The Cost of Undergraduate Education at a Research University

By Charles Schwartz
Department of Physics
University of California, Berkeley, CA 94720
updated September 11, 2005
Posted at


The actual expenditure for undergraduate education at the University of California  this year is calculated to be an average of $6,847 per student; and the current level of student fees amounts to 100% of that cost.

This result contrasts sharply with the figure $15,810 published by university officials as the Average Cost of Education in 2004-05. The discrepancy arises because I disaggregate undergraduate education from the whole bundle of academic functions, which includes all levels of education plus faculty research.

The methodology used is discussed in some detail, as is the actual calculation. The main data inputs come from official University accounting reports as well as a Faculty Time-Use Study conducted some years ago. The implications of this result are also discussed.

I. Introduction

     Does it make sense to calculate the cost of undergraduate education at a research university, as separated from the cost of the other core academic functions – graduate education and faculty research?  That question is the main focus of this paper.

     Such a separation of costs is not customarily done (or at least not customarily published) by either the private or public institutions that we recognize as research universities; and the reasons for this will be noted later on.

     So, why should anyone care about this?  

     At the leading public research universities (among which the University of California has been a leader, if not a role model), it used to be that state funds provided an almost free undergraduate education of high quality to all eligible students.  About 15 years ago, this funding pattern was changed.  Students were asked to pay a share of the cost of their education, since they received substantial personal benefits from it; but the state continued to provide a significant subsidy for their education.  Now, according to the result reported here, that subsidy has run out: undergraduate students (and their families) are now paying for all of the actual cost of what the university spends on them.  If student fees continue to rise, as everyone expects, then we will shortly be in an entirely new environment – one akin to private universities – where the undergraduate fees are required to subsidize faculty research and related graduate programs.

     That is quite a significant turnaround, de facto, in public policy; but it has not been publicly recognized as such.

     The idea of “unbundling” the research and teaching components of faculty work and reward has been suggested as a way to increase overall productivity in our universities [1].  Such suggestions, however sensible, are not expected to be readily adopted by established faculty or administrators.  The present paper, by contrast, is focused on a concrete question about existing financial expenditures; and it also brings in a new party of interest – the undergraduate students (and their parents) who now pay so much of the bills at public, as well as private, universities.

     Earlier study of college costs [2] encountered a major obstacle in trying to separate out teaching from research activities of the faculty (the so-called “joint production problem”).  In the present work, the availability of the Faculty Time-Use data overcomes that problem.

     The question posed here is primarily one of accounting, an exercise in rationality. In Section II I review the relevant basic data and terminology and in Section III I present the methodology, with further chewing over in Section IV.  Section V contains the details of the calculation; and further discussion is in Sections VI and VII.  Three appendices contain additional data.

     The question of pedagogy – Could/Should one separate faculty teaching from faculty research? – is one that can generate a great deal of heated debate; but it is of minor relevance here because our interest is in Cost Accounting, not Philosophy. Nevertheless, I state my own views on this topic: I believe there is minimal connection between faculty research and undergraduate education; however, I believe that graduate education in the PhD program is mostly inseparable from faculty research.

II. Basic Concepts, Data and Terminology

     The University of California (UC) has a three-part mission: Teaching, Research and Public Service.  All three start with the work of the faculty, and are supported by staff and institutional infrastructure.  There is no difficulty in establishing the total annual cost of UC operations – this is well monitored by the accounting offices and published in the Annual Financial Report: Expenditures of Current Funds.  More details of this accounting can be found in the Campus Financial Schedules (CFS) [3] published annually by the University of California Office of the President (UCOP).

     Both the University’s accounting and budgeting systems rely upon a set of uniform classification categories (established by NACUBO [4]), which at the top level are listed in Table 1, along with the actual expenditures for 2003-04.

