Downloaded from gas.sagepub.com by guest on September 18, 2011
GENDER INEQUALITY
AND TIME ALLOCATIONS
AMONG ACADEMIC FACULTY
SARAH WINSLOW
Clemson University
This article focuses on faculty members’ allocation of time to teaching and research,
conceptualizing these—and the mismatch between preferred and actual time allocations—as
examples of gender inequality in academic employment. Utilizing data from the 1999 National
Study of Postsecondary Faculty, I find that (1) women faculty members prefer to spend a
greater percentage of their time on teaching, while men prefer to spend more time on research,
although these preferences are themselves constrained; (2) women faculty members spend
a greater percentage of their workweek on teaching and a smaller percentage on research
than men, gaps that cannot be explained by preferences or educational and institutional
attributes; and (3) women faculty members have larger time allocation mismatches than
men—that is, their actual time allocations to both teaching and research diverge more from
their preferred time allocations than those of men. These findings shed light on how gender
inequality is both produced and maintained in this aspect of academic employment and have
implications for job satisfaction, productivity, and the recruitment and retention of current
and future faculty members, especially women.
Keywords: education; organizations; work/occupations
Although much has changed since the 1991 publication of Massachusetts
Institute of Technology’s report on the status of women in the sciences
(Massachusetts Institute of Technology 1991), more recent data indicate that
gender inequality persists in the academy. While women have nearly reached
parity with men in the receipt of doctoral degrees (Survey of Earned Doctorates
2007) and initial academic appointments, women faculty members
achieve tenure and promotion to full professor at a slower rate than men, are
AUTHOR’S NOTE: An earlier version of this article was presented at the 2009 annual
meeting of the Southern Sociological Society in New Orleans. I am grateful to Jerry Jacobs,
Dana Britton, and the anonymous Gender & Society reviewers for their insightful comments
and suggestions.
GENDER & SOCIETY, Vol. 24 No. 6, December 2010 769-793
DOI: 10.1177/0891243210386728
© 2010 by The Author(s)
769
Downloaded from gas.sagepub.com by guest on September 18, 2011
770 GENDER & SOCIETY / December 2010
underrepresented at top ranks and in administrative positions, and earn less
than comparable men (Jacobs and Winslow 2004; Krefting 2003). I explore
another component of the gendered nature of academic employment—faculty
members’ allocation of time to research and teaching and their time allocation
mismatches, or the differences between actual and preferred time
allocations.
While there are minimal gaps in total weekly work hours (Jacobs and
Winslow 2004), women faculty members spend more time teaching than
men, while men spend a larger percentage on research than women (Link,
Swann, and Bozeman 2008). While some have argued that the allocation of
faculty members’ time is largely left up to their own discretion (Massy and
Zemsky 1994), faculty members are constrained by a variety of forces—
institutional structures, disciplinary norms, and cultural expectations—in
allocating their time as they desire. A number of studies indicate that the
correspondence between actual and preferred weekly work hours is weak
for workers in general (Clarkberg and Moen 2001; Jacobs and Gerson 2004;
Reynolds 2003; Reynolds and Aletraris 2006). Moreover, insofar as the
ability to get one’s preferences met is fundamentally an indication of power,
those with more status and in more powerful positions in the academy—
historically, men—ought to have smaller time allocation mismatches.
In this article, I apply insights from research on gender inequality in paid
labor to an examination of faculty members’ actual and preferred time
allocations to teaching and research, conducting analyses of faculty time
use that focus on how gender operates as a social structure “shaping actors’
perceptions of their interests and . . . constraining choice” (Risman 2004,
432). Specifically, I add to the research literature on this topic by (1) investigating
multiple explanations for differences in time allocations, (2) examining
how preferences are themselves constrained, and (3) conducting the
first set of analyses on time allocation mismatches among faculty and potential
explanations for those mismatches.
GENDERED TIME ALLOCATIONS AND TIME ALLOCATION
MISMATCHES: POSSIBLE EXPLANATIONS
Employment-related behavior and outcomes—and any gender-related
differences in these—have a multitude of interrelated causes, ranging from
workers’ preferences and traits to workplace constraints and employers’
demands. A preference-based explanation would assert that gender differences
in time allocation derive largely from gender-differentiated preferences.
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 771
Hakim’s preference theory (Hakim 2002) posits that recent advances have
given women “genuine choices” for the first time; any residual gender gaps
in labor market experiences and outcomes are the result of gender-differentiated
preferences, with institutional constraints relevant “only at the margins” (p.
