In order to study the cost of accommodating the expected growth in undergraduate enrollment (Tidal Wave II) at the University of California, we start by identifying the key quantities in UC's planning as recently presented to the Regents. The following numbers cover the "general campus" activity of the University (excluding health sciences) over the 12-year period 1998-99 to 2010-11.

UC has estimated that this plan will require capital funding of $500 million per year for growth and upgrading of the campuses. This adds up, over the 12 years, to a total capital cost of $6 Billion.

UC officials have said that their plan will require a substantial commitment of additional resources in the annual operating budget, but no numbers have been given as yet. I can estimate this cost using two different figures for the per-student cost at UC that are found in the Governor's Budget Summary for fiscal year 2000-01. The lower number is derived from the $51.2 million budgeted for next year's enrollment growth of 6,000 FTE students. The ratio of these two numbers, called the marginal cost of instruction, is $8,533/year per FTE student . The higher number for the cost of education at UC is quoted in the discussion of student fees: $18,025/year per FTE student. [Trying to learn why these two numbers are so different, I have written to officials at UC's Office of the President (UCOP) and in the Governor's office, but have received no response.] Using these numbers to calculate the total increase in operating cost over the 12 years, assuming the additional enrollment grows linearly from zero to 63,000 FTE, we get a total operating cost of $3.2 - $6.8 Billion.

The combined cost, in capital plus operating budgets for UC, is thus estimated: between $9 Billion and $13 Billion over this 12 year period.

This is a huge amount of money and it dwarfs all other discussions about how most efficiently to handle Tidal Wave II at the University. For example, a February 1999 report from the California Legislative Analyst's office advocates the adoption of year-round operations (YRO) in order to save the state the expense of building additional instructional facilities. They estimate the saving for UC "could easily exceed $1 billion." This is significant money, but small compared to the costs we have estimated above.

The first and most important observation about the $9-13 Billion cost increase figure is that by far most of this money will be spent for the additional faculty members in UC's plan. Most of the capital cost increase will go for constructing the offices and the research facilities for the additional faculty members and for the additional supporting staff that they will require. (The cost of additional classrooms and other instructional support facilities required by students will be small by comparison.) Most of the operating budget cost increase will go for salaries, benefits, equipment and supplies needed for the additional faculty members and the additional supporting staff that they require. (The cost of additional teaching assistants and other staff directly serving students will be small by comparison.)

Therefore, if one is serious about trying to minimize the overall cost of accommodating Tidal Wave II, then one must ask the question, Can UC do with significantly fewer additional faculty members and still maintain its quality as a world famous research university? This is a heretical notion and it is why I write this paper.
 
 

UC's study of long-range enrollment issues has been centered in the Office of Planning and Analysis under the Provost and Vice President for Academic Affairs. A collection of recent documents on this subject may be found on the Internet at www.ucop.edu/planning/lrenrol.html. These reports comprise about 200 pages of text plus a number of colorful slides prepared for the regents.

Much attention is given, in these documents, to estimating the increase in undergraduate enrollments at UC over the coming decade and discussing a variety of issues about how to accommodate them. In previous years, UC's projections of the magnitude of Tidal Wave II were significantly lower than those calculated by the state's Department of Finance. Now, however, the two seem to be pretty much in agreement: UC predicts an undergraduate enrollment between 167,000 and 189,000 (year-average headcount) by the year 2010. This calculation is based upon demographics of the state's school-age population and the University's commitment to the Master Plan - accepting the top 12.5% of California's high school graduating class each year.

The greatest portion of these documents is devoted to the issue of graduate student enrollments. Here we find a very different kind of story. The University has been wanting to increase its graduate student population for a number of years but has not been able to do so for financial reasons, which I shall return to shortly. This part of the reports reads, quite frankly, like a sales pitch - proclaiming the many benefits, both economic and social, that California will reap from an increased production of PhD's. The reports acknowledge the great uncertainty in attempts to predict the future job market for new PhD recipients and they end up by simply saying that the percentage of graduate students in the overall UC enrollment should be 18%, about what it is today.