Table 1.  University of California; Current Funds Expenditures 2003-04
                        ($ in Thousands)
Public Service
Academic Support
Medical Centers
Student Services
Institutional Support
Operation & Maint. of Plant
Student Financial Aid
Auxiliary Enterprises

Source:  Reference [3]: UC Campus Financial Schedules, 2003-2004; Schedule 12-I.  These figures have not been adjusted (downward) for Expense Capitalized or for Scholarship Allowance. This table does not include expenditures for the three Laboratories operated by UC for the US Department of Energy.

     It is most important to understand the definition of the Instruction category.  This is defined in the UC Accounting Manual [5] as follows:

This category includes expenditures for most activities that are part of an institution’s instruction program, including expenditures [for] the following:
• Subject A;
• academic, occupational and vocational instruction (credit or noncredit courses), for regular, special, or extension sessions;
• departmental research and public service that are not separately budgeted; and
• expenditures for department chairpersons who are also instructors.

     The definition given in the budget is even more explicit (pages 77-79 of the Regents’ Budget for 2005-06) [6]:

[“General Campus” means excluding the Health Sciences, which are budgeted separately.]
The general campus Instruction and Research (I&R) budget includes direct instructional resources associated with schools and colleges located on the nine UC general campuses.  ...  The major budget elements and their proportions of the general campus I&R base budget are: faculty and teaching assistant salaries and benefits, 57%; instructional support, 38%, which includes salaries and benefits of laboratory assistants, supervisory, clerical, and technical personnel, and some academic administrators, as well as costs of instructional departmental supplies; and instructional equipment and technology, 5%.

     So it should be clear that spending for Instruction is really spending for all the work of faculty members (teaching and research and public service), along with supporting staff, supplies and equipment at the departmental level.  This is sometimes referred to as the “core” academic function or the “core” academic budget in the University.  Instruction, the accounting category, covers much more than just teaching.

     The next two accounting categories listed in Table 1, Research and Public Service, cover additional expenditures, using specific (often external) sources of funds, to expand programs beyond what faculty members themselves can carry out.  The next three categories (Academic Support, Medical Centers, Student Services) cover important support activities; the following two categories (Institutional Support and OMP) cover overhead services of administration and facilities management: the last two categories are either “pass-through” of funds or other self-supporting activities.

III. Methodology

     With this background of terminology, let us now see what UC’s Vice President for Budget, has provided [7] as the methodology used to calculate “the Average Cost of Education” at $15,810 per student.
(I have separated his paragraph into three sentences for later reference.)

(i) The denominator is the total number of General Campus FTE students.  
(ii) The numerator includes direct expenditures for Instruction and the General Campus share of Instruction-related expenditures, such as Student Services, Libraries, Organized Activities, Institutional Support, and Operation & Maintenance of the Physical Plant.  
(iii) Public Service, Research and Financial Aid expenditures are excluded, along with the Public Service and Research-related indirect costs of Libraries, Organized Activities, Institutional Support, and Operation & Maintenance of the Physical Plant.

This should be fully understandable, once I explain that in budget language “Libraries” and “Organized Activities” are the two components of the Academic Support category seen in Table 1.

     This talks of “expenditures”, so I will use the accounting documents rather than budget documents to get the most reliable numbers.  It says “General Campus,” so we are excluding the Health Sciences.  The direct costs come from the Instruction category, with supporting or indirect costs from the other categories mentioned (ii).  This is explicit about which categories are to be excluded (iii).  

     This makes good sense, except that the result is misleadingly named:  this is not the Cost of Education but rather the Cost of the Core Academic Functions.   This is the total cost of Undergraduate Education plus Graduate Education plus Faculty (Departmental) Research.

     Now we are ready for my question. Is it sensible to separate out from this total cost, the cost of Undergraduate Education?  Let me note and answer some anticipated objections.

Objection #1.  You can’t separate the various parts of a professor’s job any more than you can separate those parts of his/her brain.  

     In fact, this university has a very well-established procedure for evaluating individuals for appointment or promotion in the professorial ranks.  The Academic Personnel Manual [8] lays out the criteria for such evaluation in four separate domains:

The review committee shall judge the candidate with respect to the proposed rank and duties, considering the record of the candidate’s performance in (1) teaching, (2) research and other creative work, (3) professional activity, and (4) University and public service.

