THE UNIVERSITY OFCHICAGO 5 EECORDApril 3, 1973 An Official Publication Volume VII, Number 4CONTENTS49 REPORT OF THE COMMITTEE ON THE RELATIONSHIPBETWEEN THE UNIVERSITY'S ACADEMIC PROGRAMS ANDTHEIR INTENDED EFFECT UPON THE STUDENT56 REPORT OF THE COMMITTEE ON TUITION69 ON TUITION AND THE COSTS OF HIGHER EDUCATION:PROLEGOMENA TO A CONCEPTUAL FRAMEWORKTHE UNIVERSITY OF CHICAGOFOUNDED BY JOHN D. ROCKEFELLER© 1973 by The University of Chicago. All rights reserved.THE UNIVERSITY OF CHICAGO RECORDREPORT OF THE COMMITTEE ON THERELATIONSHIP BETWEEN THE UNIVERSITY'SACADEMIC PROGRAMS AND THEIRINTENDED EFFECT UPON THE STUDENTTo President Edward H. LeviJune 14, 1972It is almost a truism to state that universitiesin general are institutions which study practicallyeverything except themselves. As in the case ofmany such truisms, this one applies only in alimited fashion to The University of Chicago,which is not known particularly for a lack ofself-consciousness. The University of Chicago(henceforth referred to as the University) hasalways carried out much solid research and muchfruitful discussion about basic academic problems, but on the whole the research tends to becarried out by the specialists in education, andthe discussions tend to be confined to "administrative" or "para-administrative" circles. As forus, the other members of the University community, this apparent absence of academic interest in academic problems has reasons whichshould be mentioned, because in themselves theyconstitute an important part of the problemunder discussion. We are trained to deal withmatters toward which we have or attempt to have"critical distance." We are trained to avoid "personal," "subjective" involvement with the matters under our scrutiny. The alma mater is fartoo close to our lives to permit both the necessary "critical distance" and trie concomitantemotional neutrality. Furthermore, we operatein an institution of enormous diversity. Hereagain, our training and our daily academic workteach us to avoid generalizations, to see theparticular, to distinguish the specific. It is oftenvery difficult to see common important problemsin, for example, the programs of the New Col-Peter F. Dembowski, Chairman of the Committee, wassolely charged with the writing of this essay. legiate Division and the postdoctoral studies inInternal Medicine. Nevertheless, when we dothink about the academic problems we have incommon, we realize that in fact there existwidely accepted, unifying academic attitudes aswell as widely held academic views, but suchattitudes and views are only tacitly accepted.And our professional formation, or if you wish,deformation, makes us reluctant to engage in theanalysis of tacitly accepted attitudes and views,for they tend to become, in our minds, "truisms." And we abhor the "obvious." We mistrustthe "cliches" pertaining to the University. (Andthis mistrust is translated, paradoxically, into thequarterly commencement rites during which wetolerate large amounts of "cliches" in the perhaps subconscious hope that such rites will makeus "immune" from the "cliches" for the rest ofour time.)Occasionally a crisis (either more or less general or localized) forces us to think and to discussthe basic theoretical assumptions upon whichmuch of our work in the University is based.Crises, as necessary as they are for the generalhealth of an institution, are not really the bestoccasions to think and to formulate thinkinginto coherent statements understandable bothinside and outside the University. That is thereason we believe it to be valuable to try toformulate some basic principles concerning academic programs and their intended effects on thestudents now, during a period of comparativeabatement in the atmosphere of crisis. Whatfollows is an essay, taken here in its etymologicalmeaning, that is, an attempt to present as simplyand as unpretentiously as possible a statementof those basic, "tacit" assumptions which guidemost of the decisions concerning the Universityin one of its fundamental functions: organizingand maintaining academic programs. The essayresults, in large part, from the deliberation of a49Faculty Committee. A considerable degree ofunanimity has been reached on most aspects under discussion. Nevertheless, the writer of thisessay is solely responsible for the final statement.He is solely accountable for any errors, misunderstandings, and other inadequacies in theinterpretation of the various views of the Committee members.One of the most striking facts encounteredby anyone who studies the problems of academic programs is their great diversity. There aregraduate and undergraduate programs, "pre-pro-fessional" and "liberal" programs, professional("practical") and research-centered ("pure") programs. One academic department can carry outan "unstructured" program geared to a broad"cultural" or "humanistic" education on the undergraduate level and, at the same time, on thegraduate level it can offer a highly structured"university professor training program." Thesame area, and the same level, can containprograms which are "flexible," offering manychoices of courses to the student, and otherprograms which are quite "inflexible," offeringin effect the same "fare" to all their students.Some of the programs seem to be unchangeable,fixed forever, while others appear to undergocontinual evolution. Under such conditions itwould be imprudent to try to offer a usefuldefinition of academic programs which wouldencompass all the varieties mentioned above andwhich would not at the same time be a merestatement of generalities. Rather than risk thepitfalls of defining what the programs are, letus try to state what they do.One of the first things the program does is todefine itself. Most, if not all the programs are infact self-defining. As we shall see, they also tendto be self-confirming. Taking this into consideration, we came to the realization that very fewuseful things can be stated here as to the scopeand length of time necessary for academic programs in general. But it is very important to understand the necessity of a constant, corrective,critical reexamination of each individual program, as to the ratio of its "core" requirementsto the rest of the program, and the concomitantminimum time needed to complete it. In a self-defining program there must be some "outside"criticism.Some of the programs are centered around asingle specialist. But more often the faculty of agiven program represent a real collegium in whichthere are several persons collaborating in theprogram. Very often there are considerable dif ferences of opinion, of methodology, and oftemperament. Such differences are not necessarily pernicious; knowledge and various techniquesin acquiring and analyzing it should constantlybe tested against intellectual opposition. But inorder to preserve the program for disintegration,much patience, mutual respect, and the occasional availability of a sense of humor are needed onthe part of the collaborating faculty members.Here again the programs exist with a tacit assumption that the individual differences are lessimportant than the program itself and that thechange in the nature of the program should becarried out in such a manner as not to endangerthe existence of the program or its capacity tofunction with fairness and equity in regard to thestudent. Working within such programs requiresfrom the participating faculty a constant testingof professional and institutional ethics and loyalties. It is only a slight exaggeration to say thatprofessors are total individualists whose individualism is limited by the exigencies of commonprograms. In order to preserve the value, thedesired effect of any program, the medievalmaxim diversa sed non adversa must constantlybe considered and guarded.The methods of teaching, the degree of facultyinvolvement with teaching and even the importance attached to teaching varies widely in different academic programs. Formal courses oflectures, informal lectures, formal and informalseminars, numerous combinations of trie abovecan be found in our programs. In most, up tothe M.A. level at least, various courses or seminars carry out a major part of the intended effectof a given program: the acquisition, as systematically as possible, of the knowledge of the fieldas defined by the program, and the cultivationof the analytical and critical abilities relevant tothe field. Very complex, specific problems ofteaching lie, fortunately, outside the scope ofthis essay. We need not be detained by the widespread, but, at least theoretically, false teaching-versus-research-controversy. .What should be saidis simply that while there is a need for constanteffort to maintain and improve the standards ofteaching, courses and seminars are still the mostefficient learning devices available for groups oflearners. Those critics who are most pessimisticabout teaching and its various techniques are veryoften considerably more than that: they are pessimistic about the human capacity to learn.Almost all of the academic programs providefor group or individual examinations. The technique as well as the character of examinations50varies considerably from program to program.Their purpose is to make sure that the knowledgeof the field has been imparted to the student and/that the analytical and critical abilities relevantto the field have been acquired. Furthermore,and this is quite important, the examinations introduce the student to intensive, individualstudies. The thoughtful preparation of an examination can be a first step in attaining the majorobjective of the academic program— the creative scholarship and research. Apart from mostthoughtlessly carried out "cramming" the preparation and the actual taking of an examinationare an important, probably even indispensable,part of the general educational purpose of mostof the academic programs. Examinations function additionally as the simplest (and probablymost efficient) way of certifying the minimumrequirements of a given program (or of a partthereof) and by so doing they also fulfill therole of "guardians" of the minimum intendedstandards. They can also serve, but probably withless efficiency, as an economical "diagnostic"procedure which discovers various lacunae andshortcomings in the student's mastery of a givenfield of knowledge.Very many academic programs have (either inaddition to, or in lieu of the examinations) individual study projects. Like everything else inour programs, such individual projects vary immensely in scope and in educational importance.Whether we think about weekly written assignments, term essays, an M.A. paper, or a Ph.D.dissertation, we realize that they all contributenot only to the knowledge of the field and to theanalytical and critical abilities inherent in thefield, but, more importantly, to the major intended effect of most of the academic programs—the instillment of the habits of creative scholarship and research. While preparing and writing anindividual project, the student develops insightsinto the critical problems of the field, into thespecific methodological approaches inherent inthat field, and also into the institutional, professional, and ethical questions which are inherent in the acquisition and cultivation of knowledge of a given field. Individual projects, carriedout seriously and under propitious conditions,provide an excellent practice in the field with theconcomitant "feedback" and/or correctives offered by the teacher. Individual projects are themost significant meeting places of the two mostimportant component parts of any academic program: the teacher and the student.As can be seen, the effect of academic pro grams, at their best, is not simply an acquisitionof knowledge, techniques, and critical insights,but also the cultivation of thought habits leadingto the creative, research-minded personality. Itis obviously essential, therefore, that the studentbe not an "object," not a "consumer" of aprogram, but a true collaborator in it, in fact a"co-creator" of the program. The frequent misconceptions in this area, prevailing outside theUniversity, and occasionally even inside a givenprogram— a sure sign of its weakness— are themain causes of much needless and on the wholeunproductive public and private controversiesconcerning the student and his program. Thetacit acceptance of "consumer" ("customer") asa model for the student leads fatally to the distortion of the whole basis of this relationship.Its immediate, visible result is an anguish invariably experienced by the student who discovers that, unlike the customer, he1 is not always "right," and, more importantly, that hisrelationship with the program is far more complex than the simple, but false, customer-suppliermodel.It is our contention that the student is alwaysa "co-creator" of the program, even when he isnot fully conscious of this role. (One of thehallmarks of a good program is, of course, thathe is aware of it.) This contention stems not onlyfrom the fact that ours is not a vast "multi-university," that the ratio of student and faculty isstill very favorable, and that, on the whole, thestudent "feedback," either formal or informal,receives serious consideration. It follows alsofrom the fact that our teaching, in the broadestsense of the word, is co-determined by the student's response or lack of it. A program, say aseries of lectures in, for example, First YearHumanities "core" courses, given to a group ofdull and unresponsive students has a quite different effect on the student than the sameprogram (same teacher presenting the same subject material) offered to a lively, respondinggroup. The higher the level of instruction, themore obvious this contention becomes, but it isvery important to .realize that the responsiveparticipation by the student, in the class, outside of the class, formally or informally, is operative on all levels in all the programs. It co-determines the effect of any program.And this is the main, the real reason that wedo choose our students. Even if material restric-l8tHe" referring to a student or to a faculty membermeans, of course, "he" or "she" in this essay.51tions did not exist, even if, by some miracle, theUniversity could teach anything to anybody atanytime, we would still have to make a selectionof the students in a similar way that we select theteachers, our colleagues. The programs, all ofthem, select their students by two differentways: by admitting them; and furthermore, byoffering financial support to many of them. Here,as in so many other aspects of the University,there is great diversity in the methods and theintensity of recruitment of new students. Thus,for instance, the College, which can be considered both as a program and as an ensembleof interlocking programs, has one centralizedoffice of admissions which serves, in fact, severalprograms. In the Divisions, each program selectsits students either through its entire faculty oran admissions committee, the divisional officesof admissions acting chiefly as correspondingsecretary, coordinator, and occasionally as advisor to the admitting programs. The professionalschools have each their own machineries of admission and aid. This decentralization, reflectingthe basic intellectual structure of the University,is not without dangers which are often tacitlyaccepted, but perhaps not sufficiently understood by many. A systematic selection of faculty and students may further strengthen thenatural tendency of any program to become aself-confirming process. The program is definedin terms of existing competency and the selected members of the program, chosen for theactual or potential competency, tend to confirmthe program, which in turn may lead to "sclerosis." The two things must exist to obviate thisdanger. The first is that there must be the creative scholarship that goes beyond the existing competency. Thus, without any consideration of the ultimate results of scholarship on the"outside" world, creative, research-minded persons are indispensable for the healthy life ofpractically any program. The teaching-versus-research controversy is, as we have stated, reallyfalse, at least on the theoretical level, for creativescholarship is an inherent part of productiveteaching. The second requirement is true varietyin the students selected. Any program must be"challenged," and this challenge must itself bevaried, that is to say that the students mustrepresent different intellectual, social, and cultural backgrounds.We think that it is possible to state in generalterms the main desiderata in the selection of thestudent only if we keep clearly in mind the intended effect of the academic programs. What is sought is an individual capable of both profiting and contributing to an existing program. Heshould be capable and willing to do so. In allcases this presupposes the actual "prerequisites"and/or potential capabilities. He should be capable of acquiring, as early as possible, what weconsider to be both the sum total of educationalpurposes and the most important intended effectof all academic programs: the capacities for creative scholarship and critical, independent inquiry. The intended effect of academic programs,as we have touched upon previously, is to allowthe student to bring himself to the level of competence which would permit him to deal with thesubject matters of his field in both critical andcreative ways. This means the cultivation of attitudes (until they become habitual) of searching for those significant questions that identifyimportant problems, and of proposing, exploring,and testing the alternatives. These capacities arein fact more important than the specific contentof an academic program. They mark the pointof departure from dependency to scholarly independence. The intended effect of all the academic programs is, in the final analysis, theacquisition of the capacities of inquiry. Suchcapacities represent what used to be called prudence, for they allow the student to deal withthe very practical problem of imperfect knowledge. No information is truly exhaustive, no dataare truly complete, but creative scholarship andthe critical, independent inquiry are based uponthe responsible use of imperfect knowledge todraw tentative inferences. The acquisition of theabove capacities should also be accompanied bya sense of their relevance to other fields ofstudies and by the understanding of institutionaland professional ethics and loyalties. Both areinseparable aspects of the educational intents ofacademic programs.It should be quite obvious that the desideratain the selection of the student largely parallelwhat we seek in selecting a faculty member.There are, of course, many differences as to thedegree which these various desiderata apply tostudents in a given program. Thus, for example,the importance of institutional ethics and loyalties is perhaps more easily seen in, say, theSchool of Social Service Administration than in aprogram of the required courses in the College.But even in the second case, proper and improper acquisitions and presentation of the viewsand opinions of another person can be a seriousproblem of both academic ethics and professionalloyalties. Or to take another example, the willing-52ness and ability on the part of the student toacquire the knowledge and techniques in a program which is defined by rather strict professional goals seems to have more importance thanall the other desiderata. Nevertheless, a closerexamination of the real nature of the intendedeffect of a given professional program, such asmedicine, invariably reveals that the other desiderata have their place in it. A closer examination of the real nature of the intended effect ofacademic programs certainly exposes the fundamental unsoundness of the often expressed dichotomies such as professional versus liberaleducation. All good programs are in a sense both"professional" and "liberal," but there are, ofcourse, good programs and bad programs. . . .The real problem lies elsewhere. There are veryconsiderable differences among the various levelsand areas of the University as to the degree ofconscious motivation in the student. The College,taken as a whole (with a possible exception ofits strictly "pre-professional" programs, such as"pre-medical" ones), is quite different from theprograms in the Divisions and the Schools. Forthis and other reasons, the College presents muchmore complex problems in the selection of thestudent and the evaluation of the academic programs.We have been talking about some fundamentalcriteria in the selection of an individual student.The basic assumption has been that he is to be,to a greater or lesser extent, a "co-creator" ofthe program. But "program" means "students"(in plural). We have already stated that diversityand variety among selected students are verydesirable aspects of a program. Here again thecomparison with the faculty is helpful. It is certainly necessary to ensure variety in choosing thefaculty of a given program. A program composedof the "disciples" of a professor, a program composed of like-minded faculty will sooner or laterprove itself "sclerotic." The diversity and varietyof incoming students are, in our opinion, goodthings in themselves. This means that the searchfor different kinds of students should be carriedout irrespective of the broader social considerations, no matter how just and how pressing theseconsiderations appear at the given moment. Identical, Hke-minded students are as pernicious to aprogram as identical, like-minded faculty.In the Divisions and probably in the Schools,the theory and practice of the admission andaid procedures result in considerable diversity inthe student body. Here the criteria are moreeasily definable. The past achievements here (college degree) are the best prediction of thefuture accomplishments. The high school diploma is less "predictable." In doubtful cases, theDivisions and Schools have at their disposal theletters of recommendation, statements of purpose which, coupled with some "objective" tests(such as GRE, for example), tend to offer theadmitting faculty an indication of possibility ofsuccess in what is usually a well-defined program.The differences between the admission practicesin the professional programs result from the widevariance in the ratio of applying to admittedstudents. The greater the "pool" of applicants,the greater the current popularity of a given program, the greater the need of conscious andconcentrated efforts to ensure the variety of students.The College presents, in this, as in many otherrespects, a more complex picture. As we havesaid before, the College is a program and, at thesame time, an ensemble of programs. It is aprogram defined in terms of "pre-graduate," or"pre-professional" education. The common assumption in these programs is that even the most"pre-professional" programs must include broadeducational goals. Every good professional schoolwishes its candidates not only to be prepared forthe programs it offers, but also to be "educated,""to be able to think." This is one of the mainreasons for the required courses in the biological,physical, social sciences, and humanities. Abandoning these requirements would radically changethe very nature of the College. Both its "program" (general education) and its "ensemble ofprograms" ("pre-graduate" and "pre-professional" programs) would invariably suffer. The verycomplexity and the "multi-purposiveness" of theCollege make the selection of the student difficult. This difficulty is compounded by thenecessity to appraise not only past accomplishments, but also potential capacities. Because ofreal differences between high schools, we mustbe able to encourage the admission of the willingand bright, even if, at times, less well-preparedstudents. Otherwise we will not be able to maintain the necessary heterogeneous nature of thestudent body. Only a heterogeneous studentbody will continue to supply the University withthe "input" of new attitudes, new approaches,and new motivations so necessary for the continuous growth and health of any institution oflearning.This in turn implies some far-reaching consequences. We cannot do what is done, or ratherwhat is alleged to be done, by some big state53universities. We cannot accept, for example, thefirst 2,000 high school applicants, lecture tothem over the loud-speaker for a quarter or two,and then "weed them out" down to the desirablenumber of, let us say, 600 by a simple processof written examinations. Such "admission procedures" are not only impractical, wasteful, buton the theoretical level, educationally absurd. Wewould have to destroy the very nature of ourprograms to accomplish it. Furthermore, it wouldsurely result in a freshman class composed ofrather uniform students whose chief qualitywould be the uniform ability to "survive" certainkinds of examination. In order to assure that thisclass does represent the desirable variety of background attitudes and talents we must continuewith our complex, "multi-leveled" admission procedures and, what is even more important, wemust continue and improve the advising and thesupportive "survival" programs for the willing,potentially successful, but "weak on admission"students.The advising is, of course, part of the "functional definition" of any program. In the Divisions it is very often an integral part of the individual study project discussed above. Furthermore, the very nature of programs in the Divisions and, probably, in the Schools, makes theadvising of the formal or informal kind relativelyeasy to carry out, although, even here, there isan ever present possibility of a "communicationbreak." It must be borne in mind that such abreak can be caused by either the advisor, thestudent, or the circumstances of a given program.In the College the situation is more complex.There are advisors in various departments, who,as a rule, know quite well the offerings and conditions of the departmental programs. They tend,quite naturally, to think about the student interms of the departmental possibilities and needs.They conceive of their role as a service for thestudents who are already in the departmentalprograms, or are at least serious candidates forthese programs. But since the College is a verycomplex institution (there are some eighty-oddways of proceeding to the bachelor degree), ithas, in addition to the departmental advisors, aseparate body of College advisors. They are full-time, non-faculty employees attached to theOffice of the Dean of Students in the College.Concurrently to an important "technical" sideto their work, "bookkeeping" services and thelike, they help the student map his courseof studies and interpret the intricate situationwith which the student, confronted with many choices, often finds himself. College advisors donot serve any particular department, althoughsome of them specialize in certain broad areas(such as physical sciences, pre-medical training,etc.). As a rule they do not "promote" one oranother program, this or that educational career.Rather