The result of adding the anticipated influx of undergraduates and the hoped for buildup of graduate students leads to the result: 63,000 additional FTE students between 1998 and 2010; and this is the number quoted at the beginning of this paper.

What about the number of additional faculty members that UC will need to teach this additional number of students? This topic is not discussed at all in the UCOP reports. The first mention of UC needing "approximately 3,000 new faculty members for enrollment growth" appears (as far as I have found) in the "Regents Briefing," prepared by UCOP's office of News and Communications following the regents' meeting of October 14, 1999. This number was again given by Provost C. Judson King in his progress report to the regents on January 20, 2000, with no indication of how it was calculated. So I am left with the following surmise as to where the number comes from: divide the enrollment increase (63,000) by the faculty increase (3,000) and you get the ratio 21:1, which is close to the university's current student-faculty ratio.

In summary, all the work that the UC administration has done on studying this major issue ends up with a proposal which simply says: reproduce the status quo of UC's staffing and funding arrangements, scaled up to the expected influx of undergraduate students. If I observe that this plan is a "no-brainer", or if I suggest that the new students can be accommodated by a much more economical arrangement, I expect the first response will be something like this: The University of California is the world's greatest public research university and therefore we must be doing things right; any attempt to significantly reduce the size of the faculty, relative to the size of the student enrollment, will surely damage the quality of this great institution and such a sacrifice is not in California's best interests. Let us look into this subject.
 
 

We are aware of much concern over the quality of this nation's public schools. The most common notion is that of class size - the number of students taught (on the average) by each teacher. We recognize that a student-teacher ratio of 20 is a lot better than a student-teacher ratio of 30 (assuming that we are talking about good, qualified teachers here) because, especially for the younger children in school, the amount of time that the teacher can spend personally attending to each individual student is of primary importance for the overall quality of the education provided.

Much of this emphasis on a low student-faculty ratio still makes sense through high school, and there are many excellent colleges which continue to shape their educational mission around a rich personal encounter between teacher and student. But the modern American Research University is a very different kind of institution. I will give some more general discussion a bit later on; now let me just give you some facts for my own campus to illustrate the great variability of such ratios at UC.
 

Thus one should understand that when the UC administration talks about the overall student-faculty ratio - currently budgeted at the value of 18.7 and planned to move "closer to the historic student faculty ratio of 17.6 to one" - this is simply the bureaucratic language which they use for budget negotiations in Sacramento.

So, how does the numerical student-faculty ratio relate to the quality of the education provided to students at UC? There is no simple answer. My purpose here is to take this subject apart and examine the separate components.
 
 

I have never seen any UC document that looks at Graduate Student - Faculty ratios; so I have collected the data from authoritative sources and put it in Table 1.
 
 

Table 1.  UC Faculty & Graduate Students, 1998-99

Campus	Faculty*	Graduate Enrollment **	Ratio
Berkeley	1,406	7,315	5.20
Davis	947	3,172	3.35
Irvine	736	1,865	2.53
Los Angeles	1,520	6,815	4.48
Riverside	480	1,217	2.54
San Diego	773	2,124	2.75
Santa Barbara	852	2,157	2.53
Santa Cruz	520	964	1.85
UC Total	7,237	25,629	3.54

* Faculty General Campus FTE data for October 1998 from Table 1A of, "Annual Academic Personnel Statistical Report 1998-99," and campus-by-campus details of this data, Office of the Provost and Senior Vice President -- Academic Affairs. The definition of "Faculty" I have used here is the one given in the Glossary of this official report.
** Graduate Enrollment data from "2000/2001 Budget for Current Operations," UCOP, October 1999, General Campus Actual Year-Average FTE Enrollments, page 221.