     It is worth noting that the criteria on Teaching include the requirement that  faculty keep up with developments in their discipline - which tells us that this aspect of being a good teacher is separate from one being actually involved in research. Here is the relevant statement.

In judging the effectiveness of a candidate’s teaching, the committee should consider such points as the following: the candidate’s command of the subject; continuous growth in the subject field; ability to organize material and to present it with force and logic; capacity to awaken in students an awareness of the relationship of the subject to other fields of knowledge; fostering of student independence and capability to reason; ...

     It is further worth comparing the criteria for Professors to those for Lecturers, who are hired only to teach, not to do research.  The statement quoted above – the criteria on Teaching – is exactly the same.  One may also note that about half of all undergraduate courses throughout UC are taught by Lecturers, rather than by Professors, with no known loss of quality in the educational program. (See Table 3a.)

     Thus there seems to be little substance to this objection. Teaching is an activity which is separable from research without loss of quality, at least for undergraduate education.  At the level of graduate education, especially in the PhD programs, there is a strong case to be made that the separation of teaching from research – both on the part of the professor and on the part of the student – is much more problematic. I share that view and do not attempt to make such a separation at the graduate level.

Objection #2.  There is no objective way in which to make a meaningful quantitative separation.

     In fact, this is a well-known issue in Economics and Business Management.  The subject of Cost Accounting is well established.  In a factory that makes several products, one can assign costs to each one on the basis of raw materials; but what about the labor costs of those employees who may work on more than one product line?  The answer is: Weigh their salaries according to the fraction of their work time devoted to each product.  This approach (called Activity-Based Costing or ABC) requires the company to conduct a survey of how their employees devote their work time.

     In the case of the UC faculty, there already exists a Faculty Work-Time Study [9], ordered by the University administration about two decades ago; it was conducted by outside professionals who had extensive consultation with faculty leaders. Their results, as summarized in Appendix A, lead to the conclusion that approximately 23% of faculty work time is devoted, on average, to undergraduate education, as distinct from other work. (Appendix C contains additional information which confirms this central piece of data.)

Objection #3.  That work-time formula may be sensible for a factory, but the professors at a university are a much more sophisticated and complicated work-force, not so simply measured.

     Let’s look at an alternative measuring process. As I learned from an article in The New Yorker magazine [10], the US medical profession has been subjected to a detailed cost-accounting study on behalf of the federal government, which pays for so much of the country’s health care expenses.  A complex set of formulas has been derived to give the relative costs to be assigned to each individual medical procedure that doctors are called upon to perform.  For each task, not only the work-time, but also other factors – “mental effort and judgment, technical skill and physical effort, and stress” – are folded into the relative cost.   Certainly, professors’ efforts at graduate education and at research work involve a greater degree of those sophisticated talents and sufferings than does undergraduate teaching.  So, in every category – time spent, level of expertise required, etc. – the undergraduate education would come out as the smallest component of faculty cost; but the numerical value of the final outcome would depend on how these different elements were weighted and how small the undergraduate components are. I think a good case can be made that time spent is by far the most important factor, i.e., should be given by far the greatest weight – from the faculty’s own perspective.  This I would support by reference to faculty teaching loads as a most critical factor in recruitment and retention of research professors: teaching should not be allowed to take too much of their time away from concentration on their research work. In conclusion, I expect that such a more elaborate measurement of the fraction of total faculty cost attributable to undergraduate education would differ little from the simple time-only measure that I have used. In any case, I shall rely here upon the best data now available.

Objection #4.  UC’s Vice President Hershman has written that “any attempt to disaggregate the cost of education will be based upon questionable assumptions and that any resultant numbers will be of little value.” [7]

     Given any new result, it is proper to engage in a detailed discussion of the many technical assumptions that go into the calculation; and let us see what kind of rational conclusions emerge.  Section V gives considerable details on the calculation (none of which have been criticized to date) and I am prepared to go deeper into any related questions.  An honest evaluation of the residual uncertainties in any such calculation does not invalidate, but rather better informs, the results of those calculations. Such an intellectual discourse has been my repeated request to University officials.

