The Outcomes of a Liberal Arts Education: State of Research

Michael McPherson, senior fellow, The Andrew W. Mellon Foundation Mariët Westermann, executive vice president for programs and research, The Andrew W. Mellon Foundation

In 2017 the Andrew W. Mellon Foundation launched the Mellon Research Forum on the Value of Liberal Arts Education.  The Forum aims to connect and support researchers in the social sciences and the humanities who are generating rich data and fresh perspectives that can help us understand better the role of the liberal arts for students and for society at large.  While it is often assumed that such an education prepares individuals to be well-rounded members and leaders in our society, democracy, and economy, it is clear that the evidence of these effects is subject to growing skepticism and outright claims of irrelevance.  Mounting student debt as well as reports of underemployment of college graduates and their prolonged dependence on family support have reinforced such doubts.  The Foundation has a longstanding commitment to strengthening liberal arts education in its various institutional forms, but also believes that, given its cost, it is reasonable to ask for clearer evidence of its efficacy and sharper understanding of how it may actually work.  And if a liberal arts education turns out to be as beneficial as we believe it to be, how might it be delivered more equitably across American society?

We report here on an early undertaking of the Mellon Research Forum to give interested researchers as well as general readers an overview of the current state of knowledge about the effects of liberal arts education on students’ cognitive and psychosocial capacities as well as later life outcomes such as health and well-being, civic engagement, and economic success, among others.  These are important empirical questions about which it is hard to advance knowledge.  The seven essays assembled on this website address the question of what we know today about the history and the various potential effects of liberal arts education and also include ideas and suggestions about how researchers might go about learning more.  None of these papers claims to be definitive; they represent a kind of status report on the state of knowledge at the beginning of what we envision as a long journey of inquiry and learning.  

Liberal arts education has its share of critics, but the fact is that most American institutions of higher learning, from community colleges and small liberal arts colleges to the great public and private universities, offer some version of a liberal education to a large share of their undergraduate students.  Indeed, as Figure 1 shows, slightly more than half of students graduating with bachelor’s degrees in the US major in a traditional liberal arts subject, like history, sociology, English, or mathematics, rather than a pre-professional area such as business or nursing.  Indeed, the share of degree recipients with liberal arts majors is currently (as of 2014) higher than it has been at any time since 1975.  It is worth noting, as Figure 1 also shows, that the share of liberal arts degrees awarded in the humanities has fallen over time, while the natural sciences’ share of liberal arts majors has risen and that of the social sciences has been roughly constant.  In community colleges (see Figure 2), which often are known for their extensive vocational and technical offerings, roughly 40 percent of all Associate’s degree recipients earn their degrees in liberal arts studies, a share that has been roughly constant over time.  And almost everywhere in American colleges and universities, students, regardless of major, are expected to meet some set of core or distribution requirements that aim to ensure that graduates will be familiar with a range of arts and sciences subjects that have long been part of the liberal arts curriculum. 

 

 

 

 

Liberal arts education, broadly construed, is thus a central part of the American undergraduate experience.  It is striking that, in a country where there have always been voices pressing for a cheaper and more narrowly vocational kind of education for most students, there remains a persistent demand for and commitment to an educational experience that is deeper, more lasting, more integrated, and more devoted to expanding students’ intellectual and social capacities than vocational training in one area of expertise aims to provide.  This ambition touches students from all parts of our society, regardless of race, gender, or social and economic background.  It is a mistake to think that liberal education in this broad sense has been marginalized in the US.

This broad commitment to principles of liberal arts education influences not only the education of undergraduates but also the way in which colleges and universities go about fulfilling their important public roles.  They are generally supportive of freedom of thought and expression. Colleges and universities are often hosts of candidate forums and community discussions; they invite controversial speakers to events open to the local community (even when such invitations are flashpoints for public or campus disagreement in their own right); and they defend the academic freedom of faculty who express controversial views.  Traditions of open sharing of ideas and of the pursuit of knowledge for its own sake as well as for career purposes, all of which are consonant with a liberal arts perspective, have helped make American universities among the most productive and admired in the world. 

