I read with great interest Kevin Kiley’s October 10 Inside Higher Ed piece -- "Starting to Worry"  -- which uses Smith College’s very interesting and valuable "Futures Initiative" planning process as a "take-off" to rehearse again a narrative about what's wrong with higher education in America. While that article focused on elite residential liberal arts colleges, there is a push across all sectors of higher education to consider a radical shift away from proven modes of teaching and learning, with arguably the greatest pressure coming on institutions without the resources of Smith and others discussed in that piece.
I agree that these are very challenging times for American higher education. I agree that getting our strategy right for the future is urgent. I agree that higher education is expensive and becoming more so. I disagree vehemently, however, with proposals for reform of American higher education that jettison what we know works in teaching and learning for untested -- or, where tested, proven less effective -- restructurings that are tantamount to betting the farm. Here are my arguments and evidence.
1. You must begin with aims and objectives.
Missing from Smith’s Futures Initiative (though perhaps unstated because implicit in its mission) and Kiley’s article is a statement of what the aims and objectives of higher education should be. It is impossible to assess the efficiency of a system, or whether it costs too much, when its goals are unspecified. We are hearing today a literal cacophony of commentary from business leaders, economists, journalists, and even a growing number of higher education leaders that the education Americans need for the 21st century must stress inquiry and analysis, critical and creative thinking, integrative and reflective thinking, written and oral communication, quantitative literacy, information literacy, intercultural understanding, and teamwork and real-world problem-solving, as well as knowledge and competence in specific fields of learning.
These are the skills and learning necessary for success in today’s occupational system; they are even more important for the global occupational system to which we are moving. They are as essential for community college students as they are for students seeking four-year degrees. We must ensure that community college students are not left behind in the competition for jobs in the new economy because they do not have the higher-order skills it requires. Proposals that focus on giving community college students only short-term job training have the potential to do just that.
2. These learning goals are hard to accomplish. They are even more difficult to accomplish through systems that seek efficiency through substitution of capital for labor.
Research shows that the learning outcomes essential for the 21st century are most fully achieved by students in colleges characterized by high levels of student engagement — when, in other words, students are challenged academically; experience high levels of purposeful, active, and collaborative learning; enjoy quality interactions with faculty around their academic work; experience the enrichment of such opportunities as well-designed internships, collaborative research with faculty members and study of other cultures (abroad or in the U.S.); and benefit from supportive campus environments. The extent to which campuses provide these kinds of engagement for their students is measured by the National Survey of Student Engagement (NSSE).
But very importantly, a 2010 Wabash College-University of Iowa collaboration shows that NSSE benchmark scores, with controls for school-level pre-college scores, are strongly correlated with direct measures of learning similar to several of the Association of American Colleges and Universities' "Essential Learning Outcomes."  In my view the most important findings Arum and Roksa report in Academically Adrift  are not that so much of higher education produces limited student learning, but that where expectations of students are high and students major in liberal arts and sciences disciplines, students work hard and make significant progress toward the outcomes and levels of learning they need for 21st century challenges. In a book that rightly chronicles the failures of higher education, we also find further evidence of what works.
For those who have given their lives to teaching undergraduates these careful, systematic assessment results are consistent with experience — that is, teachers know academically challenging, high student-engagement learning environments work because they have seen the results with their own eyes. We have no comparable evidence that these learning outcomes can be achieved effectively strictly through on-line/distance learning systems or through "reforms" that result in decreasing student engagement.
3. High student-engagement learning environments are also efficient and cost-effective.
At a time when colleges are criticized for low graduation rates and are urged to move toward distance learning, why is no one paying attention to the reality that colleges of all kinds — private and public, small and large, liberal arts colleges and multi-purpose universities — that have created learning environments characterized by high levels of student engagement have the highest four-year graduation rates? Campuses characterized by high levels of student engagement have high four-year graduation rates because students are more successful academically and are more highly motivated to learn and to persist in college — they leave college less often for academic reasons, and they graduate at higher rates because they love what they are doing and value the personal educational transformation they experience. In these environments students are inspired to become lifelong learners (see results from Wabash-Iowa study referred to above), motivated to teach themselves new things — especially the new things required by an evolving and changing workplace.
