|
The
Gender Equity Report
[note:
you may print a PDF file of this report by clicking HERE]
Executive
Summary
| Part I: Faculty Census | Part
II: Professional Status of Penn Women Faculty | Part
III: Salary Analyses | Part IV: Faculty Survey
Data | Conclusions | Data
Sources | Committee Members
Reply
from the President and the Provost, also in this issue
Executive
Summary
The
Gender Equity Committee was established in June, 2000, by Provost
Robert Barchi and Faculty Senate Chair Larry Gross at the recommendation
of the University Council Steering Committee. The charge was to
undertake a systematic review of the status of women faculty at
the University of Pennsylvania. Four subcommittees were formed to
complete the review: one on faculty census using fall 1999 data,
one on professional status of women faculty, one on faculty salaries,
and one on a survey of faculty regarding their quality of life.
The
census data show that women averaged 23.8% of the faculty in 1999.
While comparison with the prior ten years indicates an upward trend
in the hiring of women, there are indications that this increase
has slowed or ceased in some areas. The data show considerable variability
in the proportion of women among the various schools, much of it
reflecting differences in the availability of women in the doctoral
pool. There were also marked unexplained differences within groups
of departments that should be drawing from pools of similar size.
Examination of Ph.D. pool data suggests that in many schools and
departments the number of women at the rank of Assistant Professor
or Associate Professor is fairly consistent with the numbers in
the doctoral pool. The School of Medicine, which has more than half
of all of Penn's women faculty, reached expected numbers in clinical
departments but had lower than expected numbers of women Assistant
and Associate Professors in basic science departments. In the Natural
Sciences in SAS, the Wharton School and the Basic Science Departments
in the Dental School, the number of women was less than 60% of that
expected based upon the pool. The census data also show that the
proportion of men in the faculty increases with rank while the proportion
of women does not; thus, women are 35% of all Assistant Professors,
23% of all Associate Professors, and only 15% of Full Professors.
This pattern appears to be the result of both loss of some senior
faculty women to other universities and the hiring of relatively
large number of men at senior faculty rank. A Harvard survey comparing
the numbers of women at Penn with numbers at other universities
shows Penn at or better than median rank among Medical School Full
Professors and untenured Associate Professors as well as non-Medical
Assistant Professors. However, Penn was below median rank in the
remaining five categories, and ranked lowest for tenured Associate
Professors in both the medical and non-medical areas.
Recommendation
We
recommend that the University work with the deans to develop policies
about hiring practices to ensure change, particularly in those
departments whose hiring of women is not consistent with their
numbers in the pool. These policies should also aim at countering
the male bias in the hiring of senior faculty and at keeping more
tenured women at Penn.
Half
of all women faculty at Penn are in the School of Medicine and almost
half of these are Assistant Professors-Clinician Educator. This
large group of junior faculty women is experiencing particular difficulty
reconciling their professional responsibilities with the demands
of family and home life, resulting in an unusually high resignation
rate.
Recommendation
We
recommend that policies affecting the retention and promotion
of Assistant Professors-Clinician Educator be evaluated for their
disproportionate impact on faculty women.
With
respect to the professional status of women faculty, the committee
determined that at the more junior ranks women had more research
space per grant dollar than men, but women Full Professors averaged
somewhat less space per grant dollar than their male colleagues;
in both SAS science departments and the School of Medicine, senior
women faculty had about 85% of the space assigned to males. Women
have been funded by the University Research Foundation in proportion
to the number of grants they have submitted and, in some years,
in a greater proportion than the number submitted. Women faculty
hold administrative positions at the school level (dean and sub-dean)
in proportion to their numbers, but they are under-represented among
department chairs. They also tend to be under-represented as holders
of endowed and term chairs. When measured against their proportion
of the total faculty, the proportion of women winning the Lindback
teaching award was roughly comparable to their proportion in the
faculty.
Recommendation
The
University and the deans should work together to develop policies
that assure that women achieve leadership positions and scholarly
rewards in schools and departments consistent with their interests
and capabilities.
A
statistical analysis of salary by gender among faculty of the same
rank and cognate disciplines was carried out, controlling for time
since degree, for whether the faculty member was first hired as
a full professor (for comparisons of professor rank only) and, in
some instances, for specific department. This analysis indicated
that women had slightly lower salaries than men in most of the groupings;
however, only a few of these differences were statistically significant.
Recommendation
The
equity of faculty salaries in all schools should be reviewed with
special attention to salaries of women faculty.
The
survey of non-medical faculty (men and women) yielded a number of
important similarities in terms of the work they do and their satisfaction
with space. However, the women reported being less satisfied with
their jobs and the majority reported feeling that men faculty were
more advantaged at Penn. The survey of medical faculty also showed
males and females being similar in the work they do; but, here too,
most women faculty (and 25% of the men faculty) felt that male faculty
were advantaged. Comparison of non-medical and medical school data
suggested that faculty in the School of Medicine were more concerned
about the impact of gender than faculty in the other schools. Finally,
a comparison of the non-medical data with the 1999 National Survey
of Post-Secondary Faculty showed that Penn faculty are more satisfied
with their salaries and benefits and the quality of students. It
also showed that Penn faculty, especially women faculty, put in
much longer workweeks than do faculty at most other institutions.
In other respects, Penn faculty opinion closely mirrors that of
faculty nationally.
Recommendation
The
University and the deans should work together to find ways to
alter the environment in which many women and some men perceive
men to be advantaged. It is also important for the University
of Pennsylvania to make a major, visible commitment to efforts
intended to create an environment friendly to women.
Executive
Summary | Part I: Faculty Census | Part
II: Professional Status of Penn Women Faculty | Part
III: Salary Analyses | Part IV: Faculty Survey
Data | Conclusions | Data
Sources | Committee Members
Part
I: Faculty Census
Women in the Standing
Faculty
Our
information on composition of the Standing Faculty and Standing
Faculty-Clinician Educator is derived from fall 1999 data, and a
summary is presented in Table
I. More detailed information on faculty composition for each
department in the University, listed by gender and by the percentage
of women, for fall 1988, 1994 and 1999, is provided in Appendix
I-A (GRAPH 1 | GRAPH
2 | GRAPH 3 | GRAPH
4 | GRAPH 5 | GRAPH
6 | GRAPH 7)
In
1999, 50% of the women faculty (and 52% of the total faculty) was
in the Medical School, while SAS contained approximately 20% of
the women faculty (and 20% of the total faculty). With the exception
of the Law School, all schools have made considerable gains in the
number of women faculty in the past decade. The total faculty at
Penn in 1988 was 17% women, while in 1999 it averaged 23.8% women.
