I. Review of the 1995-96 Report and the Provost's Response
In our report for 1995-96 we presented a detailed ten-year review of the salaries paid to Penn faculty. We compared salaries (and total compensation) to those at peer institutions and to the changes in the consumer price index. On average, the faculty was doing at least as well as or better than most other institutions, and were well ahead of the growth in the consumer price index. However, when we examined intra-university equality of salaries and salary raises, we noted that among faculty of identical rank, within schools and broad disciplinary areas, inequalities were substantial. We also noted that some fraction, perhaps 10-15% of the faculty in some areas, had received raises less than the rate of growth in the consumer price index. The available data did not permit us to indicate the extent to which inequalities were inequities, that is that they were not based on experience, productivity, or market forces.
We made a series of recommendations to the administration meant to address some of those concerns.
Our recommendations may be described as having two thrusts: increasing the amount of information available to individual faculty, and establishing additional principles of compensation having to do with acceptable floors for salary increases over more than one year. If they had been adopted these recommendations would have reduced discretion of administrators to set salaries, both because individual faculty would be better informed about their status relative to others, and because administrators would be further constrained in giving raises below a fixed floor.
In response to our recommendations, the provost agreed to implement a policy of ad hoc examination of the salary histories of faculty who were outliers to make sure that they were not underpaid by unfair accumulation of year-to-year decisions. He rejected all of our other recommendations, in particular refusing to provide any additional information to individual faculty, using the CPI as a basis for defining the floor for raises, or defining a multi-year salary policy.
II. Introduction to the 1996-97 Report
The current report is meant to serve as a focused update to the fuller 1995-96 report. It has three sections: the first provides the most recent information about average salary increases compared to the consumer price index and to peer institutions. The second revisits the discussion of equality of salaries with some additional data. The final section addresses a particularly disturbing pattern reflecting the pattern of small raises given to faculty in the School of Arts and Sciences.
A. The pattern of overall increases
We begin with average results for the entire University excluding the Medical School.
Increase for Faculty continuing in rank from 1995-96 to 1996-97
Mean Increase Median Increase Full 3.8% 3.1% Associate 4.3% 3.5% Assistant 5.0% 3.5%
Increase for Faculty continuing in rank from 1986-87 to 1996-97
Mean Increase Full 72% Associate 74% Assistant 79%
We do not have publicly available data that compares faculty by discipline. However, private data supplied by the University does make comparisons within four broad disciplinary classifications [engineering, humanities and social sciences, sciences, management] among an elite group of public and private research universities. These data are current only to the 1995-96 academic year.
These data show that in each disciplinary area faculty at the University have done better than the average faculty at other peer institutions over the ten-year period ending in 1995-96. The rates of increase in salaries for Wharton and natural science faculty were particularly better than the rates of increase for faculty elsewhere, while humanities and social sciences faculty did somewhat better and engineering slightly better than comparable faculty elsewhere.
B. The inequality among faculty is substantial.
However, the degree of inequality is not increasing over time.
In last year's report we noted that there was a large difference between the best paid and least well paid faculty among us, even if we ignore differences across schools. We examined data among full professors within 13 broad disciplinary areasten schools of the University (Annenberg, Dental, Education, Engineering, Fine Arts, Law, Nursing, Social Work, Veterinary Medicine and Wharton) and three areas in SAS: humanities, natural sciences and social sciences. We found that the best paid 20% of full professors in each area were paid about 75% more than the least well paid. The absolute level of inequality ratio varied from the least unequal school, where the best paid full professors made about 50% more than the least well paid, to the most unequal schools where they made 100% more.
It was difficult to provide a confident interpretation of these inequality numbers. One interpretation says that the use of broad groupings risks including faculty for whom the University competes in sharply different markets in a single category. Thus the inequality might be market-driven and acceptable. Another view notes that the level of inequality is present even in quite small schools, where the idea that different markets have been blurred is less telling. This view is less accepting of the idea that individuals of similar rank who are asked to do similar work should be compensated at such extremely different rates. We do not have sufficient information to choose sides between these views. However, both sides expressed concern that there might be an increasing level of inequality of salaries over time, within groupings. This might reflect an evolving policy of investing more in star faculty at the cost of the salaries of other faculty. On the basis of anecdotal reports some feared this might be the case.