Alongside the coalition arguing that the skills and learning I've just described are essential for the 21st century there is, of course, another narrative arguing that a more narrow, initially less-expensive to provide, first-job-focused kind of higher education is what we need. What is left out of this argument, however, is the recognition that a narrow, vocational education geared to particular jobs requires its recipients to be retrained for the next job at significant cost to the job-holder, employers, and taxpayers who fund federal and state job programs. This vocational training is not intended to and does not succeed at helping students become lifelong learners, motivated to teach themselves new things. The graduate’s education/training, though less expensive to provide, depreciates. In contrast, students educated in high-engagement learning environments for the high-level thinking and skills needed in the 21st century are constantly educating themselves in their current jobs and for their next jobs because they have the curiosity of lifelong learners and skills that transcend the particular. This kind of education actually appreciates in value as its beneficiaries become more valuable in the marketplace — good for them and good for the rest of us who benefit from their improved productivity.
No one has estimated the enormous size of the depreciation cost borne by individuals educated narrowly and vocationally and by those who must then subsidize their retraining. These costs do not figure in to the debate we are having in America about college costs, and political leaders of both parties are shaping a public discourse that argues for narrow, short-term educational solutions. It is an enormous challenge to break through an anxiety-fueled policy and public dialogue that too often treats college as a 21st-century version of trade school.
But that is not all. I have written elsewhere about how the increased efficiency of high four-year graduation rates lowers the cost of higher education for all stakeholders — students, families, colleges, states, the federal government, and the American public, which subsidizes higher education through gifts and taxes. If a student finishes in five, six or more years, instead of four, the costs of delay will be immediate — in additional tuition, and the self-multiplying costs of not earning a salary in a full-time job (economists call these costs "opportunity costs") — as he or she loses ground professionally and personally.
Indeed, my analyses comparing the costs to families when a child attends a public college or university (because public institutions have significantly lower four-year graduation rates and lower levels of student engagement as measured by NSSE) show that if one takes this additional cost of delay in entering the job market into account, there are no scenarios where attending the average public out-of-state institution is cheaper than attending the average private institution if one doesn’t graduate in four years but does graduate in five. And only if one assumes the student works half-time in the fifth year is it cheaper for a non-aided student to have attended an in-state public institution! Only students from high-income families benefit for certain from low in-state public sector tuition if they do not graduate in four years, and then only if the student works at least half-time while finishing college.
Public and private institutions that charge lower tuitions are able to do so by not funding learning environments of high engagement. Ultimately, however, I believe these choices cost students, families, and American employers much more than is saved.
4. We must truly come to understand the cost-drivers in American higher education if we are to get our strategy for the future right.
In my view, Archibald and Feldman’s Why Does College Cost So Much?  is the very best book on the rising cost of American higher education ever. They take what they call an "aerial view" rather than looking at the cost issue from the institution out to explain the rise in higher education costs over the last 50 years or so. Higher education is expensive because a very high (roughly two-thirds) proportion of employees must be and are highly educated. Especially since the 1980s, demand for highly educated workers has grown faster than supply, driving the cost of employing them up.
At the same time, rapid advances in technology in support of teaching and research deepen, enrich, and even change the nature of what can be learned and must be learned rather than allow the substitution of capital for labor. Experience with these technologies is necessary for graduates to be both competitive and effective in the jobs our economy has and needs now. Having these technologies available for teaching and research has added new layers of cost. This explains why in higher education, contrary to the case in other sectors of the economy, new technology rarely (though not never) allows substitution of capital for labor to lower costs.
But then Archibald and Feldman make this most intriguing argument, introducing to higher education the notion of "standard of care":
If new technologies seem to increase the quality of care, patients and physicians alike will want to use them. As a result, the new technologies are adopted despite the fact that they increase the cost of providing care.
We don’t use the language “standard of care” in describing higher education, but this may be a mistake. At colleges and universities, the chalk, paper, pen, and test-tube world has been replaced by wired buildings, laptops, high-tech classrooms, and pulsed laser systems in physics labs, together with the specialists needed to make the systems work. This change has its roots in the fact that the outputs of higher education (both research and graduates) are the inputs for other industries. This forces higher education institutions to educate students to a standard influenced in part by those who will hire its students and part by the past developments that have come out of their own research activities.... These industries and agencies have changed the way they operate by adopting a series of technological advancements, and higher education must adapt or they will in effect be guilty of educational malpractice.
Archibald and Feldman show that higher education’s costs have grown like those of other service industries, such as dental services, where a high proportion of employees are highly educated and where the role of technology in the production of the service has changed as it has in higher education. Indeed the pattern of cost increases in dental services from 1970 to the present is identical to higher education.