they explain the conditions, the prerequisites, etc., of a given program, while the studentis making his choice and help him to keep him inthe proper course of study once he has made hischoice.Working full-time with the students, Collegeadvisors have acquired a thorough, practicalknowledge of various programs and also of themain, current problems encountered by studentsin the various College programs. The facultywould do well to avail themselves of their knowledge, especially in the planning of new programsand in the renewal of the old.Advising on any level is particularly importantin a University such as ours where there existsa large degree of flexibility and a great varietyof programs. We must find ways of maintaining,and whenever necessary, of strengthening theexisting advising system. In the College, andprobably in certain cases in the Divisions andSchools, there are two groups of students inparticular need of all forms of academic advice:(1) the "weak on admission," that is to saypoorly prepared students, culturally underprivileged, etc., (2) the students who are not sure asto the direction of their academic career. Thesecond group is probably larger than the first anddoubtless includes members of the first group. Itis very important to take the problem of "undecided" students seriously. As early as possibleupon their reaching the University, they shouldbe advised (in the broadest sense of the word)not on "what they should be interested in" (oneoften hears questions to that effect), but ratheron "what there is, in a given program, to beinterested in." Students in both groups requirenot a change in the accepted standards— doublestandards are ultimately wrong both for the students and the programs— but advice, encouragement, and, if necessary, supportive, "survival"sub-programs. We should take care to see to itthat such assistance is given on a permanent basisand given in such a way as not to produce afeeling of inadequacy, helplessness, and despondency on the part of the students.All the academic programs involve not onlythe relationship between the faculty, subjectmatter, and the student, but also the relationshipbetween the student and his peers. This relation-54ship is very important and surely plays an enormous role in the effect of a given academic program. A five-minute talk between two studentsabout a common reading, or about a lecture,often makes this reading of this lecture trulymeaningful, and thus allows the student to continue active participation in the program. Student-peer relations are obviously crucial in overcoming the problems of poorly-prepared orpoorly-motivated students. The peer relationsfunction doubtless as the most important sourcesof support for the intellectual in everyday as wellas in emotionally stressful situations. And whileone hears a considerable number of complaintsabout the quality of various relationships in theUniversity, one seldom encounters Explicit criticism of the student-peer relationship. We are surethat if could be much improved, especiallyamong the undergraduates. Unhealthy forms ofmisunderstood competition do exist among thestudents, which in turn make the tasks of the"weak on admission" emotionally and intellectually even more difficult. Somehow we mustconvince ourselves and our students that ouracadernic programs are not for academic supermen, and that academic excellence should notbreed academic arrogance. To many incomingstudents their fellows here often appear arrogant.Arrogance in the peer seems to be far more contagious and damaging than arrogance encountered in professors.Student-peer relations are very important fromanother vantage point: they transgress the boundaries of programs. Talking about the College,we used the terms "ensemble of interlockingprograms," or simply "ensemble of programs."These terms apply grosso modo to the Universityas a whole. Since the intended effect of the academic programs is not only the systematic acquisition of knowledge- and the professional andcritical abilities that it entails, but also theacquisition of those abilities which encourage aperson to continue to develop, to "think on hisown," therefore the general cultural atmosphereof the University is of prime importance to thereal effect of all the programs. This considerationmust be kept in mind in the selection of thefaculty and students. We are not only selectinggood specialists and potentially good, futurespecialists, but also the persons who can profitfrom our cultural atmosphere and contribute toit. Whether the life in the University residences,the University theater, concerts, public lectures,exhibitions, films, student publications, variousorganizations and clubs, etc., should properly be called "academic programs" is a moot point.Each case has different merits and must bejudged differently. But the University does educate through these multiple and diverse activities.The "extracurricular" are, in fact, "curricular"because they contribute to the necessary culturalatmosphere without which the final effect ofmost, if not all, programs would certainly bepoorer. Our peculiar geographic situation, andcertain realities which make downtown appear tobe far away from the campus, compel us furtherto till our own cultural garden.Perhaps the best way of considering the difference between "curricular" and "extracurricular" is by distinguishing the different degrees ofresponsibilities. "Extracurricular" activities arebasically student-centered and student-run. Onthe whole, they are planned for immediate use.The role of the faculty in these activities is. andshould be secondary. This brings us to the consideration of a final very complex problem: thestudent responsibility and role in the shaping ofthe various aspects of the academic programs.Since the application of the false customer-supplier model to the student-program relationship is prevalent, especially outside of the University, much more must be done to correct thisview. Students must be made far more aware oftheir possibilities and responsibilities not only toprofit but also to contribute to a given program.It is very important that they become aware ofall the implications of being to a large extent"co-creators" of programs. It is to be expectedthat many will resist the conscious acceptanceof such responsibilities. Blaming the "establishment" for all the ills and failures, consideringoneself as being "programmed" by it (in the IBMsense of the word), is much easier than becominginvolved in a creative, active relationship with anacademic program.There are two aspects of this relationship. Inthe first place, the individual student has a fargreater degree of opportunity than he sometimesrealizes to shape and to improve the real natureof his own program. We believe that, on thewhole, our students have a certain sense of freedom and, concomitant with it, a certain sense ofresponsibility in selecting, modifying, and developing their own individual programs. (Thisagain would apply to a different degree in different programs.) In other words, many of ourstudents, are in fact what we have called "co-creators" of their own individual programs. Butthere is a second, larger aspect of the relationshipbetween the student and the academic program.55Except for certain rare, very specialized, and advanced learning situations, our academic programs are, by their very nature, collective. Beingcollective means that they have to be planned fora certain span of the future, and that changesproposed in them affect usually not the studentswho propose the changes, but the future members of the program. It is a common experienceof members of various curricular committees thatthe students, on the whole, are not impressed byarguments that certain aspects of a programwhich they wish to change have been proposedand introduced only a short time ago by anothergroup of students. One finds quite often that astudent or a group of students has great difficultyof imagining what other students thought abouta given aspect of the program, or what they willprobably think two years hence.In order to break through these enchantedwalls of the "everlasting now," we must do muchTo President Edward H. LeviJanuary 31, 1973What should be tuition policy for The Uniyersityof Chicago?This is an important question for the University. The tuition structure has a significant bearing on its financial position and the compositionof its student body. The subject draws strongreactions from students, particularly when tuition changes are under consideration. Facultymembers seem eager to express their views. EvenThis statement on tuition policy does not rest on extensive research. A scattering of writing on the historyand nature of tuition charges was perused, but littlematerial of relevance to the present undertaking wasfound, and no effort was made to develop a full bibliography on the subject. Several major universities wereasked about their own tuition policies, and their responses, along with published statements of a few otherinstitutions, were reviewed. No empirical studies werepursued, though it was recognized that such studiesmight be helpful in establishing policy in the future. Afew modest efforts were mounted to locate empiricalwork on tuition policy done elsewhere, but nothing more to explain the nature of the academic programs first of all to ourselves, and then to ourstudents. Only then can we expect far greaterand more responsible participation in the never-ending process which is the maintaining, expanding, and improving of the common academicprograms.Peter F. Dembowski (Chairman)Winston A. AndersonBenjamin S. BloomPhilip C. HoffmanPeter MeyerMichael J. MurrinNorman H. NachtriebTetsuo NajitaMargaret K. RosenheimCharles W. WegenerRichard WeisbergIzaak Wirszupalumni have on occasion taken more than acasual interest in the matter. Throughout theinstitution a belief exists that its tuition policysignals much about the character of the University.Various aspects of its tuition program havebeen discussed inside the University from timeto time. These largely piecemeal efforts have notbeen tied together sufficiently. Some aspects ofthe tuition program, moreover, seem to havebeen taken for granted over the years. A criticalreexamination of tuition policy, broadly conceived, might contribute to better understandinghelpful for this endeavor was turned up. Assistance wasderived from faculty memoranda dealing with variousaspects of tuition; and consideration of the theoreticalunderpinnings of tuition policy was furthered by questions raised by Marc Nerlove in his piece "On Tuitionand the Costs of Higher Education: Prolegomena to aConceptual Framework" [see page 69]. From the startit was assumed that a clean slate approach would not beappropriate for the task at hand. The University ofChicago, of course, has in effect had a tuition policy,even though the outlines have not always been clear.Prudence seemed to require keeping that policy in mindas this statement was being developed.REPORT OF THE COMMITTEE ON TUITION56and shaping of the University's program.Although questions of tuition policy havelong been discussed at private universities, thereseems to have been little theorizing about theproblems. This shortcoming is easily accountedfor. Private institutions differ so much amongthemselves that common denominators relevantto tuition are probably lacking. In some respectsalmost every private university is unique. Evenif the differences could be covered over bysimplifying assumptions, an overriding complication would remain: almost all private universitiescompete directly with state schools whose tuitionpolicies are fixed mainly by political bodiesmotivated largely by consideration of publicpressures. Various relationships can presumablybe established between tuition levels and enrollment patterns in universities viewed in the aggregate. Such generalizations, however, would notnecessarily be valid for private universities as agroup and are unlikely to be of help in formulating policy for any particular institution. Themore unusual the institution, the more insulatedit will be from analysis on a grand scale.The task of dealing with tuition at The University of Chicago, or any other private university,obviously cannot and need not await the development of theory about the role of tuition in oursociety. Common-sense requires that in establishing tuition programs each institution take intoaccount its own character, aspirations, and circumstances—including its competitive situation.These are the factors The University of Chicagoneeds to consider closely in arriving at an appropriate tuition policy.Any practical discussion of tuition policy mustbe anchored in economic matters and deal withfinancial affairs. It would be a serious error, however, to think that financial strength can make auniversity distinguished. It would be equallywrong to think that a university can be eminentwithout being financially strong. While this statement on tuition policy necessarily emphasizeseconomics and finance, the special character ofThe University of Chicago is dominated by valuesand qualities that are at some distance from themarketplace. All that follows is to be understoodwith this important consideration always inmind.Tuition is usually equated with the standardfees or charges that a school calls upon studentsto pay for pursuing their programs of study.Such a concept of gross tuition revenues is usefulfor many purposes. But it is not sufficient in thesearch for tuition policy because universities commonly provide students with various offsetsto tuition charges, offsets that may significantlyalter the impact of required fees. Tuition policymust encompass those offsets; thus it needs todeal with both gross tuition fees or revenues andnet tuition amounts that students are called uponto pay for carrying on their courses of study.The earlier sections of this statement are concerned with gross tuition. The latter ones dealwith net tuition.IWhy should private universities ever charge anytuition fees?The most obvious and compelling answer tothis threshold question is that virtually all ofthem desperately need the tuition income inorder to keep operating as private organizations.Revenues from endowment and gifts are far fromadequate to support their undertakings. Government funds, especially those available on a basiscompatible with the character and aspirations ofa particular institution, are relatively small andtheir continuation is uncertain. Most private universities could easily arrange to be absorbed instate systems of higher education; indeed, manyhave already done so. Such a move obviouslywould not preserve the distinctive qualities of aprivate organization. Nor would it put an end totuition fees, since all public universities in thiscountry now charge substantial sums for tuition.Another reason often advanced for a private(or even a state-supported) university to demandtuition fees is that higher education correlateswith increased life-time net earning capabilities.Although the evidence is subject to differing interpretations, the prevailing view is that advancededucation in general contributes to earningpower. The potential increase, as well as therealization of it, obviously will vary greatlyamong courses of study and among students.But the averages show that, up to a point, a widegulf exists between the earning power of thosewho reach one level of educational attainmentand those who attain a higher level. Under theseconditions, not to charge tuition fees appears toresult in conferring potential economic gains onindividuals without requiring them to contributedirectly to the economic support of the endeavor. Such a result would seem to violate abasic sense of fairness.It has been suggested that the economic advantages of a higher education call not for charging tuition fees but merely for encouraging57alumni to contribute voluntarily to their benefactor school if and when they personally reapthe economic benefits. This sanguine expectationis at least as impractical as it is uplifting. Theproblem of a transition period alone condemns itto failure: a large gap, probably averaging somewhere between 10 to 30 years, exists betweencompleting formal higher education and benefiting economically from it. A gap of even a fewyears between paying operating expenses andreceiving revenues would surely put almost allprivate universities in impossible financial straits.Moreover, past experience strongly indicates thatmany beneficiaries of tuition-free private education cannot be counted on to share their goodfortune with the schools they attended. Eventhose who become charitably disposed are quitelikely to support projects or institutions thatengage their attention in later years— includingschools attended by their children and grandchildren. There is, in addition, no persuasive evidence that, among alumni who in the long runare more or less similarly situated, scholarship recipients on average tend to contribute more totheir schools than do their tuition paying classmates. No incompatibility exists between charging tuition while students are in school and alsoencouraging them to support the institutionwhen they are in a position to undertake that responsibility. This by now is a very familiar combination.A related reason occasionally mentioned for aprivate university to charge tuition fees is tocarry out the expectations of past donors andnot to repel potential future contributors. Individuals directly or indirectly donate to a privateuniversity for any number of reasons. There is acommon-sense basis for believing that when auniversity acts in fiscal ways that seem indefensible or self-defeating, it will chill the interestsof possible benefactors. Prediction in this areaadmittedly is not highly informed, but it mightbe instructive to reflect on an extreme move. Itis at least a fair guess that many actual or prospective supporters would be dismayed to learnthat a university, especially in times of financialstringency, embarked on a program of eliminating tuition charges, thereby relieving the mostimmediate beneficiaries from having to pay forkeeping the venture going. Granted that tuitionpolicy should not be skewed by views held bypossible donors, a private university neverthelesswould be well advised not to act in ways thatare seen by its supporters as being financiallyreckless. Still another reason suggested for private aswell as public universities to charge tuition is thatthe imposition of fees probably screens out manyapplicants who are not sufficiently motivated tojustify using up scarce resources on them. This,it must be granted, is a conjecture. In the case ofwealthy families the barrier is apt to be insignificant. For the poor, the fees might keep awaynot only the unmotivated but also the cautiousor the financially worried or economically embarrassed. In the middle, however, it seemsplausible that there is more than a remote connection between tuition fees and motivation.Tuition fees likewise can have the effect ofencouraging students to expedite their formalstudies. Post-secondary school education is costlyto students in terms o'f income that is foregone,but such opportunity costs are not always veryvisible. The periodic payment of tuition fees canact as a strong reminder that higher educationcomes at a substantial price. Those students whorespond to economic pressures can be expectedto seek out ways of speeding up their programs.Although one might easily conceive of situationsin which hurried learning would be unsound educationally, there appears to be considerable slackin the system today. Tuition charges are likelyto reduce rather than increase that slack, if onlyby stimulating student pressure on institutions toreconsider the traditional time spans for variousstudy programs. Such moves, either by individualstudents or by the school, not only wouldeconomize on the use of valuable resources butmight well produce both better education andmore resourceful citizens.A major function of tuition fees is to allocateavailable places among interested applicants. Involved in this point is the standard supply anddemand relationship. If the price rises, the quantity demanded will decline; and if the price falls,the quantity demanded will increase. The situation in a private university, however, is not thatsimple. One may assume that a private university,taking into account its goals and resources, typically wishes to attract those it regards as themost promising or the best-qualified students.The process of determining which students aremost promising is never easy; many factors maybe relevant, and there is the pervasive consideration that judging which students are best forthe university depends importantly on judgingwhether attendance at that institution is in thebest interests of the particular students. By relying on the price mechanism alone a particularinstitution might obtain a student body with58greater wealth or easier access to loans, thoughnot necessarily one it conceived of as mostqualified. Quality control of some sort is thusneeded to fulfill its admissions objective. A tension may then develop between quality controland the play of the price mechanism, but not anirreconcilable clash. The quality control can seta minimum standard while the tuition price isallowed to allocate places above that level.Such a combination, however, may still turnout to be unsatisfactory. Using a minimum quality standard is not equivalent to filling all placeswith the student group best qualified in the eyesof the institution. If spaces are to be allocatedsolely on this latter criterion there would seem tobe no role for the rationing function of the pricemechanism, except under one condition. A poolof potential students may be thought to exist atthe bottom margin who are equally qualified andwho, therefore, could be sorted out by way ofthe tuition charge.This view of demand conditions perhaps isrealistic for some, if not all, private universities. The accompanying prescription nonethelessmight not be wholly acceptable. Admissions personnel, especially in graduate areas, usually believe they can rate applicants in some rank orderbased on performance or background or otherfactors, and therefore would be quick to denythe assumption of inter changeability. The generaluniversity community is not only likely to concur but to add that choosing among students ongrounds of ability to pay is incongruent with theprofessed nature and goals of higher education.The upshot may well be that both those whothink in economic terms and their critical colleagues have grasped part of the truth. Tuitionfees in fact will serve an allocative functionamong some undefined segment of at least minimally qualified applicants; even so, few in auniversity will think it wise to view narrowlytuition charges as being imposed in order to perform that role.IIIn light of these possible functions, what shouldbe the role of tuition fees at The University ofChicago?Surely, given the institution's history and proclaimed goals, fees should not be set explicitlyor primarily with a view to allocating placesamong applicants. As a leader in higher education, the University must rely heavily on qualitycontrols in selecting its students. Further, to re main financially viable, substantial revenues mustbe collected through relatively high tuitioncharges. Constructive consideration of tuitionpolicy at the University necessarily must proceedfrom these two points of departure.In setting tuition fees the University cannotescape market forces. State-supported and otherprivate universities provide alternative programsin the field of higher education. As its tuitionrises relative to levels at institutions regarded bypotential applicants as acceptable alternatives,the University becomes comparatively less attractive by becoming relatively more costly. Reducedattractiveness means that it will draw a studentbody that, under the criteria it has developed,is less promising on average than the optimumgroup. Any significant movement in that direction is likely to generate other changes. In manyacademic areas the University's ability to attractand retain outstanding faculty members woulddecline because student quality is regarded as ahighly important condition of research and teaching. A falling-off in faculty distinction wouldfurther reduce the relative attractiveness of theUniversity to 'students, and this in turn wouldfurther handicap the University in maintainingthe excellence of its faculty. Quality of facultyand quality of students invariably go together ina great university. A tuition schedule that pricedthe University out of the market for qualitystudents would soon alter the character of theplace.Does this reasoning, it might be asked, work inreverse? Would the University be able to improvethe quality of students— and in so doing increaseits ability to assemble an outstanding faculty—by drastically reducing tuition charges? This isan intriguing challenge. All other things beingequal, a reduction in price should lead to an increase in applications for admission. A possibleoffsetting factor is that adoption of an outstandingly low tuition schedule might be construed asan admission of serious institutional weakness.Quite likely the use of this strategy would alienate potential donors and other friends of theinstitution. But of decisive importance is the boldfact that tuition revenues are indispensable topay the salaries commanded by top-flight facultymembers and to maintain the physical plant andbasic student services. The University is simplynot in a position to move along such a coursewithout destroying itself financially and therebyending its existence as a private organization.All these considerations suggest a touchstonefor tuition policy: The University should seek59an equilibrium that will produce maximum revenues without causing the quality of students tofall below the very high level required to maintain the institution as a leader in education. Thisvery general formulation recognizes that it maynot be possible both to maximize tuition revenues and to attract the most promising studentbody imaginable. It therefore attempts to strikethat balance best calculated to promote thefuture of the institution as a great private university.1The suggested formulation of policy guidelinesinvites a number of important questions deserving of attention. In general terms these concernthe usefulness of "fair price" standards, the roleof tuition revenues in financing the institution,the relationship between tuition charges andother sources of revenue, the choice betweenuniformity and diversity in tuition charges, andthe relationship between size of the student population and total tuition revenues.IllIf maximizing revenues is an important ingredientof a tuition policy, will the price charged students be "fair"?This is a treacherous question. It implies thatprices set by operation of market forces may beunfair and assumes that there is an alternativestandard of fairness. The