To be able to answer this question, we looked at these inequality ratios over time. Overall there was no evidence that there was any increase (or decrease) in the level of inequality. The following table presents this ratio for the previous five years. The figures reported are the average inequality ratio across the 13 groupings, weighted by the number of faculty in each grouping. The numbers reported are the ratios of the total salary paid to the top 20% of full professors to the salary paid to the bottom 20%.
Inequality ratio for full professors by year (weighted average)
1992-93 1993-94 1994-95 1995-96 1996-97 1.74 1.82 1.711 1.74 1.75
The relative constancy of the inequality ratio was matched in most of the thirteen groupings, at least when there were enough individuals included in the estimations to make the ratios meaningful. However, in one of the larger groupings, where there were enough full professors included to expect some stability in these estimates, there was evidence of an increase in relative inequality (SAS-social science from 1.84 to 2.02). This increasing inequality, in the context of poor overall raises in SAS, raises troubling issues about policy in that school. We turn to that issue next.
C. A Special Problem in the School of Arts and Sciences
Thus far we have reported about salary gains, on the average, across the University. While the average gain was less than typical in recent history, it could be described as adequate, especially given the low rate of inflation. However, hidden behind the figures for salary raises "on the average" there is some distinctly bad news. The faculty of the School of Arts and Sciences received very low raises in 1996-97. Faculty continuing in rank in the rest of the University received mean raises of 4.5%; in the School of Arts and Sciences mean raises for faculty continuing in rank were only 3.4%. Even worse, the median raise in SAS was just slightly more than 2.5%, less than the Philadelphia CPI change.
In the previous report we used the Philadelphia CPI to define a floor; any raises below that level we considered to be very low raises, since they represent an absolute decline in purchasing power. [We recognize that some may argue that the CPI overestimates increases in the real cost of living; absent any other generally accepted measure we have no choice but to continue to rely on that as our criterion.]
In 1995-96, about 15% of the University faculty received raises at 3% or below. Since the Philadelphia CPI was a little under 3%, we estimated that perhaps 10% of the faculty were below our floor. In 1996-97 the picture was far worse.
Of the 13 groupings defined previously, 6 gave above CPI raises to more than 90% of their faculty; 4 more gave raises above CPI to between 71-88% of their faculties. However, all three SAS areas gave less than half of their faculties raises which met the CPI:
Altogether, and largely because of these poor raises given in SAS, about 35% of the entire faculty received raises lower than the CPI.
We do not have 1996-97 comparative data for peer institutions; the comparative data described above ends in the previous year. Thus we do not have systematic data to show whether this decline in real salaries for so many of our faculty was matched in peer institutions. However, we do note that the poor raises seem to reflect the ongoing economic crisis in SAS, and thus it is not likely to be merely a reflection of external market forces. And thus it immediately raises the specter of peer institutions raiding our faculty.
We do need to note that this is a sharp departure from the pattern of the previous years. It is not yet a trend and the absolute level of salaries in SAS does not appear to be low, yet. However, any continuation of this pattern into subsequent years surely will signal a crisis for SAS.
It may be worth noting here that the low salary allocations in SAS were not merely the result of a low total pool for salary increases. In fact, as we noted above the salary pool in SAS for raises for continuing faculty (3.4%) was just slightly below the provost's guideline of 3.5%. If all money allocated to raises is included, including the faculty who were being promoted, the total pool available for raises was 3.6%, slightly above the provost's guidelines. However, the administration of SAS chose to allocate these raises in a manner that may be seen to have exacerbated the effects of this moderate shortfall in the total pool. The dean outlined her policy in a memo dated May 20, 1997. Twenty-nine percent of the money available for salary increases (1% of the 3.5%) was reserved for the small number of faculty being promoted, and for those who received outside offers which SAS decided to match. An additional 14% (0.5% of the 3.5%) was reserved for "rewarding exceptional merit." That left barely 2% for ordinary raises.