IV. But …

     But, my colleagues will say, surely the research work of this faculty must contribute some value to the undergraduate education. If not, how are we, in the research university, giving anything better than what undergraduates can get in an ordinary four-year college?

     My first (mechanical) reply is this: This differential is already built into my calculation, since the faculty at a top research university will get a higher salary than those at a lesser-ranked institution.  Of course, professors at a top research university spend relatively less of their time at undergraduate teaching than do those at a liberal arts or other four-year college. (So, although I do not have the data to prove it, it may well be true that the unit cost for undergraduate education is less at UC than at, say, CSU.)

     But, still, you may ask about some additional cost that ought to be added to my calculation of undergraduate education, in order to account for the “added value” that research activity of the professor brings into the classroom.  This requires further explication.

     I have already noted (see data in Table 3a) that almost half of all undergraduate classes throughout UC are taught by Lecturers, who are hired only to teach and not expected to do research. There is no lack of quality found here; so “research activity” is not essential at the undergraduate level. But I must do some more economic discussion to make this issue clear.  Let me offer some model situations for comparison.

Model 1.  Suppose I own a business, which has been successful, and you want to buy it.  We can probably agree on the value of its tangible assets; but what about the intangible values of reputation, brand name, etc.?  There we can only argue and reach, or fail to reach, some compromise.

Model 2. Suppose I own a business that has several products to sell. How do I decide what price to attach to each of them?  I can use the methods of Cost Accounting (as described above) to apportion my actual cost, direct cost plus indirect cost, for each product; and then I can add some amount of profit to each one in any way I choose.  Of course, market competition may be an important factor in my setting of prices.  Also, it is up to me to decide if I might want to make one product subsidize the cost of another. This is private business and I don’t have to justify my prices to anyone.

Model 3.  Suppose I want to create (or already own) a private research university; and I have to decide what amount of tuition to charge undergraduate students. This is analogous to Model 2: the money paid by those students should not only cover the actual cost of the education I provide for them, it must also support the faculty research and graduate (PhD) programs. (Professional schools, medicine, business, law, are other graduate programs where I charge high fees to cover the total cost of maintaining their faculties.) This is, again, private business and I don’t really have to justify my prices to anyone.  There is no profit motive here (however, see the critique by Sowell [11]) although there may be market/competitive forces that influence my choice of tuition charges.

     Do you see anything above that might be applicable to the situation of a public research university? I do not.

     The problem addressed in this study is the following:  Given the known actual expenditures of this institution, we want to disaggregate, as honestly as possible, the total cost of core academic functions so as to determine the actual cost of undergraduate education.  In the private sector models, the owner can load on all kinds of charges to be taken in as profits or cross-subsidies; but here we have a given total cost and the separate parts of this must add up to 100%.

V. The Calculation

     We now follow the methodology described above, starting with the outline given by UC’s Vice President for Budget and taking the portions appropriate for undergraduate education.  This work starts with the data shown in Table 1 and then digs deeper into the official UC accounting records (CFS 2003-04 [3]) for more details.

     First, we will extract the portion of expenditures for Instruction that go to undergraduate teaching; then we will add in a portion of Academic Support (Libraries, etc.) and a portion of overhead expenses (Inst. Supp. and OMP). Finally we will add in expenditures for Student Services (not covered by other types of fees) and compare the result, scaled “per student”, to the current level of mandatory student fees paid by California resident undergraduates.

Step 1.  Using CFS Schedules B for each campus, remove from the Instruction category the total expenditures for University Extension ($198 million) and also all expenditures by the Health Professions ($1,055 million), Law Schools ($39 million) and Business Schools ($111 million). This leaves us with an adjusted expenditure for Instruction of $1,559 million.