Excluding Nursing, which is 97.9% women, the range goes from a high
of 50% in Social Work to a low of 6.4% in Engineering. Several schools
are well above the University average: Social Work (50%), Graduate
Education (45.2%), Fine Arts (32.1%), Arts and Sciences- Humanities
(32%). However, three areas fall well below the University average:
Wharton School (12.9%), Arts and Sciences-Natural Sciences (13.3%),
and Engineering (6.4%).
There
is considerable variation in the proportion of women not only among
the various schools but also within a given school or division.
Some of this variation may be due to differences in the proportion
of women obtaining doctorates in specific disciplines, but it also
is observed in disciplinary groupings that might be expected to
have similar proportions of women. Figure 1 (below) illustrates
the range for some of these groupings. Engineering and clinical
medicine are fields with great variability in the percent of women
entering individual doctoral and specialty programs and would be
expected to show large differences in the proportion of women within
these groupings. In SEAS one department, Computer and Information
Sciences, has a high of 13.6% women, but four of the seven departments
contain no women faculty. The clinical departments in the School
of Medicine range from 43.5% women in Emergency Medicine to no women
in Orthopaedic Surgery. Some clinical departments were well above
the School's average of 23% women: Emergency Medicine (44%), Family
Medicine (40%), Pediatrics (36%) and Rehabilitation Medicine (36%).
However, four departments with women faculty fell well below the
average: Radiation Oncology (5%), Surgery (9%), Neurosurgery (11%),
and Otorhinolaryngology (14%). The largest department, Medicine,
contained 21% women.
To
gauge the equity of the current status of women on the faculty,
an attempt has been made to compare percent of women Assistant and
Associate Professors in the various schools with their respective
pool size (Table II ).
The pool size is determined from the percentage of doctorates awarded
to women in a given discipline between 1985 and 1995. While some
disciplines expect a period of post-doctoral training before entry
into faculty positions and others do not, it is reasonable to assume
that most Assistant Professors and a majority of Associate Professors
in 1999 will have received their Ph.D. degrees between 1985 and
1995. Engineering, Annenberg and the clinical departments in the
Dental School have well above the number expected from the pool
data. However, the Natural Sciences division of SAS, the Wharton
School and the basic sciences departments in the Dental School have
less than 60% of expected numbers based upon the pool.
Figure 1.
Analysis by department of percent women standing faculty for selected
divisions at Penn. Departments containing fewer that 10 faculty
have been omitted.
Distribution
Among Ranks
As
of 1999, women comprised 35% of the Assistant Professors in the
Penn faculty, 23% of the Associate Professors, and 15% of the Full
Professors, i.e. the proportion of women decreases with increasing
rank. AAUP data [Academe 86(2), 2000] show that in private
universities women are 37% of Assistant Professors, 30.5% of Associate
Professors and 14.5% of Full Professors. Thus Penn is about even
with the national data for Assistant and Full Professors, but below
the national average for women Associate Professors. Figure 2 (below)
displays representative data on the distribution of men and women,
by rank, for our largest groups of faculty: SAS, Wharton and the
clinical faculty of the Medical School. Within the clinical departments
of the Medical School, women were 18% of senior faculty (Full and
Associate Professors) on the clinician educator track, and they
were 9% of the senior faculty on the tenure track.
Figure
2. The percent
of women at each rank in the School of Arts and Sciences, the Wharton
School, and the clinical departments of the School of Medicine.
It
might be plausible to attribute the high proportion of male Full
Professors to large numbers of men hired and promoted many years
ago, since it is a phenomenon that appears to cut across all private
universities. However, Penn women constitute a smaller proportion
of Associate Professors than Assistant Professors throughout the
University, except for SEAS where they were 7.1% of Assistant Professors
and 12.5% of Associate Professors. We therefore considered several
other explanations for a pattern of declining percentage of women
with increasing rank: 1) men Assistant Professors are promoted to
tenure in greater proportion than women; 2) a greater fraction of
women leave voluntarily due to failure to retain tenured women faculty
relative to men and/or a greater fraction of women Assistant Professors
resign before attaining tenure; 3) new hires at tenured ranks are
proportionally fewer among women than men. Data previously reported
by the Provost's Office (see
Almanac vol. 44 (6); Sept. 30,1997) indicated that the
first possible scenario, i.e. men are promoted to tenure at a greater
rate than women, does not hold.
Based
on information from SAS, it appears that scenario 2 explains at
least some of the problem. Of the 24 men and 16 women who resigned
from SAS in 1997-2000, 54% of the men and 69% of the women were
tenured. Ten tenured women left from Humanities and Social Sciences
departments, representing a loss of over 17% of the tenured women
in these areas. In contrast, resignations of tenured men from these
departments represented only 4% of the men with tenure. Thus, tenured
women in the Humanities and Social Sciences of SAS left in disproportionate
numbers; most of these went to other institutions. In the School
of Medicine, both men and women Assistant Professors-Clinician Educator
are leaving because they experience a high degree of discouragement
(see Part IV of this report), but the resignation
rate among female Assistant Professors-CE (16%) is significantly
higher than that of male Assistant Professors-CE (9%). We also examined
59 confidential questionnaires returned to the Faculty Senate office
by faculty who resigned during the period 1997-2000. While the response
rate to this questionnaire was poor, the replies supported a conclusion
that women resigning from Medical School faculty appointments were
primarily Assistant Professors-CE while those resigning from non-medical
areas were more likely to be tenured faculty. A summary of the findings
is presented in Appendix
I-B.
Analyses
of new hiring patterns support the third scenario as also contributing
to the problem, i.e., appointments of men directly to tenured ranks
far exceeds that of women. Table
III presents data on the numbers of men and women Full Professors
in each school, and the proportion who were hired at that rank.
SAS shows roughly equal proportions of women and men hired as Full
Professors. Four additional schools with a total of 53 Full Professors
(Annenberg, Grad Education, Fine Arts, and Social Work) appear to
have hired senior women at a rate higher than that for senior men
(although the total number of faculty hires is small). In contrast,
6 schools, representing over 60% of all Full Professors at Penn,
show marked differences in the route by which men and women achieved
that rank.
Figure
3A (below) presents information on the number of men and
women faculty newly hired with tenure for the period 1986-1996,
and Figure 3B (below) is the same information for hires during
the past three years. Data compiled annually by the National Research
Council indicate that the proportion of women receiving doctorates
in science and engineering fields has been increasing steadily over
the past 30 years, and in the humanities it has been least 40% women
since the late 1970s. It is therefore notable that while the percentage
of women who might be hired as senior faculty should be increasing,
the number hired at Penn had decreased in the most recent three
years for all schools represented. The most dramatic decreases occurred
in the Natural Science departments of SAS and in Wharton, where
no senior women were hired in the last four years compared to 12%
in the preceding eleven years.