All that said, we need to acknowledge that the far-reaching but rather generic commitment to liberal arts education we have described manifests itself in very different ways in the teaching of undergraduates in different parts of the US higher education system.  Some of the variation arises from differing beliefs or convictions about essential elements of a liberal arts experience.  Some institutions, for example, emphasize curriculum:  what content should be taught and learned in liberal education?  Others may describe liberal education in terms of its intended learning outcomes for the student:  what should a liberal arts education enable a graduate to know and to do?  Others may emphasize the cultivation of a certain type of character, although, as defenders of the liberal arts have noted, many institutional leaders now shy away from emphasizing “character development” as an intended outcome of undergraduate education.  The way a liberal education experience is provided also depends heavily on the resources available to the institution, the backgrounds, high school preparation, and interests of the students, and the circumstances of education, which differ greatly, for example, between an urban commuter campus and a residential campus in a small-town setting, or between a flagship land grant institution and a private research university.

Does all this variation matter?  Does it, in particular, matter to the nature of the educational experiences undergraduate students have, and are those varied experiences reflected in turn in differences in students’ later lives?  Does, for example, an instructional program built around small classes, heavy on discussion and writing assignments, have a measurably different effect on a student’s later life capacity for critical thinking than does a similar program built around large lecture courses and multiple-choice exams?  Does including critical reading of serious fiction in an undergraduate program increase the likelihood that graduates will prove in later life to be more empathetic toward people who are different from themselves?  Also of great interest are questions of whether particular kinds of liberal arts experiences have similar influences on all students, or do variations in students’ backgrounds, interests, and capacities influence what they get out of a particular kind of educational experience?  Many of us have opinions about the answers to questions like these, opinions that bring with them the weight of lived experience, but as the papers collected here show, systematic empirical evidence to support these opinions is often disappointingly thin.  One of the central aims of the Mellon Foundation’s initiative is to complement our informal judgments on these matters with more evidence-based findings.

We have no illusion that any research initiative, however well-funded and well-staffed, will provide definitive answers to these subtle and many-sided questions, nor that empirical findings should supplant (rather than inform) the wisdom and judgment of practice of those who have long labored in these fields.  But we do have every reason to think we can make progress on these and related questions, and we have confidence that striving for better answers to these questions is well worth the effort.  Americans remain committed to the idea that an ambitious and meaningful education is the birthright of every American, and more knowledge of how such an education works can only help to make that idea a reality. 

One specific reason for our optimism about this initiative is a recent turn in social science investigation of education and in other areas of scholarship.  Several of our authors have noted in their papers that one of the limitations of available studies in their fields of interest is that social scientists have tended to conceptualize education as a quantitative rather than a qualitative variable. That is, social scientists, including economists, sociologists, and psychologists, have focused more on tracking “how much” education students acquire than on “what kind of” education they experience.  The level of educational attainment — for how many years did you go to school — was relatively easy to measure and to compare across people.  Measures of educational attainment have proved indeed to be a powerful explanatory variable in studies of subjects as varied as life expectancy, income, criminal behavior, and voting.  Recently though, education researchers have increasingly found evidence, especially in pre-college education, that the characteristics of the schools one attends can matter as much as or more than how long one has been enrolled there. Here are a few striking examples of such findings for K–12 education:

  • Attending an integrated school instead of a segregated one makes a substantial positive difference to how much black students learn (Rucker Johnson, forthcoming);

  • In the early grades, schools with smaller classes (other things equal) produce more learning among their students (Tennessee Star experiment);

  • Attending a school where the primary grade teachers use empirically tested techniques to teach reading produce many more fourth grade students who can read fluently than other schools do (Emily Hanford; Catherine Snow); and

  • Other things equal, schools that spend more money per student produce stronger learning results among students than others do (Kirabo Jackson and Rucker Johnson).

The list goes on. This “qualitative revolution” in elementary and secondary school research is in fact part of a broader trend throughout the social sciences to downplay focusing on averages in favor of focusing on variations, the factors that produce them, and the consequences they bring.  Improved analytical tools and much richer data sets have made such work more feasible, and we expect that a significant share of the work that will be undertaken in this initiative will draw on these improved — and still rapidly improving — tools and data sources.  We suspect that studying and learning from the consequences of variation among colleges about how they go about the work of education (instead of lumping very different kinds of institutions together and measuring average results) will become a growing trend in studies of higher education. Our interest in the varieties of liberal arts education practiced in the US is one important dimension along which variation can be studied, and we hope the Foundation’s support for this line of research will prove a stimulus and an encouragement to this larger trend. 