Step 2.  Estimate, using various data sources, the portion of this $1,559 million paid to Lecturers and Graduate Student Instructors for undergraduate courses ($180 million.)  Then take 23% of the remainder, for the fraction of faculty work-time devoted to undergraduate teaching.  This gives us the Direct Cost of Undergraduate Instruction = 180 + 0.23x(1559-180) = $497 million.

Step 3. After removing the Health Sciences component ($549 million) and some other unrelated items in the Academic Support category, we are left with expenditures for: Libraries ($234 million), Computer Support ($43 million), Academic Administration ($197 million).  I estimate that half of the Library and Computer Support expenditures may be charged to undergraduate education, along with one-quarter of that for Academic Administration (Deans’ Offices).  This gives an amount $188 million, adjusted Academic Support, to be added to the result of Step 2. (Comment: From the Budget [6], page 177, I learn that the Library budget allots 57% to acquisitions and 38% to reference-circulation. The first is mostly for research journals while the second probably serves mostly undergraduate students. This leads me to the 50-50 estimate used above.)

Step 4.  Combine the expenditures for Institutional Support and Operation and Maintenance of Plant, and distribute this uniformly, as overhead, over all the listed categories, excepting Medical Centers and Student Financial Aid, in Table 1. This gives an overhead rate of 10.7%.  Adding this to the sum from Step 2 and Step 3, gives us the Total Cost of Undergraduate Instruction (2003-04) = $758 million.  We divide this by the total number of undergraduate students (158,230 budgeted FTE for 2003-04) to get the Unit Cost of Undergraduate Instruction = $4,790 per student.

Step 5.  The total expenditure for Student Services needs to be reduced by $109 million, the amount recorded for Student Health Services, because this is paid for by non-mandatory fees for Health Insurance. The net amount should then be increased by the 10.7% for overhead and then divided by the total number of students (graduate and undergraduate) for whom these services are provided: net cost = $1,858 per student.

FINAL RESULT: Add the results of Step 5 and Step 4 to get
The Unit Cost of Undergraduate Education = $6,648 per student (for 2003-04)

     For the current academic year (2005-06) there is a 3% increase in overall budget, so we can project this result: Unit Cost = $ 6,847 per student in 2005-06.
The average mandatory fees this year for resident undergraduate students at UC = $6,817 per student. This is 100% of the actual Unit Cost as calculated above.

     It is obvious that there are some questionable estimates in this calculation and therefore there is room for better data to lead to a more precise determination of this cost. Perhaps the most sensitive question would be: How reliable is the 23% figure derived from the Faculty Time-Use Study?  Some data shown in Table 3b indicates that not much has changed in faculty teaching over a recent decade’s time; and the additional data in Appendix C does confirm this critical number. One might note that the  Time-Use survey excluded faculty who were on sabbatical leave, and that would imply a downward adjustment from 23% to 21%.  There is certainly room for further study, and perhaps negotiation, on a number of small points. However, I believe we are in the right ballpark with the result stated above.

VI. Discussion

     This result, that undergraduate fees at UC are now at the level of 100% of the actual cost of undergraduate education, is new and unsettling.

     One reason this is new (and a surprise to many) is the awkward history of relations between public research universities and their publics, the legislators and taxpayers of the state that pays their bills.  Faculty know that we are hired for, promoted for, and motivated by our research work first and foremost.  Our academic administrators know that the legislators and the taxpaying public are primarily interested in the undergraduate education we offer.  To avoid having to resolve that mismatch of understandings, budgets are presented in a big package format: undergraduate teaching and graduate teaching and faculty research are bundled into one big ball – the I&R budget.

     One can say that this was a harmless lack of candor, since all of those functions, at the public research university, were seen as “public goods” supported by public money.  What is new is the recent rapid rise of student fees.  During the budget crises of the early 1990s, it was quickly said, and accepted, that students should bear some portion of the cost of their education, since they gained individual benefit from it.  A public good was transformed into a private good.  