Figure
3. Percent
women among appointments of new faculty to tenured faculty positions.
Actual numbers of women/total new hires are shown in parentheses.
3A. Upper figure: New tenured appointments from fall 1986 through
fall 1996.
3B. Lower figure: New tenured appointments from fall 1997 through
fall 2000.
Hiring Patterns
for Junior Faculty
We
examined whether the proportion of women among newly hired Assistant
Professors during the period 1986-1996 was consistent with their
increased share of Ph.D.s awarded. In SEAS, the proportion of women
hired as Assistant Professor during this period was twice that expected
from the national pool of Ph.D.s in Engineering; unfortunately,
most of these women had left the faculty by 1999. Appointments of
Assistant Professors in clinical departments in the School of Medicine
have also exceeded the estimated availability pool, although this
group has experienced a simultaneous high resignation rate (see
above). In the Humanities division of SAS, the percent of women
among newly hired Assistant Professors for the period 1986-1996
was comparable to the proportion of women receiving doctorates in
the Humanities during that period, and the combined number of Assistant
+ Associate Professors in 1999 (Table
II) suggests that retention of these women has not been a major
problem. However, the percentage of women among newly hired Assistant
Professors was slightly lower than expected in the Social Science
departments of SAS, and it was markedly lower than expected in the
Natural Science departments of SAS and the basic science departments
of the Medical School. More detailed information on hiring patterns
and availability pools for SAS divisions, the School of Medicine,
and Wharton may be found in Appendix
I-C .
Trends in the
Proportion of Women Faculty
The
steadily increasing number of women receiving Ph.D.s should be gradually
but steadily reflected in significant increases in total faculty
women. As noted previously, almost all schools have shown considerable
increases in faculty women since 1988. Figure 4 (below) presents
the percentage of women faculty in SAS (Humanities, Social Sciences,
and Natural Sciences) and for the Medical School, Wharton and Engineering,
for the period from 1988 to 1999. Much of the increase in faculty
women occurred during the first half of this period, and the increase
has slowed or stopped in more recent years in Wharton, Engineering,
and the Social and Natural Sciences divisions of SAS.
Figure
4. Changes in percent of women faculty during the past 12 years.Numbers
in parentheses = absolute number of women.
A. Percent women in the School of Arts and Sciences.
B. Percent women in the School of Medicine, Wharton, and SEAS.
Comparative
Data
Benchmark
data from a group of 16 Ivy-plus schools provides a context within
which Penn's gender issues might be viewed. Highlights of the Harvard
Survey of '99-00 provide that context.
With
respect to the proportion of women among non-medical school faculty,
Penn ranks close to the median (9th out of 16 institutions) for
Full Professors, last (14/14) for tenured Associate Professors,
and in the middle (8/16) for untenured Assistant Professors. For
comparisons of medical school faculty, the data set consisted of
all full-time faculty including research and clinical faculty. With
respect to the proportion of women faculty in medical schools, Penn
ranked in the upper half (4/11) for Full Professors, last (8/8)
for tenured Associate Professors, 3rd out of 10 for untenured Associate
Professors, and in the middle (6/11) for Assistant Professors.
These
data suggest that with respect to women Full Professors and junior
faculty hires, Penn is neither better nor worse than its peers.
Our low rankings for tenured Associate Professors suggest that Penn
has major problems in retaining meritorious younger faculty women,
either prior to or after the tenure process.
A
separate report was recently prepared by a committee of the Medical
School, comparing the proportion of faculty women in each department
to the national averages compiled by the American Association of
Medical Colleges in 2000. In most cases, Penn departments were near
the national average. Two departments were markedly higher than
the national average (Emergency Medicine and Microbiology), while
five departments were at least 30% lower than the national averages
(Physiology, Pharmacology, Orthopaedic Surgery, Otorhinolaryngology
and Obstetrics and Gynecology).
Summary of Census
Findings
The
upward trend in the hiring and percentage of women faculty at Penn
is encouraging. However, there are some areas where the trend has
slowed, and others where it has even reversed, in spite of the increasing
percentage of Ph.D.s being awarded to women. While retention of
tenured women faculty seems to play a role in some disciplines,
a major factor appears to be the hiring patterns within selected
departments. It is apparent that some departments at Penn are responding
to the growing numbers of women scholars with increased hiring rates,
while others are not. If Penn is to progress in gender equity with
respect to the percentage of women on the faculty, it is obvious
from the data that it must be done at the level of the individual
departments where recruiting and hiring decisions are made. Most
of the departments at the lowest end of the range represent disciplines
in which the proportion of women Ph.D.s is markedly higher than
is reflected in hiring patterns. The faculty of these departments
should be made aware of their position, and the potential impact
on Penn's academic mission.
In addition to increasing the numbers of faculty women, it is clearly
important to rectify the low number of senior faculty women. Many
departments in the University have relatively few women Assistant
Professors in the pipeline, and promotions to more senior ranks
in the near future will therefore tend to be male. If the fraction
of women at tenured ranks in these departments is to improve within
a reasonable time frame, it will require an exceptional effort to
appoint women as Associate or Full Professors. Our data suggest
that, throughout the University, greater attention must be paid
to gender equity when hiring directly to senior faculty positions.
The
benchmark data indicate that our record in hiring and/or promoting
women faculty is mixed at best. The fact that Penn ranks last in
comparison with its peers in the numbers of tenured Associate Professors
who are women is particularly disconcerting, and should be taken
into consideration at all stages in hiring and promotion decisions.
Because data from the Harvard study are aggregated and do not provide
information for specific disciplines, it is not possible to determine
if Penn's relatively low rankings reflect University-wide problems.
Our departmental analyses suggest that they may result from an unusually
low number of women in a sub-set of Penn departments relative to
comparable departments in peer institutions. Alternatively, it is
possible that Penn has developed extraordinarily large departments
in those disciplines where women are poorly represented.
Data
presented above also indicate that we are disproportionately losing
faculty women in the humanities and social sciences once they achieve
tenure. While there are many possible causes for this, the fact
that these women are leaving for other institutions at a rate markedly
higher than their male counterparts suggests that Penn should place
a greater emphasis on counteroffers designed to improve retention
of tenured women faculty.
Executive
Summary | Part I: Faculty Census | Part
II: Professional Status of Penn Women Faculty | Part
III: Salary Analyses | Part IV: Faculty Survey
Data | Conclusions | Data
Sources | Committee Members
Part
II: Professional Status of Penn Women Faculty
Allocation of
Research Space
The
committee collected data on research laboratory space assigned to
science faculty for FY2000. Faculty without assigned research space
were omitted from the analysis. Information on space allocations
was obtained from SAS, the School of Medicine, and SEAS. In SAS,
we examined space allocations only for Chemistry, Biology, and Psychology,
because Math and Physics had no women faculty who required laboratory
space. Data from SEAS were not sufficiently detailed to permit a
valid analysis.