There are a variety of possible strategies that a budget with a median salary raise below CPI permits. If all faculty receive equal below CPI raises any faculty who can leave will be sorely tempted, and there will be rapid attrition among the most skilled of our faculty. On the other hand if SAS meets offers for faculty who are in demand elsewhere, but does so out of the common salary pool, then there will be great exacerbation of already existing inequalities, and a likely sharp reduction in morale.
III. Conclusions and recommendations
The Senate Committee on Administration met regularly during 1996-97, and we continued our engagement with the major issues that also occupied us in 1995-96.
Reappointment of Deans
We spoke at length with the Provost about his concerns regarding the proposal made by SEC at the end of 1996 to revise the procedures prescribed in the Handbook for Faculty and Academic Administrators for the reappointment of deans. He preferred the existing term limits, the existing role of student representatives in the process, and the continued direct consultation of the President and Provost with knowledgeable colleagues and officials. Although the committee had recommended the shortening of terms, we did not feel that this or any of the other areas of disagreement between the Provost and SEC were matters of great significance. We chose instead to focus the discussion on means for insuring the rigor of reappointment reviews, which was the central purpose of the original recommendations that the committee forwarded to SEC. In this regard, we were generally satisfied by the Provost's description of a review procedure that was carefully integrated with the periodic review of schools and the schools' strategic plans.
We were assured that the element of outside comparison would be part of the review process.
Evaluation of Teaching/Post-tenure Review
We note that SEC has not yet considered the recommendations regarding the evaluation of teaching that we made on April 13, 1996. At the request of the Faculty Senate Chair, we continued our discussion of the broader question of post-tenure review, seeking data about present practices from all of Penn's schools and from other Ivy League universities (Brown, Cornell, Dartmouth, Harvard, Princeton, and Yale). These data are quite consistent: most of Penn's schools require the presentation of annual faculty reports, and no Ivy university has a formal post-tenure review process beyond that. The committee supports the present use of annual reports, and, after much discussion, its majority strongly opposed recommending new policies for post-tenure review. We will continue to collect information on this subject.
In fulfillment of the agreement made with the Executive Vice President in January 1996, the committee met regularly with Mr. Fry to discuss the administrative restructuring project, with an eye to insuring appropriate faculty involvement. These discussions were wide-ranging and collegial. In addition to maintaining a general overview of administrative restructuring, we heard more detailed reports on student housing, the bookstore, dining services, and the restructuring initiative as it affects the several schools. At the time of this report, we are engaging Mr. Fry in a discussion of the measurement of administrative size.
In accord with our charge to monitor affirmative action in faculty hiring, the Senate Committee on the Faculty has spent a good part of 1996-97 analyzing and discussing the information provided in the Affirmative Action Report for the Current Standing Faculty prepared by the Office of Planning Analysis and Institutional Research, based on data available as of Fall 1995. The primary focus of that report is a comparison of the percentage of Assistant Professors hired at the University for the period 1985-1995 who are women to the percentage of women in the reported national pools for the period 1984 to 1994. The tables in the following pages summarize the information from that report on which our conclusions are based and compare the figures from the 1995 report to those in the comparable report for 1988. Like the information in those reports, our analysis pertains only to members of the standing faculty.
First, it should be noted that any assessment of progress in affirmative action is limited by the available information. It is almost impossible to draw meaningful conclusions concerning members of ethnic and racial minority groups because our principal source of information about the available pools from which faculty members are hired, the National Research Council's figures on Ph.D.'s awarded in various fields, do not specify racial categories for non-citizens. As a consequence, we have been forced to focus mostly on the hiring of women. Second, the task of evaluating the success of individual departments and schools by comparing the rate of hiring with the proportion of women in the available pools is made more difficult by the fact that the pools are differently defined in different areas: for example, for most SAS departments, the pools are defined as all Ph.D. recipients as reported by the NRC; for the Law School, the pool is defined as those law students who show an interest in law teaching by signing up with the Association of American Law Schools Teaching Registry; for the clinical departments of the Medical School, the pool is defined as M.D.'s employed in U.S. medical schools. Finally, the information gleaned from this and comparable reports should be supplemented by an analysis that would track changes and progress over time, and we urge the University to make such a longitudinal study a regular part of each annual report on Affirmative Action.