As an early step in developing this initiative, the Mellon Foundation decided to commission a set of papers aimed at helping readers understand what the current state of knowledge was regarding the impacts of a liberal education on a variety of later-life outcomes of students.  We encouraged our carefully chosen writers to go beyond reporting on the current state of knowledge to share their ideas on how further progress might be made.  We asked our authors to focus on their fields of expertise, which include health and well-being, economic success, civic engagement and democratic participation, cognitive development, and social-emotional development. It is important to stress that the authors were encouraged to express their own judgments and points of view. The essays are not intended to be thorough literature reviews but rather the judgments of these scholars on what are the most illuminating and instructive of the available studies. The authors received feedback from multiple expert readers at or selected by the Mellon Foundation, but the papers register the authors’ own views and not those of the Foundation. 

Our authors faced a challenging task. For many students, college is a valuable and rewarding experience in its own right, but for most students (and almost all parents) the most important aspect of college lies in the effects it is expected to have on a student’s later life. If the alternatives are college vs. no college, social scientists have a reasonably good handle on some of the answers.  (Notice that this question falls neatly into the framework of conceiving education as a quantitative variable: how do people with 14 or 16 years of education compare with those who have 12?)  If you attend and complete a college program (BA, AA, or certificate) the likelihood is that you will earn a higher income, experience less unemployment, enjoy better health, and participate more actively in civic and political activities than a similar person (similar as far as we can tell) who did not attend college.  None of these outcomes is guaranteed, but on average this is the powerful tendency.  Explaining what it is about college that produces these results and establishing whether the results are really owed to college or are due to differences between the students that researchers weren’t able to identify — these are much more controversial questions. 

But once we venture beyond the familiar territory of averaged outcomes and quantities of schooling to try to learn how results vary depending on the content or character of the education provided, we are in largely unexplored — or at best much less explored — territory for empirical research.  The primary question becomes: What difference does it make which or what kind of college you go to?  Or, to be a bit more precise, what difference does it make whether you go to a college that emphasizes a liberal arts experience or one that does not?  Our authors do a terrific job of identifying the considerable difficulties in getting reliable answers to this sort of question.  Among the obvious challenges is identifying and keeping track of similar students who attended different kinds of colleges, identifying in a consistent way the characteristics of colleges that make them more or less liberal arts-oriented, and then following students long enough to allow differences in their later life outcomes to be observed.   We also expect the differences among students who went to different sorts of college to be smaller and harder to measure than the outcome differences between students who did or didn’t go to college. 

Most of our authors rightly asked us how we were defining liberal arts education.  This is a topic that engaged and even vexed the initial advisory committee for the Mellon Research Forum initiative on the liberal arts for many hours.  To help the committee achieve a workable consensus on how a liberal arts education may be defined for purposes of research, we asked Harry Brighouse, professor of political philosophy and education at the University of Wisconsin – Madison and a member of the committee, to provide a frame of reference for how the features of a liberal arts education may be teased out and differentiated from other types of education.  His essay is posted here.  We also asked Robert A. McCaughey, professor of history at Barnard College and a scholar of American higher education, to accomplish the challenging task of writing a thumbnail history of liberal arts education from the 18th century to the present.  Both of these papers provide valuable introductory reference points for the research review essays on specific types of outcomes of liberal arts education. 

The first of our review essays tackles what are from some perspectives the biggest questions: how does the character of a student’s undergraduate education help to shape his or her longer-term health and personal well-being?  Carol Ryff, a distinguished psychologist at the University of Wisconsin – Madison, took up these questions. In a very thoughtful essay that includes a searching examination of how well-being is best understood, Ryff also helps us to see why certain experiences associated with liberal arts education, like close reading of literary fiction, might be expected to have lasting effects.  She is, however, frank to admit that available evidence to test this and like hypotheses is very limited, partly because the research tradition has focused on the amount and not the kind of education people receive.  Besides providing a clear account of the current state of the evidence on her central questions, she also contributes a thought-provoking and encouraging treatment of the potential for further work. 