question is associatedwith the pervasive view, expressed over the yearsin many different ways and in many differentcontexts, that the only proper price is a justprice.A central difficulty with this doctrine is thatthe concept of a just price is highly amorphous.At times it seems to be built around the thoughtthat whatever is the "traditional" or long-established price is the just price. While this conception might be understood in a society in whichchange is so slow as to be imperceptible, the ideahas little applicability in our rapidly changingworld. At other times the concept of a just priceappears to turn on the notion that no one shouldgain an "unconscionably large" profit in sellinggoods or services. To do so, it is felt, is to takexThe recommended guideline obviously is not precise.A degree of ambiguity is recognized and is regarded asnecessary. The position taken in the statement is thatvarious factors are to be considered in setting tuitioncharges and establishing tuition offsets. People will differ over how these factors should be weighted. Prudential judgments rather than precise mechanical formulasare ultimately required in implementing any soundpolicy guideline. unfair advantage of another's misfortune or toexploit unfairly one's own position of dominance—as in the case of a monopolist. Again, theseideas seem to translate into operational rulesonly in a slow moving society in which all profitmargins tended to be stable around some widelyrecognized norms.It is most improbable that anyone at TheUniversity of Chicago, which operates at a largedeficit,2 associates tuition charges with generating excess profits. More likely the notion of a faircharge embodies an idea related to costs ratherthan profits. The essence of it is that no studentshould be called upon to pay more than the costassociated with his attendance— and this shouldbe the limit even though the institution as awhole runs at a deficit. "Fairness," in brief, ischarging an individual no more than the costsallocable to him.Any such doctrine is almost certain to be adistraction in setting tuition charges.To begin with, there is no single principle onwhich costs can be allocated to particular students. Sharp differences in views will arise overtreatment of expenditures connected with constructing buildings, maintaining the physicalplant, conducting research, running student housing and other student facilities, stocking andoperating the library, protecting and improvingthe neighborhood, and so on and on. Equallygreat divergences will develop over the questionwhether allocations to individual students areconcerned with marginal costs, average costs, orsome other conception. In the face of these manydebatable issues, an effort to use a cost, basisin setting tuition fees turns out merely to shiftthe discussion from what is a fair price to whatis a fair allocation of costs. To neither questionwill there be any answer that commands generalsupport.The whole doctrine, however, should be rejected even if agreement on a proper allocationof costs to students could by some magic bereached. At the heart of the matter is a decisivequestion: why is it improper for the University,which operates at a deficit, to charge at least2The concept of "deficit" in the case of a private university is not the same as deficit in the case of a businessenterprise. In the university situation the concept mustbe viewed broadly. For purposes of this statement, theconcept of deficit is that normal anticipated revenuefrom all sources is insufficient to pay for total expenditures. A deficit, in this sense, may be made up for byincreased unrestricted gifts from donors or withdrawalsfrom endowment that would not otherwise be authorized but for the necessity of covering the deficiency.60some students more than the costs associatedwith their attendance? Is it wrong, in otherwords, for the institution to hold down the sizeof its deficit by collecting tuition revenues inexcess of allocable costs, however computed?Pursuing such a course, it is submitted, producesno injustice. The University in no sense is in theposition of a monopolist; the fact is that whatever fees it charges must stand the test of marketcompetition with other educational institutions.The fees it can obtain reflect the relative valuesthat many individuals place on the goods andservices available to them in the market. If inthe process of attracting the quality of studentit seeks, the University were to hold down itsprices below market levels, a few fortunate individuals who became the beneficiaries of theaction would end up saving more or consumingmore of other things, while the University suffered a diminution in its resources. Can it seriously be contended that this outcome would represent an advance in fairness? From the viewpointof society, considering the important contributions made by the leading private universities,such an outcome seems less than appealing.IVWhy single out tuition in an effort to relieve thefinancial pressures being experienced by theUniversity?The often-made point that a university is theseller and students are the buyers of education isnot only simplistic but is of no assistance inmeeting the query. Even if the asserted relationship were realistic, there is no reason to believethat a private not-for-profit organization, especially one that depends heavily on support of acharitable nature, ought to aim at maximizing itsnet income along the lines appropriate for abusiness enterprise in a market economy. Thepurpose of the University is not to make a profit(as unlikely as that outcome may be) nor tominimize its loss as defined in business terms.Were the University "to cherish such a goal, notonly would many financial supporters be sure tolose interest, but its general reputation as a greatseat of research and learning might suffer. Theeducational venture is much more than a buyer-seller relationship. Additional insight is to begained by viewing a university as an academiccommunity in which the students are an important part. Through formal higher education students do achieve a transformation that is ofpersonal and perhaps economic value to them; they also, as students, do contribute to the advancement of knowledge in society and tocreating the atmosphere in which the transformation itself can occur. In this sense students areproductive factors in the educational process aswell as consumers.The appropriate response to one who questions looking to tuition as a means of ameliorating economic pressures must accordingly take ina wider horizon. Given the financial facts of itslife-, the University must continue strenuous efforts to increase its revenues from all sources(including tuition fees), provided only that thenew funds are not subject to use limitations thatare incompatible with its tenets and goals. Theseefforts must include obtaining more gift support,improving investment performance of the endowment, and locating more government funds onterms compatible with the over-all program ofthe institution. At the same time the Universitymust intensify its efforts to eliminate unnecessary costs and prune peripheral activities whosecontributions to the vitality of the place do notjustify their continuance.VWould it not be feasible for the University toreduce tuition charges without increasing thedeficit merely by enlarging the student body?This is a question to which the general formulation of tuition policy is not addressed becausesize should be only a collateral aspect of tuitionpolicy. The number of students clearly has aprofound bearing on the character and tone ofthe institution. For that reason the matter ofsize should be determined directly, on the basisof all relevant considerations, and not be treatedsimply as a by-product of tuition goals. Size andtuition charges are of course related. At a particular tuition level the University might not beable to attract the number of qualified studentsit had independently determined to aim for. Insuch an event it would be necessary to decidewhether the institution would be better servedby accepting a lower enrollment or by adoptinga reduced tuition level.The importance of not letting the need fortuition revenues dictate size policy should beunderscored. At present this concern is unrealistic. No one now argues that a combination ofhigher tuition fees and a smaller qualified studentbody would enlarge gross tuition revenues; andevidence is lacking that a major reduction intuition would increase gross tuition revenues byattracting a much greater number of qualified61students. But even if revenues could be so improved, a change in the size but not the qualityof the student group would have many far-reaching effects on the University— effects on operating costs, on educational design, on living conditions, and on the intellectual climate— whichin themselves might be undesirable. These consequences should be weighed carefully on theirown terms. It goes without saying that an expanded enrollment at reduced tuition would beall the more unacceptable if it were achieved bylowering the average quality of the student population.VICould not the University reduce tuition chargeswithout enlarging the deficit by cutting back onvarious expenditures or by increasing incomefrom other sources?Although an affirmative answer to this question is formally correct, approaching tuitionpolicy from the implied perspective would go fartoward undermining the strength of the institution. This is not to deny that the scope of itsprograms, the size of its faculty, the extent of itsstudent facilities, and the dimensions of its physical plant are all directly tied to the financialresources available to the University. It wouldbe folly, however, to consider contracting oreliminating various activities specifically in orderto reduce tuition levels. Determining such matters as faculty salaries, research expenditures,new construction projects and building maintenance budgets are difficult and sensitive tasks.Use of a large portion of available funds is restricted by grantors or by past commitments. Areduction in a wide variety of expenditureswould in no sense enable the University to preserve its over-all budgetary condition while lowering tuition.Despite such considerations, some in the academic community continue to argue that it isunfair to demand high tuition fees when largesums are spent by the institution for purposesnot to their liking. The ultimate ground for rejecting this contention should be clear. Theartistry required to achieve and preserve excellence in the University cannot be divorced fromthe skill entailed in allocating funds to get themost out of whatever financial resources areavailable (as well as to best attract additionalresources). Tuition policy should be set in thelight of this challenge— not the other way around. VIIShould tuition charges be pegged at a single levelthroughout the University or should there be differences in fees among the various academicareas?For more years than not the general practicewas to maintain one regular fee level for all partsof the College and another regular level for allpost-college areas. Recently the single standardfor graduate work again has been somewhat relaxed and seems to be under mounting criticism.The vital issue is whether moving away from asingle standard represents sound policy.The case for diversity is anchored in economicconsiderations. The regular fee is the full farefor tuition in the sense that no student is calledupon to pay more. Due to significant differencesin supply-demand conditions, the sustainablefull-fare varies substantially from one area toanother. A full-fare level that serves to maximizetuition revenues in one area may well cause embarrassment in other areas, either because it istoo high and results in too few qualified studentsor because it is too low and results in loss ofpossible tuition revenues from those areas. Onlydiversity in regular fees permits proper economic adjustments to occur.This economic argument for diversity is notbased on the notion that total or marginal costsper student differ among the academic areas inthe University. Such costs indeed vary greatly,and in some high cost areas* it might be goodpublic relations to inform students and others ofthe full institutional outlay associated with particular patterns of study. But these outlays arelargely determined on a basis independent of demand conditions. The economic justification fordiffering fees is that, given whatever capacitiesexist for enrollment of students, demand conditions applicable to the various academic areasdo differ and therefore a variety of price levelsis called for in order to carry out the recommended main directive of tuition policy.A uniform standard tuition charge has theadded disadvantage of restricting experimentation in changing fees. A judgmental error inmodifying the uniform rate can have disturbingrepercussions throughout much of the University,while an error in setting the fee for only one areawill tend to have immediate consequences forthat area alone. This localization of detriment,along with the presumably superior knowledgeeach area possesses (or should possess) regardingits own market conditions, suggests that considerable weight should be attached to the rec-62ommendation of each area as to the appropriatelevel of its own standard charge.The counter-argument for a uniform tuitioncharge is based not on economic grounds but oninstitutional concerns. Although these are expressed in numerous ways, the underlying themeseems to be that differential rates, especially ifcoupled with considerable autonomy on the partof areas in setting them, will tend to cause theUniversity to become less cohesive and less balanced. The fear runs in two directions. Someareas, operating under favorable demand conditions, will be able by raising tuition to increasetheir relative command over resources and thusbe in a position to enlarge their faculties andresearch programs (and perhaps increase salaries).Other areas, operating under unfavorable demandconditions, will seek to preserve or even augmenttheir student populations by lowering tuitionfees. Having "rich" areas and "poor" areas is not,the contention goes, a sound basis for maintaining the over-all strength of the institution. Theseworries are heightened by a related suspicion. Anability to vary tuition fees will further encouragesome areas to indulge in the undesirable practiceof admitting more qualified students merely toincrease tuition revenues, without giving properconsideration to the other effects of enlarging aparticular student group on the campus.While such concerns are no doubt deeply felt,they very likely rest on some misapprehensionsabout the dynamics of allocating institutional resources. In every private university there are"rich" and "poor" areas, and that has alwaysbeen the case at The University of Chicago. Partly this is a result of donor preferences; partly itcan be traced to the persuasiveness of administrative heads in obtaining funds for certain areas;partly it is due to past events. To a considerableextent, however, it is also based on recognitionby the central administrative authorities thatareas differ both in their strengths and in whatthose strengths can contribute to the over-allwell-being of the institution. An important aspectof area strength is the power to attract qualifiedstudents. The budgetary process not only mustsomehow accommodate all these considerationsbut must also prevent any one of them fromseriously distorting the ultimate allocation offunds. Although academic strength and fund-raising ability must be reckoned with as highlysignificant considerations, they must not be elevated to being the controlling factors in makingthat decision. Academically weak areas may needto be strengthened; new areas may need to be developed; separate areas may need to be combined or redefined— these and a host of otherfactors are to be taken into account. A wiseallocation of funds must, in short, always stemfrom a comprehensive and somewhat detachedconception of the University as a whole.Multiple tuition charges, determined withheavy reliance on area recommendations, verylikely would bring some new strains into theprocess of allocating funds. But there is littlereason to think that such tensions would differin kind or intensity from those long experiencedunder a uniform fee structure for graduate instruction. In any case the central administrativeauthorities must ultimately determine the composition of the budget. They have the ultimateresponsibility to shape the University; and thestrength of the University will always be associated with the capabilities of the central authorities. Ironical though it may seem, much of theirproperly treasured maneuverability is protectedby the very complexity of the process and thegreat variety of factors that rightfully have aplace in influencing them as decision-makers. Anadded factor in the mixture— in this case multipletuition rates— might well be a blessing in disguise.Whether this prospect proves correct, a planof encouraging each academic unit to be enterprising in recommending its own tuition level islikely to produce greater total revenues for theUniversity. An increase in over-all revenues is initself an unmitigated good. In a very real sense,moreover, what is good for the institution as awhole is good for its parts. The best strategyfrom this viewpoint might well be a combination.Considerable latitude in tuition matters wouldbe allowed those areas desiring to be enterprising,while a uniform level for all others would befixed by the central authorities.One administrative point should perhaps benoted about this possible resolution of the pluraltuition question. Plural fee levels will presentoperational problems growing out of the soundeducational policy of allowing students to takecourses concurrently in different academic areas.But such problems seem manageable. It is onlynecessary (1) to adopt the principle that thegoverning fee is that which pertains to the academic area in which a student is registered,(2) to use the average of applicable fees where astudent is registered for a joint degree program,and (3) to establish a separate standard chargefor all students-at-large.A somewhat more difficult hurdle in implementing a multiple tuition program is deciding63what parts of the University are to be regardedas separate units for this purpose. Clearly theCollege and each of the professional schoolsought to be so treated. The case for viewing eachof the graduate divisions as separate is almost asstrong. Until differential pricing has been testedon these levels, however, prudence would suggestnot extending it to each graduate department orto divisions of the College. At this time it isdoubtful whether multiple pricing could be successfully adapted for use at such echelons in theUniversity.Because diverse tuition fees might give rise tomisunderstandings and uneasy feelings inside theacademic community, it is important that specialcare be taken in presenting or managing any suchprogram. No implications or conclusions aboutthe intrinsic worth of an educational activityought to be drawn from student enrollmentfigures; nor is there any single or easy way ofcalculating the contribution made to the totalacademic community by investment in any particular activity. Nevertheless there is reason tobe concerned that diverse tuition fees may bemistakenly viewed as an official valuation of thework offered in various parts of the University.Some will contend that the institution is publiclyannouncing that some of its parts are better thanothers; and it will be thought that some membersof the faculty are acting as though tuition differentials indeed had that significance. The harmony and integrity of the place could easily beimpaired by such reactions. One way to minimizethese concerns is to emphasize, again and again,what a diverse tuition policy is in no sense meantto impeach: differing market conditions in themselves say nothing about the quality of any educational process or its value to the University.VIIIRegardless of whether tuition charges are setuniformly for all post-college work or set differentially for each academic area, should the feevary with the financial resources of the particularstudent?This question usually is phrased in terms of agraduated fee schedule. The charge would bestepped upward, depending on the income of thestudent or his immediate family. Under someproposed versions of the "progressive" fee plan,the tuition charge would be predicated upon income as defined for federal tax purposes.Only one or two colleges have begun to use agraduated fee program and the results are not yet in. The main hurdle to assessing any suchplan is that the arrangement is not to be compared merely with a flat tuition fee, but ratherwith a flat charge as modified by aid in the formof scholarships.This is the only useful comparison because agraduated fee plan either is intended as a substitute for the traditional tuition scholarshipprogram (or at least is designed to dovetail withit in the sense that the amount of any scholarshipaid is to be determined in light of the fee levelapplicable to the particular student). Thus, thereal question is whether a graduated fee plan isbetter than a traditional scholarship program.Apart from eye-catching rhetoric, which mightimpress some potential applicants, the only advantage of graduated fees seems to consist instating publicly and mechanistically the terms onwhich reductions from the top charge will bemade. There might be some virtue in trying tominimize the discretionary role of administratorsin awarding tuition subsidies or discounts. However, this same goal could be achieved with lessfanfare merely by maintaining a flat tuitioncharge and announcing in advance the incomecriteria that will be applied in awarding scholarships.On the other side of the ledger is the loss offlexibility in granting offsets to regular tuitioncharges. Although income has become the mostimportant determinant in granting tuition dispensations in many academic areas, it surelyought not to be the exclusive consideration. Putting aside an applicant's academic achievementand promise, such economic factors as abilityto borrow funds, family wealth (as contrastedto income), future earnings prospects, and accessto other means of defraying tuition costs mightwell be assigned a part in assaying the totalsituation. The notion of "income" is itself quiteambiguous, and the version developed for federalincome tax purposes is surely not the most fittingfor administering a scholarship program. An additional problem in relying on income is thenecessity of deciding whose income is to becounted in measuring an applicant's potentialresources. All these uncertainties— plus the importance of taking account of the academic aswell as economic factors— strongly tilt the scalesin favor of retaining a great deal of leeway indifferentiating among individuals in pricing education at a private university. Both fairness andthe goals of the institution will be promoted byconsidering each case on its own merits. Administrative discretion in this endeavor is far from64being an unmitigated evil.When applied to students who are financiallyable to pay the regular flat tuition fee, an attempt to graduate charges in accordance withfinancial ability would be almost certain to misfire. The problems of defining and ascertainingability to pay are much more aggravated in dealing with persons in the upper income and wealthranges. Indications are lacking, moreover, thatthe well-to-do in our society look upon the pricefor education as being somehow different fromthe prices for other goods or services. They maybe expected to react adversely to paying a tuitionfee higher than that charged others whom theyregard as comfortably situated. In this connection it is to be remembered that the progressiveincome tax is not a favorite of the wealthy nor isit easily enforced.It is now appropriate to turn from gross tuition to net tuition. The concern shifts to suchmatters as identifying various types of offsets(such as scholarships and financial assistance)that enter into the calculation of net tuition,choosing among alternative types, and establishing criteria for awarding offsets among competingstudents.IXWhat items should be treated as offsets to grosstuition charges?It is generally understood that a scholarshipaward made by the University out of its ownresources is a reduction in tuition charges. A fulltuition scholarship is equivalent to a zero tuitionfee. There is no economic difference betweennot charging a student any tuition fee at all andrequiring a student to pay the full tuition feewhich is then refunded to him immediately.Other types of benefits to students, however,are not so easily classified. The University doescharge less than full-cost prices for housing itsstudents in residence halls; it does charge less-than-market rents for occupancy of certain married student housing units; it does provide transportation services at less-than-cost; and it doesemploy students at rates that sometimes differfrom the market value of services received. Thesearrangements affect the cost of attending the institution. For purposes of analyzing tuitionpolicy it nevertheless seems practical to ignorebenefits of this type which, generally speaking,are not designed to effect individualized adjustments to the regular tuition fee schedules. Instead they are primarily concerned with carrying out other University policies— such as providingamenities of life, protecting against crime, fostering an environment congenial to those seeking aneducation, or encouraging students to live nearbythe places of instruction and research.More difficult is the question whether grantsof cash aid are to be treated as a component oftuition. These stipends apparently are never givenby the University in place of a scholarship, butare only paid on top of a full tuition remission.One might regard such stipends, whether theyare labeled fellowships or something else, asnegative tuition: in effect the recipient is paidto attend the institution. On another view stipends might be regarded as offsets to living andeducation costs other than tuition, so that theyserve as substitutes for reduced charges on suchitems as housing and food. (Indeed, not long agoany cash award in the College over full tuitionhad to be applied to room and board in a University residence hall.) Unlike subsidized goods furnished to students in kind, however, stipends aredesigned primarily to facilitate enrollment ofparticular students regarded as especially desirable. Although the stipends may tend to promotegeneral university policies to diversify the student body, these grants serve directly to financethe education of the recipients. On this basis, itseems more realistic to bring stipends under therubric of tuition.Similar reasoning should apply in the case ofstudent loans made by the University at lower-than-market interest