It is evident that some progress has been made over the last decade in increasing the representation of women on the Penn standing faculty: as Table 1c shows, in the seven-year period from 1988-1995, the percentage of women on the faculty increased from 17.1 to 22.3. Many schools and departments have hired women faculty members at rates that match or exceed the proportion of women in the relevant pool. Where the hiring of members of minority groups is concerned, the numbers are small, but there has been slight improvement, most markedly in the clinical departments of the Medical School. But Penn still falls far short of having achieved a truly representative or diverse faculty. What progress there has been remains modest for several reasons. In some areas, the number of new faculty members hired over that period is low. There are many fields in which the number of women and members of minorities in the available pool remains small. And, finally, not all areas of the University have shared in this progress: in particular, basic science departments in the Medical School and natural science departments in SAS have tended to hire fewer women than would be expected given the size of the relevant pool.
The committee strongly affirms the urgency of continued and expanded efforts to hire more women and minority faculty throughout all parts of the University. Building a broadly diverse faculty is essential, not only to provide all of our students with strong role models, but also for the quality of intellectual life at the University, which can only be strengthened by continued expansion of the range of perspectives and experiences represented on the faculty. The administration must make this a top priority and should take vigorous steps to ensure that attention to the hiring of women and minority faculty becomes, not simply a matter of bureaucratic procedures, but an integral element in the way schools and departments go about hiring. The provost should issue through the deans explicit statements to everyone involved about the central importance of hiring more women and members of minorities. Questions of diversity should be addressed at every stage of a search, starting at the point at which a position is requested and authorized. Specifically, diversity should be considered in the initial identification of the specialty and subfield in which a search will be conducted. In many disciplines, particular subfields have significantly fewer women and minority group members, and there is a danger that, by defining a position in terms of a traditional strength or of the interests of the person being replaced, a school or department may miss a valuable opportunity for increased diversity. We recognize that departments and schools are under heavy pressure to achieve excellence by maintaining and extending their established strengths, and we hope that this conflict can itself be discussed and addressed at all levels of the University. Finally, it should be required that chairs' letters proposing individuals for appointments include discussion of how the candidate's presence would help to broaden the diversity, as well as to enhance the intellectual strength, of the department.
Notes to Tables 1a and 1b
* Percentage of women among Ph.D.'s awarded, from National Research Council data, with several exceptions. For the Law School students who register with the Association of American Law Schools Teaching Registry and thereby express an interest in the teaching of law are counted. For the Dental School clinical departments enrollees in Advanced Dental Education Programs, compiled by the Department of Educational Surveys of the American Dental Association, are used. For the Medical School clinical departments M.D.'s employed in U.S. medical schools in 1994 are counted. For the Veterinary School clinical departments the figures represent graduates of veterinary medical school programs, from American Veterinary Medical Association reports.
Expected number of female hires, based upon the total number of Assistant Professor hires during 1985-95 and the pool percentage figure.
This questionnaire is voluntary and confidential. The data will be reported in a way that will not identify individual respondents. The results will be analyzed by the Senate Committee on the Faculty which is charged with monitoring the treatment of faculty at the University. If you have questions about this questionnaire, please call the Faculty Senate Office at (215) 898-6943. You are also invited to discuss your experience at Penn in a confidential interview with a faculty member from a panel selected by the Faculty Senate Executive Committee. If you would like to arrange such an interview, please contact the Faculty Senate Office at (215) 898-6943.
NOTE: You are invited to write further comment on the back of the questionnaire and to attach additional pages.
Volume 43 Number 34
May 13, 1997
Return to Almanac's homepage.
Return to index for this issue.