The current narrative, which says that higher education costs have risen inexorably as a result of such things as mismanagement, misplaced pandering to student non-academic demands (e.g., recreation facilities, "luxurious" residence halls) in order to recruit students, and declining faculty productivity, calls for controlling revenue through limiting tuition increases and state appropriations to force efficiencies on institutions unable to do so on their own. With Archibald and Feldman, I am not prepared to say there are no inefficiencies — that all higher education capital investments support student learning, or that all faculty are strong and productive. Indeed all of us in positions of leadership in higher education, like our counterparts in the for-profit sector of the economy, must work constantly to spend our resources as wisely as possible. But pursuing the implications of that narrative without understanding the real cost drivers causes us to take our eyes off the ball and virtually ensures a long-term decline in college and university student learning outcomes.
5. Faculty time available for and devoted to undergraduate education must be protected and expanded. It is the key resource for success in the kind of education our students need.
Only if one ignores the systemic costs of not structuring teaching and learning properly does it appear more expensive to create and sustain learning environments characterized by pedagogies of engagement. At the same time, where will the faculty time come from to increase the extent to which students experience the learning environments we know work?
In my view the real elephant in the room is not the high cost of pedagogies of engagement — because as I have shown, they are both effective and efficient if you think systemically — it is the cross-subsidy of research and graduate education with undergraduate tuition that is now so deeply embedded in much of our system of higher education. Resources from undergraduate tuition (and, in the public sector, state subsidies for undergraduate education in addition) that could make possible the adoption of more successful undergraduate teaching and learning environments are instead subsidizing graduate education and research at large numbers of American universities which, in the aggregate, have a large fraction of the nation’s undergraduate enrollment. Unless we face up to the choice we are making as a nation not to change how we finance research and graduate education, we will not accomplish the transformation of undergraduate education in America that is absolutely necessary for 21st-century success.
I believe we absolutely must face up to this choice. Too many undergraduate students at research universities — even the most elite research universities — and wannabe research universities all across America are being taught at very low salaries by graduate students (even first-year graduate students with essentially only an undergraduate degree) and adjunct faculty members, neither of whom have any incentive or resources to create the teaching and learning environments that we know work best. These salary savings fund low teaching loads and high salaries for faculty teaching graduate students so that they have adequate time for research. NSSE data show that among all American colleges and universities, undergraduate students at the research universities least frequently experience pedagogies of engagement and high-impact practices and least frequently encounter high faculty expectations that they develop three critically important kinds of higher-order thinking: analytical thinking, integrative thinking, and reflective thinking.
This is unacceptable. But I know in my heart that we will continue to accept it because it is essentially not discussable. Research and graduate education are both critically important. They need to be funded on their own merits. As someone whose teaching and leadership career has been almost completely in selective liberal arts colleges I do not have practical suggestions for how to pull this off, but I believe the first step is to name the problem, and that is what I have done here.
If we can’t or won’t tackle the elephant in the room, what might we realistically hope to do? In the private liberal arts colleges I know best the HERI Faculty Survey shows that an average of about a day a week of faculty time is spent in shared governance (e.g., faculty councils, senates, and advisory committees of many different kinds). Since the same survey shows that faculty in these colleges work 48-50 hours a week during academic terms, that represents roughly 17 percent of the work time of the entire faculty. Are we getting outcomes from that time that are as important as the augmented learning outcomes we could achieve if some or even most of that time were reallocated to student learning? What is it that is most essential for us to achieve from faculty involvement in shared governance, and can we achieve it with less faculty time invested?
While I was president, the St. Lawrence faculty wrestled with this issue multiple times because they too believed that more of their time should be allocated to student learning. But they were not able, within our shared governance system, to come to a place where a significant reallocation could happen. If I could do it over again I would give much more leadership and support to their efforts. All of this, of course, was prior to the great recession. This issue is on the table again at St. Lawrence with much more urgency. I hope for a different and better outcome.
I do not have all of the answers to the question of how we can conserve our most precious resource — faculty time — for its highest and best use, but what I am arguing for is a greatly heightened effort to do so across all of American higher education. We need a vigorous reinvestment in what we know works in student learning rather than the kind of disinvestment we are seeing today — a disinvestment that is tantamount to betting the farm on unproven replacements that are less effective, less efficient, and ultimately more costly.
Daniel F. Sullivan is president emeritus of St. Lawrence University and chair of the Association of American Colleges and Universities Presidents' Trust.