     The new question, not previously asked, is this: Is there some logical, moral or political limit to how high student fees (tuition) could be raised at a public research university?  A fee level at 100% of the actual cost of their education seems to provide a most significant boundary.  Beyond that point, undergraduate students (and their parents) are being forced to not only pay for everything they get, but must also subsidize the other functions – graduate education and faculty research – which undeniably provide broad benefits to the whole public, beyond the individual student.  One could call this selective taxation, privatization, or any other name you wish; but such a forced subsidization is something that deserves  serious debate as a matter of public policy.

     While the calculation done here is for the University of California (averaging over all its campuses) the methodology should be applicable to any major research university.  The expenditures data and student numbers would be different for another school, although the NACUBO accounting categories provide a uniform standard.  In addition, I would think that the data from the UC Faculty Time-Use Study is probably reasonable to use at comparable institutions because faculty teaching loads are a most sensitive benchmark in the competition for top-quality research faculty.

VII. Financing the Universities -- The Big Picture

     The leading research universities in this country are usually divided into two groups: Private and Public. The self-selecting Association of American Universities has 34 Public institutional members (including six UC campuses) and 26 Private members in the US.  It is instructive to note in what ways these two camps are alike and are different.

     The overall work situations of faculty members are pretty much the same, since these institutions compete with one another to hire the best research professors.  This means that teaching loads (in each discipline) are pretty much the same; and salaries are not very far apart.

     Looking at graduate students in the PhD programs, again one sees pretty much the same financial and working conditions, regardless of the Private/Public status. The basic financial fact here is that these students pay little or nothing in tuition or fees in those disciplines with substantial external research funding.

     The features mentioned above were true in the good old days when state tax revenue was generously applied to supporting the Public universities; and they continue to be true today in a very different public-funding climate.  The features mentioned below have changed dramatically since the early 1990s.

     Looking at the elite professional graduate programs – medicine, business management and law – the students are charged very large tuitions/fees; and the rate at the Public universities is approaching that at the Privates. Privatization is well under way in this sector.

     Finally, we look at undergraduate education.  What was once nearly “free” in the Publics is now rather costly, but still priced far below the level charged by the Privates. The “fees” charged to undergraduate students at the Public universities have been rising rapidly and one may well ask how far this “convergence” may proceed before the significant distinctions between Public and Private disappear.

     First, we must ask: What is the basic difference – philosophical or sociological – between the two camps?  Foremost, there is the contrast of exclusivity (at the Privates) and inclusivity (at the Publics) in regard to their acceptance of academically qualified undergraduate students.  Therefore, one may well worry that as undergraduate fees at the Publics continue to increase, there may be a substantial shift in the economic class background of the undergraduate student body; and this may lead to more rapid erosion in state funding; and this in turn will accelerate the rise in student fees; etc.  Such a transition – the Privatization of undergraduate education at the Public universities – will have profound implications for the stratification of American society: access to top quality college education will be more and more reserved for those students from more affluent families, thus preserving the class status of those families.  The American Dream of class mobility through educational opportunities will shrink or nearly disappear. (There are important financial aid programs for lower income students; their adequacy is the subject of a very separate debate.)

     With that likely picture in mind, we ask: Is there some principled line that could be drawn to limit the level of undergraduate student fees at our leading Public universities?  A most concrete answer is the focus of the present study: Undergraduate student fees should not exceed the actual average cost, expended by the institution, for undergraduate education, as that activity of the university may be disaggregated (through objective cost accounting) from the whole bundle of academic missions.

     The consequences of adopting such a principle would be something quite unsettling to university administrators, who would have to recast their entire mode of financial management: they would lose the pliant “cash cow” that eager students (and their families) now provide as alternatives to state appropriations.       Faculty members, as well, might see this as quite a threat: Who will then pay the salaries for our research endeavors?

     On the other side, the consequences of not adopting such a principle would seem to imply the sure erosion, and perhaps actual loss, of that guarantee of access to all future students who are academically qualified, regardless of their financial status.

     A tough choice is there to be addressed; and that should be done openly and honestly.  Our duty is to not allow it to remain hidden.