Note the wide variation in the ratios of Table 1 from one UC campus to another - almost a factor of three from lowest to highest. (If one looks instead at total student enrollments, the student-faculty ratios for the individual campuses vary only slightly - they are all within 10% of the overall UC ratio.) Next, lets look at Table 2, which has comparison data for other research universities.
 
 

Table 2.  Comparison of Graduate Student-Faculty Ratios

University of California	3.5
Average of 4 Public Comparison Universities*	4.9
Average of 4 Private Comparison Universities**	5.4

Sources: Letter from L. Hershman, UCOP 2/10/94; and CPEC Report 93-2, p.48
* Illinois, Michigan, SUNY-Buffalo, Virginia.
** Harvard, MIT, Stanford, Yale.

Despite the time lag between these data sets, we can make some general observations. Berkeley is close to the Private comparison universities (ratios 5.2 and 5.4); UCLA and the Public comparison universities are a little bit behind them (ratios 4.5 and 4.9); the other 6 UC campuses have ratios that are significantly lower (1.85, 2.53, 2.53. 2.54, 2.75 and 3.35). What does this mean? The schools that everyone recognizes as the very best research universities have the highest ratio of graduate students to faculty, and this contradicts the commonplace notion that high quality of education correlates with a low student-faculty ratio. To resolve this we must understand the business of graduate study in the PhD program.
 
 

The modern American research university came into being after the second world war, fueled by the military's enthusiasm for scientific research and development. Here is how it works. Professors are hired and promoted primarily for their abilities to conduct research. They seek research contracts and grants from outside agencies (government, private foundations, industries) and use this money to hire graduate student research assistants, postdoctoral fellows and other support staff as well as for equipment and other costs of their research project. The university handles the paperwork of these research contracts and grants, tacking on an overhead cost, much of which ends up as discretionary funds in the hands of the administrators. Teaching of undergraduate students is necessary to provide the income (tuition in the private universities, state appropriations in the public ones) which pays for the professors' salaries and the supporting infrastructure of the university.

Thus one can understand that research professors want lots of graduate students (that is, want someone to provide the money with which to pay for their research assistants) and want to spend as little time as possible teaching undergraduate students (distracting them from their research work, which is the primary basis for the professor's reputation and advancement.) One also understands that the presidents and chancellors of these research universities see their duty in preserving this set of priorities (research first, teaching second) but must speak publicly in a different voice.

Graduate students (in UC's general campus enrollments) are conveniently divided into three categories in the UCOP reports:

This last group, those in the PhD programs, are essential to the research mission of the University. They serve three invaluable functions, in chronological order:
1) As Teaching Assistants (called TA's or GSI's), they teach the discussion and laboratory sections that accompany lecture courses, they grade papers, keep course records, consult with the students, etc. -- all under the supervision of the professor who is in charge of the course. This is valuable training for the graduate students' own professional development; and, most importantly, this arrangement allows the professor to fulfill the university's undergraduate teaching obligation while spending as little time as possible away from research activity.
2) As Research Assistants (called RA's or GSR's), they work on research projects under the direct supervision of a faculty member, leading up to their dissertation. This is the core of graduate students' training to become scholars/researchers in their own right; and this apprenticeship is highly valued for increasing the research productivity of the professor and/or the professor's research group.
3) As new PhD's, they go out into the professional world on their own, bringing credit to the reputation of their university, their department and their mentor. (Professors are evaluated for promotion first on their output of research papers and second on their output of PhD's.)

From this understanding of the role of graduate students in the PhD program, one sees that the professors are, with few exceptions, eager to have more of them (and, of course, want to attract those of the highest possible calibre.) Money is critical in determining the number of graduate students that enroll in any high quality department on any campus. While undergraduate students pay the university for their education, students in the PhD program are offered payment by the university in a variety of ways: Fellowships (mostly from university endowments and government programs, also from private foundations); Teaching Assistantships (mostly from State funding); Research Assistantships (mostly from the faculty's extramural research contracts and grants); fee waivers and fee remissions by the university (where this money comes from is unclear to me.) Of all these, the amount of outside research funds is the most important consideration when the faculty of each department decides each year how many new graduate students they want to admit.
 