In
the SAS departments, women faculty at the ranks of Assistant or
Associate Professor tended to be allocated research space that was
more than their male colleagues. This was true whether one looked
at the net square feet (NSF)/faculty or at grant dollars/net square
feet. However, a different picture emerged when we analyzed space
allocated to faculty who were Full Professors. As had been suggested
by the MIT report on Women Faculty in Science, there was a consistent
pattern of senior women faculty in science departments having less
research space. In each of the three SAS departments where women
Full Professors had research space (Biology, Chemistry, and Psychology),
the women had fewer square feet assigned to them, with women's space
averaging 63% of that for male Full Professors.In the School of
Medicine, space allocated to women Assistant and Associate Professors
was also a mixed picture. At the Assistant Professor level, men
had 90% of that for women, and there was little difference between
men and women who were tenured Associate Professors. Among Full
Professors who were tenured, the net square feet allocated to women
was 94% of that for men.
Because
these differences in space at the Full Professor level might reflect
gender differences in research support, we obtained information
on external grant funding. In the SAS departments, 85% of the male
Full Professors and 90% of the female Full Professors had grants
to support their research; however, the average size of women's
grant income tended to be less than that of the men. For this reason,
the differences between male and female Full Professors in each
department decreased when the net square foot of research space/$1,000
grant income was calculated. Nonetheless, women Full Professors
in Biology, Chemistry and Psychology averaged 84% as much research
space per grant dollar as their male counterparts. In the School
of Medicine, tenured women Full Professors averaged more grant income
than their male peers. The difference in research space allocation
for men and women therefore increased when normalized for grant
income; men Full Professors in the Medical School averaged 1,950
sq.ft /$500,000 while women Full Professors averaged 1,660 sq.ft/$500,000;
thus women Full Professors had 84% as much research space per grant
dollar as their male counterparts. The picture was different with
respect to more junior faculty; in both the Medical School and most
of the SAS departments, women Assistant and Associate Professors
tended to have more space per grant dollar than men.
University Funded
Research Grants
The
funding available from internal sources to faculty for research
grants varies from school to school. In so far as we have been able
to collect information on these grants, we can find no evidence
of gender inequity in the allocation of University-funded grants.
The most important general source of funds is the University's Research
Foundation. This is an important source of research support, and
its procedures are well-established and effective. The competition
is public, the rules are clear, and the selections are made by a
faculty committee. Over the past three years, women have constituted
between 23% and 34% of all applicants, and they have been awarded
between 22% and 38% of all the grants.
We
have also collected data on the awarding of the McCabe Fund grants
by the Medical School. These funds to support biomedical and surgical
research are intended for junior faculty with tenure track appointments
who have received limited amounts of external funding. Approximately
25% of tenure-track Assistant Professors in the Medical School are
women. Since 1996, women have been awarded approximately 25 % of
the grants each year, and they have been between 14% and 24% of
the faculty nominated annually for the grants.
Administrative
Positions
Superintendence
of school administrations below the level of the Dean has been entrusted
to Associate, Deputy, and Vice Deans. Since most schools other than
the School of Medicine have only a handful of such positions, one
case constitutes a significant proportion of the total number. In
the largest schools, women have held mid-level administrative positions
roughly in proportion to their number in the senior faculty. In
the Medical School 12-13% of the Deans and sub-Deans are women and
in SAS women now hold two of seven such positions. Both the Wharton
School and the Law School have one female working in such a position,
while women hold all of these positions in the Nursing School. However,
over the past five years, the Schools of Social Work, Veterinary
Medicine, Engineering, and Fine Arts have not had a female in any
of these jobs.
In
the academic year 1999-2000, women were 6 of 92 or 6.5% of the department
chairs in the seven schools surveyed (Table
IV). This proportion is roughly half the expected number
based on the proportion of women among Full Professors within the
seven schools (12.5%). Data on the distribution of department chairs
by sex covering the period since 1995-6 show two dissimilar trends.
In 1995, five of the schools at the University (SEAS, GSFA, MED,
VET and DENT) had no women serving as chairs. Since that date, there
have been 1-2 women chairs in the Medical School and one in the
Graduate School of Fine Arts. An opposite trend can be seen in both
SAS and Wharton, where the proportion of female chairs dropped since
1995. In SAS, where the proportion of female Full Professors has
risen from 15% to 19% since 1995, the proportion of female department
chairs has dropped from 22% to 12%.
We
conclude that, in the largest schools, women have held mid-level
administrative positions roughly in proportion to their number in
the senior faculty. The position of department chair, which often
rotates among tenured faculty, is one in which women are currently
under-represented throughout most of the University.
Endowed Chairs
and Term Professorships
While
the overall difference is small, women do not occupy endowed chairs
in proportion to their numbers in the faculty. The University of
Pennsylvania has, at present, 288 endowed professorships awarded
to faculty, 86% of which are held by men and 14% held by women.
The total Penn faculty is currently 76% male and 24% female, but
the proportions of women among Associate and Full Professors (the
primary recipients of endowed chairs) decreases to 17.4%. At one
extreme, the School of Veterinary Medicine (15% of whose tenured
faculty is female) has assigned no endowed chairs to females; nor
have the Dental School, the Graduate School of Fine Arts or the
Engineering School. In several of these faculties, however, both
the number of endowed chairs and the number of female faculty is
quite small. In the two largest units of the University, SAS and
the School of Medicine, women are under-represented among the holders
of endowed chairs. In these two schools, both Full Professors and
Associate Professors are recipients of endowed chairs. Women hold
only 14 % of the 73 endowed chairs in SAS, although they constitute
19.5% of the Associate and Full Professors. In the Medical School,
women occupy 10% of its 88 endowed chairs but are 14.4% of the Associate
and Full Professors. At the other end of the spectrum can be found
the Nursing School and the Graduate School of Education, where 100%
and 75% of the endowed professorships are held by women.
Term chairs are
awarded in significant numbers only in SAS, and to a lesser extent
in the Wharton School. They are frequently awarded to Assistant
Professors. Figures from SAS show that women are currently over-represented
(33%) among the holders of term chairs if compared to their share
of the total standing faculty (24%), but not if compared to their
presence among Assistant Professors (40%). In the Wharton School,
one of 9 term chairs is held by a woman; females constitute only
13% of the total Wharton faculty, but 21% of the junior faculty.
Teaching Awards
Our
committee used the Lindback awards as a proxy for the recognition
of teaching excellence. When measured against their proportion of
the total faculty, the proportion of women winning this award was
roughly comparable to their proportion in the faculty.
