Research Universities: American Exceptionalism?

By Henry Rosovsky, Professor of Economics, Emeritus, Harvard University

Editor’s Note: These remarks were prepared for the Carnegie Corporation/TIME Summit on Higher Education, September 20, 2013.


I. A Paradox

Domestically, American higher education is the subject of almost unprecedented criticism. “Too expensive and inefficient and not a good investment” is a common conclusion. Students are said to be unprepared for the job market. Higher education is accused of being too permissive in tolerating low faculty productivity and in resisting the technological revolution. In general, the current “business model” is judged unsustainable: some think that we are riding on the road to self-destruction. The United States confronts great social and economic problems, yet—in Arthur Levine’s gloomy words—“public and opinion leaders alike view [universities] as more of a problem than a solution.”1

But in international discussions and evaluations of higher education, American universities are frequently called “the envy of the world.” Not by any means all our universities. Indeed not very many, but some—and that is my point.


In the United States, it makes no sense to speak about “higher education” or “universities” in general—yet it happens all the time. (The December 1, 2012, issue of The Economist provides a recent example. The headline announced: “Not what it used to be: American universities represent declining value for money to their students.” In the text there is little recognition of the tremendous diversity of higher education in the United States.) The label “American universities” has little meaning when our country is home to more than 4,000 tertiary institutions, ranging from those that might actually be the envy of the world to those barely distinguishable from high schools— with a tremendous variety in between.

At the top of our higher education pyramid we find the public and private research universities with their special role of creating and maintaining knowledge, training graduate students in arts and sciences and professional schools, and offering a liberal education to undergraduates. According to Jonathan Cole, there are about 125 diverse universities that fit this description and they “…are able to produce a very high proportion of the most important fundamental knowledge and practical research discoveries in the world. It is the quality of the research produced, and the system that invests in and trains young people to be leading scientists and scholars, that distinguishes them and makes them the envy of the world.”2 These 125 universities play a less singular role in undergraduate education. As Cole again points out, some American liberal arts colleges are able to offer undergraduate education of equal quality. I agree, but the nature of the educational experience is different: for undergraduates, the research university might be compared to life in a big city with a great diversity of inhabitants—undergraduates, graduate students, professional school students, and faculty reflective of that diversity—and the liberal arts college comparable to a more homogeneous and community-oriented small town populace. Each has its own advantages for undergraduates.

“Top of the pyramid”—my sole focus here—does not mean that institutions below the top are less worthy, less deserving of private or public support, or less essential in the national scheme of higher education. Nor does it imply that the current storm of criticism is irrelevant for research universities. I completely understand the need for controlling costs and expanding capacity. But it does mean that criticisms have to be as differentiated as the range of institutions: unless that happens, inappropriate remedies may damage a sector of American higher education where we are using accepted but necessarily questionable measures with the potential to lessen our status as world leaders. All systems of international university rankings agree that U.S. universities dominate the top twenty or thirty places. (Twenty-two out of thirty in the Times Higher Education survey and twenty-three out of thirty in the Shanghai Jiao Tong ranking; both in 2013.)

In the United States, it makes no sense to speak about “higher education” or “universities” in general—and yet it happens all the time.

It is unlikely that American dominance is accidental, but a convincing explanation would have to be extremely complicated. History, wars, culture and customs, and resources are all involved. But all the institutions at the top of the American educational pyramid—and some others as well—share six characteristics closely associated with high quality. (My initial preference was to call these “necessary conditions,” but that seemed a bit too rigorous.) Their absence would preclude—or make it much more difficult—for research universities to achieve the highest quality, not just in this country but anywhere else. Indeed, their partial or total absence abroad helps to explain why there are relatively few foreign—especially non-Western—institutions represented at the top of the accepted surveys.3 None of the six characteristics is wholly unambiguous; all are blurry. But is not difficult to detect their presence or absence.

II. Six Characteristics of Quality

Shared governance. First, these institutions all practice shared governance: the trustees and president conditionally delegate educational policy to the faculty. That would primarily include curriculum and the initial selection of those who teach, are admitted to study, and do research. The administrative style is collegial rather than top-down, faculty sharing authority in specified areas with appointed administrators and trustees, the latter holding final authority. This is a distinctly American form of shared governance, which relies on a strong executive. Presidents, provosts, and deans possess and exercise considerable authority over budgets, institutional priorities, and many other matters of consequence. This may be contrasted with the so-called “continental model” that features what, in its purist form, can only be described as “participatory democracy”—faculty elections of rectors and deans, and policy decisions sometimes placed in the hands of assemblies based on the principle of parity: faculty, students, and employees sharing authority. In my opinion, this form of governance has been a great obstacle to progress, and while it is very difficult to generalize, it seems that even continental practice is moving toward greater executive authority.