rates— meaning rates that donot fully take account of all relevant risks aswell as the prevailing return on riskless investments. Unless it is borne by the government oranother third party, the spread between marketrate and bargain rate represents an expenditure ofresources in that this difference would otherwisebe available to the University. To the extent thatbargain loans can be said to finance payment oftuition fees, the spread is the equivalent of areduction in those fees. In this respect, it shouldbe noted, a tuition deferral is equivalent to atuition loan at a zero rate of interest. To theextent that bargain loans can be said to defrayother costs (including living costs) associatedwith education, the spread is equivalent to student assistance through cash payments. If, assuggested, the latter is properly viewed as negative tuition, the whole of the spread in studentloans should be looked upon as part of thetuition structure.While the line between net tuition and othercosts of attending school may not be sharp, these65observations locate it adequately for purposes ofdiscussing tuition policy.XShould the recommended tuition policy— that ofseeking the equilibrium that will produce maximum revenues without causing the quality ofstudents to fall below the very high level requiredto maintain the University as a leader in education—apply in establishing net tuition levels?The immediate question is easily put to rest inlight of the previous discussion. Tuition policymust erribrace tuition offsets as well as tuitionrevenues. Reflection will quickly reveal that thecase stated for maximizing tuition income (whilemaintaining the quality of students) is to be readas the case for maximizing net tuition revenuesafter deducting all offsets. Likewise, the casestated for allowing or encouraging a diversity oftuition fees among the academic areas is to beread as the case for diverse net tuition levels.Indeed, the University has long had widely diverse average net tuition level's among academicareas, reflecting market conditions. The essenceof the matter can be summarized very simply.The use of tuition offsets enables the Universityto obtain differential net tuition payments fromstudents in the same academic area. It is thisflexibility that permits the University to takeproper account of the academic promise andthe financial circumstances of particular individuals in competing with other institutions fordesirable students.Because the competitive situation differs markedly from one academic area to another, thebest combination of tuition fee level and tuitionoffset level will also vary. In some areas a relatively higher fee and a higher average offset willbest carry out the recommended goal of tuitionpolicy; in others a combination of a relativelylower fee and a lower average offset will bestserve to advance this end; in still others it mightbe best to charge relatively a higher fee and dowithout tuition offsets altogether.In short, tuition policy should be administeredwith net tuition in mind. There is, for thatreason, nothing to be said on how much theUniversity should provide in tuition offsets unlessthe matter is coupled with treatment of grossrevenues from tuition fees.XIShould all scholarships and other tuition offsets not specifically funded by outside sources beawarded by the University solely on the basis ofthe recipient's economic need?Benefits exclusively for the needy is a mostappealing slogan. As an operating principle, however, it would not be in the best interests of theUniversity, since it is directly at odds with therecommended tuition policy of attracting thedesired number of qualified students while producing the maximum tuition revenues.The "need only" principle would be compatible with this touchstone policy if all universitiescharged approximately the same tuition fees, ifthe cost of living were measured on the samescale by all institutions, and if all adhered exclusively to a common criterion of need in awardingscholarships and other tuition offsets. These conditions do not now exist and are most unlikelyto evolve in any foreseeable future. Under prevailing circumstances, "price" is a major factorin the competition for students among thenumerous state and private institutions offeringcomparable academic programs. Were The University of Chicago to adopt the "need only"principle it would lose qualified students toschools giving assistance independently of needand to schools that defined need more generously. In some academic areas the resulting deterioration of student quality could be grave, evendevastating.In this light the question, restated in modifiedbut realistic form, is: how should the Universitydetermine its mix of assistance based on needand assistance based on academic considerations?A choice must be made, though perhaps by default, whenever there are both needy and non-needy qualified applicants (or students alreadyenrolled) in an academic area who will not cometo (or remain in) the University unless they areawarded scholarships or other assistance. If allthese individuals were of equal caliber, the casefor allowing need to control the awards would bevery strong indeed. The situation becomes troublesome where the non-needy contenders areclearly superior. Then, the issue is whether theUniversity should compete for these superiorstudents at the expense of not subsidizing themore minimally qualified contenders who areneedy. Another form of this same issue ariseswhere, given a fixed number of student spaces,the qualified students are willing and able to paytheir own way but the superior students holdout for tuition offsets.The awkwardness of dealing with this issuemight encourage some to deny that it is genuine.66Their claim in effect is that no reliable methodhas been found for distinguishing between superior and qualified individuals at an institutionthat, under any circumstances, admits only veryhigh caliber students. But the issue will not goaway so easily. In many areas of the Universityit is said that a significant difference among thequalifications of applicants for assistance can beestablished. Where this condition is thought toexist, and where it appears that the quality differential will likely be felt in the level of teaching, or classroom discussion, or research work orany work or any other aspect of the educationalenterprise, the University ought to favor the candidate with the greater potential. Even if hisrefusal to pay his own way may not be commendable in the eyes of some, the strength ofthe University relates to his intellectual capabilities rather than to his point of view on exploitinghis market position.XIIOn what basis should the University select amongalternative types of offsets to tuition fees, assuming that some predetermined total amount isavailable for that purpose?The avenues for effectuating tuition offsets include scholarships, tuition deferrals, stipends,low-interest loans made by the University, andlow-interest loans made by other lenders as aresult of University subsidies. Included also areloans (made either by the University or outsidelenders) having special features that at presentcannot be reflected in an interest rate. Illustrativeof these features are rights to forgiveness of thedebt or postponement of repayment under specified conditions. The question is one of priorities,dealing with how the University should rankthese various devices in utilizing whatever fundsit has allocated to accomplishing a reduction ofregular tuition charges for particular students.The most suitable basis of choice is to maximize the effectiveness of these earmarked dollarsin carrying out the tuition policy. This meansthat, all other things being equal and operatingwithin a given budget for student aid, everyeffort should be made to utilize the funds so asto provide students with the greatest total dollaramounts to cover the underwriting of their education. Under this principle the priorities, statedin reverse order, would in general be: (1) stipends, (2) scholarships, (3) deferred tuition,(4) low-interest loans by the University, and(5) low-interest loans by other lenders subsidized by the University. The explanation for this ranking is fairly clear. Most costly to the University isa dollar of stipend because it is disbursed immediately, thereafter produces no income for theinstitution and never is to be repaid. A dollar ofa scholarship award is equally costly only if itis assumed that the recipient could be replacedwith an equivalent full-tuition-paying student—which may or may not be a realistic alternative.Low-interest loans made by the University aredefinitely less costly; repayment is required andthe institution is out-of-pocket only the differential between the interest rate actually chargedand the rate that would obtain on the market inarm's length transactions. (Deferred tuition is ineffect a loan at no interest, and is more costlythan a loan at a low rate of interest.) Subsidizedlow-interest loans made by other lenders areprobably even less costly to the University inthat a commercial lender is likely to be appreciably more efficient in processing loan transactions. In a competitive market the cost savingthrough efficiency will likely equal or exceedthe subsidy required by the lender for handlingthe loans.The main thrust of this "dollar stretching"approach is to reduce direct student assistanceand scholarship aid and to- encourage students toborrow funds to finance their education. Movingin this direction would be in keeping withchanges already under way in society generally.The availability of low-interest loans made orguaranteed by the federal government (and tosome extent by state governments) is Increasingnoticeably. Banks and other private lending institutions are more willing to lend money to students provided the schools they attend bear partof the risk. Foundations have begun to experiment with subsidized student loan programs.These and other happenings are strong testimonythat, in the years immediately ahead, greaterreliance will in general be placed on loan arrangements of one sort or another to finance studentsin obtaining their formal higher education. Students today are often reluctant to accumulatelarge long-term indebtedness. This resistance islikely to decrease as the practice becomes morecommon and better understood.Not much can be said at this time aboutrecently adopted or proposed plans for incorporating various relief provisions in loans. Theseescape hatches take the form of specifying conditions under which, in whole or in part, therequired principal repayments or the requiredinterest payments will be extended, modified, or67cancelled. The common denominator is that,whatever agency carries the loan (and whether itis insured by government), the school undertakesto bear the actual cost of providing the relief.On initial analysis there is reason to think thatsuch use of funds should rank high on The University of Chicago's priority list, especially considering that its commitments will not eventuatein disbursement of money until long in thefuture. It must be conceded, however, that thetotal cost of these underwritings is uncertainand that prediction is difficult because experience with the arrangements is wholly lacking.Much more hazardous are loan plans, optionalon the part of students, that key the size of repayment installments to the debtor's income atthe due date. Several features of these plans aredistinctly unattractive from the viewpoint of theeducational institution. The well-to-do studentsand the optimists who anticipate having relatively large incomes would not find it in their self-interest to elect the variable repayment option.If they refuse to participate in the plan, aheavier burden of repayments necessarily has tobe imposed on the others, including those whoall along expect to enter low-paying fields ofemployment. In some cases the plan might beblamed for discouraging students from seekingmore remunerative employment, the point beingthat the variable repayments provision couldhave the disincentive effect commonly associatedwith a tax (especially a progressive levy) on income. The plans, moreover, require that participants submit evidence of their income in makingrepayments (unless they pay the ceiling amount,assuming the plan provides for one). Experiencereveals that students are sometimes reluctant tosubmit income data to a university. It is a goodguess that alumni, away from campus for manyyears, are likely to be even more hesitant on thisscore. Some might even be slightly suspiciousthat the financial information would somehowend up in the development office. Sensitivitiesmight be all the higher if copies of one's federalincome tax returns had to be attached.In looking at the panoply of student aid devices as a matter of priorities, it is important tobear in mind that the various alternatives are notcompletely interchangeable. Some individualsmight seek a scholarship but will not accept alow-interest loan. Others might seek a stipend inaddition to a full scholarship. Still others mightseek to borrow funds from the University butnot from an outside lender, believing that a commercial lender would be less generous if repay ment becomes impractical. Where students cannot be persuaded to accept an arrangement otherthan their first choice, a priorities doctrine hasno room to function. In such instances the University is confronted with a situation not unlikethat faced when a scholarship or other tuitionoffset is sought by one who has the wherewithalto pay for his own education. The decision, tobe taken against the background of competition for students, once again ought to turn onwhether the particular student is clearly superior,from the University's viewpoint, to other students who are amenable to receiving aid in aform less costly to the institution. In such circumstances he should not be awarded financialaid unless there is a persuasive case showing thathe is superior. A parallel approach is likewiseappropriate in choosing between those qualifiedstudents who, not being able to carry additionaldebt obligations, seek scholarships and thosequalified students who are willing and able totake on additional debt burdens. Stretching dollars in these situations is, after all, an exercise infairness as well as prudence.XIIIDoes the recommended tuition policy discriminate unfairly against the poor, particularly thosehandicapped by cultural disadvantage?The answer to this question is no. But in orderto reach it, the related matters of admissionspolicies and tuition policy must be disentangled.A firm consensus exists at the University thatit is unwise to increase the proportion of studentsfrom culturally disadvantageous backgrounds byjudging them on lowered admissions and performance standards. There is an equally strongconsensus that, in terms of its own character andaspirations, the University would definitely benefit by increasing, in numerous academic areas,the proportion of environmentally handicappedstudents who are qualified under the generalstandards currently being applied. A more precisestatement of the foundation for this positionmight be helpful. Diversity in the student population is important in fulfilling the goals andcarrying on the educational functions of theUniversity. Therefore, under some circumstances,an individual from a certain economic, cultural,ethnic, or racial background might be regardedfrom the University's vantage point as beingmore desirable than others who in all otherrespects are equally qualified.Given such a directive on admissions, the68tuition policy here being advocated would notfurther handicap the poor or the culturallydisadvantaged. They would, under appropriatecircumstances, be viewed from the University'sstandpoint as having the potential to raise theaverage quality of the student body and accordingly would not be crowded out under therecommended system of priorities. These students, it should be noted however, would not befavored over others who, taking all factors intoaccount, are clearly more promising. To preferthe better rather than the best— for reasons extraneous to the functions of the institution-would not be in harmony with the character ofthe University."Who is the expert in perfecting the human andsocial qualities? I assume from the fact of yourhaving sons that you must have considered thequestion. Is there such a person or not?""Certainly," said he."Who is he, and where does he come from?"said I, "and what does he charge?""Evenus of Paros, Socrates," said he, "and hisfee is twenty guineas."Plato, The ApologyWhat role can and should tuition play in theeconomics and financing of higher education inthe United States? Central to my approach toMarc Nerlove is Professor in the Department of Economics. His essay is reprinted with permission from theJournal of Political Economy, Volume 80, Number 3,Part II, May/June 1972. Copyright 1972 by The University of Chicago.The work on which this essay is based was supportedunder a grant from the National Science Foundation toThe University of Chicago and under a grant for thestudy of the economics of population and family decision-making from the Rockefeller Foundation. Someof the work was done during August 1971 when I wasa visiting professor at the University of British Colum- XTVTaken together, these positions constitute themain outlines of a tuition policy for The University of Chicago. That policy admittedly has itsdrawbacks, but the ideal is not within reach. Itis a policy that comports with the financial position of the University while being faithful to theUniversity's character and aspirations. No otherpolicy is likely better to serve one of the fewdistinguished American private universities.Walter J. Blum (Chairman)Julian R. Goldsmith Ralph M. LernerJohn E. Jeuck Robert H. PalmerWilliam H. Kruskal Karl J. Weintraubthis question is the belief that tuition is not now,but should be, a price in an economic sense. Bythis I mean that tuition should perform the function of rationing the available supply of highereducational services and of allocating scarce resources between the sector producing higher educational services (including graduate educationand basic research) and the rest of the economy,as well as among different types of institutionsof higher education, and even among differentparts of the same institution. To say this doesnot constitute a case against subsidizing highereducation in a variety of ways, however. I believe that, at a given level of tuition paid by indi-bia. John Perry Miller graciously supplied a compendiumof materials on the Yale Tuition Postponement Option.For helpful preliminary discussions on matters coveredby this essay, I am indebted to Robert W. Fogel, Edward H. Levi, Walter Y. Oi, and T. W. Schultz. RobertMcC. Adams, Denis De Tray, Richard B. Freeman, Stanley Fischer, Robert J. Gordon, Zvi Griliches, D. GaleJohnson, Harry G. Johnson, Reuben A. Kessel, EdwardH. Levi, T. Paul Schultz, T. W. Schultz, George J. Stigler,Virginia D. Thurner, and James Tobin have commentedextensively on an earlier draft. Needless to say, none ofthese generous people is responsible for such errors asremain, and many do not share the opinions expressed.ON TUITION AND THECOSTS OF HIGHER EDUCATION:PROLEGOMENA TO A CONCEPTUAL FRAMEWORKBy Marc Nerlove69vidual students, the demand for higher educational services will tend to be below socially optimal levels, and that, at the "price" representedby a given level of tuition, a less than sociallyoptimal level of resources may be devoted tohigher education unless society also subsidizesbasic research.Loan programs, especially the proposed "Educational Opportunity Bank" and the "Yale Tuition Postponement Option," in which there isso much current interest, are considered in a sequel to this paper.1. Is Tuition a "Price" in an Economic Sense?Should It Be?The role of prices is, on the one hand, to rationavailable supplies of a good among those who demand it and, on the other, to allocate scarce productive resources to the industry and /or firmswhich supply the good. Does the tuition chargedby American colleges and universities for undergraduate and other post-secondary educationnow perform, or could it conceivably perform,the allocative functions of a "price" in aneconomic sense? Since the time of Adam Smitheconomists have been aware that, under fairlystringent assumptions to be sure, free marketsand prices which perform these functions willlead to socially optimal results. The chief sourcesof difficulty which arise have to do with so-calledexternalities, with uncertainty which cannot becovered by insurance, and with certain intergen-erational welfare comparisons which may not beadequately made in current-generation markets.Such barriers to effective functioning of marketsand prices are especially important in analyzingthe demand for and supply of products producedby the higher educational sector.The higher educational sector in the UnitedStates is exceptionally complex and produces abewildering variety of products. This sector consists of many different types of "firms," whichrange from proprietary vocational establishmentsand two-year junior colleges to four-year liberalarts colleges and the great public and private universities with their penumbra of associated professional schools and research institutes. Amongits products are narrowly directed vocationaltraining and undergraduate education— varyingin quality from little better than what is offeredin many secondary schools to the equal of anything offered anywhere in the world in any era-graduate education (also of varying quality),applied and basic research, and public service andextension.1 Despite the underlying complexity of reality, the analysis that follows makes use ofthe concept of "units of undergraduate education" of constant quality, as if it were really possible to put a credit hour at Podunk State on acomparable basis with one at The University ofChicago. The special problems posed by graduateand professional education are neglected, although the framework can and should be extended to cover these types of educational services.The major public and private universities ofthis country produce research and graduate education jointly with undergraduate education. Toa lesser extent this is also true of four-yearliberal arts colleges, although the nature of theresearch done is a little more difficult to define.Many institutions of higher education produceno research of any discernible value at all, although, in some cases, other social functions maybe performed. For such institutions, it is relatively simple to assess the extent to which tuitionand student fees cover the costs associated withthe production of the good conveyed; in general,of course, these costs are not covered by tuitionand fees. They are supported, in part, throughgifts, through income from endowments that arethe result of past gifts, and by the general public,directly through taxes or indirectly through taxexemptions. For institutions in which undergraduate education and research and graduateeducation are jointly produced, it is much lessclear how to measure the extent to which thetuition and fees charged undergraduates (apartfrom student aid) fail to cover the costs of providing that education. Many would regard thecommodity supplied by these institutions asqualitatively rather different from that suppliedby institutions which produce nothing else. Thatthey fail to cover costs, however, is the inescapable conclusion reached by all who have examined the problem. The tuition and fees chargedby public institutions are quite consciously setbelow any plausible estimate of what it costs toeducate a student; the same is as true of high-quality liberal arts colleges as it is for the majorprivate and public universities. The rationing, byselective admission, of places in both public andprivate institutions to those who seek attendancesuggests that tuition does not cover the costs ofproduction of the quality of commodity de-An impression of the complexity of the industry, itsincredible diversity of products, and the role which thesector plays in American society as a whole can befound in Jencks and Riesman (1968); Bolton (1969)presents a more quantitative view.70manded, nor does it serve the traditional economic function of rationing the available supplyamong those who would consume it or of callingforth additional supplies to meet these unmetdemands.2 Moreover, because a substantial number of universities charge students substantiallyless than cost, they may restrict entry to theindustry, reduce in some dimensions the diversityof the products that are produced, and perpetuate misallocations of resources among the"firms" comprising the higher educational sector.3Tuition and fees are not the only costs to astudent of obtaining an undergraduate education.Such an education is, in fact, a commodity, partly capital good and partly consumption good,which is produced by a student using his owntime and a variety of services supplied by thehigher educational sector. Thus, the demand forthe services of the sector is a derived demand,that is, a demand for a factor of productionrather than for a good valued for itself; and thisis the case irrespective of whether an undergraduate education is regarded primarily as aninvestment in human capital or as somethingwhich contains significant quantities of currentconsumption. The costs associated with the timeinvolved are the earnings foregone because thestudent must devote much of his time to hisstudies, leaving less for paid work and reducinghis wage during those hours he does work belowthat which he might have received in full-timeemployment. The foregone-earnings componentis not easy to measure because employment opportunities differ by region, race, sex, and other2 O'Neill (1971) confirms, for the institutions includedin the U.S. Office of Education surveys, Bowen's (1968)findings for major private universities, namely, that,even netting out expenditures for "organized" (separately funded) research, instruction and departmental research costs are rising far faster than tuition and fees.O'Neill's study is notable for her attempt to includecapital costs (rarely taken into account) and to makesome adjustments for quality differences in output. Thepapers by Jenny and Wynn (1969a, 19696) attempta detailed analysis of expenditures in a group of small,private, four-year liberal arts colleges and provide further confirmation, although they do not allow forcapital expenditures.3 Harry Johnson points out that diversity may be increased in some directions because the products produced are not subject to market discipline. The point isthat once the market tie is broken, decisions as to whata university or universities will produce lie with a smaller, more homogeneous group than would otherwise bethe case. Imposition of the particular set of standardsand values of this group on society as a whole must tendto reduce freedom of choice. characteristics, and