 

Table 3. Extramurally Funded Research per Faculty FTE 
(General Campuses) 1998-99 

Campus	Faculty*	Funds ($000)**	Ratio
Berkeley	1,406	225,603+	$160,000+
Davis	947	100,036	$106,000
Irvine	736	48,139	$  65,000
Los Angeles	1,520	114,997	$  76,000
Riverside	480	31,741	$  66,000
San Diego	773	182,082	$236,000
Santa Barbara	852	73,798	$  87,000
Santa Cruz	520	32,700	$  63,000
UC Total	7,237	809,096	$112,000

* Same as in Table 1.
** Data from UC's "Campus Financial Schedules": Expenditures of Restricted Funds for Research, excluding Health Sciences.
+ One should add some portion of the $357 million budget of the Lawrence Berkeley Laboratory, which supports the research of many UCB faculty and graduate students.

Table 3 shows data on the per-faculty amount of this outside ("sponsored research") funding at UC. The ranking of these ratios, comparing one campus to another, is similar to the comparisons in Table 1 (with the notable exception of UCSD). This supports the proposition that graduate students go to where the research money goes.

The UCOP plans to increase graduate enrollments are thus dependent primarily upon the federal government's overall funding levels for academic research, something outside the control of the University or of the State of California.

Finally, what can we conclude about the relation between plans to increase graduate student enrollments and a need to increase the number of faculty? As far as the PhD programs are concerned, there is, for the most part, no relation! (Consider, in particular, the UC campuses with the smaller per-faculty numbers of graduate students and of research funds. If one could increase their level of research funding and increase their number of graduate students, would the faculty members now present on those campuses want to increase their own number and thus dilute the research benefits coming to each of them?)
 
 

Now we turn to look at the relation between the expected increase in undergraduate enrollment and the need for additional faculty. The real issue at stake, when UC officials talk of the quality of this institution, is the ability to attract and retain the very best faculty members. So the administration's prime directive, in planning for Tidal Wave II, is to keep the faculty happy.

At a research university, like UC, faculty are hired and promoted primarily upon the basis of their record and potential as researchers and scholars. While the official public position stresses the three-fold mission - teaching, research and public service - one should not imagine that these three components carry equal weight in the university's decisionmaking processes regarding faculty members.

Here is how the Contemporary Culture of Research Universities is described in the "Report of the Universitywide Task Force on Faculty Rewards" [UCOP 1991]:

Foremost in the minds of faculty (beyond that which is self-imposed) is the pressure to be a productive scholar in the sense of discoverer and reporter of new knowledge -- that pressure which is captured in the popular expression "publish or perish." ...
Faculty caught up in this system have generally gone along with it. They have contributed to the exaggerated credit assigned to the scholarship of discovery at the expense of the scholarship of integration, application, and teaching -- little of which carries the financial consequence or peer recognition of sponsored research.

A crucial item in the competition to attract and hold stellar research professors is the number of courses they will teach each year. Lets look at the data in Table 4.
 
 

Table 4. Average expected courseloads for faculty in the major discipline groups

Discipline Group	Courseload at the University of California	Courseload at the Comparison 8 Universities
Humanities	4-5 qc	4 sc = 6 qc
Social Sciences	4-5 qc	4 sc = 6 qc
Mathematics	4-5 qc	4 sc = 6 qc
Engineering & CS	3-4 qc	3 sc = 4.5 qc
Physical Sciences	3-4 qc	3 sc = 4.5 qc
Biological Sciences	3 qc	2 sc = 3 qc

From: University of California, "Undergraduate Instruction and Faculty Teaching Activities, Seventh Annual Report to the Legislature," March 1999; available at www.ucop.edu/planning/ and the cited portions are on pages 45-6.