A major component of education is real as opposed to virtual encounters between students and teachers to encourage participation and critical thinking.

More than a decade ago, I had the opportunity to study universities in developing countries all over the world while preparing a report for the World Bank and UNESCO.4 Problems and issues varied enormously depending on economic conditions, political system, history, etc. But those who were in charge of universities almost always agreed on one point: poor systems of university governance were the greatest obstacles to institutional improvement—more so than inadequate financing or anything else. Of course, poor governance meant many different things but certainly included interference by ministries of education, unclear lines of authority and perhaps, most important, barriers to faculty input or initiative. It would be a mistake to believe that poor governance applies only to the developing world. Similar obstacles have slowed quality growth in European and American higher education.

What makes shared governance so important? There are many possible answers, but these are among the most frequently mentioned: universities are extremely complex organizations in which centralized decision-making does not achieve the best results; in universities the proportion of self-motivated people is large and to capture the full measure of their “creative juices” requires a sense of ownership. Susan Hockfield, former president of MIT, puts it very well: “Faculty travel the frontiers of their disciplines and, from that vantage point, can best determine future directions of their fields and design curricula that bring students to the frontier. No academic leader can chart the course of the university’s discipline independent of the faculty.”

These reasons apply in general to organizations in which profession-based authority is important, a good example being law firms and large consulting firms. Shared governance may frustrate administrators intent on implementing rapid change, but a slower pace may also lead to wiser choices and certainly has not—in light of university histories—prevented fundamental changes. (It should be added that the current use of adjuncts, offering over 50 percent of instruction in many universities, has surely undermined the integrity of shared governance. A corps of instructors in which half are employed on a yearly basis and without rights or sense of ownership will not be doing much creative thinking about the future.)

Academic freedom. Second, despite periodic challenges, American research universities enjoy academic freedom—“the right of scholars to pursue their research, to teach, and to publish without control or restraint from the institutions that employ them”—and, in addition, all rights granted to inhabitants of this country, especially those associated with the First Amendment.5

Merit selection. Third, admission of students and selection and advancement of faculty is based on merit measured by recognized and accepted institutional standards. Some form of prior achievement would define merit: assuredly not an issue devoid of numerous ambiguities. One cannot ignore legacies, affirmative action, athletic scholarships, and similar deviations from the simplest notions of merit for students, such as scores on a standardized national test. Similarly, gender, race, and old-boy networks can create other deviations from a straightforward standard for selecting and promoting faculty. Nevertheless, objective measures of merit remain at the very least the first approximation.

©Andreas Rodriguez/iStock/Thinkstock

Significant human contact. Fourth, a major component of education is now and is intended to remain significant human contact: real as opposed to virtual encounters between students and teachers to encourage participation and critical thinking. In his 2012 Tanner Lectures, William Bowen calls this “minds rubbing against minds.” The proportions may change over time but the basic principle has to be retained: it has to be part of liberal education for undergraduates who need guidance and contact in making choices, and it is a self-evident part of the mentor-mentee relation for those aspiring to reach a Ph.D. Leon Wieseltier, in language that is both valid and vivid, captured the spirit of this characteristic extremely well in a recent New Republic essay:6

When I look back at my education, I am struck not by how much I learned but by how much I was taught. I am the progeny of teachers; I swoon over teachers. Even what I learned on my own I owed to them, because they guided me in my sense of what is significant. The only form of knowledge that can be adequately acquired without the help of a teacher and without the humility of a student is information, which is the lowest form of knowledge. (And in these nightmarishly data-glutted days, the winnowing of information may also require the masterly hand of someone who knows more and better.)

One might quarrel with some specific phrases, but it is not easy to imagine these sentiments being addressed to a screen. Few would deny the great value of digitization, virtual course materials, or occasionally flipped classrooms but they remain complementary rather than primary.