because many students dowork part-time, thus reducing the amounts foregone. There are other costs associated with goingto college that are not due to tuition and fees norto the time involved; for example, the extent towhich board, room, and incidentals exceed thatwhich would have been spent had the studentnot sought a higher education is a cost attributable to going to college. Various estimates areavailable that suggest that the foregone-earningscomponent may amount to more than half thetotal cost paid by a student attending a prestigious and expensive private institution and, ofcourse, a correspondingly greater fraction for oneattending a zero-tuition or low-tuition publicinstitution (Hansen and Weisbrod 1969, pp. 41-54; Schultz 1971, pp. 82-95, 102-15). Other expenses account for a much smaller, but presumably nonnegligible, fraction of the total (Hansenand Weisbrod 1969, pp. 49-51). Because substantial other costs, principally foregone earnings,are incurred by a student in the production ofthe human capital and consumption representedby a college education, the level of such costsand variations therein will affect the derived demand for the services of the higher educationalsector and, of course, the consumption of college education and investment in human capitalwhich takes place over time.Tuition clearly does not now function as aprice in an economic sense throughout the highereducational sector, although it may do so for asmall and rather unimportant subsector that doesnot include any of the major private or publicuniversities, or indeed public institutions of anykind of most private, four-year colleges. Whethertuition should be a price in an economic sensedepends on the role tuition might play in themarket determination of demand and supply andhow and why a market mechanism might breakdown from a social point of view.2. Tuition and the Demand for UndergraduateEducationWhat would be the effects of using tuition toration the available supplies of different undergraduate services produced by the higher educational sector among those who demand thoseservices? In particular, for a given structure andlevel of tuitions and fees, will the demand for the"products" of the higher educational sector beoptimal from a social point of view? Or will thedemand be too great or too small and will thecomposition of demand for the different types of71products be socially undesirable? Are there socially undesirable effects from using nonmarketmechanisms to allocate available supplies?If undergraduate education were purely a consumption good and if a student's consumption ofit made only himself better off and no othersin society, the problem of the adequacy of demand from a social point of view would be easyto analyze. College education is not merely acurrent consumption good, however, but, in part,a durable producers' good enabling those whohave acquired it to achieve both higher pecuniaryand higher psychic incomes. Moreover, there maybe important benefits accruing to others insociety of the current generation or of futuregenerations, benefits which are not captured bythe individual educated. In addition, the production process by which an individual transformsthe services of the higher educational sector, hisown time, and his own abilities into human capital "in place" and current consumption is, atbest, poorly understood, thus further complicating the analysis. Recent discussions of the problem among economists have emphasized the investment character of a college education, havede-emphasized the current consumption component, and do not appear to have considered fullythe issues raised by the notion that the demandfor the services of the higher educational sectoris, in fact, a derived demand for a factor of production.The Importance of Understanding the Technology of the Educational Transformation: Notusing tuition as a means of allocating resourcesamong "firms" producing different products orproducts of different qualities in the higher educational sector may lead to restriction of diversity within the sector or to diversity in undesirable directions. Moreover, since demand isrationed by other means, students themselves donot have the opportunity, assuming they dounderstand the transformation process, to combine inputs in the most productive way. A certain number of hours of class per week for fouryears, with summers off, and in such and suchcombination of subjects, may simply not be thebest way for everybody, or perhaps anybody, toachieve the desired transformation of himself.There may be, to be sure, a substantial informational gap: neither a student nor his parents norhis advisors may be in a position to judge thevarious products offered or the manner in whichthose products will interact with his particularabilities and interests. Informational gaps of this sort lead in other areas such as medical care tononmarket control by licensing and other formsof certification. It is not clear, however, that thetransformation process is much better understood by those who do the licensing (accreditation) in the higher educational sector than bythose who utilize the services produced by thesector, although it is possible that the "gatekeepers" who actually ration available suppliesmay have superior knowledge.The question, it would seem, is whether anyuseful function may be served by not allowingfull freedom of the choice of which types ofhigher educational product are purchased, evenfrom a "suitably" restricted menu. This is a question not of the level of aggregate demand butof its composition. If it is assumed that thesevere informational problems connected withthe nature of the product have been solved, thereis only one argument which I find cogent: a substantial part of what a prestigious private university or college has to offer may depend on thesort of students it can attract. Thus, the consumption of the product of such an institutionby a student with the qualities desired (say exceptional intellectual or athletic ability) enhancesthe quality of the institution's output and, hence,its attractiveness to similar students. Perhaps it isonly by charging a relative tuition too low toration the supply of places among potential occupants that the institutions can ensure an adequate supply of the sort of student who generates the appropriate externalities. Scholarshipsand other forms of student aid also may be usedfor such effects and to achieve those mixtures ofstudent types believed to enhance the quality ofthe education produced. Indeed, it is possiblethat a regime of high tuition and substantialstudent aid could simultaneously enhance boththe selectivity of the institution and its revenue,but it is doubtful that an individual institutioncould achieve such a happy state of affairs unilaterally.That a higher education, indeed any educationat all, is produced, in a sense, by each studentby means not only of the services of the highereducational sector but also by inputs of his owntime, efforts, and abilities, is an important consideration not only in the analysis of the composition of the derived demand for educationalservices but also of their level in aggregate. Opportunities for alternative employment may significantly affect the costs of acquiring a collegeeducation, especially for low-income groups. Uncertainty about the results of going to college,72stemming from the individual's uncertainty abouthis own abilities, the quality of the educationservices he may receive, or the effect of a highereducation on his own income prospects may reduce demand below socially optimal levels. Quiteapart from the compositional aspect of demand,the aggregate demand may be expected to bebelow a socially optimal level at any level of tuition sufficiently high to ration the supply ofservices provided by the higher educational sector.Effects of Imperfections in Capital Markets: Inhis now classic paper in which he made the firstsuggestion of an income-contingent loan-repayment program for the financing of higher education, Friedman (1955, reprinted 1962) pointedout that an important reason for the insufficiency of demand is a particular sort of imperfection in the capital market:Investment in human beings cannot be financed on the same terms or with the sameease as investment in physical capital. Itis easy to see why. If a fixed money loan ismade to finance investment in physicalcapital, the lender can get some securityfor his loan in the form of a mortgage orresidual claim to the physical asset itself,and he can count on realizing at least partof his investment in case of default byselling the physical asset. If he makes acomparable loan to increase the earningpower of a human being, he clearly cannotget any comparable security. In a non-slavestate, the individual embodying the investment cannot be bought and sold. ... Aloan to finance the training of an individualwho has no security to offer other than hisfuture earnings is therefore a much lessattractive proposition than a loan to finance the erection of a building: thesecurity is less, and the cost of subsequentcollection of interest and principal is verymuch greater. [1962, p. 102]Such reasoning lies behind many of the existingand proposed tuition-loan programs.4In view of the substantial magnitude of theforegone-earnings component, existing loan or4Hartman (1971) describes existing and proposed programs and analyzes their effects on existing capitalmarkets, the distribution of their benefits, and variousaspects of alternative terms and implicit elements ofsubsidy. A subsequent paper of my own will deal withincome-contingent loan-repayment programs in moredetail. loan-guarantee programs are too restricted intheir terms, length of repayment period, and sizeof loan to offer many lower- and middle-classyoung people an opportunity to borrow againstfuture earnings to the extent necessary to, achieveadequate levels of demand even for the educational services of virtually free public institutions.Some part of a college education may be purepresent consumption, or future psychic income.To the extent that it does not, on the average,result in higher pecuniary earnings to those whoobtain it, it is unreasonable to expect the capitalmarket to provide the wherewithal to finance iton terms different from those through whichother forms of consumption are financed. Butclearly a large part of a college education doesrepresent investment in human capital whichdoes result in higher financial rewards to thosewho acquire it. The degree to which rates ofreturn to this form of investment exceed thoseon other assets (apart from risk) is indicative ofthe gap between actual demand and a sociallyoptimal level. Because of the special problemswhich the riskiness of the investment in humancapital raises, however, it is not clear with whatrate of return the returns to a college educationought to be compared in order to assess the extent to which capital-market imperfections reduce demand below optimal levels.Effects of Risk and Imperfections in Information Flows: The process of obtaining a collegeeducation is not only costly but quite risky aswell. A particular student does not really know,in any precise way, the quality of the serviceshe will receive from an educational institution.Moreover, these services are only one of severalinputs to the educational transformation. Thestudent is also likely to be at least partiallyignorant concerning his ability to make use ofwhat the higher educational sector provides and,more important, of the value of the product produced by those inputs and with his own effortsand abilities. Some types of education are narrowly vocational in character; much professionaleducation at the postgraduate level in medicine,law, business, and engineering may also be of thisnature, as recent events bring home. The demandsby the economy for particular sorts of educatedpeople are constantly changing; to the extent thata college education specifically "fits" an individual for a particular profession or occupation orspecifically prepares him for postgraduate training which does so, an element of substantial riskis introduced into the calculation of the worth of73an investment in a college education.5 To the extent that a "liberal" education, a perfection of"the human and social qualities" which may beinherent in every human being, enables the student to adapt more flexibly to the economic environment in which he may find himself over hislifetime, the acquisition of an education reducesthe economic risks to which the student may beexposed. Certain kinds of education may be regarded as an "insurance policy" against unemployment, against technological obsolescence ofone's skills, and against low income. There arereally two types of risks involved in the acquisition of a higher education: (1) the risk that eachstudent faces, of not having the requisite abilitiesto benefit, and (2) the risk that what he doesacquire will not provide him with higher incomeor better employment opportunities. Even if thelatter risk were minimal, the former would stillexist.There are substantial numbers of people wholike to gamble, not only at roulette wheels andracetracks, but on commodity exchanges, onstock markets, and in the foundation of newenterprises and the production of new products.On the average, however, people are risk-averse.Higher average rates of return must be paid onmore risky, as compared with less risky, investments; people do purchase insurance and, ingeneral, they must be compensated for bearingrisk, although some groups may be willing tobear risks at lower costs than others. This hasimportant consequences for answering the question of whether the aggregate demand for theservices of the higher educational sector will belarge enough from a social point of view at anylevel of tuition that serves to ration availablesupplies. (Risk is also an important, but not theonly, reason why the market mechanism will notcall forth a socially optimal volume of research,both basic and applied.)In the absence of insurance against acquisitionof unwanted skills or lack of ability to completean educational transformation to the fullest ex-For a careful analysis of the dynamics of the marketfor college-trained manpower see Freeman (1971). Freeman (pp. 202-26) presents evidence that students dohave realistic income expectations, do consider foregoneearnings in their educational plans, and do alter theircareer choices on the basis of additional informationregarding income prospects and other occupationalcharacteristics during the course of their career preparations. This is not to say that students consider onlypotential pecuniary benefits, but rather that they appearto be quite conscious of the "trade-offs" betweenmoney income and other career benefits. tent (and thereby failure to acquire the maximum number of "units" of a college educationeven though purchasing the requisite number ofinputs from the higher educational sector), ratesof return to investments in college education willhave to be higher than to other, less risky investments." In calculating such rates of return, thecurrent consumption component should be deducted from the cost and the nonpecuniary income stream valued and added to the futurestream of money income; existing calculations donot make these adjustments and thus tend tounderstate the true rates of return to private individuals investing in college educations. Is it optimal from a social point of view that the rate ofreturn to investment in higher education shouldbe so high, or does the presence of substantialrisk suggest that, in the absence of insurance orsubsidies, aggregate demand will be deficientfrom society's standpoint? Do existing or proposed loan programs even begin to cope with thisproblem?Two aspects of risk must be considered: Someindividual risks can be pooled so that the aggregate risk is much less than the individual risk.This reduction in variance for sums of independent events is, of course, just what insurancecompanies sell. To the extent that the risks associated with investment in a college education canbe pooled away in this sense, they are privatebut not social risks; if individuals are not madeto bear them, they will, on the average, investmore in their educational transformations, andthe allocation of resources among investments inhuman capital, investments in physical capital(and knowledge), and current consumption willbe more socially optimal. Observed rates of return to different forms of investment will tend tobe closer to one another after adjustment fornonpecuniary elements. Uncertainties associatedwith the way in which time, abilities, and theinputs provided by the higher educational sectorare combined to produce the educational transformation are perhaps largely of this "poolable"character. The risks associated with society'schanging demands for skills and college-educatedmanpower are not primarily, however, of a"poolable" sort, since the individual events whichcombine to form the aggregates are not independent of one another. These are social risks whichmust be borne by someone, although not necessarily solely by the individual students who areexposed to them. Such risks do require that ratesof return to investments subject to them shouldbe higher than those on riskless investments in74order to achieve an optimal allocation of resources in the economy as a whole. But whoshould bear the risk on society's behalf and whycannot an individual risk-averse student transfersome of the risks for a price to someone or somegroup less risk-averse than he? As Tobin andPugash (1971) so neatly put it: "When bothparties stand to gain, there is room for a deal.This is the simplest and most fundamental principle of economics."6The existence of insurance markets in connection with the uncertainties associated with theeducational transformation would permit both areduction in risk due to pooling and a transfer ofrisk from those less willing to bear it to thosemore willing to do so. Unfortunately, except ina very rudimentary sense in connection with theYale Tuition Postponement Option, such insurance markets do not exist. Arrow (1971, p. 184)comments:The nonexistence of markets for the bearing of some risks in the first instance reduces welfare for those who wish to transfer those risks to others for a certain price,as well as for those who would find itprofitable to take on the risk at such prices.But it also reduces the desire to render orconsume services which have risky consequences; . . . these commodities are complementary to risk-bearing. Conversely, theproduction and consumption of commodities and services with little risk attachedact as substitutes for risk-bearing and areencouraged by market failure there withrespect to risk-bearing.Thus, the failure of society to provide a meansfor pooling and shifting the risks associated withobtaining a college education reduces demandat any tuition below a level which would besocially optimal. Why has such a failure occurredand what bearing do these reasons have on thepossibility of providing nonmarket alternatives?Not all the risks of fire and death, which itmight be socially desirable to provide a mechanism for shifting, can be shifted through insurance.Limitations on the size of policies, direct controls over those insured, insistence on elementsof coinsurance, and the bankruptcy and limitedliability laws are all nonmarket responses to theabsence of an appropriate market mechanism.At Chicago, we would regard it as the second simplestand most fundamental principle, the first being thatthere is no Such thing as a free lunch! The most important reason that the market failsto provide insurance or, when it does provide it,limits it in ways which prevent full shifting ofrisk, is the so-called moral hazard. This is thedanger that the insurance policy itself changesincentives and affects the probability of the contingency against which the insurance is sought."The problem is that the insurer, or morebroadly, the risk-bearer cannot completely definehis risks; in most circumstances he only observesa result which is a mixture of the unavoidablerisk, against which he is willing to insure, andhuman decision. . . . But the insurance policymay . . . lead to a motive for increased loss, andthen the insurer or risk-bearer is bearing sociallyunnecessary costs" (Arrow 1971, p. 142). In thecase of an investment in a college education, themoral hazard that the insured will take a job thatis not monetarily remunerative after graduationis compounded by the difficulties of assessingintrinsic abilities to begin with and the outcomeof the educational process at the end. This problem is especially complicated for women. (Pace!)Income-Contingent Loan-Repayment Programs:The existence of a severe moral hazard and thediscrepancy between the information possessedby an individual student and that which could beknown to an insurance company are presumablyreasons why the market economy has not provided institutions for shifting the risks involved,but this does not mean that the same moralhazard would exist for society as a whole, orthat at least partial insurance would not be desirable from a social point of view. Indeed, aform of coinsurance or limited liability Friedmansuggested in his seminal paper is the basis forall income-contingent loan-repayment programssuch as the Educational Opportunity Bank (U.S.Panel on Educational Innovation 1967; Shell, etal. 1968; Shell 1970) and the Yale Tuition Postponement Option Plan (Brewster 1971; Tobinand Pugash 1971; Yale University 1971): "Thedevice adopted to meet the corresponding problem for other risky investments is equity investment plus limited liability on the part ofthe shareholders. The counterpart for educationwould be to 'buy' a share in an individual's earning prospects; to advance him the funds neededto finance his training on condition that he agreeto pay the lender a specified fraction of hisfuture earnings. In this way, a lender would getback more than his initial investment from relatively successful individuals, which would compensate for the failure to recoup his original75investment from the unsuccessful" (Friedman1962, p. 103).Of course, it is only to the extent that pecuniary benefits are obtained from a college education that an income-contingent loan-repaymentscheme can contribute both to the problemsarising from imperfections in the capital marketsand to those arising from risks that cannot otherwise be shifted. To the extent that the gainsfrom a college education are in the form offuture psychic income or current consumption,neither imperfections in the capital market northose in the risk-bearing market can be correctedin this way. Moreover, the difficulty which prevents a market solution to the risk-shifting problem is certainly not removed entirely by an income-contingent loan-repayment program, especially one operated by a single university. Thosewho have poor income prospects, either becauseof self-recognized lack of ability or because theyfeel likely to choose low-paying occupations onaccount of the value they place on the nonpecuniary benefits associated with such occupations,are more likely to borrow under such schemes tofinance their educations than are those with exceptionally good income prospects because oftheir economic and social background, self-recognized ability, or prospective choice of occupation. This is the problem of adverse selection,which should be carefully distinguished from themoral hazard. The former is ex ante; the latteris ex post.The insurance afforded by an income-contingent loan-repayment program may greatlyalter the incentives of the women who participate in the program to participate later in thelabor force, since their pecuniary income is moreheavily "taxed" under such a program than itmight otherwise be.7 There may be a good reasonfor society to subsidize the education of women(beyond any level at which it might be desirableto subsidize higher education generally), so thatthis particular moral hazard might well be considered bearable by society as a whole.One fear that has been expressed concerningThus, an income-contingent loan-repayment programmight, after allowing for the effects of increased education of women, have the unfortunate social consequence of increasing fertility. There is considerableevidence that, ceteris paribus, anything which diminishesthe incentives for a woman to work outside the homeenhances the probability that she will have a largefamily. In the absence of such a change in incentives,one might expect increased education for women to leadto a decline in fertility, since such education enhancesthe income possibilities of labor-force participation. the income-contingent loan-repayment programsis that they may alter the incentives for alumnigifts. Such gifts already are subsidized by thefederal government, since they are deductible onfederal tax returns. It is not possible to say howthe repayment of a fixed, but small, percentageof one's income, part of which would also be deductible as interest, would alter the incentive ofan alumnus to donate to his alma mater. Largedonors, who have buildings or professorshipsnamed after them, presumably would be affectedrelatively little; moreover, the evidence is thatsmall donors may contribute rather little overtheir lifetimes.Income-contingent loan-repayment plans (seeHartman 1.971, pp. 61-92) have been much discussed in the last few years. Many ingenious devices have been suggested for dealing with thespecial problems of women, with the problem ofadverse selection and with the moral hazardinvolved in the insurance element, and for introducing various elements of subsidy deemeddesirable. A sequel to this paper deals with someof these schemes. Suffice it here to say: (1) Theproblems of administering the programs, enforcing whatever repayment plan is adopted, andensuring that the magnitude of the program andthe size of the loans granted thereunder are sufficient to remove imperfections in the capitaland risk-bearing markets for nearly all studentswho would otherwise be prevented from utilizinga socially optimal level of higher educationalservices are greatly mitigated by a programnational in scope. Beyond "leading the way,"what can be accomplished by an individual private university is very small indeed. (2) Theproblem of adverse selection facing an individualinstitution is much greater than that facing society as a whole, and the possibility for reductionof aggregate risk through pooling of individualrisks is much less. (3) The element of subsidyand the question of whether it is desirable to subsidize the utilization of the services of the highereducational sector at all is quite separable fromthe issue of the desirability of introducing income-contingent loan-repayment plans on a largescale. Without such plans it is clear that, at anytuition level representing the cost of supplyingthose services, demand will be below sociallyoptimal levels. An economist would say that,without such a program and if there were