In the cited source document, the UC data are given as so many "quarter courses" (qc) per year while the data for the other universities are given as so many "semester courses" (sc) per year. I have converted the sc numbers in this table to equivalent qc numbers so that one may make a comparison (apples to apples). Taken at face value, these numbers show that faculty courseloads at the comparison 8 universities are larger than those at UC by 33%, 29% or 0%, depending on the discipline group. The UC report explains why it chose to compare apples and oranges, as follows:
"A direct conversion of semester courses to quarter courses is problematic, in part because semesters are of different lengths. In addition, quarter courses may be more intensive than semester courses, in some cases meeting longer hours and covering more material per week."
I find this explanation not very persuasive (and lacking in any quantitative value). Nevertheless I will allow some reduction of the percentage discrepancies noted in Table 4 and conclude this section with the estimate that UC faculty courseloads are, on the average, around 20% lower than those at our comparison universities.
 
 

If one is thinking about asking faculty to teach undergraduate students a bit more, it will be useful to know how their present work activities are distributed. From the "University of California Faculty Time-Use Study 1983-84" we learn about the average UC professor's work week. See Table 5.
 
 

Table 5.  UC Faculty Time-Use Survey 1983-84

Principal Activity	Hours/week
Instructional Activities	26.0
Original Research/Creative Activities	23.2
University Service	6.6
Professional Activities/Public Service	5.5
All UC-related Activities	61.3

 

Details of Instructional Activities	Hours/week
Regularly Scheduled Courses (classes, labs, fieldwork, giving exams)*	5.1
Supervising Independent/Special Study (197-199, 297-299, etc. courses)*	2.5
Non-credit Instruction (seminars, colloquia, symposia)*	0.7
Student Advising	3.2
Giving Oral Exams	0.2
Course Preparation	10.1
Other Instructional Activities (preparing & grading exams and papers, confer with TA's, letters of recommendation, reading theses, committee discussions)	4.1
Total Instructional Activities	26.0

 

Further Breakdown of "In-Class" Hours/Week by Level of Instruction

Type of "In-Class" Instruction	Lower Division	Upper Division	Graduate Level
Regularly Scheduled Courses	1.1	2.3	1.8
Supervising Independent/Special Study	0.0	0.5	2.1

Source: UC Archives, 308kd.tu, Bancroft Library, Berkeley Campus.
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".

The total "Instructional Activities" time accounted for 26/61.3 = 42% of the average work week; but the time spent in directly teaching students (including both regularly scheduled courses and independent study courses) was only 7.8 hours/week (13% of the total work time), and this time was evenly divided between undergraduate courses and graduate courses. Of all the time spent in Instructional Activities, 70% was accounted for in class preparation, student advising, and other supporting and incidental work.
 
 

Using the data and the analysis presented above we come to a number of conclusions/suggestions on how the University might accommodate Tidal Wave II with far fewer additional faculty members than the 3,000 planned by UCOP -- thus saving several $ Billions in overall cost.

More graduate students.

More undergraduate students(43% increase).

I will offer the estimate that this leaves us with something like a 30% increase in the number of courses (mostly at the upper division level) that will have to be provided. Based upon the facts discussed earlier in this paper, there is reason to believe that this can be accomplished with a significantly smaller percentage increase in the size of the faculty and no substantive damage to the quality of the University.

This approach does imply some increase in average faculty teaching loads, but we have noted that those numbers for UC appear to be anomalously low when set beside those for our comparison institutions. In addition, let us consider how this added "burden" may be mitigated, and perhaps even turned to advantage.

Recommendation: I would suggest that the Regents direct the President of the University, working with administration and faculty leaders on each campus, to go back to the drawing boards and plan to accommodate TWII within the limit of an additional 1,000 faculty members. I am not saying that I know the optimum number, but this exercize will be most useful in figuring that out.