional doctorates," "research doctorates," and "other doctorates."
According to the IPEDS reporting guidelines, a research doctorate is defined as a "Ph.D. or other doctor's degree that requires advanced work beyond the master's level, including the preparation and defense of a dissertation based on original research, on the planning and execution of an original project demonstrating substantial artistic or scholarly achievement." Professional doctorates, by contrast, are "… conferred upon completion of a program providing the knowledge and skills for the recognition, credential, or license required for professional practice."
Since over 50 percent of institutions still reported the old doctoral categories, we converted the new doctoral degree categories into the old categories in the 2010 classification update, that is treating doctoral degree–research/scholarship as research doctoral degrees, and doctoral degree–professional practice and doctoral degree–other as professional degrees.
Previous editions of the Carnegie Classification used three-year averages for some data elements. This can smooth out year-to-year fluctuations, especially for smaller institutions, but it also increases the influence of outdated data that may no longer be accurate due to changes in institutional policy, programs, or infrastructure. The classifications are inherently retrospective, and increasing the leverage of older data reduces precision for institutions that are changing.
We used the most recent single year data to maximize the timeliness of the classifications. In some cases we have used up to three years of data at an institution's request, when the single-year data represents an aberration.
In the institution display, we include footnotes for cases where the classification is based on a small number of observations (degrees or students) and cases that are close to category borders. We also note cases where classifications are based on up to three years of data.
Because this is a more complex system, the Foundation no longer publishes classification listings. The online system provides for institution lookup, list generation, and custom listings, and any listing can be rendered in printer-friendly format or downloaded in CSV format (which can be opened by spreadsheet programs such as Excel). Downloaded listings include IPEDS UNITIDs and all classifications for listed institutions. In addition, the Downloads page includes a comprehensive download file.
The Downloads page has a complete spreadsheet file with all institutions and classifications, as well as most of the underlying data and IPEDS UNITIDs. The Downloads page also includes information from previous editions of the Carnegie Classification.
The Carnegie Classification has always analyzed existing national data to assess what institutions do. Consequently, we have been limited to information that is available for all institutions. Retaining the classification’s traditional focus on empirical indicators of institutional activity drawn from the available national data, we wanted classifications that would answer three key questions: what is taught, to whom, and in what setting? We wanted to call special attention to the instructional program, and the shift to multiple classifications enabled us to do so in far greater detail than would have been possible in a single classification. In focusing on the student population and institutional setting, we also bring in characteristics that have been used by other analysts to represent important institutional differences.
The Carnegie Classification has always identified institutions whose programs typically center on a single field or set of related fields (for example, seminaries, free-standing law or medical schools, schools of art, etc.). Among the classifications introduced in 2005, only the enrollment profile includes special-focus institutions because these institutions may be exclusively undergraduate, exclusively graduate, or they may enroll both undergraduates and graduate students.
Special-focus institutions are inherently distinctive, and multiple classifications have less relevance for this group than they do for institutions offering a range of programs serving a variety of purposes. In effect, their specialized nature trumps the need for more elaborate classification.
The instructional program classifications are derived from national data on degree conferrals by field of study. Program concentrations are inferred from the degree field information. This inference is reasonable at the 4-year level, but much less so at the 2-year level because the majority of community college students—including those who may have achieved their educational objectives—do not receive any formal award. As a result, classifying community colleges on the basis of the minority of students who receive a credential would not be representative of the colleges’ offerings. One of the biggest gaps in the national data on colleges and universities is the absence of comprehensive, institution-level data on curricular emphases at community colleges.
Published versions of the Carnegie Classification have always listed institutions by state and control (public or private), but control has not been an explicit classification criterion except for the Associate's Colleges categories in the Basic Classification. Indeed, one reason the Carnegie Classification was developed in 1970 was because routinely separating public and private institutions often ignored important similarities in mission and function.
Classification listings can be filtered by control (or other attributes), and any listing can be sorted by control (click the column heading).
Classification is different from ranking, and the Carnegie Foundation does not rank institutions. Teaching quality is very important, and it is at the heart of many Carnegie Foundation programs, but it is not something that can be reliably assessed at a distance on the basis of available quantitative measures such as faculty salaries or instructional expenditures. Apart from the question of appropriate measures, it is not clear that teaching quality is best assessed at the institutional level, rather than at the department or classroom level.
U.S. News uses the Basic classification to define its comparison groups (see Ranking Category Definitions on the U.S. News Web site). For a critique of this approach, see Hidden in Plain View published in Inside Higher Ed.
Because the organizing framework of the U.S. News comparison groups is based on a conception of regional and national markets, and because this information is not a factor in our classifications, the Carnegie Foundation believes that U.S. News should develop an approach to defining comparison groups that is better suited to its analytic framework.
Questions about the U.S. News rankings should be directed to Robert Morse, Director of Data Research at U.S. News, at firstname.lastname@example.org, or by phone at 202-955-2389.
The timeline for the next round of applications for the elective Community Engagement Classification has not yet been set. We encourage you to continue to check our website for updates or sign up for the classification mailing list.
[Jump to General Questions]
This is quite common if you are looking for similarity across most or all of the classifications. Because the classifications are highly disaggregated, the number of possible combinations across all of the classifications exceeds the total number of institutions. Try selecting fewer classifications, or use the Custom Listings tool to identify and combine categories of interest within each classification.
The Size and Setting classification provides a good proxy for identifying institutions with similar institution enrollment - this can be done by clicking the check box at the row of "Size and Setting."
The instructional program classifications are based on the record of degree conferrals. For most institutions, we used degree data from 2008-09, but in some cases we averaged data over two or three years to enhance stability. Because the classifications are based on the degree record, they do not include programs that had not yet produced graduates and extant programs that did not produce graduates.
The Carnegie Classifications are retrospective, time-specific snapshots of colleges and universities. Inevitably, some changes such as program additions will be too recent to be captured in the data from which the classifications are derived. Any number of institutions might be classified differently using newer data, and to maintain the comparability of classifications we would have to reclassify all institutions. For this reason, we do not make case-by-case reclassifications based on newer data.
Name changes: Please send an email to email@example.com containing a link to the official announcement of the name change on the institution’s Web site (please specify “name change” in the subject line). Alternatively, fax a copy of the announcement to Classification at 650-326-0278. Please include a contact name and telephone number for confirmation.
Name corrections: Name corrections will only be accepted if (1) they come from the office of communications, public information, institutional research, or the chief executive officer; and (2) the requesting official’s full name, title, and contact information are clearly identified. Correction requests can be sent by email to firstname.lastname@example.org (please specify "name change" in the subject line) or by fax on official letterhead to Classification at 650-326-0278. Email requests must be sent from an institutional email account.
A small number of schools offer graduate programs that enroll students in the summer only or at remote sites, and as a result these schools show no graduate enrollments in the fall term (the basis of the enrollment profile). While we could adjust the enrollment profile to take this apparent contradiction into account, by leaving the apparent contradiction in place it is possible to identify these institutions easily.