Executive
Summary | Part I: Faculty Census | Part
II: Professional Status of Penn Women Faculty | Part
III: Salary Analyses | Part IV: Faculty Survey
Data | Conclusions | Data
Sources | Committee Members
Part III: Salary
Analyses
We
analyzed whether a faculty member's gender affected his or her salary
in the 1999-2000 academic year. We examined whether there is any
difference in salary by gender, after adjusting for potential gender
differences in characteristics (i.e., experience, rank, degree,
and department).1
We analyzed academic base salary for all faculty other than those
in the School of Medicine (SOM). For SOM faculty, we included the
actual payments (academic base and clinical fees) to tenure-track
and clinician-educator faculty for the fiscal year 2000. For all
schools other than SOM, we analyzed actual compensation. The data
collection task was particularly complicated, however, in SOM. SOM
assembled data from the University payroll and from the Clinical
Practices payroll. When available, salary from the Veterans Administration
and other external salary were included. Faculty members who were
known to have salary from the Veterans Administration or other external
sources and for whom these data were unknown were excluded by SOM
from the data that they provided to the Committee. SOM also excluded
department chairs, for reasons that are not clear. The salary data
from the SOM were provided in an "index form." All payments
to individual faculty were entered as the proportion of the average
salary of all other faculty in the data set. For all SOM faculty,
the average is 1.00. This data presentation made it impossible for
us to consider compilations involving basic sciences throughout
the University.
We
conducted analyses based on the actual compensation where available
and using the salary index for SOM. The analyses were performed
separately by rank and by school-cognate fields. There were separate
analyses for Assistant Professors, Associate Professors, and Full
Professors in each of the following groupings: School of Arts and
Sciences (SAS) sciences, SAS humanities, SAS social sciences, Wharton,
Veterinary Medicine, Dental Medicine, the Law School, SOM basic
science departments, SOM clinical departments, and a grouping of
faculty in Annenberg, Graduate School of Fine Arts, Graduate School
of Education, and School of Social Work. In all analyses, we compared
faculty with equal time since terminal degree. For Full Professors,
we also controlled for whether the faculty member was hired into
that rank from outside the University. In SOM, we also controlled
for whether the highest degree was an MD and whether the professor
was in the clinician-educator track. In SOM clinical departments
and in Wharton, we controlled further for whether the professor
was in a high-pay, low-pay, or medium-pay department. For SAS Social
Science, we controlled for whether the professor was in economics.
For a small sample of the clinical departments in SOM, we added
controls for the amount of clinical work.
Of
the 27 separate "field-rank" regression analyses, women
received less salary than comparable men in 21 estimations and more
than comparable men in 6 estimations. The magnitude of the differences
when women earned less ranged from 0.1% to 14.8%; the magnitude
when women earned more ranged from 0.1% to 27.8%. None of the differences
where women earned more represented a pattern sufficiently strong
to yield a statistically significant difference, and only a few
of the differences where women earned less were statistically significant.
For many of the regression analyses, the estimates of gender differences
are imprecise because there were few women faculty to include.2
If we exclude Associate Professor comparisons, which are made among
a particularly small and heterogeneous grouping of faculty, we have
19 separate "field-rank" regression analyses. Within these
19 analyses or comparisons, women receive less salary than men in
14 of the estimations, with magnitudes remaining between 0.1% and
14.8%. Women receive more salary than men in 5 of the estimations,
with magnitudes between 0.1% and 5.7%. Alternative analyses performed
separately for each rank but aggregating all faculty (across divisions)
within SAS produced similar results; that is, there were small percentage
differences by gender that were not statistically significant. Both
analyses find similar salary differences by gender and few differences
that are statistically significant.
Executive
Summary
| Part I: Faculty Census | Part
II: Professional Status of Penn Women Faculty | Part
III: Salary Analyses | Part IV: Faculty Survey
Data | Conclusions | Data
Sources | Committee Members
Part
IV: Faculty Survey Data
In
addition to the quantitative data collected by the committee, a
survey of the perceptions of faculty on many aspects of quality
of professional life was conducted. The results indicate that while
men and women faculty are similar with regard to academic performance
and productivity, women faculty have some concerns that are not
shared by their male counterparts.
A. Non-Medical
Survey
In
the fall of 2000, a nine-page questionnaire was sent to 1,093 members
of the standing faculty at Penn in all schools other than the Medical
School. After the distribution of a reminder notice and a second
wave of questionnaires, 483 responses were received by February,
2001. The response rate was 57.2% for women faculty (276 responses)
and 39.4% for men faculty (207 responses). In the interest of brevity
we refer to this group as "non-medical"; however it should
be noted that it also includes faculty from the Schools of Nursing,
Veterinary Medicine and Dental Medicine. Women are concentrated
in the ranks of Assistant and Associate Professor and have been
at Penn fewer years than their male colleagues. Consequently, after
examining the data for gender differences, all of the measures were
adjusted to take rank and years at Penn into account. More detailed
information on the questionnaire and responses are available on
the web as Appendices.
In
a number of important respects, men and women were closely matched:
- Women
are as likely to obtain grant support as men, although men are
more likely to be principal investigators.
- Women
publish as many articles and books as men once rank and years
at Penn have been taken into account.
- Women
teach as many students as men and advise more dissertations.
- Women
and men are generally satisfied with the space they have allocated
to them, although data on actual space allocations may raise some
issues for certain subgroups of faculty.
- Women
and men both work very long hours (nearly 60 hours per week on
average).
There
are many important gender disparities, in addition to the response
rate, revealed by this survey:
- Women
faculty are less satisfied with their jobs than are men.
- A majority
of women respondents report feeling that men faculty are advantaged
at Penn, and nearly half of women respondents report that women
are disadvantaged at Penn.
- Over
half of women faculty believe they are paid less than comparable
men in their departments, although an analysis of salaries suggests
that the modest gender gap is principally due to differences in
rank and school.
- One
fifth of women report having encountered unwanted sexual comments,
attention or advances by a superior or colleague.
- Nearly
half of women faculty raise concerns about getting appropriate
credit for their work (compared with almost a third of men who
voice the same complaint).
- Women
faculty raise a number of specific equity concerns, including
access to students and the allocation of clerical support.
- Women
are less likely to obtaining mentoring from male colleagues, from
colleagues in their departments and from their department chair
than are men.
- Women
reported serving on committees more frequently than their male
colleagues.
- Time
pressure and work-family issues abound among Penn faculty, but
are felt more strongly by women than men.
- Most
women faculty report feeling unsafe in one or more locations around
campus.
Selected
items related to the above conclusions are presented in Appendix
IV-A, IV-B and IV-C
. Each section of the faculty questionnaire also included
an open-ended question that solicited additional comments and suggestions
from faculty. Women faculty were more likely to offer responses
to these questions than men (typically about 1/4 of female respondents
and about 1/10 of male respondents offered additional comments).