430). Among university faculty, although there is evidence that women have
higher motivations for teaching (Bailey 1999) and men are more heavily
oriented toward research (Schuster and Finkelstein 2006), strictly preferencebased
approaches, such as Hakim’s theory, fail to acknowledge the extent
to which men’s and women’s preferences, far from being solely a reflection
of individual desires, are constrained by structural factors.
A number of studies indicate that cumulative employment experience and
job characteristics shape an individual’s employment attitudes and work time
preferences (Clarkberg and Moen 2001; Kan 2007; Reynolds 2003; Reynolds
and Aletraris 2007; Risman, Atkinson, and Blackwelder 1999). Several
scholars argue that workplace culture and, specifically, the reward structure
of a given occupation or employment context, shape preferences for work
time (Clarkberg and Moen 2001; Kan 2007; Reynolds 2003; Reynolds and
Aletraris 2007). This suggests that faculty members at institutions that differentially
reward research over teaching may report a preference for allocating
a greater percentage of their time to research than faculty members at
other institutions. To the extent that women are disproportionately likely to
be located at less research-intensive institutions, women’s higher preferences
for teaching and lower preferences for research may be the result of genderdifferentiated
reactions to workplace climates and reward structures. Similarly,
women may prefer to spend more time on teaching because they are
more likely to be instructors and lecturers, positions for which commitment
to teaching is central to occupational success.
Significant attention has been paid to the ways work–family constraints
shape employment attitudes and preferences, particularly, preferences for
hours worked. While it is reasonable to assume that those who experience
the greatest amount of tension in negotiating work and family ought to prefer
to reduce the amount of time they spend in paid labor, the evidence that
marital and parental status affect work hour preferences is mixed (Reynolds
and Aletraris 2006). While McRae (2003) argues that women’s employment
preferences are shaped by child care demands, others (Reynolds 2003;
Reynolds and Aletraris 2007) find that neither family structures associated
with work–family conflict (e.g., single- versus dual-earner parents) nor reported
levels of work–family conflict are associated with preferences for fewer
hours at work. Nonetheless, family status may affect faculty members’ time
allocations—and in gender-specific ways. Specifically, women faculty may
Downloaded from gas.sagepub.com by guest on September 18, 2011
772 GENDER & SOCIETY / December 2010
prefer to spend a greater percentage of time on teaching because, unlike
research commitments, teaching obligations have fixed start and end times
that may more easily accommodate children’s schedules.
The contours of academic careers, including the use of one’s time, are
further shaped by educational credentials and occupational positions—
namely, degree level, degree field, and academic rank. Women, on average,
have credentials and positions associated with greater teaching obligations—
they are less likely to have doctorate degrees (Schuster and Finkelstein 2006);
are overrepresented in education, health fields, and the humanities (National
Center for Education Statistics 2009); and are overrepresented in the teachingoriented
ranks of instructor and lecturer and underrepresented among full
professors (Schuster and Finkelstein 2006). Moreover, one’s educational
credentials and position in the academic hierarchy are also sources of status
and power. This suggests that men faculty members, who are more likely to
have PhDs and be full professors, ought to have smaller time allocation
mismatches than women, who, with less status and power, are less able to
get their preferences met.
Perhaps the most immediate constraint on time allocations, particularly
teaching, are the demands placed on any given faculty member in the form
of course and student loads. Those who teach more classes and/or students
likely spend more time on teaching and, given evidence indicating that time
spent on teaching takes away from time spent on research (Fox 1992), less
time on research. This likely contributes to differences in faculty members’
time allocations. Women often have heavier student loads (Hart and Cress
2008) and mentoring and advising responsibilities (Bird, Litt, and Yong
2004). Moreover, there is evidence that students expect a more intensive,
time-consuming teaching approach from women faculty members (Sprague
and Massoni 2005).
Institutions of higher education may also be compelled to encourage
time allocation patterns that reflect their institutional missions (Link, Swann,
and Bozeman 2008), thus creating another structural constraint on faculty
time allocations. Teaching time is greatest—and increasing most rapidly—at
liberal arts institutions, while the emphasis on research is greatest at research
universities (Milem, Berger, and Dey 2000). To the extent that women are
disproportionately located at teaching-intensive institutions and men at
research institutions, institutional location may explain at least a portion of
the gap in time allocation. Similarly, faculty members may be compelled
to allocate their time in accordance with their perception of the relative
importance of teaching and research at their institution over and above the
effects of institution type. Previous research indicates that individuals’
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 773
allocation of time to work is substantially shaped by the reward structure
of employment (Reynolds 2003). If women are less likely to be at institutions
that value research over teaching, they would be expected to spend a smaller
percentage of time on research and a larger percentage of time on teaching.