nosubsidies, the entire demand function lies belowthe welfare-maximizing demand function: atleast some people can be made better off by anincome-contingent loan-repayment program with-76out, at the same time, making any member ofthe current generation worse off.8Externalities of Higher Education and the Casefor Subsidies to Students: If a college educationimproves the economic productivity of an individual and if he can capture that benefit throughhigher earnings, he has an incentive to acquirethose skills. The individual and the social interestcoincide; beyond ensuring that market imperfections do not impede his quest, society has nointerest in further rewarding him in order to induce him to use still more scarce resources thatmight be devoted to other valuable endeavors.But if a college education for one person contributes to the welfare of others in a way whichcannot be appropriated by that individual in theform of higher current consumption or futureearnings, both psychic and pecuniary, then removal of market imperfections will not ensurea socially optimal level of demand for the servicesof the higher educational sector. We call suchnonappropriable, "third-party" benefits externalities or external effects.The externalities of elementary school education are manifest and universally obvious. Nonewould deny that literacy and a reasonable arithmetic facility are essential to the functioning ofAmerican society and that such social benefits gofar beyond any individual's ability to capturethem in the form of higher earnings. The external benefits of secondary education are perhapsless obvious and some would even go so far as tosay that since the era of the "melting pot," whenthe secondary (and elementary) schools served ahomogenizing function and, at the same time,provided a mechanism for social mobility, therehave been few externalities of any magnitude. Itis still more difficult to identify the externalitiesof college education. (Note that we must distinguish here between the external effects of producing inputs for college education in the highereducational sector— for example, research— andexternalities in the utilization of those inputs.)The kinds of externalities that one hears muchabout are benefits to society as a whole in suchforms as training for citizenship and for community leadership, creation of a group with"human and social qualities" that may serve as8Because of possible fertility effects, future generationsmay be made worse off. There are reasons to believe,however, that providing the present generation withmore education will tend to make future generationsbetter off, so that the net effect on future generationsis not clear. an example to others, providing a mechanism forsocial mobility, or ensuring the continuation ofsocial and cultural values.9 Indeed, it is difficultfor me to believe that such externalities are notimportant in view of the substantial amount oflong-standing public and private subsidization ofhigher education. Yet it is probably true, asFriedman (1968, p. Ill) suggests, that there hasbeen "no serious attempt to identify true external effects systematically in such a way as topermit even a rough estimate of their quantitative importance— of how large a subsidy, if any,can be justified on these grounds." Admittingthat we are far from being able to measure theexternal benefits of post-secondary education isnot, however, the same thing as denying thatsuch benefits exist; it is wrong to conclude that"until this is done, the demand for subsidy in the'public interest' must be regarded as specialpleading pure and simple." Nor can one argueon this basis for eliminating such subsidies asthere are. The lack of measurement is not evenan argument for decreasing the level of subsidy.It is an argument for trying to find out what theextent of the externality is and why there hasbeen so much subsidization in the past.Higher education will undoubtedly continueto receive substantial public and philanthropicsubsidies. It is important that such subsidies aremade in a way which only increases the utilization of higher educational services and the resources devoted to their production and does notdistort the allocation of resources in other wayswhich may be socially undesirable.One of the major sources of confusion inthinking about subsidies to higher education isthe redistributive effect which such subsidiesoften have. We believe that more equality of income is a good thing, but, perhaps harking backto an earlier era, equality of opportunity is stillmore important. To the extent that the acquisition of a college education increases the futureincome of one who acquires it, subsidizing himto do so in one way increases his income relativeto what it would have been in the absence of asubsidy, although not necessarily in relation tothat of others. Such subsidies may, therefore,have the effect of reducing income inequality,although it is likely that they will in fact increase9 Harry Johnson points out that higher education mayhave negative externalities in the form of greater socialinstability. Of course, what is negative in the short runmay be positive in the long run; the capacity for changemay be enhanced by a certain amount of instability,although too much might destroy society altogether.77inequality because the overall returns to education will be reduced. There are other ways to reduce income inequality, and it is clear thatpublic support of higher education may haverather different effects in this direction than areintended by those who argue for such support.Even with progressive general taxation, for example, it is clear that public support of low- orzero-tuition public institutions redistributes income primarily from low- and very high-incomegroups to middle-income groups.10 Public support of undergraduate education in private institutions (say, through the subsidy of buildingsof various sorts) not only may distort the sector'suse of resources, because of the selective natureof the subsidy, but tends even more to redistribute income away from those in society whohave least. The point is that there may bemore efficient ways to redistribute income thanthrough subsidies to higher education, and theissues associated with income distribution shouldbe separated in our thinking from those associated with the desirability of subsidies to thesector.Equality of opportunity is a thornier problem,as Harry Johnson has pointed out (in this Supplement). To some extent inequalities of opportunity to acquire the larger lifetime incomestreams afforded by access to higher educationare simply the result of imperfections in thecapital and risk-bearing markets, and, to theextent that they are, the solution to the problemof increasing equality of opportunity is not tosubsidize students but to improve these markets.Inequality of opportunity results not only fromdifferences in material inheritance but also fromthe unequal inheritance of genetically determinedability and from more elusive non-material elements in the family backgrounds of children.Very few would argue either for the equalizationof genetically determined opportunities or forthe means which would be necessary to createmore equality in this dimension, although anargument could be made for compensating thosewho were unlucky enough to be born with lesswit and intelligence than average.11 The difficulties in distinguishing between inherited andacquired abilities are manifold— indeed, the dis-10 Some of the subsidy may be recaptured throughhigher taxes on the higher incomes resulting from theincreased education of those subsidized.11 See Denison (1970). It is terribly difficult for anacademic to argue this position, but, of course, thatdoes not make it wrong. tinction is becoming increasingly blurred— but Itake it for granted that the nonmaterial elementsin the family backgrounds of children areimportant to their opportunities and, more particularly, to their ability to engage in the educational transformation.Although modern welfare economics does notappear to have satisfactorily resolved the inter-generational problem or recognized the social, asopposed to individual, interest in future generations, society has many institutions for assuringits own continuity and the welfare of its members yet unborn. The chief among these is thefamily.Intergenerational Externalities and the Case forSubsidizing the Education of Women: More than50 years ago, Frank Knight (1921, pp. 374-75)wrote some words about the interest of societyin the continuity of the social order, the role ofthe family, and the possibility of equality ofopportunity, which are worth remembering:The personal powers of individuals . . . obstinately resist generic separation frommaterial goods in their economic bearings.Innate ability, in the sense in which thereis such a thing, is inevitably hereditary,and nothing can be done about it exceptto modify the conception of the individual's property rights to his own powers.But culture in all its subtle significance,as well as education and training in theircruder forms, are also more or less transmissible and more or less subject to voluntary bestowal, and the factor of personalinfluence or "pull" can by no means beleft out of account. The significance ofcontrol over these things is very great. . . .It seems that real equality of opportunity,a true merit system, is hardly conceivable,and that no very close approach to such aconsummation can be expected in connection with the private family. Plato, ofcourse, recognized this fact, which most ofhis modern successors have a tendency toblink.The ultimate difficulties . . . centeraround the problem of social continuityin a world where individuals are bornnaked, destitute, helpless, ignorant, anduntrained, and must spend a third of theirlives in acquiring the prerequisites of a freecontractual existence. The distribution ofcontrol, of personal power, position, andopportunity, of the burden of labor and ofuncertainty, and of the material produce ofsocial industry cannot easily be altered. . . .78The fundamental fact about society as agoing concern is that it is made up ofindividuals who are born and die and giveplace to others; and the fundamental factabout modern civilization is that it is dependent upon the utilization of three greataccumulating funds of inheritance from thepast, material goods and appliances, knowledge and skill, and morale. Besides thetorch of life itself, the material wealth ofthe world, a technological system of vastand increasing intricacy and the habituations which fit men for social life mustin some manner be carried forward to newindividuals born devoid of all these thingsas older individuals pass out. The existingorder, with the institutions of the privatefamily and private property (in self as wellas goods), inheritance and bequest andparental responsibility, affords one wayfor securing more or less tolerable resultsin grappling with this problem.Many things have changed since Knight wrote,but in American society the family is still thecentral institution ensuring continuity of thesocial order. Women have achieved far greaterequality with men and participate in the laborforce to a much greater degree and on betterterms than they did 50 years ago, but for all thestrident protest of the women's liberation movement, the woman remains the center of thefamily and its most important component in theintergenerational transfer of acquired ability. Thefamily and its central figure are, at the sametime, the most important source of social continuity and the greatest source of inequality ofopportunity. Evidence is accumulating that theeducational attainment of the mother is amongthe most important determinants of a child'sfuture academic and economic success. The education of women enhances the educability of thenext generation, and greater equality amongwomen should tend to reduce the inequality ofopportunity inherent in the intergenerationaltransfer of acquired ability through the familyand to increase the efficiency of the family as aninstitution which helps to maintain the continuity of the social order. Needless to say, theextent to which elementary, secondary, and higher education contribute to these goals is noteasily quantified, nor, to my knowledge, hasanyone come close to measuring these effects.In the modern theory of family decision-making, as developed in economics, the welfare ofthe children and other members of the family isassumed to enter the utility function of a single decision-maker (not always the husband andfather!), thus obviating the necessity for assuming a "family utility function" with all the concomitant problems of social utility functions ingeneral. It might be argued— and has been— thatthe benefit to succeeding generations and tosociety in terms of social continuity is internalized; that is, it "accrues in large part to theparents in terms of satisfactions" (Schultz, inthis Supplement). Since no one has yet measuredthis effect, it is just as easy to argue that a substantial part is not internalized and that a socialoptimum would not result if each family simplymaximized its own welfare subject to a set ofmarket prices. Indeed, there is good reason tosuppose that each individual's concern for othersdiminishes with distance in both time and space,and that the reason for many institutions ofsociety is precisely to ensure that the interestsof future generations will be adequately guardedby the present.12By withdrawing from market activities to devote herself to her family, a woman may deprivesociety of the benefit of certain other sorts ofexternalities which higher education may gen-ate. These must be offset against the intergenerational externalities which may be generatedwithin the family. However, many of the sorts ofexternalities of higher education which comereadily to mind do not depend upon marketactivities for their realization, and these wouldnot disappear if a woman did not participate inthe labor force.The traditional role of women in Americansociety is changing, and perhaps the roles of menand women are becoming less differentiated; but,to the extent that the woman remains the centerof the family and its prime component in thetransmission of some acquired abilities and of animportant part of culture, I believe a case can bemade for subsidizing the education of womenover and above any subsidy deemed appropriatefor the education of men. To the extent thatboth men and women play an equal role in theintergenerational transfer that takes place withinIt is possible, of course, that a part of society investstoo much in future generations. Wealthy families maybequeath too much material capital to their progenyand force their children to undertake larger than optimalinvestments in human capital. This does not mean,however, that the investment of society as a whole isoptimal or that it is properly distributed betweenhuman and nonhuman capital. Low-income parents, inparticular, may be prepared to invest far less thanwould be socially desirable in the education of theirchildren.79the family and to the extent that individualutility maximization cannot guarantee the adequacy of this transfer, subsidy to the educationof both men and women would seem to bedesirable from the point of view of society. Asindicated above, what is being made is a case forall education, not just higher education. The contribution of the latter to the social goals outlinedhere has not been measured; it may be slight orgreat. Moreover, if the skills acquired throughhigher education are used later in the market andto enhance the income, pecuniary and psychic,of the family unit, lesser intergenerational externality exists and there is less of a case for a subsidy, differential or absolute.3. Tuition and the Supply of UndergraduateEducational ServicesWhat kinds of supply effects may be expectedfrom the use of tuition to allocate scarce resources between the higher educational sectorand the rest of the economy and among "firms"in this "industry"? Does it make sense to usetuition to allocate resources within colleges anduniversities as well as among them? What is thenature of production in the higher educationalsector? How do research and undergraduate education interact, and what is the role of differentdisciplines? All of these are questions connectedwith the supply of undergraduate educationalservices.13Product Diversity and Accreditation: There areimportant similarities between medical care andthe services provided by the higher educationalsector to those seeking an undergraduate education. In both cases, the service provided is informational in character. While it is true that onecan envisage situations in which the value of apiece of information is well-defined even thoughthe information itself is not known, in mostsituations the value of the information is notprecisely known until the information itself hasbeen acquired. This may be the case with medicaland other professional services. One of the mostimportant responses to uncertainty of this sortin the professional sphere is licensing; in highereducation this takes the form of accreditation,which is basically a form of quality controldesigned to compensate for the lack of a market13 Recall that the simplification of counting undergraduate educational services in quality-adjusted units hasbeen adopted. mechanism for shifting the risks and uncertaintiesassociated with the purchase of an informationalcommodity. Just as the uncertainty about theprospects of medical treatment is socially handled by rigid entry and licensing requirementsfor physicians, so the uncertainty about highereducational services is socially handled by moreor less rigid accreditation requirements for specific programs and restrictions on the entry ofcolleges and other institutions of higher education into the sector as defined by those requirements.As I pointed out in the first section of thisessay, one of the consequences of not allowingtuition the role of a price in an economic sensemay be a restriction in the variety of servicesoffered by the higher educational sector. Essentially the argument is that the decision on whatis to be produced is confined to a narrower groupwith more homogeneous preferences than societyat large.14 In a market context, the preferencesof the consumers of the product or products areexpressed through their effects on prices, butwhen prices cease to function so as to rationdemand among consumers, they also cease tofunction in such a way as to allocate resourcesamong firms producing different products or toprovide incentives to produce goods desired byconsumers. Restriction of the decision to a smallgroup may lead to greater diversity than wouldbe desirable from a social point of view if thepreferences of the group are sufficiently bizarre,but, on the whole, it is plausible that productdiversity will be diminished in at least somedirections.Because of the informational nature of theproduct, a great deal of uncertainty is bound tobe associated with the use of higher educationalservices. To some degree it may be possible toshift a portion of the risk through income-contingent loan-repayment programs, but, for thereasons discussed in detail in the previous section,it will never be possible to shift all of the risks.because so much of the return to the individualis nonpecuniary. Moreover, moral hazards existon the supply as well as the demand side; evenwere society as a whole to assume a large part ofthe burden of risk, direct controls would nodoubt still be imposed. Use of tuition as a pricein an economic sense would undoubtedly in-14The same point also is made in a somewhat differentcontext by Buchanan and Devletoglou (1970, p. 39);they argue that the quality and characteristics of theproduct will be chosen so as to reflect maximization ofthe utility of its producers rather than of its consumers.80crease the diversity of products offered by thehigher educational sector, and the right sort ofloan programs would mitigate some, but not all,of the uncertainty; accreditation and some restriction of entry might still be desirable from asocial point of view, although much could bedone to increase informational flow^ withoutrestricting entry.Widespread below-cost pricing and accreditation interact in a most unfortunate way toexacerbate the lack of diversity in Americanhigher education and to force homogeneity in aparticularly inappropriate direction. The recentcases of Parsons College and Marjorie WebsterJunior College illustrate this point. The case ofMarjorie Webster Junior College is particularlyinstructive. Marjorie Webster, a two-year institution incorporated and run as a business organization in Washington, D.C., had periodicallyrequested accreditation from the Middle StatesAssociation of Colleges and Secondary Schools.It had been repeatedly refused review on thegrounds that the association grants accreditationonly to "a non-profit organization with a governing board representing the public interest."Finally, Marjorie Webster brought suit against theassociation in the U.S. District Court for theDistrict of Columbia and was granted an injunction forbidding enforcement of the association'srule (Manne 1969; Koerner 1970).15 The initialdecision was reversed on appeal and, sad to say,the Supreme Court refused to review it, but thearguments and the judge's opinion in the initialcase are most instructive in connection with thisdiscussion. Middle States argued that the "profitmotive . . . threatened the integrity of theproprietary institution by diverting it from thebasic responsibilities of an educational institution." The judge's reply as reported by Manne(1969, p. 8) and Koerner (1970, p. 53) was:Educational excellence is determined notby the method of financing but by thethe quality of the program. Middle States'position, moreover, ignores the alternativepossibility that the profit motive mightresult in a more efficient use of resources,producing a better product at a lower price.Additionally, an efficiently operated proprietary institution could furnish an excellent curriculum whereas a badly managed non-profit corporation might fail.For a detailed discussion of the parallel in the supplyof medical services and medical' education, see Kessel(1970). Defendant's assumption that the profitmotive is inconsistent with quality is notsupported by the evidence and is unwarranted. There is nothing inherently evil inmaking a profit and nothing commendablein operating at a loss.Unfortunately, as reversal of the lower court'sdecision illustrates, there may be an inevitableconflict between the delegation and trust implicitin the social institution of accreditation and licensing designed to reduce the uncertainty aboutthe quality of the product and profit maximization. This is as true in medicine or law as it is ineducation, and it is doubtful that the conflictcan be entirely removed by any program designed to lead to greater diversity in the highereducational sector. Greater knowledge of whatactually takes place in the educational transformation could, however, lead to more enlightenedaccreditation policies and a greater diversity inAmerican higher education.Joint Production in Higher Education: It isassumed for the sake of this argument that onlytwo products are produced by the higher educational sector, basic research and graduate education, on the one hand, and undergraduateeducation, on the other. Variations in qualityare here taken account of by counting higherquality research or undergraduate education as alarger number of units than that of lower qual-ity.16Graduate education and basic research seem tome to be inextricably intertwined, so that treating them as Marshall's knife blade and handle-perfectly complementary— as a first approximation does little violence to the reality. Applied research is of a somewhat different character, andSuch simplification not only ignores some incrediblydifficult measurement problems but also implicitly assumes that different types or qualities of educationalservices or research services are perfect substitutes forone another (for example, that some number of yearsat low-quality college x is equivalent to one year atHarvard). Not only does the assumption of perfect sub-stitutability do violence to my intuitive sense of whatis involved in the educational transformation (differenttypes of experience are valuable to different sorts ofpeople), but it is not obvious that the marginal rates ofsubstitution between years x and years at Harvardought to be independent of the amounts of each whichare available. If there were a lot more Harvard yearsthan there are, college x years might be worth relativelymore. Still, one has to draw the line somewhere in thedegree of complexity which can be introduced, and Idraw it here for the time being.81much, especially the more applied, can be successfully accomplished without the presence ofstudents. Certain research organizations such asthe RAND Corporation and the Brookings Institution are attempting to incorporate rudimentarygraduate programs in their existing researchstructures; others, such as the National Bureauof Economic Research, Battelle, and StanfordResearch Institute have long had university ties;and the large governmental research units tend tofunction most effectively in the natural sciences,where the institution of postdoctoral training ismost firmly established so that they do in effecthave students. This suggests that there is a substantial area of basic research and perhaps somepart of applied research (which, after all, shadesoff rather gradually from the pure) where graduate students are such an important part ofthe process that their training cannot really betreated separately from the ongoing researchprogram of which they are a part.Professional education in such fields as medicine, law, engineering, social work, or businesspresents more difficult problems. For a longtime, training in many of these fields was carriedout separately from university programs in graduate education and basic research, but it is nowconcentrated in university professional schoolsor university-related schools.17 This suggests,too, a certain complementarity. Such complexities are largely ignored in the discussion whichfollows, although the analysis does suggest a wayof looking at the problem of additional jointproducts such as professional education, appliedresearch, and, perhaps research administration,which is frequently undertaken by individualuniversities or consortia of universities.That basic research (and the graduate education associated with it) is typically a jointproduct produced with undergraduate educationis a fact which needs to be explained and mayeven provide some argument for subsidizing certain kinds of undergraduate education (Kaysen1960, pp. 56-57; Weisbrod 1962, p. 120). Iargue below that there is indeed a strong case forpublic subsidy of the basic research which takesplace in the higher educational sector, even atcertain types of institutions not usually thoughtof as being in the "basic-research circuit," so toHistorically, of course, many colleges and universitiesin the United States grew out of professional schools,especially schools for the training of individuals for thelaw and for religious ministries. speak, but that this does not necessarily entaila subsidy to undergraduate education and mayactually make such education more costly tosociety.Universities such as Johns Hopkins, which wasoriginally founded primarily for the purpose ofdoing graduate education and basic research,have, over time, gradually incorporated undergraduate programs of substantial size relative tothe program originally emphasized. Conversely,Harvard and Yale, where undergraduate education was once paramount, have become increasingly active centers of basic research and graduateeducation. To some extent these developmentsmay have been due to the changing pattern ofsupport for higher education in this country; tosome extent they are surely due to the complementarity among different fields, the natureof basic research in the humanities, and thecomplementarity between basic research (andthe graduate education associated with it) andthe production of undergraduate educationalservices.One of the central reasons why a universityis more than a collection of departments is theexistence of strong complementarities amongfields and disciplines, both with respect to research and with respect to undergraduate teaching. It is difficult to conceive of good researchbeing done, for example, in political science orsociology without strong programs in anthropology, economics, and history, or in economicswithout strength in statistics and mathematics, orin astronomy without corresponding activity inphysics. This is even more the case with researchand training which takes place in professionalschools, and, although there are many greatuniversities without an extensive set of associatedprofessional schools, the reverse feedbacks aretoo often neglected by those who attempt tounderstand the nature and working of a university. Thus, a variety of research and teachingactivities are mutually supportive and more efficiently carried on (using fewer scarce resources)together than apart.There is some difficulty in distinguishing truejointness in production from scale effects. Thevarious activities of a university use commonfacilities, such as buildings, libraries, and computers. One of the reasons that different disciplines are pursued and both research and undergraduate teaching are done together is to gainmore efficient use of these common facilities. If,for example, the scale of operation of any oneactivity were sufficiently large, it could be that82any supposed complementarity would disappear.I think, however, there are deeper reasons forjointness in production.In the natural sciences, especially the physicalsciences, it is remarkable how frequently seniorscientists of great eminence teach the most elementary undergraduate courses. The sameas trueto a lesser degree in the biological and socialsciences.18 It may well be that surveying one'swhole field quite broadly, or trying to makeclear the structure of a discipline to the "uninitiated," or showing how a discipline wouldattack a particular scientific, social, or philosophical problem, is inherently "educative" tosomeone engaged in basic research. To this extentresearch and undergraduate instruction are complementary in production and best carried onwithin the same institution and by the samepeople. (This is true in a gross sense but, asargued below, will not be true at the margin,where, indeed, research and the provision ofundergraduate educational services must be substitutes if resources are allocated efficiently.)The Special Nature of the Humanities: Whatappears to be true of the sciences, to a greateror lesser degree, depending upon which particulardiscipline is considered, is, in a sense, fundamentally true of the humanities. I can claim no expertise within this group of disciplines beyondthat afforded by a good— and by now somewhat old-fashioned and outmoded-general, liberal education, but it has always seemed to methat the humanities are concerned, in one wayor another, with the meaningful reinterpretationof man's past and his works to each succeedinggeneration. They are concerned in a very fundamental sense with the intergenerational transmission of culture and of values. Directly orindirectly, therefore, they are deeply bound upwith undergraduate education and the "educational transformation."19 Modern culture contains a large component of science, too, ofcourse, and, to the extent that natural and socialscientists are concerned with making the scientific part of our heritage meaningful to succeeding generations, they too are humanists in mysense. The undergraduate college is to thehuman-18 Curiously, casual observation suggests such participation in undergraduate teaching is rarer among distinguished humanists. If such is indeed the case, it is notconsistent with my view of the nature of the humanitiesas expressed herein.19For a similar view, see Roush (1969, p. 641). ist what the laboratory is to the natural scientistand the survey and the census are to the socialscientist. It is impossible to conceive of basicresearch in the humanities being carried outwithout a considerable direct or indirect involvement in the whole process of undergraduateeducation. This does not mean, of course, thatthe two must be carried out by the same individual, or even at the same institution, but only thatmeaningful research in the humanities cannot beseparated from undergraduate education somewhere in the system.The Case for Subsidizing Basic Research: As remarked above, information has a number ofpeculiarly uncomfortable properties from aneconomic point of view and cannot easily beconsidered as a commodity in the ordinary use ofthat term. Some information results in benefits,pecuniary or nonpecuniary, to the producer himself; a number of institutions (the patent system,copyrights) have been developed in our economyto increase the scope of appropriability of thegains from the production of new knowledge. Ingeneral, the more applied a piece of research, thegreater the appropriability or potential appropriability of its benefits. New knowledge, especiallyof a more basic character, has benefits, however,which are so widely diffused throughout societythat it is difficult, if not impossible, to invest theproducers of such knowledge with appropriateproperty rights in the information.20 It is whateconomists would call a "public good," a commodity the consumption of which by one individual does not reduce the amount available forthe consumption of others; indeed, in the case ofknowledge, individual consumption may, in asense, enhance the availability for other persons.Fundamental or basic research is, almost bydefinition, an input for further research (Arrow1971, pp. 154-55): "To appropriate informationfor use as a basis for further research is muchmore difficult than to appropriate it for use in^Hirschleifer (1971) distinguishes between "foreknowledge" and "discovery" and shows that, in general, noincentive need to be provided to those who obtain theformer type of information and that, indeed, there maybe socially undesirable overinvestment in the acquisitionand dissemination of such knowledge. "Discovery," onthe other hand, creates speculative opportunities whichmay or may not enable the discoverer to capture partof the value of the new knowledge. In the case ofbasic research the uncertainties are so great, and thechains by which the discoveries produce real world effects so long, that it is doubtful that any speculativeopportunities provide sufficient incentives to engage inthe activity.83producing commodities; and the value of information for use in developing further informationis much more conjectural than the value of itsuse in production. . . . Thus basic research, theoutput of which is only used as an informationalinput into other inventive activities, is especiallyunlikely to be rewarded [by the market mechanism]."Again, basic research in the humanities presents special problems just because such researchis so intimately tied up with" the educationalprocess and the intergenerational transmission ofculture and of values. To the extent that suchresearch is not immediately rewarding to its producer, however, there would appear to be an evenclearer case that its benefits cannot be privatelyappropriated.From a social point of view, moreover, itwould also seem that various devices which havebeen introduced in order to invest individualswith certain property rights in knowledge mayhave the effect of reducing the general welfare.Once basic knowledge or information has beencreated, the social product will be greater if allhave access to it merely for the costs of transmitting it. Artificial barriers to its use designedto convey some part of the social benefit of newknowledge to its producers and so give them anincentive to invest scarce resources in its acquisition can only have the effect of reducing theultimate benefit to society as a whole.21 But,without such incentives, private individuals, motivated only by the direct satisfaction of discovery, will tend to invest too little in theproduction of new basic knowledge. Clearly, it isin the public interest to subsidize collectivelythe acquisition of new knowledge, especially thatof a fundamental or basic character. To somedegree, private philanthropy may fill the gapbetween what individuals would do for their ownsatisfaction and what would be socially optimal,but such an exact coincidence of public andprivate interests would be fortuitous. Indeed, thebenefits of basic research to society may be sogreat and so uncertain in character and extentthat reliance on individuals' maximizing theirown satisfactions may be exceedingly undesirable from a social point of view.Not only are the gains from basic research21 As Hirschleifer (1971) points out, overinvestment inthe acquisition and dissemination of "foreknowledge"or opinion can easily occur. The discussion here, however, is restricted to the type of knowledge generallyacquired within the educational sector, and one hopesthe speculative component is not large. largely unappropriable by those who do theresearch, but the process itself is a highly riskyone (Arrow 1971, p. 138): "The outcome of anyresearch project is necessarily uncertain, and themost important results are likely to come fromprojects whose degree of uncertainty to beginwith was greatest. The shifting [and presumablypooling as well] of risks is thus most neededfor what is very likely the most profitable ofactivities from society's point of view."Both the "public" nature of the good produced and the riskiness of the research process,coupled with the uninsurability of these risks,suggest that too little basic research will becarried out in an economy which relies purely onthe market mechanism. This may be true as wellfor much applied research. Moreover, institutional devices for making the gains from basic research privately appropriable not only do notmitigate the deficiency in the activity due to itsriskiness but also have the unfortunate featureof reducing the use of the fundamental knowledge gained below socially optimal levels. It isnot surprising, therefore, that the great bulk ofbasic research has been carried on in universities,in nonprofit institutions, by the government, andby private individuals who derive satisfactionfrom the process itself. The role of colleges anduniversities has been an especially important one,in part because of the complementarity betweenteaching and basic research. Such complementarity does not, however, imply that a sociallyoptimal level of basic research activity will beundertaken by the higher educational sector inthe absence of subsidy, nor does it implythat subsidies directly for research or indirectlythrough subsidies to students have equivalenteffects or benefits to society.The Complementarity of Research and Teachingand the Effects of Subsidizing Research: Thatbasic research and graduate training and the provision of undergraduate educational services arecarried on within the same institution and frequently by the same individuals, as alreadynoted, suggests that the two activities are complementary in production in the sense that fewerscarce resources are needed to produce givenamounts of the two goods if they are producedtogether rather than separately. Indeed, there issome doubt that certain kinds of basic research,for instance in the humanities, can be producedindependently of undergraduate education at all.But this does not imply that the provision ofundergraduate educational services should there-84fore be subsidized. I argue here that if societydoes pay for basic research so that a sociallyoptimal quantity is produced, it may well havethe effect of making the provision of undergraduate educational services more expensive,but this increase in cost is desirable from society's point of view. A few diagrams are helpfulin the elucidation of this argument. Figure 1shows the various combinations of research andgraduate education, on the one hand, and undergraduate educational services, on the other,which can be produced with two different levelsof resources in the higher educational sector.REGION OF SUBSTITUTABILITYPositive Prices for Both UndergraduateEducation and Research andGraduate EducationUNDERGRADUATE EDUCATIONAL SERVICESFigure 1: Combinations of undergraduate educationand of research and graduate education showing theregions of complementarity and the region of substi-tutability.The boundaries of the feasible regions for givenlevels of resources are called transformationcurves; the curve T V represents a lower levelof resources than the curve T T' . The sections2 2of the transformation curves that bend outward,for example, between the abscissa and point Bor between the ordinate and point A on thecurve TV represent a region in which more ofboth teaching and research can be obtained without using additional scarce resources; that is,they are regions in which teaching and researchare complementary. If "firms" (educational institutions) in this country behave optimally, wewill never observe them operating to the left of apoint such as A or below a point such as B, sinceto the left of A more teaching can be obtainedas a by-product of research without any furtherexpenditure of resources, and, conversely, belowB more research can be obtained as a "free" byproduct. Nonetheless, the existence of such regions in the transformation curves explainswhy we will generally observe teaching and research being done together.Even if society pays nothing for basic research,some of it will be undertaken in connection withundergraduate education, since more of the latter can be obtained from a given volume ofresources if some of the former is also produced.Remember that I have adopted the simplificationthat both research and undergraduate educational services are measured in "quality-adjusted"units, so that what is here expressed as an increase in the quantity of teaching may simplyrepresent an increase in the quality of teaching.Thus, at a zero "price" for research, the sectorwill operate at a point along the transformationcurve such as B, where some quantity of researchis nonetheless produced. This research, however,is of positive benefit to society as a whole; collectively, society should therefore be willing topay for it. Moreover, it has been argued abovethat unless society does pay for it, too little willbe produced from a social point of view.The slope of the straight line tangent to thetransformation curve at points A, B, or C represents the ratio between the prices paid for teaching and for research. Thus, at A the slope iszero, implying that nothing is paid for teachingbut that there is a positive price for research;conversely, at B the slope is infinite, implyingthat nothing is paid for research but that thereis a positive price for teaching; at C both pricesare positive. // a price per unit of research is setin such a way that it represents the value tosociety of an additional unit of research, then therelative prices represent the socially optimaltrade-offs between teaching and research.22 Asocial optimum, then, is achieved when the priceline is tangent to the transformation curve, for atsuch a point the quantity of teaching whichwould have to be given up to obtain an additional unit of research for given resources (themarginal rate of transformation between teaching and research) is exactly equal to the relativevalue which society places on the two products.Thus, if society subsidizes the higher educationalsector to produce research in a socially optimalway, the sector will always operate at a pointsuch as C, where teaching and research are sub-If these trade-offs depend on how much of eachproduct is produced and consumed, the lines may becurved. It may be argued, however, that they will beconvex when viewed from the origin, and thus thatnone of the conclusions reached by approximating suchtrade-offs by constant "price" ratios is altered.85stitutes in the sense that for a given level ofresources devoted to higher education more research can only be obtained by providing fewerundergraduate educational services, and conversely.23In general, of course, a subsidy to basic research and graduate instruction carried out in thehigher educational sector will cause additionalresources to flow into the sector. The extent towhich this occurs will determine whether moreor fewer undergraduate educational services willbe supplied with than without such a subsidy toresearch. Figure 2 illustrates the two possibilities.RESEARCHANDGRADUATEEDUCATIONPrice Line in the Absence ofa Subsidy to ResearchPrice Lines WhenResearch IsSubsidizedB A CUNDERGRADUATE EDUCATIONAL SERVICES If no subsidy to research is provided, the level ofresources drawn into the higher educational sector will be sufficient to produce combinations ofteaching and research along the transformationcurve T V ; in fact, a quantity OA of undergraduate educational services will be produced.If research is now subsidized, the price line willbe twisted from its original vertical orientationand more resources will be drawn into the educational sector, shifting the transformation curveoutward. If very few additional resources aredrawn in, the transformation curve will shiftoutward only a very little, say to T2 V ; at theparticular relative price for teaching and researchrepresented by the relationship between the research subsidy per unit and tuition, less undergraduate educational services— an amount OBrather than OA— will be provided. Of course, ingeneral the provision of less teaching will causea movement along the demand function forundergraduate educational services, so that, in afree market, tuition will rise, twisting the priceline toward the vertical a bit so that somewhatmore would be provided by the sector. On theother hand, if a great many resources flow intothe higher educational sector as a result of thesubsidy to research, the transformation curvewill shift outward by a lot, say to TV; at therelative price for teaching and research reflectedin the slope of the price line, more of both, anamount of teaching represented by OC ratherthan OA, will be supplied. The greater supply ofundergraduate educational services would, in afree market, however, cause a movement alongFigure 2: Illustration of two possible effects on thesupply of undergraduate educational services resultingfrom a subsidy to research.This is simply an expression of the well-known principle that when only two products are produced, wewill only observe profit-maximizing firms producing at apoint at which the two are substitutes. If three or moreproducts are produced, however, it is possible to observetheir production at points where certain pairs maystill be complementary, in the sense that more of bothcould be produced for a given level of resources, butonly, of course, by giving up some of the third good.A three-good model would be appropriate if, for example, we were to treat graduate education and researchas separate activities, but then it would be difficult toargue that research and undergraduate teaching werethe complementary pair. The addition of professionaland preprofessional educational services to the productmix, however, does complicate matters quite a bit.Nonetheless, the conclusions reached on the basis of thetwo-commodity model hold, in broad outline, andshould provide a firm basis for subsequent analysis ofmore complicated models. TUITION tRATE D\ )\^y *yE j^^^ J^>^^^ y /S^ - s, — ¦ "— "D'UNDERGRADUATE EDUCATIONAL SERVICESFigure 3: Illustration of the possible effects on tuitionand the equilibrium quantity of undergraduate educational service when basic research is subsidized. Supplycurves: SiS\ = unsubsidized research; S2S'2 = subsidizedresearch, resources in inelastic supply; S3S'3 = subsidizedresearch, resources in elastic supply.86the demand function so that tuition would fall,twisting the price line more toward the horizontal; the sector would operate so as to produce anamount between OA and OC.Figure 3 illustrates how the interaction between tuition and the research subsidy wouldfunction in a free market situation. The curveS S' is the supply curve of undergraduate services when research is unsubsidized; the upwardslope of S S' reflects the -need for higher levelsof tuition to call for a greater supply becausescarce resources must be drawn into the highereducational sector from other productive uses inorder to produce the additional quantity. If thedemand for undergraduate educational servicesis represented by the curve DD' , it can be seenthat the market equilibrium tuition rate will beOE and the equilibrium quantity produced bythe sector and used by students will be OA. Fortuition rates below OE more undergraduate educational services will be demanded than aresupplied; and for tuition rates above OE thesector would be prepared to supply more thanstudents would want to use. When research issubsidized, more resources will be drawn into thehigher educational sector, but now undergraduateeducational services must compete within thesector for the use of these resources. Whetherthe supply curve for undergraduate educationalservices will rise or fall in relation to the supplycurve when no subsidy is provided to researchdepends upon two factors: (1) how easy it is tosubstitute research for teaching within the sectorand (2) how easily additional resources are drawninto the sector.If the resources used by the higher educationalsector are very scarce and/or exceptionally valuable in alternative uses, they will be in inelasticsupply to the sector. It will require a large subsidy to research to attract additional resources,and the more easily research can be substitutedfor teaching, the more resources from the inadequately supplemented total will be drawnaway from teaching in the sector. When resourcesare in inelastic supply, the supply curve for undergraduate educational services will lie abovethe curve for unsubsidized research, as S^S'' 2 2does. When resources are easily drawn into theeducational sector, because either the right sortare not particularly scarce or because these resources are not exceptionally valuable in alternative uses, they are in elastic supply. Then a smallsubsidy to research will draw forth a great manyadditional resources; even though some of theseare diverted to research, so many additional re sources are available that more teaching andmore research are both produced. In the case ofan elastic supply of resources to the educationalsector, the supply curve for undergraduate educational services will he below the curve for unsubsidized research, as S S' does. Of course,the subsidy may alter the whole shape, not justthe position of the curves as I have drawn themin Figure 3; and it is possible for the curve forsubsidized research to lie above that for unsubsidized research for part of its length and belowit for the rest.24If resources are in inelastic supply, so that thesupply curve becomes S £', the quantity of undergraduate educational services supplied will notfall to OB but only to OB' because tuition willrise to OF, causing some substitution away fromresearch. Conversely, if resources are in elasticsupply, subsidizing research will cause the supplycurve to shift downward and to the right; butthe quantity of undergraduate educational services will not rise to OC but only to OC, sincethe tuition rate will fall to OG, causing some ofthe increase to resources to be drawn out ofteaching and into research.I must candidly admit that I do not knowwhat the elasticity of supply of resources to theeducational sector is or how easy it is to switchresources in the educational sector from teachingto research (at least at the margin where theswitch counts). The substitutability of resourceswithin the sector depends on the productionfunction of the various outputs produced by thesector. Provosts may have some intuitive notionof what this function is like for their own institutions, but I doubt that anyone has a clear graspof the function for the sector as a whole. Myown guess is that at the margin it is in fact easyto substitute, since many of the same peopletend to do both undergraduate teaching andresearch. However, I would also guess that quality resources are not easily drawn into highereducation in the short run, but in the long runSince we would expect resources to become increasingly scarce as the sector expands, they might be inelastic supply for low levels of tuition and subsidy andin inelastic supply at high levels. This would have theeffect of making the unsubsidized curve rise morerapidly and might cause the subsidized curve to risestill more rapidly, so that the curve for subsidized research would start out below the unsubsidized curvebut would cross it at some point and continue aboveit. The analysis of the effects of a subsidy would thendepend on the point at which the demand curve crossedthe unsubsidized supply curve, above or below its intersection with the subsidized supply curve.87they are; hence, subsidizing research in the shortrun will put a strong upward pressure on tuitionrates, but in the long run it will not and mayeven tend to reduce them somewhat per constant-quality unit of undergraduate educationalservices supplied. The answer to the question ofhow market pricing of educational services wouldinteract with subsidies to research depends heavily on quantitative knowledge which we simplydo not now have.25If society subsidizes basic research, such asubsidy may have the effect of making undergraduate education more expensive. Does thisgive rise to an argument for subsidizing the production of undergraduate educational services aswell? Provided society does not subsidize basicresearch excessively, I think the answer is unequivocally "no." The resources used by undergraduate educational activities are in this casemore expensive from