About half of all respondents added comments to at least one of
these questions. In addition, several respondents requested an interview,
and a summary of these interviews is presented in Appendix
IV-D.
B. School of
Medicine Survey
In
January 2000, as part of the Faculty-2000 project, all School of
Medicine faculty were asked to complete a questionnaire to examine
academic job satisfaction and reasons for changing institutions
or for leaving academia altogether. A total of 938 faculty (70%)
in the school responded to this survey. We are grateful to Dr. Judy
Shea, Department of Medicine, for sharing summaries of data which
we then sorted by gender. To maintain consistency, this questionnaire
also formed the basis for the survey sent to non-medical faculty
described in section A.
Women
faculty in the School of Medicine were more likely than men faculty
to be Assistant Professors and to be in the Clinician Educator track.
In the total sample, men had been faculty members at Penn for a
longer period of time and held their current rank for a longer period
of time than women. No differences in duration at Penn and at rank
were observed within the Assistant Professor subgroups.
In
several of the areas, responses from the medical faculty were similar
to those obtained from non-medical faculty:
- Women
are as likely to obtain grant support as men, although men are
more likely to be principal investigators.
- Women
publish as many articles and books as men once rank and faculty
status have been taken into account.
- Women
and men both work very long hours (60 hours per week on average).
Men averaged 5hrs/week more than women (62 vs. 57); but this difference
was statistically significant only for faculty at the level of
Assistant Professor.
- A majority
of women respondents report feeling that men faculty are advantaged
and women faculty are disadvantaged at Penn.
- 40%
of women faculty (and 26% of men faculty) believe they are paid
less than their peers.
- Nearly
half of women faculty raise concerns about getting appropriate
credit for their work (compared with almost a third of men who
voice the same complaint).
There
were also several areas of gender differences that appeared to be
unique to School of Medicine faculty women. Women were far more
likely to report that their professional responsibilities and workload
impacted on family and personal time:
- Women
at all ranks reported spending twice as much time in childcare
as men.
- Women
were far more likely to have a spouse/partner who worked full-time
than their male peers (90% vs 48%). Among male faculty, 30% had
spouses who were full-time homemakers, while only 3% of women
faculty had spouses/partners who were at home full-time.
- Women
were also disproportionately impacted by meetings scheduled before
8 a.m. or after 5 p.m., by the absence of on-site childcare and
by the absence of part-time positions.
|
Total Medical
Faculty Assistant Professors |
|
Men |
Women |
Men |
Women |
Difficulty with meetings before
8 |
41%
|
64% |
48% |
69% |
Difficulty with meetings after
5 |
34% |
59% |
38% |
63% |
Absence of on-site childcare |
15% |
35% |
23% |
46% |
Absence of emergency childcare |
19% |
38% |
27% |
45% |
Absence of part-time positions |
9%
|
34% |
10% |
38%
|
All differences
between men and women responses were statistically significant,
p<0.01 |
Women
Assistant Professors in the School of Medicine perceived that they
were less likely to be promoted, and more likely to leave their
academic positions for private practice or jobs in industry.
Aggregated
data suggested that women were significantly more dissatisfied with
their jobs than men; however, analysis by rank indicated that this
was because both male and female Assistant Professors-Clinician
Educator were more dissatisfied than other faculty, and the majority
of women were in this category.
C. Comparison
of Medical and Non-Medical Responses
Since
many of the questions were identical in both the School of Medicine
and non-medical questionnaire, it was possible to directly compare
the responses of the two groups of faculty.
1.
Perceptions of the status of women and men faculty. Faculty
in the School of Medicine, both male and female, appeared more concerned
about the impact of gender differences than faculty in other schools
("non-medical"). Responses to the question: "At Penn,
do you perceive that any of the following groups are either disadvantaged
or advantaged?" are summarized below.
Who is Advantaged? Who
is Disadvantaged?
Respondents:
|
Men |
Women |
Men |
Women |
Non-Medical Men |
18.3% |
13.3% |
5.0% |
15.2% |
Women |
60.3%* |
2.8%* |
0.0% |
46.2%* |
|
|
|
|
|
Medical School Men |
13.7% |
7.0% |
0.9% |
25.9% |
Women |
59.4%* |
1.4%* |
0.5% |
64.4%* |
|
|
|
|
|
* Indicates that
gender gap is statistically significant, p<0.05 |
It
is noteworthy that while 26% of the men in the medical school thought
women faculty were relatively disadvantaged, only 15% of men in
other schools shared that perception.
2.
Perceptions concerning relative rank and salary compared to peers.
When asked to compare their salary with that of faculty within their
department with equivalent training, responsibility and accomplishment,
women in both the School of Medicine and the "non-medical"
group were more likely than men to perceive that their rank and
salary were inappropriately low.
|
Compared
to My Peers
|
My Rank Is |
My
Salary Is |
Respondents: |
Higher |
Lower |
Higher |
Lower |
Non-Medical
|
Men
|
7.8%
|
9.4%
|
14.0%
|
30.4%
|
|
Women
|
2.7%
|
21.0%
|
9.0%
|
55.4%*
|
Medical School
|
Men
|
8.9%
|
13.2%
|
13.0
|
26.0%
|
|
Women
|
3.8%
|
17.1%*
|
2.7%*
|
39.9%*
|
* Indicates that gender gap
is statistically significant, p<0.05
|
D. Data from
Other Institutions
Eight questions
included in this survey were designed to match questions asked in
a 1999 National Survey of Faculty conducted by the National Center
for Educational Statistics. A comparison of the responses of Penn
non-medical faculty to the national results is presented below.
Comparison
of Penn Faculty Survey
(excluding Medical School)
With the 1999
National Survey of
Post-Secondary Faculty
|
|
Penn Faculty
Questionnaire
|
1999
National Survey
of Post-Secondary
Faculty |
|
Men
|
Women
|
Men
|
Women
|
Satisfied
with job overall (%)
|
86.4%*
|
78.6
|
86.1
|
83.3
|
Satisfied
with salary (%)
|
70.8%*
|
60.3
|
58.4*
|
53.3
|
Satisfied
with benefits (%)
|
86.7%
|
89.0
|
67.9*
|
61.3
|
Satisfied
with quality of undergrad. students (%)
|
85.2%*
|
91.2
|
71.6*
|
74.0
|
Satisfied
with quality of graduate students (%)
|
82.2%*
|
92.5
|
83.6*
|
84.3
|
Think
female faculty are treated unfairly (%)
|
10.8%*
|
30.2
|
10.0*
|
29.5
|
Think
minority faculty are treated unfairly (%)
|
11.9%*
|
24.1
|
11.1*
|
20.3
|
Average hours
worked per week
|
59.5
|
58.9
|
48.2*
|
43.3
|
* Indicates
that gender gap is statistically significant, p<0.05
|
Faculty
at Penn exceed national faculty benchmark data in a number of important
respects. Penn faculty are more satisfied with some aspects of their
jobs, e.g. their salaries and benefits and the quality of both undergraduate
and graduate students. Penn faculty, especially women faculty, put
in work much longer work weeks than do faculty at most other institutions.