In this way, gender gaps in time allocation may be partly explained by gender
gaps in institutional reward structures. Moreover, if research is valued over
teaching, institutions may privilege those who spend a larger percentage of
their time on that activity. To the extent that allocating one’s time as one
desires is a manifestation of that privilege, the reward structure of an institution
may contribute to any gender gaps in time allocation mismatches.
An additional aspect of workplace culture important for understanding
gender differences in time allocation and time allocation mismatches is the
treatment of women faculty members number of studies demonstrate that
women faculty feel less included, respected, valued as researchers, and likely
to be taken seriously than men (Cress and Hart 2009; Fox 2001, 2010). Cress
and Hart (2009) found that just more than half (57 percent) of women feel
that gender equity exists on their campus; this contrasts sharply with the 90
percent of men who feel similarly. Hart and Cress (2008) further demonstrate
how a seemingly unfair allocation of teaching responsibilities affects time
expenditures, arguing that men are more likely to teach “vanity courses with
small enrollments” (p. 182), leaving women disproportionately responsible
for large, core courses. Thus, women and men faculty members may spend
different amounts of time on teaching and research because resources, opportunities,
and expectations are unfairly distributed. Moreover, women may
experience greater time mismatches because of discriminatory actions—
whether they be subtle or direct—that leave them less able to align their
actual and preferred time allocations.
Finally, just as child care and scheduling concerns may shape women’s
time allocation preferences, so too might they affect women’s actual time
allocations. A large body of research indicates that parenthood differentially
affects men’s and women’s employment experiences, including time spent
in paid labor (Kaufman and Uhlenberg 2000). At the same time, scholars
have argued that faculty members allocate their discretionary time to research
(Massy and Zemsky 1994). To the extent that women with children have
less discretionary time than men and childless women, they likely allocate
a smaller percentage of their workweek to research. Marital and parental
status may affect men’s time allocations less insofar as men are less likely
to be responsible for family care needs.
Assumptions about gender and family obligations are also embedded
in the structure of organizations (Acker 1990). A number of scholars have
Downloaded from gas.sagepub.com by guest on September 18, 2011
774 GENDER & SOCIETY / December 2010
argued that the scheduling of work and structure of workplaces are built
around a model of an “ideal worker” who is relatively unencumbered by
responsibilities outside of paid labor, best represented by a man with a
wife at home to manage family responsibilities (Williams 2000). There is
substantial evidence to suggest that academia is built on a male worker
norm. The likelihood of successfully marching through the lockstep life
course of a traditional academic career is much greater for those not encumbered
by family demands (Mason and Goulden 2002), and women faculty
members report being looked down upon as less qualified or committed
academics because they have families (Cress and Hart 2009). Applied to
time allocation, the institutionalization of the male worker norm suggests
that those who fit such an ideal ought to be more likely to get their time
allocation preferences met. Women’s lower likelihood of fitting this mold
could to be a factor in any greater time allocation mismatches they
experience.
METHOD
The Data
The data were drawn from the 1999 National Study of Postsecondary
Faculty (NSOPF), a nationally representative survey of college and university
faculty administered by the National Center for Education Statistics (NCES)
of the U.S. Department of Education. The NSOPF is a multiwave, crosssectional
survey that has been administered four times to date: during the
1987–88, 1992–93, 1998–99, and 2003–4 academic years. The analyses
presented here utilize the 1999 wave because it is the most recent wave with
complete data on actual and preferred time allocations. For the present analysis,
I restricted the sample to full-time faculty members not on sabbatical
during the fall 2008 semester; all employment-related information refers to
this semester. This resulted in a sample size of 12,510.1
Measures2
Actual time allocations. The analyses focus on teaching and research
because (1) they are the two largest components of the faculty workweek
and (2) additional analyses indicate no systematic differences between men
and women in time allocations to the other NSOPF-measured components
of the faculty workweek (professional growth, administration, service, and
consulting).3
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 775
•• Percentage of time spent on teaching: The NSOPF instructs respondents
to report the percentage of their workweek they spend in the classroom,
preparing for class, and advising and mentoring undergraduate and graduate
students.