Preservation of culture. Fifth, all these universities consider preservation and transmission of culture to be one of their missions. This would include representation of the humanities in curriculum (mandatory for undergraduate liberal arts), as well as, for some, more specialized activities including research and language studies, and the maintenance of libraries and museums. Preservation of culture applies as much to MIT, Caltech, and Purdue as it does to the more traditional Yale and the University of Wisconsin. Indeed, many “polytechnics”—certainly including the ones mentioned here—have been the source of major innovative scholarship in the humanities and social sciences. The history of science and economics are excellent examples. It is a simple fact that our most prominent universities specializing in science have programs and/or departments that transcend traditional definitions of science. But why? Because they believe that this both improves the education of their students and the research of the faculty. Interdisciplinary approaches in all fields have been gaining favor for many years and that may be the most powerful driver of all.

Nonprofit status. Sixth—and finally—all research universities operate on a not-for-profit basis. If maximizing profit or increasing shareholder value were the goal, all the previous conditions become unwelcome obstacles and inefficiencies that could not be tolerated by a competent management. But this condition is not as cut and dried as it may seem. Decisions in not-for-profit universities can be influenced and possibly distorted by considerations of revenue. For example, activities that generate research or operating funds in return for certain privileges obtained by a funder may require exclusive access to specific scientific results for a limited period of time. In this sense, no research university today is purely not-for-profit. None, however, is mainly directed by the business aims of outside supporters.

The six characteristics are neither canonical nor subject to rigorous mathematical proof. They are based on my (I believe uncontroversial) reading of our historical experience.

III. Understanding and Misunderstanding the Quality Requirements

Many academics will consider a listing of these characteristics individually familiar, obvious, and of little interest. Non-academics, on the other hand, may have a quite different reaction. The list could easily be interpreted as a plea for the status quo, typical of the academic establishment that stubbornly resists all change.

Both perspectives are wrong. The characteristics of quality are almost never considered as a system even though the absence of any one of them will affect the integrity and quality of a research university. Faculty wishing responsibly to exercise rights of shared governance should have the whole group clearly in mind.

Turning to the non-academic perspective, none of these characteristics, singly or as a group, make—to use the term beloved by our critics—disruptive change impossible. This is an important point because, I think, it runs counter to widely held beliefs.

For example, tenure is perceived to be an obstacle to change. It may indeed be desirable instead to adopt a system of long-term contracts—particularly because federal law prohibits mandatory retirements. Faculties are aging and so are their ideas, in turn raising costs and keeping out the young. But it is not the enumerated characteristics that stand in the way of change. To take the most relevant, in the American tradition, employment contracts have never been within the purview of shared governance. Faculties don’t determine their own pay or conditions of employment; these are in the hands of the administration—even when union negotiations are involved. A main barrier to change has to be the fact that—noted by Bowen in his second Tanner Lecture—that competition between non-profit peer institutions currently drives up cost. No ambitious and quality-centered research university can afford, on its own, to abandon tenure and move (say) to long-term contracts. Only an understanding with peers would make it possible and that is illegal. Bowen wonders if some collusion would now be in the public interest.

Internal and external critics have suggested various other cost-cutting measures. For example, raising teaching loads or a rising student-faculty ratio—I do not necessarily suggest either one—would lower cost. A three-year bachelor’s degree would have the same result. More machines, fewer humans, and an increase in online learning (MOOCS) may also decrease expenses. Again, these may be good ideas or not— but respecting the six characteristics does not prevent their implementation (so long as shared governance is clearly understood not to be participatory democracy).

Shared governance does, from time to time, increase the burden of administrators. Bowen, in his Tanner Lecture, asks if shared governance is suitable for a digital world in which decisions about educational policy can frequently go beyond individual professors or departments and need to include a great mix of constituencies. As he suggests, individual or groups of faculty should not have veto power over change. Have they ever in a well-administered institution? Bowen is right: the definition or concept of shared governance may have to change with the times, while the principle of faculty voice and participation is vigilantly maintained. The important words are sharing combined with good leadership. The notion that research universities are “unchanging” has always struck me as bizarre. Our products are education and research, and the vital element is not the format or setting (the bottle) but the content (the wine.) And that is forever changing.7

Conditions that are at the core of what it means to be a university are, for many people, counterintuitive, especially for those with a background primarily in business. This was vividly illustrated by the recent events at the University of Virginia where a few board members, mainly from the private sector, believing that the new president was making changes too slowly, engineered her abrupt dismissal after three years on the job. It seems to me that this kind of coup would not be considered good practice even for a private corporation, but for an institution in which shared governance was the assumed norm it proved to be disastrous. The UVA board may have acted within its legal authority, but the total absence of consultation created a faculty-student revolt that forced a reversal of the original action. All emerged worse off.