     My recommendation is that the leaders of our public (and private!) research universities publish disaggregated calculations of what they actually spend on undergraduate education, as I have done for UC.  This may make their jobs more difficult – restraining the impulse to raise student fees to cover the escalating cost of research – but it will be more honest.  It will also help to preserve those principles of access and affordability, which have formed the basic mission of our public universities.

     In recent years there has been a general call for more “transparency” (and, of course, more honesty) in the financial reports issued by large corporations throughout this country.  The present work is a contribution along those lines directed at the institutions of higher education.


     I am grateful to Patricia Cross, John Douglass, Laura Nader and Brian Pusser for their generously given comments and advice on earlier drafts of this paper.


Many UC documents are available on the Internet, as cited below.

This paper is an expansion of the earlier report posted on my website,
Financing the University – Part 7 (11/28/04)

[1] James S. Fairweather, “The Mythologies of Faculty Productivity,” The Journal of Higher Education, Vol. 73 No. 1, 2002

[2] Gordon C. Winston, "A Guide to Measuring College Costs," Chapter 3 in Middaugh, Michael F. (ed.)  Analyzing Costs in  Higher Education:  What Institutional Researchers Need to Know.  Jossey-Bass/NDIR, 2000.

[3] UCOP, Campus Financial Schedules:

[4] NACUBO = National Association of College and University Business Officers

[5] UCOP, Accounting Manual; Chapter U-751-17:

[6] UCOP, Budget for Current Operations:

[7] Letter from Vice President Larry Hershman, dated December 21, 2004:

[8] UCOP, Academic Personnel Manual; APM 210:

[9] Institute for Research in Social Behavior, “University of California Faculty Time-Use Study 1983-84,” University Archives, Bancroft Library, UC Berkeley 308kd.tu

[10] Atul Gawande, “Piecework: Medicine’s money problem,” The New Yorker magazine, April 4, 2005, page 44.

[11] Thomas Sowell, “The Scandal of College Tuition”, Commentary 94:2 August 1992, page 23.

[12] University of California, “Undergraduate Instruction and Faculty Teaching Activities; Ninth Annual Report to the Legislature”, April 2001:

[13] UC Ladder Rank Faculty by Discipline, October 2004.

[14] UC Non-Senate Instructional Unit Memorandum of Understanding; Article 24 – Instructional Workload.

Appendix A.

Table 2a.  UC Faculty Time-Use Study 1983-84
Principal Activity
All UC-related Activities

Instructional Activities
Original Research/Creative Activities
University Service
Professional Activities/Public Service

Table 2b. Details of Instructional Activities
Details of Instructional Activities
Regularly Scheduled Courses (classes, labs, fieldwork, giving exams)*
Supervising Independent/Special Study (197-199, 297-299, etc. courses)*
Non-credit Instruction (seminars, colloquia, symposia)*
Student Advising
Giving Oral Exams
Course Preparation
Other Instructional Activities (preparing & grading exams and papers,

confer with TAs, letters of rec., reading theses, committee discussions)
                   Total Instructional Activities

Table 2c. Further Breakdown of “In-Class” Hours/Week by Level of Instruction

Lower Div.
Upper Div.
Graduate Level
Regularly Scheduled Courses
Supervising Independent/Special Study

Source:  Reference [9].
This survey sampled 100% I&R FTE faculty, excluding those in health sciences and law.
*This counts time only when the professor was present.
Contact with graduate student research assistants was counted under “Research” time, although such students generally registered in a course ”295. Research”.

     I take 1/3 of the University Service work time from Table 2a (2.2 H/W) and add this to the 26.0 H/W listed there for Instructional Activities.  This sum (28.2 H/W) amounts to 46% of the total University work-time. Now from Table 2c I see that faculty teaching time is equally divided, on the average, between teaching undergraduate classes and teaching graduate classes.  Lacking any better data, I will take this 50-50 split and apply it to the total of faculty teaching work-time.  The result: UC faculty devote, on average, 23% of their work time to undergraduate instruction.