•• Percentage of time spent on research: The NSOPF defines research as
preparing articles or books, attending or preparing for conferences or
speeches, performing or exhibiting fine or applied arts, seeking funding,
and reviewing proposals.
Time allocation preferences. To measure time allocation preferences,
I use faculty members’ direct reports of the percentage of their workweeks
they would prefer to allocate to instructional activities and research
(defined above).
Time allocation mismatches. These measures—one each for teaching and
research—reflect the mismatch between respondents’ actual and preferred
time allocations, created by subtracting actual from preferred time allocations.
A positive mismatch indicates that a faculty member would prefer to
spend more time on that activity than he or she currently does; a negative
mismatch indicates that a faculty member would prefer to spend less time
on that activity.4
Sex. Respondents were asked to indicate whether they were male or female.
This is a dummy variable with female coded 1 and male coded 0. In my discussion
of the results, I use the terms sex, male, and female since they reflect
responses to this survey measure. In the conclusion, I use the term gender
when arguing that distinctions between male and female respondents reflect the
operation of gender at the individual, interactional, and institutional levels.
Controls.
•• Highest degree: This is a dummy variable coded 1 if the respondents’
highest degree is the highest terminal degree in his or her field (PhD, MD,
JD, etc.) and 0 otherwise.
•• Rank: Respondents are classified as being a (1) full professor, (2) associate
professor, (3) assistant professor, (4) instructor or lecturer, or (5) another
rank or at an institution with no ranking system.
•• Discipline: Respondents are classified as having one of the following
primary disciplines: humanities (reference), biological sciences, physical
sciences, medicine, nursing, other health fields, architecture/engineering,
business, computer science/math, social science, education, vocational
fields, and other fields (see Jacobs and Winslow 2004).
Downloaded from gas.sagepub.com by guest on September 18, 2011
776 GENDER & SOCIETY / December 2010
•• Institution type: Respondents are classified as being at one of the following
types of institutions (Carnegie classification; categories with two tiers
combined): research, doctoral, comprehensive, liberal arts, two year, or
other (theological seminaries, medical and other health profession schools,
business schools, schools of engineering and technology, and law schools).
•• Value of research over teaching at institution: This is a dummy variable
coded 1 for agreement (agree or strongly agree) and 0 for disagreement
(disagree or strongly disagree) with the statement, “At this institution,
research is rewarded more than teaching” (NCES 1999, 25).
•• Number of classes: This measure reflects the total number of classes the
respondent taught in fall 1998 (individualized instruction and labs excluded).
•• Average class size: This is a NSOPF-created variable measuring the average
number of students in each of a respondent’s classes.
•• Number of students receiving individual instruction: This is the number
of students in independent studies or other one-on-one instruction
situations.
•• Number of student committees: This is a measure of the number of undergraduate
and graduate thesis, dissertation, or exam committees on which
the respondent served during fall 1998.
•• Discrimination: While the NSOPF does not contain a direct measure of
personally experienced discrimination, it does contain a measure of perceptions
of the treatment of women at one’s institution. This is included as a dichotomous
measure coded 1 for disagreement (disagree or strongly disagree)
and 0 for agreement (agree or strongly agree) with the statement, “Female
faculty members are treated fairly at this institution” (NCES 1999, 25).
•• Family status: The NSOPF asks directly about marital status but not parental
status. However, the survey does ask respondents to indicate how many
dependents they have (a dependent is defined as someone who gets at least
half of his or her financial support from the respondent). I utilize a created
variable in which respondents are categorized as being (1) married with
dependents, (2) married without dependents, (3) single without dependents,
or (4) single with dependents. Given literature suggesting that the ideal
worker is exemplified by a married person with someone at home managing
family responsibilities, those who are married with dependents are
utilized as the reference group in the analyses.
Analysis Plan
The analyses are presented as a series of regression models predicting
(1) preferred time allocations, (2) actual time allocations, and (3) time
allocation mismatches (see Reynolds and Aletraris 2007). The first set of
models allows for an examination of how preferences are constrained, possibly
in gender-specific ways. The second set considers a series of possible
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 777
explanations for faculty members’ actual time allocations and differences in
time allocations. The third set of models, examining mismatches, is presented
alongside parallel analyses of actual and preferred time allocations, allowing
one to determine whether a given factor affects the gap between preferred
and actual time allocations because it affects preferred time allocations,
actual time allocations, or both (see Reynolds and Aletraris 2007 for an
extended discussion of this methodological approach). In discussing the
results, I focus on differences between men and women and the ability of
other measures in the models to explain these gaps, highlighting the relationship
between other measures and the dependent variable as warranted. In
addition, I present pooled analyses using sex of the respondent as a covariate.