Shared governance is perhaps the classic source of “misunderstandings,” but it is not by any means the only one. Academic freedom is a perpetual sore point, especially when it comes to the expression of political opinion by faculty. To take one more example, preservation of culture may be seen by those exercising sound business judgment as an entirely discretionary luxury when it is, in reality, an integral part of research universities.

Harvard President Drew Faust framed the issues eloquently in a recent address at Boston College:

Universities are a set of institutions unlike any others in our society. Certainly our budgets must balance, our operations must be efficient, but we are not about the bottom line, not just about the next quarter, not even about who our graduates are the day they leave our walls. Our task is to illuminate the past and shape the future, to define human aspirations for the long term. How can we look past the immediate and the useful… to address the larger conundrum of: How shall we best live? What do I want to be today—and tomorrow? To discover not only the ways in which human civilization plans to get somewhere, but to ask the question, Where does it—and where should it—hope to go?8

Those are not questions likely to arise in many corporate boardrooms but they should be raised regularly among university trustees.

IV. Addressing the Present Moment

We come now to some of the real difficulties of the moment. To fulfill their role in society—creating knowledge and educating graduate and undergraduate students— the university community makes assumptions that may not always be, and almost certainly are not now, obvious either to the trustees who are their governors or to the wider public. For example, the characteristics associated with quality can be seen as pleas for special privileges. In business or in government neither the freedom of expression nor a voice in governance is the practice. Decisions are largely profit-based or necessarily political.

Universities are a set of institutions unlike any others in our society. Our task is to illuminate the past and shape the future.
—Drew Faust
President, Harvard University

Another reality to consider is that American universities only rarely have written constitutions or long-lasting traditions of common law. The guarantors of their privileges and practices are trustees, most of whose life experiences have been in private business, admittedly a category possibly so broad as to be largely meaningless. (Currently, around 50% of trustees come from “business,” 22% from professional service, and 13% from education.) Furthermore, in the case of state universities appointment to positions of governance can be political, frequently in the hands of governors, and sometimes subject to state elections.

At a time of contentiousness and criticism current practices raise questions: do those who constitute the court of last resort understand the unusual entity with which they have been entrusted? When trustee initiative is necessary and appropriate and when it is not? Do we do enough to prepare trustees for their responsibilities? Are those who make the appointments more concerned about the candidate’s ability to read balance sheets than their appreciation of university values? Or do we look primarily at the capacity of potential trustees to make large donations? Or are those who have the power of appointment primarily interested in a candidate’s political affiliation? The same point can be made about faculty. We take great care to examine research credentials and—these days, and that is a major and welcome change—we look more closely at teaching capacities. But do we do anything to prepare faculty to participate productively in shared governance? Both of these tasks will grow in urgency as the American research university—“the envy of the world”?—navigates very stormy seas predicted by nearly all observers.

1 “Today’s Unpresidential Presidents,” The Chronicle of Higher Education, October 26, 2012.

2 The Great American University, 2009, p. 5

3 For a very recent confirmation of this point, one need only look at Michele Lamont and Anna Sun’s op-ed, “How China’s Elite Universities Will Have To Change,” in The Chronicle of Higher Education, December 14, 2012.

4 Higher Education in Developing Countries: Peril and Promise, 2000.

5 The Columbia Encyclopedia, 6th edition, 2001.

6 December 31, 2012

7 A brief digression. In The Great American University, previously mentioned, Jonathan Cole suggests a list of thirteen items under the title “What Makes Great Research Universities,” p. 109. There is very little overlap with our list—the main common point being academic freedom—because what I call “characteristics of quality” all pertain to internal governance and, subject to constraints, are controlled by the university and ultimately by the trustees. And that becomes very consequential when trustees and their responsibilities are considered.

8 “Scholarship and the Role of the University: Remarks at the Boston College Sesquicentennial,” October 12, 2012.

Posted in Article, The Future Tagged with: , , , , , , ,