In other respects, Penn faculty opinion closely mirrors than of
faculty nationally. Women are less satisfied with their jobs overall
and with their salaries. Women are less likely to agree with the
statement that 'women faculty are treated fairly' and 'minority
faculty are treated fairly.' On these two items, Penn faculty--both
male and female--closely match the national averages.
Executive
Summary | Part I: Faculty Census | Part
II: Professional Status of Penn Women Faculty | Part
III: Salary Analyses | Part IV: Faculty Survey
Data | Conclusions
| Data Sources
| Committee Members
Conclusions
The
first official analysis of the status of faculty women at Penn was
the Cohn Committee report [Almanac April 13, 20, and 27,
1971] that was commissioned by the University Council Faculty Affairs
Committee in May 1970. Since that time, periodic updates have been
prepared by the Office of Institution Research and Planning Analysis
and provided by the Provost to each department. These reports have
focused on the distribution of faculty women by department and rank
as well as ten year rolling histories of faculty hiring patterns
for each department in the University.
The
picture that emerged from the Cohn Committee study was similar to
that seen at other universities in the early days of affirmative
action: women were only 7% of the standing faculty at Penn and were
concentrated at the rank of Assistant Professor. Most strikingly,
there were no women Full Professors in the College and many departments
had no women faculty. The present study confirms that there has
been marked progress in several important respects: the proportion
of women faculty University-wide has risen to 24%, women now comprise
18% of the Full Professors in the College, and there are relatively
few departments without women faculty. Nonetheless, it is clear
that major problems still exist.
- Women
are still concentrated at the rank of Assistant Professor in most
disciplines. Major contributing factors are that faculty hired
as Full professors are predominantly male, and resignations among
senior faculty in SAS are disproportionately female.
- In
most of the departments in the School of Medicine, Wharton, and
the Natural Sciences in SAS, the proportion of women Assistant
and Associate Professors is considerably below that predicted.
-
Benchmark data indicate that the number of women Associate
Professors in both the medical and non-medical area is low, and
in some departments there are no women Assistant Professors who
might be considered for promotion to Associate Professor.
- The
number of department chairs who are women is substantially lower
than women's representation among senior faculty.
- Although
not statistically significant within individual groupings, our
analyses of salaries by gender, rank, and cognate discipline groupings
found that women's salaries are slightly lower than men for the
majority of groupings.
- There
are small but persistent differentials with respect to research
space for Full Professors in the sciences and medicine.
The
responses to the faculty survey provide additional important evidence
about faculty perceptions of these differences.
- Although
women faculty appear to be as productive as men faculty with respect
to publication, teaching and research support, they are significantly
less satisfied with their jobs and many issues related to working
conditions.
- Both
male and female faculty perceive that women faculty at Penn are
disadvantaged when compared to men.
- Dissatisfaction
with working conditions is a particular problem among the large
number of women Assistant Professors with clinical responsibilities,
whose work loads and family responsibilities are difficult to
reconcile.
In
1997, the Senate Committee on the Faculty reviewed affirmative action
in faculty hiring [Almanac,
May
13,1997]. They concluded that while some progress had been made
and many schools and departments had "hired women faculty at
rates that match or exceed the proportion of women in the relevant
pool.....Penn still falls far short in having achieved a truly representative
or diverse faculty." This committee re-affirmed the urgency
of continued and expanded efforts to hire more women and minority
faculty throughout Penn, and made several recommendations:
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.
Since
our committee has not attempted to analyze hiring and promotion
processes, we cannot judge whether the requested procedures are
being followed. However, our results indicate that the desired outcomes
have not been achieved. It is therefore time to consider additional
measures.
Our
evidence suggests that the problems reside primarily in individual
departments rather than at the University-wide level. Those actions
that occur centrally, e.g. Provost and Trustee approval of appointments
and promotions, designation of Deans, Research Foundation awards
show no signs of gender inequity. In contrast, those actions which
are the provenance of individual departments or faculty groups,
e.g. selection of new faculty hires, nomination of new department
chairs, establishment of a congenial work environment, have resulted
in many problems. This conclusion is supported by the census data
for individual departments; groups of departments that should be
hiring from a pool of applicants with similar percentages of women
show wide variation in the proportion of faculty women. Clearly,
some departments have been more successful in hiring and providing
an appropriate environment for women faculty than other departments.
With respect to senior faculty hires, however, the under-representation
of women appears to be surprisingly widespread. Because departments
select the faculty hired at senior levels, the underlying causes
for the under-representation must lie within the individual departments
that carry out the selection and initiate the hiring process.
Therefore,
institutional change can only arise from new policies to influence
decision making within departments. The policies must deal with
both departmental hiring practices and creating change in those
environments where women faculty may feel unwelcome or undervalued.
Specific policies must be designed with input from the Deans, who
must play a major role in implementing them. They should address
the following issues:
- How
can departments with longstanding and large deficiencies in hiring
of female faculty be directed to immediately alter their hiring
practices? This might require budgetary constraints on problem
departments and rigorous performance evaluations of department
chairs and Deans. Re-instituting the practice of annually publishing
data on faculty composition by department may influence those
departments that are unaware of their discriminatory practices.
Publishing reports targeting those departments which seem incapable
of hiring women in proportion to their availability might be more
effective.
- How
can all departments increase the proportion of women among senior
faculty hires? It is possible that improved awareness plus
greater scrutiny of proposed senior appointments will motivate
departments to avoid gender stereotypes and "old-boy"
networks, but it is also likely that more stringent controls will
be required in some departments and schools. Departments with
disproportionately few women Assistant Professors should probably
be targeted for particular attention.
- How
to keep more tenured females at Penn? Deans and department
chairs should recognize that high caliber female faculty in many
fields are very likely to receive outside offers. It is therefore
important that department chairs and Deans work to head off dissatisfactions
and be prepared to raise salaries substantially in response to
outside offers.
- How
can we include all schools in a review of the equity of faculty
salaries, with special attention to the salaries of women faculty?
- How
can we deal with the particular problems of junior women faculty
in clinical departments who have workloads incompatible with family
life? Policies in clinical departments and clinical practices
must be adjusted to ensure that it is possible for women as well
as men to retain full time faculty positions.