society's standpoint, whichnow includes the provision of basic research aswell as all the other alternatives to which theresources used in the higher educational sectorcould be put; and these resources should bevalued accordingly even if, on other grounds,subsidies to undergraduate education are provided. The proper way to determine the subsidyto basic research constitutes, however, a gravedifficulty. Presumably the subsidy should be setso that the expected rate of return to dollars invested in the last project undertaken is just equalto the rate of return on the next best alternativeproject, public or private. The measurement ofsuch return, however, and proper allowance forrisk from a social point of view raise problemswhich seem far from solution. It is possible thatresearch will, under these circumstances, be overly subsidized, so that some case might be madefor a compensating subsidy to undergraduateeducation if, indeed, subsidizing basic researchcaused the cost of obtaining such an educationto rise. Too much at this point is unknown,however, for continued rational speculation.Public and Private Institutions: Tuition, ofcourse, does not function in anything like theway it would have to function in order to allocate resources between teaching and research inSince there has been a very large expansion of highereducation in the past two decades and, at the same time,only a modest increase in the relative incomes of professors and researchers, it would seem that the laborsupply is quite elastic. Labor, of course, accounts for alarge part of the resources used by higher education.I owe this point to D. Gale Johnson. the manner described above. Not even in privateinstitutions does it represent the value of resources used in teaching at the margin. Thesituation is far more serious for the allocation ofresources among different types of institutions,particularly between the so-called public institutions, which charge little tuition or none at all,and private colleges and universities, which aregenerally subsidized in other ways and may receive a large measure of public support but whichof necessity charge a good deal more for the undergraduate educational services they provide. Because most support is channeled through institutions which pass the subsidy along in the form ofbelow-cost charges to those lucky enough toqualify for admission, the public sector of highereducation has grown spectacularly relative to theprivate sector. In 1950 approximately 50 percentof undergraduate enrollment was in publicly controlled institutions; by 1967-68 the percentagehad risen to nearly 70 percent (O'Neill 1971,P. 71).Former Chancellor Lawrence Kimpton of TheUniversity of Chicago is reported to have oncesaid that it is difficult to sell a product whenthey're giving it away down the street fornothing. Not only has public support of highereducation through markedly below-cost tuitionat publicly controlled institutions effectively redirected scarce resources into the public sectorof the industry, but it has surely had a depressingeffect on the level of tuition private institutionshave been able to charge. The downward pressureon tuition at private institutions has had severalsocially undesirable effects.Most obviously, of course, rapid developmentof the public sector in the form of extremelylow-tuition institutions has contributed to theserious financial condition in which private institutions now find themselves, both by holdingdown the revenues of private institutions and byincreasing their costs by bidding up the prices ofscarce factors of production. Much of the bidding up of prices would have occurred as theresult of increased support to the sector as awhole in any case, but the manner in which thatsupport has been directed has had the effect ofvirtually strangling the private part, which iscaught between rising costs and its own inabilityto raise the prices it charges for its product.Much has been said about the advantages of apluralistic system of higher education in America; it is believed by many that private institutions are more innovative and freer from politicalcontrol and can supply a more diverse, if not a88better, product. While this is not necessarily thecase, to the extent that a pluralistic system doeshave value to society, destruction of the privatesector is socially undesirable.26Another, more subtle, effect of the downwardpressure on tuition has to do with the nature ofthe product produced by the private sector.Major private institutions typically "rebate" alarge fraction of the tuition they receive to students in the form of aid. Partly this may be doneout of sheer altruism— the desire that at leastsome poor but bright boys will have the opportunity to attend a Harvard or a Yale— but inpart it is done to achieve a particular type ofstudent body. A large part of what happens during the educational transformation, or at leastsome kinds of educational transformation, comesabout through the contact of the student withothers of similar age but dissimilar backgroundand ability. Intelligence has, at least until recently, been thought to be the most importantcharacteristic of a student who generates largeexternal effects of this sort in other studentsaround him. Since bright people are always inshort supply, private institutions tend to "bid"for them by selectively reducing the costs ofattending college. By holding down tuition levelsat private universities, public support of highereducation through publicly-controlled institutions has reduced the possibility of such pricediscrimination and thereby the ability of privateinstitutions to supply the distinctive productthat results from the externalities generated bycertain types of students. To be sure, these samestudents may continue to generate externalitiesin the public sector, but scale effects are surelyimportant here, and it may be difficult to generate the kind of critical mass needed for theeffect in a large public institution. Recent educational experiments in the public sector are,however, directed rather explicitly to solution ofthis problem.4. Policies and Directions for Research: SomeConclusions"I felt that Evenus was to be congratulated if hereally was master of this art and taught at sucha moderate fee. I should certainly plume myselfand give myself airs if I understood these things;but in fact, gentlemen, I do not."Plato, The Apology'The undesirable redistributive effects of public sup- I must confess that when I began this study I hadhoped that a policy paper, in the best sense,would emerge, but now I find what is largely aresearch proposal has come out instead. This section summarizes the main problems on whichfurther research appears most needed and drawstogether some tentative conclusions of the preceding analysis for public policy.Policies: There is a clear case for a national loanprogram with an income-contingent repaymentplan to mitigate the worst effects of imperfections in the capital market and the market forrisk-bearing. Without such a program, nationalin scope, students cannot finance higher education on terms comparable with investment inphysical capital, nor can they insure themselvesagainst the risk of a low lifetime income fromwhich to repay an ordinary loan. Without such aprogram, the demand for undergraduate educational services will be below the level desirablefrom society's standpoint. Programs on a university-by-university or college-by-college basis cannot begin to solve the problem of demanddeficiency due to market imperfections.Loan programs by themselves, however, maynot increase demand sufficiently in view of possible externalities of higher education. To theextent that further subsidies are found sociallydesirable, they should go directly to the studentsand not, except under most unusual circumstances, to institutions for disbursement or forgeneral support of undergraduate educationalfacilities at below-cost tuition. Special consideration should be given to the possibility of subsidizing women's education so long as the privatefamily remains a central institution of our society and women continue to play the centralrole in the family that they have in the past.Such subsidization has as its goal greater equalityof opportunity among members of succeedinggenerations, as well as their enhanced educability.It is in the best interest of society and privateinstitutions of higher education that there be fullmarginal cost pricing of undergraduate educational services. Publicly supported and controlledinstitutions will lose from such a policy, althoughtheir losses may be substantially mitigated by removal of demand deficiency due to marketimperfections. A policy for all institutions ofequal prices for equal services should be en-port through low tuition have been mentioned earlier.As indicated, these effects may consist largely of an income transfer from low-and high-income groups tomiddle-income groups.89couraged, leaving open, however, the possibilityof systematic price discrimination among individual students through selective aid, if such apricing policy is necessary to achieve a certainquality of product. The aid given to some students, however, should be fully covered by highercharges to their fellow students, who benefitfrom the higher-quality education the presenceof those aided makes possible.Accreditation agencies need to be controlled,publicly or by institutions of higher educationthemselves, so that they do not stifle imaginativeand innovative solutions to higher education'sproblems and prevent the immense diversity ofwhich American higher education is capable fromoccurring.Public support for basic research in all disciplines (not just those momentarily fashionable)through individual project support or throughinstitutional support seems warranted; however,the appropriate level of support will depend onthe expected benefits to society of such research,and these benefits are not now measurable withany degree of confidence, let alone precision.Finally, it would seem both appropriate anddesirable for the major national research universities to mount a sizable research program designed to discover answers to some of the questions about the nature of education and itsproduction which, unanswered, preclude moredefinitive policy recommendations.Directions of Research: Two and a half milleniaago Socrates strove to understand the nature ofthe educational transformation. It left him puzzled and his contemporaries in worse condition.What was it beyond training in specific skills thathappened to a human being? It is easy to understand the training of a horse or a dog, but the education of a person is not so well defined. Whatare the contributions of inherited abilities, familybackground, previous training and education,surroundings, teachers, fellow students, time, andeffort? What differences do the educational services provided by a college or university make? Isthe transformation impeded by high student-faculty ratios for some people but not for others?Are particular educational environments betterfor some people and different ones better forother people? Are there different kinds of educational transformations? Is what happens quantitatively measurable along a single axis, or areseveral dimensions needed to quantify the result?Can it be quantified at all? We have made littleprogress in the last 2,500 years in answering these questions. It is possible that many ofthem are unanswerable; yet these questions lie atthe very heart of the economic issues discussedin this essay and are central to many social issuesof fundamental importance.A related series of questions concerns the roleof the family in the intergenerational transmission both of acquired ability and of values andculture. Do we understand this process? Can wedisentangle genetically determined abilities andcharacteristics from acquired ones? It used to bethought that the answer to this question wassimple; now we know better. To what extentdoes a model based on the maximization of asingle "decision-maker's" utility function, containing the utility of consumption of otherfamily behavior, especially its behavior withrespect to numbers of children, investment inthem, and allocation of the woman's time? Ifsuch a model does account for observed familybehavior, can it also be used to measure the degree to which the current generation's investmentin their children falls short, if it does, of whatthe next generation would have invested in themselves had they been in a position of control?How does the investment take place and what isthe role of each parent? Have these roles beenchanging, and how will such changes as have beentaking place affect the intergenerational transferand the extent to which it is optimal from asocial point of view? Can we formulate a clearand consistent set of principles for determiningthe social interest and its possible divergencefrom the private interests of the current generation? Economists working on growth have beenseeking such a set of principles for some time,but I do not find an adequate formulation in anyof the literature with which I am familiar.What are the externalities of higher education?Can they be measured? This is Friedman's challenge. Economists have made some progress inrecent years in identifying and measuring theexternal effects of elementary and secondaryeducation, but the externalities of higher education, if any, are much more subtle and elusive.It seems certain that they are of a generally non-Two examples of the sort of thing which has beendone in recent years are Astin (1968) and Grilichesand Mason (1972). The Astin paper is relatively naivefrom a statistical point of view, overlooking the seriousproblems arising from the simultaneous determinationby several relationships of a number of variables involved in the analysis. The Griliches-Mason paper ismore sophisticated statistically, but the authors encounter severe measurement problems in the data theyanalyze.90economic character, and their elucidation andmeasurement may require quite different ways oflooking at human and social behavior than thosewhich economists find congenial. Unless, however, the external benefits to society can bemeasured and valued, even if only crudely, thecase for subsidies to higher education remainsweak.Another area in which both conceptual formulation and measurement are required is in theanalysis of the production function and processin the higher educational sector. It may be necessary to distinguish among different types ofinstitutions, and it is certainly necessary to distinguish a variety of outputs of different qualities. Can the quality differences be quantified?What are the trade-offs, for instance, betweenresearch and graduate education, undergraduateeducational services, professional educationalservices, and applied research? How do different disciplines interact within the same institution? Do the trade-offs differ from onediscipline to another, or is there a commonstructure? How can the different outputs bemeasured? What -is the relationship of theseoutputs to the inputs of such scarce resources assenior faculty, junior faculty, secretarial andclerical support, administrative services, libraryand computer facilities, laboratories and equipment? Numbers of classroom hours at variouslevels (graduate, undergraduate, seminar), numbers of doctoral dissertations supervised, degreesawarded— these are readily available; some universities collect data on such faculty outputs aspapers and books published, outside lecturesgiven, and editorial work. While individual dataof this sort are likely to be quite unreliable("noisy"), departmental statistics may be better.Inputs at the departmental level-types of faculty, supporting services, expenditures in variouscategories— can be assembled. Allocation of overhead facilities does cause problems however; howare the library and computer facilities to becharged, and what account needs to be takenof the outputs of other departments in the inputsof a particular one? Could sufficient comparability across some category of colleges anduniversities for a given discipline (say, for thetwenty major departments of economics) beobtained to permit estimation of a multiple-product production function?28 Despite obvious28 For an example of the estimation of a productionfunction where "output" is defined in terms of a vectorof quantities rather than a single homogeneous corn- imperfections in the degree to which these different measures represent actual outputs andinputs and lack of comparability in terms ofquality, is it clear a priori that such an investigation would produce results of little or no value?Finally, what are the characteristics of thesupply of factors of production to the highereducational sector? Are faculty resources in elastic or inelastic supply to the sector? How docareer choices depend upon expected incomes?What are the rates of response to changes inrelative income expectations? Recent work onthe supply of doctorates and the demand forholders of the doctorate by other sectors of theeconomy is largely deficient in that it neglects orgives little consideration to price effects. Freeman's work (1971, esp. pp. 160-79) is a notableexception in this respect. We are closer to answers here than to any of the other questionsraised above, but the answers to this set of questions need to be put together with an analysis ofthe production process in the higher educationalsector in order to understand the effects ofsubsidies to research, full-cost tuition, and thechanging structure of the demand for trainedmanpower in the economy as a whole.ReferencesArrow, Kenneth J. Essays in the Theory ofRisk-Bearing. Chicago: Markham, 1971.Astin, A. W. "Undergraduate Achievement andInstitutional 'Excellence.'" Science, August 16,1968, pp. 661-68.Becker, Gary S. Human Capital: A Theoreticaland Empirical Analysis with Special Referenceto Education. New York: National Bureau ofEconomic Research, 1964.Bolton, R. E. "The Economics and Financing ofHigher Education in the United States." In TheEconomics and Financing of Higher Education inthe United States: A Compendium of Papers,pp. 1-104. Congress, Joint Economic Committee,91st Cong., 1st sess., 1969.modity, see Eads, Nerlove, and Raduchel (1969). Notethat, although this paper involves estimation of the costfunction rather than the production function itself, theShepard-Uzawa duality theorem permits inference aboutthe production function from the cost function. Despitea number of quite difficult statistical problems, the results do allow an analysis of the trade-offs betweendifferent "bundles" of outputs, defined in terms oftypes of aircraft used and a number of separate characteristics of local air service.91Bowen, William G. The Economics of the MajorPrivate Universities. Berkeley, Calif.: CarnegieCommission on Higher Education, 1968.Brewster, Kingman. "Statement on the YaleTuition Postponement Option." Press release.New Haven, Conn., February 6, 1971.Buchanan, J. M., and Devletoglou, N. E. Academia in Anarchy. New York: Basic, 1970.Denison, Edward F. "An Aspect of Inequalityof Opportunity." Journal of Political Economy78 (September/October 1970): 1195-1202.Eads, George; Nerlove, Marc; and Raduchel, William. "A Long-Run Cost Function for the LocalService Airline Industry: An Experiment in Nonlinear Estimation." Review of Economics andStatistics 51 (August 1969): 258-70.Freeman, Richard B. The Market for College-trained Manpower: A Study in the Economicsof Career Choice. Cambridge, Mass.: HarvardUniversity Press, 1971.Friedman, Milton. "The Role of Government inHigher Education." In Economics and the PublicInterest, edited by Robert A. Solo, pp. 123-44.New Brunswick, N.J.: Rutgers University Press,1955. Reprinted as chap. 6 of Friedman, M.Capitalism and Freedom. Chicago: UniversityChicago Press, 1962.Friedman, Milton. "The Higher Schooling in tAmerica." Public Interest, no. 11 (Spring 1968),pp. 108-12.Griliches, Zvi, and Mason, William M. "Education, Ability, and Income." Journal of PoliticalEconomy, 80, no. 3 Part II (May/June 1972):S74-S103.Hansen, W. Lee, and Weisbrod, Burton A. Benefits, Costs, and Finance of Public Higher Education. Chicago: Markham, 1969.Hartman, Robert W. Credit for College: PublicPolicy for Student Loans. New York: McGraw-Hill, 1971.Hirshleifer, J. "The Private and Social Value ofInformation and the Reward to Inventive Activity." American Economic Review 61 (September1971): 561-74.Jencks, Christopher, and Riesman, David. TheAcademic Revolution. New York: Doubleday,1968.Jenny, H. H., and Wynn, G. R. "ExpenditureExpectations for Private Colleges." In The Eco nomics and Financing of Higher Education in theUnited States: A Compendium of Papers, pp.440-64. Congress, Joint Economic Committee91st Cong., 1st sess., 1969. (a)Jenny, H. H., and Wynn, G. R. "Short-Run CostVariations in Institutions of Higher Learning."Ibid., pp. '261-94. (b)Kaysen, Carl. "Some General Observations onthe Pricing of Higher Education." In HigherEducation in the United States: The EconomicProblems, edited by S. E. Harris, pp. 55-60.Cambridge, Mass.: Harvard Univ. Press, 1960.Kessel, R. "The A.M.A. and the Supply of Physicians." Law and Contemporary Problems 35(Spring 1970): 267-83.Knight, Frank H. Risk, Uncertainty and Profit.Boston: Houghton Mifflin, 1921.Koerner, J. D. "The Case of Marjorie Webster."Public Interest, no. 20 (Summer 1970), pp.40-64.Manne, Henry G. "Scholars vs. Profits." Barron's,August 25, 1969, p. 1.O'Neill, June. Resource Use in Higher Education.Berkeley, Calif.: Carnegie Commission on HigherEducation, 1971.Orwig, M. D., ed. Financing Higher Education:Alternatives for the Federal Government. IowaCity, Iowa: American College Testing Program,1971.Plato. "The Apology." In The Last Days, ofSocrates, translated by Hugh Tredennick, pp. 45-76. Baltimore: Penguin, 1959.Roush, G. J. "What Will Become of the Past?"In "The Future of the Humanities." Daedalus98 (Summer 1969): 641-53.Schultz, Theodore Vf. Invest men t in Human Capital: The Role of Education and of Research.New York: Free Press, 1971.Shell, Karl. "Notes on the Educational Opportunity Bank." National Tax Journal 23 (June1970): 214-20.Shell, Karl; Fisher, Franklin M.; Foley, DuncanK.; and Friedland, Ann F. "The EducationalOpportunity Bank: An Economic Analysis of aContingent Repayment Loan Program for HigherEducation." National Tax Journal 21 (March1968): 2-45.Tobin, James, and Pugash, James. "The Eco-92nomics of the Tuition Postponement Option."Yale Daily News, February 10, 1971.U.S. Panel on Educational Innovation. Educational Opportunity Bank (Zacharias CommitteeReport). Washington: Government Printing Office, 1967.IN MEMORIAMDr. Fred L. Adair (M.D. Rush '01), the MaryCampau Ryerson Professor Emeritus of Obstetrics and Gynecology, died February 13, 1972.A. Adrian Albert (S.B. '26, S.M. '27, Ph.D. '28),'the Eliakim Hastings Moore Distinguished ServiceProfessor of Mathematics and former Dean of theDivision of Physical Sciences (1962-71), diedJune 6, 1972.William Benton, Trustee and former Vice-President, died March 18, 1973.Dr. William Bloom, the Charles H. Swift Distinguished Service Professor of Anatomy andBiophysics, died May 11, 1972.Kendall Cady, Director of Housing and RealEstate, died December 12, 1971.Kenneth X. Carey, University architect, diedJanuary 10, 1973.Kenneth P. DuBois, Professor of Pharmacologyand Director of the Toxicity Laboratory, diedJanuary 24, 1973.Dr. Lloyd A. Ferguson (M.D. '60), AssociateProfessor of Medicine and Assistant Dean ofStudents in the Division of the Biological Sciences and The Pritzker School of Medicine, diedJanuary 1, 1973.Elizabeth R. Graves (Ph.D. '40), a former staffmember of the Metallurgical Laboratory at theUniversity, died January 6, 1972. She was withthe group of scientists who started Los AlamosScientific Laboratory in 1943.Arthur B. Hall, Life Trustee of the Universitysince 1951, died January 5, 1972. He was nameda Trustee in 1932.Nellie X. Hawkinson, Professor Emeritus of Nursing Education, died October 7, 1971. Weisbrod, Burton A. "Education and Investmentin Human Capital." Journal of Political Economy70, suppl. (October 1962): 106-23.Yale University. "1971-72 Tuition PostponementOption Plan," and "Questions and Answers: TheYale Tuition Postponement Option." Brochures.New Haven, Conn., July 1, 1971.Dr. Hans H. Hecht, the Blum-Riese Professor ofMedicine and Physiology and Chairman of theSection on Cardiology, died August 12, 1971.Dr. Eleanor M. Humphreys (M.D., Rush, '29)Professor Emeritus of Pathology, died December28, 1971.Frank H. Knight, the Morton D. Hull Distinguished Service Professor of Social Sciences andPhilosophy, died April 15, 1972.Stephanie Shambaugh Kramer, a former landscape consultant for the University and a member of the Women's Board and the Board of International House, died February 20, 1973. Shewas the wife of Ferd Kramer, Life Trustee.Dr. Alex E. Krill, Professor of Ophthalmologyand in the Committee on Genetics, died December 8, 1972.Simeon Leland, former Professor and Chairmanof the Department of Economics, died December 7, 1972.Stanley B. Langrand, Assistant Comptroller, diedAugust 1, 1972, after a 25-year affiliation withthe University.Col. William J. Mather, Assistant Secretary of theBoard of Trustees and formerly Bursar of theUniversity, died February 3, 1973.Maria Goeppert Mayer, former Professor of Physics and in the Enrico Fermi Institute, died February 20, 1972.A. Wayne McMillen (Ph.D. '37), Professor Emeritus in the School of Social Service Administration, died March 9, 1972.Samuel Henry Nerlove, Professor Emeritus in theGraduate School of Business, died February 13,1972.93Carl A. Nighswonger, Director of ChaplaincyServices at the University Hospitals and Clinics,and Research Associate (Assistant Professor) inthe Divinity School and the Division of theBiological Sciences, died May 13, 1972.W. R. Rossman, Assistant Comptroller, died December 8, 1972.Edward L. Ryerson, Life Trustee of the University, died August 2, 1971. He was named a Trustee in 1923 and was Chairman of the Board ofTrustees from 1953-56.Mrs. Nora Ryerson, member of the Women'sBoard and widow of Edward L. Ryerson, diedDecember 30, 1971.Susanne Schulze, Professor Emeritus in the School of Social Service Administration diedJanuary 27, 1972.Keith C. Seele (Ph.D. '36), Professor Emeritus ofEgyptology and an authority on Near EasternStudies, died July 24, 1971.Henry F. Tenney (Ph.B. '13, S.O. '15), LifeTrustee of the University, died September 1,1971. He was named a Trustee in 1945 and laterserved as Vice-Chair man.Bernard Weinberg (Ph.B. '30), the Robert M.Hutchins Distinguished Service Professor of Romance Languages and Literatures and in the College, died February 13, 1973.George W. Wheland, former Professor of Chemistry, died December 29, 1972.94THE UNIVERSITY OF CHICAGO RECORDOFFICE OF THE VICE PRESIDENT FOR PUBLIC AFFAIRSRoom 200, Administration BuildingHXwa<NHHooXI— (nossmoooop*OQ0oONo zT I om S c 3J3 £? TJ2 > «*1 w sPOSTAGAIDiO,ILLINTNO.31 o(Q»3N-tO m U¦* 7Z OCO 3