A good deal of empirical evidence indicates, as noted earlier, that more years of education are associated with better career opportunities and higher earnings.  We asked Catharine Hill, a leading economist of higher education and President Emerita of Vassar College, to report on what we know about whether different kinds of education, and liberal arts education in particular, make a difference to individual economic success. Hill, finding only limited evidence from past studies of this question, goes beyond her assignment to develop new preliminary findings of her own, drawing on a valuable new data set developed by Raj Chetty and his colleagues as a resource.  Her paper thus extends our knowledge as well as doing a fine job of pointing the way to further research. 

More years of education are associated not only with higher income but with more intense and more effective engagement with civic and political affairs.  Diana Hess, Dean of the Education School at the University of Wisconsin – Madison, Paula McAvoy of North Carolina State University, and David Campbell of the University of Notre Dame explore the evidence about whether liberal arts education has a distinctive impact on civic involvement beyond what other forms of education do.  While underscoring the limited amount of existing work directly focused on this question, the authors point out that civic engagement is an inherently social and interactive activity, suggesting that the discussion-oriented and interactive approach to instruction often pursued in the liberal arts may be particularly suited to acquiring both skill and interest in civic activity.  They suggest that newly emerging data sets and growing interest in questions about effective civic activity bode well for future work in this field. 

We conclude our brief overview of this collection of papers with two pieces focused more directly on development of skills and capacities that are often seen as important to the liberal arts, and whose effects on students could be evident even while they are receiving the education.  We asked Henry Braun, a leading psychometrician at Boston College, to inform us about the relationship between education in the liberal arts and important cognitive capacities like critical thinking and writing ability — the kinds of intellectual performances that could at least in principle be assessed through tests and written work.  Braun makes clear that developing reliable instruments for measuring these capacities is challenging in itself, even prior to investigating their relationship to liberal arts education.  He is able to identify some illuminating findings from recent work and provides a useful assessment of prospects for further investigations in this area.

Recent decades have seen a growing emphasis on a set of capacities and dispositions complementary to the cognitive capacities Braun reports on, that are important in education and in life.  This cluster of qualities, which include interpersonal capacities like the ability to work well with others and intrapersonal qualities like perseverance and self-control are often grouped together under the somewhat awkward label of “non-cognitive capacities,” even though their likely interaction with cognitive development is intuitively evident.  Camille Farrington, a developmental psychologist at the University of Chicago, provides us with a helpful roadmap of this emerging field of inquiry, which has been pursued mostly in middle and high schools and has produced evidence that certain educational interventions can foster mindsets conducive to learning and other types of achievement.   She guides the reader toward the limited evidence that now exists on the implications for liberal arts education, and examines the potential for new research. 

What does this collection of essays tell us about the current state of knowledge about the influences of liberal arts education on students’ cognitive and psychosocial skillsets and the later life development of graduates?  Readers will of course need to answer those questions for themselves. In our view, the papers make clear that, while there are certainly promising and insightful studies in all the areas examined, systematic evidence at this point about the lasting impact of a liberal arts education is thin, perhaps surprisingly so.  We have a relative dearth of evidence in some of the areas discussed here in part because these questions have not been a center of attention for most social scientists who work on education, and in part because of limitations on available data and analytic techniques.  Fortunately these limitations are rapidly easing. It is also important to acknowledge that the kinds of questions our authors have examined are inherently very difficult and subtle.  Still, these papers encourage us to recognize that there are promising paths forward for new research, building on increasingly rich data sources and new analytical tools (“big” data, predictive analytics, and even artificial intelligence among them).  The questions our authors raise are compelling not least because of their intrinsic interest.  At the same time, the commitment of most American colleges and universities to provide some form of liberal arts education to most of their undergraduates requires examination as to how effective these efforts are now, and how they can be made more effective.  We believe that studying the range of problems and questions raised by these papers, and by this initiative more generally, warrant attention from some of the best minds working in the social sciences and humanities.