Appendix B

Here is some UC Data on Formal Instructional Activities, over a period of time.
Primary Classes include only unit-bearing, regularly scheduled course offerings.

Table 3a. Primary Classes – Number in Thousands
All = Classes taught by All Instructors.
RRF = Classes taught by Regular-Rank Faculty only.

Table 3b. Primary Classes Taught by Regular-Rank Faculty – per FTE

U+G = Undergraduate and graduate classes.
U = Undergraduate classes only.

Table 3c. Independent Study Enrollment by Regular Rank Faculty – per FTE


Source: Data from Tables on pages 100-104 of Reference [12].

Appendix C

     Here is further information that lends confirmation to the main result of the Faculty Time-Use Study, namely, that UC faculty spend, on average, 46% of their work time at teaching and half of that is directed to undergraduate instruction.

Item 1.  The average teaching load for regular rank faculty at UC is 4 quarter courses per academic year. The data come first from Reference [12], page 109, where there is an official listing of the formal teaching loads for UC faculty in various disciplines: Humanities, Social Sciences, and Mathematics – 4 to 5 quarter courses per year; Engineering and Computer Science, and Physical Sciences – 3 to 4 quarter courses per year; Biological Sciences – 3 quarter courses per year. Further data on the distribution of UC faculty by discipline [13] allows me to calculate a weighted average; the answer is 4.0 quarter courses per year.

     Next we search the University’s Academic Personnel Manual (APM), Reference [8], and other UC documents for relevant statements of employment policy.

Item 2.  According to UC’s agreement with Non-Senate Faculty [14], the normal teaching load for full-time Lecturers is 9 quarter courses per year. Compare this to the average teaching load of regular rank faculty (hired to teach and do research), which is 4.0 quarter courses per year; and we conclude that regular faculty are considered to spend 4/9 = 44% of their work at teaching.

Item 3. APM 661 says that faculty who teach during the summer session are to be paid 17%, 19%, or 22% of their academic year salary rate for teaching two courses during the summer 6-, 7- or 8- week session.  Using the 4.0 quarter courses per year average faculty workload, this leads to the inference that teaching, for regular faculty, on average, accounts for 34% to 44% of total faculty pay.

Item 4. APM 200 says that emeritus faculty who are recalled to teach will be paid at the rate of 1/9 of their base salary at retirement for teaching each regular quarter course.  With the UC average faculty teaching load of 4.0 quarter courses per year, this leads to the inference that teaching, for regular faculty, on the average, accounts for 4/9 = 44% of total faculty pay.

Item 5.  APM 740 is about Sabbatical Leaves. The rate tables of accrual of credit for sabbatical leaves shows that a full-time faculty member, after working for 6 years, can have leave of 2 quarters at full pay or three quarters at 2/3 pay. An alternative arrangement is Sabbatical in Residence, which requires teaching one standard course throughout the year and getting full salary.  This equates the teaching of one course for a year (3 quarter courses, which is ¾ of the normal average faculty teaching load of 4 quarter courses per year) with 1/3 of annual faculty salary.  Restated, this equates the full normal average faculty teaching duties to 4/9 = 44% of total annual faculty salary.

     There is impressive consistency in these several items of UC policy and practice; and that figure 44% is remarkably close to the value derived from the Faculty Time-Use Study.

     Finally, we seek further data on the separation between undergraduate and graduate teaching efforts of regular rank faculty. Data in Appendix B is useful here.  From Table 3b (for the most recent year) we find that regular rank faculty teach Primary Classes to undergraduates 2.5/4.9 = 51% of the time; and Table 3c shows that they teach undergraduate students in Independent Study Classes 4.6/14.8 = 31% of the time.  So for their total teaching load we need to take some weighted average of these two percentages to say what percentage of total teaching effort goes to undergraduates.  One may argue about exactly what weighting to use; but the result will doubtless be something a bit less than 50%.

     This collection of additional data lends strong confirmation to the basic method used in this paper to disaggregate the cost of undergraduate education from other academic functions at UC.