- How
can we alter environments perceived by most women and many men
faculty as unfriendly to women? The existence of a "chilly
climate" for faculty women is a problem facing most universities.
While it is evident in the responses to our faculty survey, it
is not unique to academia, its existence cannot be proven, and
the causes are thought to be an accumulation of assumptions (often
subconscious) about the capabilities and contributions of professional
women. Given these characteristics, it is unlikely that publicity
or sensitivity training can produce significant change. Therefore,
the best remedy may be to increase the proportion of faculty women,
reaching levels where women make a major contribution to establishing
the climate and where senior women faculty can provide supportive
environments.
Executive
Summary | Part I: Faculty Census | Part
II: Professional Status of Penn Women Faculty | Part
III: Salary Analyses | Part IV: Faculty Survey
Data | Conclusions
| Data Sources
| Committee Members
Data
Sources
- Penn
faculty census: University
of Pennsylvania Office of Institutional Research and Analysis
- Availability
pools: University
of Pennsylvania Office of Institutional Research and Analysis
- Affirmative
Action Reports for Current Standing Faculty, Fall 1992 and Fall
1997; National Science Foundation/Science Resource Studies of
Earned Doctorates
- Benchmark
data:
- Demographic
Survey of Faculty and Administrators at Select Institutions
- 1999-2000,
Harvard University, May 2001
- Research
space and funding:
- SAS
department information from each department, Fall 2000;
- School
of Medicine data from the Office of Planning and Reporting,
- School
of Medicine, Fall 2000
- Administrative
positions, Research Foundation Awards, Lindback Awards, and Term
Chairs: Office
of the Provost
- Endowed
Chairs: Development
Office
- Salary
data: University
of Pennsylvania Office of Institutional Research and Analysis
- Resignation
information: Office
of the Provost and Office of the University Faculty Senate
Executive
Summary | Part I: Faculty Census | Part
II: Professional Status of Penn Women Faculty | Part
III: Salary Analyses | Part IV: Faculty Survey
Data | Conclusions | Data
Sources | Committee Members
Gender Equity Committee
- Fay
Ajzenberg-Selove, Professor of Physics
- Elizabeth
Bailey, Professor and Chair, Public Policy & Management
- Jill
Beech, Professor of Medicine (New Bolton Center)
- Mildred
Cohn, Professor Emeritus of Biochemistry and Biophysics
- Susan
Davidson, Professor of Computer & Information Science
- Vivian
Gadsden, Professor of Education
- Jeane
Ann Grisso, Professor of General Internal Medicine
- Jerry
Jacobs, Professor of Sociology
- Risa
Lavizzo-Mourey, Professor of Geriatrics
- Phoebe
Leboy (co-chair), Professor of Biochemistry (Dental)
- Lynn
Hollen Lees, Professor and Chair, History
- Barbara
Lowery (co-chair), Associate Provost; Professor of Nursing
- Janice
Madden, Professor of Sociology
- Paul
Shaman, Professor of Statistics
NOTES
1
There are several reasons why salary levels differ among faculty
members. If one faculty member has a more highly compensated degree,
or more experience, or has a specialty that other universities or
the broader labor market compensates more highly, he or she is more
likely to receive a higher salary. To quantify gender differences,
it is necessary to control for any systematic differences
between men and women faculty in these characteristics. There is
one important aspect of faculty characteristics that determines
whether they must be included in the analysis of salary differences
by gender: the characteristics are systematic by gender after
the inclusion of all of the other characteristics included in the
study. Therefore, it is only necessary that the analyses
compare equivalently qualified groups of men and women. Any
characteristics that affect salary that are possessed by equivalent
proportions, or in equal intensity, by both men and women after
controlling for any characteristic differences already included
in the model or analysis, cannot affect the size of the gender disparity
and, therefore, cannot affect the "true" level of gender
disparity in salary. The salary analyses are not designed to identify
the particular salary to be paid to particular faculty. Although
it is difficult to imagine a situation where a statistical model
would be used to set salary, such a model would have to include
all relevant qualifications for which any faculty
differ. In that way, a model that is designed to set individual
salaries is fundamentally different from a model that is designed
to determine differences in salaries across groups of faculty defined
by a characteristic, such as gender. In fact, adding characteristics
that do not differ between the genders (even though they do differ
among faculty within each gender) to the analyses may render them
less powerful and more likely to lead to erroneous conclusions.
Salary differences that cannot be explained by differences in credentials
are suspect if they are also associated with gender.
We use regression
analyses. We regressed annual salary on gender and other faculty
characteristics. In general, the regression coefficient for gender
provides an index of the impact of gender on salary after adjusting
(or controlling) for the effect of the other faculty characteristics
included in the regression equation. When a characteristic is described
as being "controlled," the statistical analysis is effectively
comparing outcomes for men and women faculty that are equal or equivalent
with respect to the characteristic. For example, when time since
degree and degree are "controlled," the model is comparing
the average difference salaries for men and women faculty that have
the same degree and have the same time since receiving their degree.
2 The other characteristics
included in the analyses--time since degree, MD degree, clinician-educator,
department, hired in as a full professor--were generally statistically
significant and of the expected sign and magnitude. The adjusted
R2 for the equations ranged from .01 to .83. If we exclude the associate
professor equations, which represent particularly problematic fits,
the adjusted R2 ranged from .10 to .55. These ranges are within
those typical for similar analyses in the published literature.
Almanac, Vol. 48, No. 14, December 4, 2001
|
ISSUE
HIGHLIGHTS:
Tuesday,
December 4, 2001
Volume 48 Number 14
www.upenn.edu/almanac/
President
Rodin has named a Philadelphia lawyer and Penn alumnus as the
new vice president and chief of staff. |
The Gender
Equity Committee's Report on the status of women faculty
at Penn concludes that problems reside primarily in individual
departments rather than at the University-wide level. |
The President
and Provost reply to the Gender Equity Report and indicate
an effort to work more closely with the deans to correct departmental
inequities. |
The University
Council Open Forum will include topics of concern to various
constituencies including staff and students. |
Fire and
Emergency Services has a new director
with decades of experience. |
The Division
of Public Safety's Advisory Board makes recommendations concerning
enforcement of PENNCard policies. |
Dr.
Thomas McNair Scott, a pioneering pediatric researcher and
professor, dies at the age of 100 after an extensive career. |
The new
Faculty/Staff Directory is
out and its cover celebrates the 125 Years of Women at Penn. |
The
report of the Ombudsman compares the cases of conflict handled
by that office over the past several years. |
Retirement
Seminars will be held this week for faculty and staff who
want to prepare for the future. |
Flu
shots will be available to faculty and staff; registration
is required. |
Penn's
Way weekly raffles are underway; the deadline for the next
one is Friday |
|