Where relevant, I discuss the results of analyses (not presented) separated
by sex. All regression analyses are adjusted for potential clustering, given
that the 12,510 respondents are located at 813 institutions.
RESULTS
Table 1 focuses on faculty members’ preferred percentage time allocations.
As will be the case in all tables, panel A presents the results for teaching while
panel B presents the results for research. The first model indicates that females
faculty members prefer to spend approximately two-and-one-half percent of
their workweek more on teaching than males. Descriptive analyses indicate
that female faculty members would prefer to spend 52 percent of their workweek
on teaching, while male faculty members would prefer to spend just
more than 48.6 percent of their workweek on teaching.5 The gap in teaching
time preferences is rendered nonsignificant in the second model with the
inclusion of controls. Stepwise inclusion of the other covariates (not shown)
indicates that this is driven by female faculty members’ lower likelihood of
having a doctorate degree. Those with a doctorate, a category in which female
faculty are underrepresented, have lower teaching time preferences than do
those without a doctorate degree.
The effects of other covariates on the sex gap in teaching preferences
essentially offset one another. Women are more likely to be instructors and
lecturers and to be at institutions other than research institutions, both of
which, as predicted, are associated with greater teaching preferences. Contrary
to the expectation that child care demands may raise preferences for
teaching among those (particularly, female faculty) with children those
without dependents, whether single or married, prefer to spend a greater
percentage of their time on teaching than those who are married with
Downloaded from gas.sagepub.com by guest on September 18, 2011
778 GENDER & SOCIETY / December 2010
TABLE 1: Ordinary Least Squares Regression of Preferred Percentage Time
Allocations to Teaching and Research
Panel A: Teaching Panel B: Research
Variable Model 1 Model 2 Model 1 Model 2
Female 2.6164** 0.1682 –5.3231** -0.9144*
Highest degree –2.9366** 10.2840**
Rank
Full — —
Associate 0.9789 0.2151
Assistant 0.6855 3.4239**
Instructor/lecturer 4.4619** –2.3308**
Other rank/No ranking system –8.9771** 0.6958
Field
Humanities — —
Biological sciences –2.3279* 9.2278**
Physical sciences 2.5225* 1.1850
Medicine –11.1122** –8.6163**
Nursing 2.5280* –4.5395**
Other health fields –4.5759** –2.9415**
Architecture/Engineering 2.8364** 1.6857
Business 2.2931* –2.4865**
Computer science/Math 5.1208** –1.8334**
Social science –1.2188 –0.5823
Education 0.9002 –5.2087**
Vocational fields 3.1189 –2.3207*
Other –2.7238** –2.9108**
Institution type
Research ——— ———
Doctoral 6.7111** –7.1583**
Comprehensive 12.3944** –13.8566**
Liberal arts 16.7755** –15.9977**
Two year 21.4674** –9.4673**
Other 3.3804* –8.2736**
Research valued over teaching –0.2682 1.0080
Teaching commitments
No. of classes 1.2583**
Average class size 0.1502**
No. independent students 0.0456**
No. of student committees 0.2237**
Women treated unfairly –1.1512* 0.1236
Family status
Married with dependents ——— ———
Married without dependents 2.3890** –0.8433*
Single without dependents 1.1583* –0.1098
Single with dependents 0.5356 –0.1992
SOURCE: 1999 National Study of Postsecondary Faculty (National Center for Education Statistics 1999).
*p < .05. **p < .01.
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 779
dependents. Female faculty members are more likely than male faculty
members not to have dependents. At the same time, female faculty are more
likely to perceive that their institutions value research over teaching and/
or treat women unfairly, both of which are associated with lower teaching
preferences. Additional analyses indicate that the treatment of women at one’s
institution interacts with sex such that female—but not male—respondents
have lower preferences for teaching when they are at institutions they perceive
as treating women unfairly.
The first model of panel B indicates that women prefer to spend just more
than five percent less of their workweek on research than do men; specifically,
male faculty members prefer to allocate just less than one-quarter of
their workweek to research, while female faculty members prefer to spend
just less than one-fifth of their workweek engaged in research. Controlling
for all other measures in the second model, the sex gap is reduced to less
than one percent. Stepwise models not presented indicate that a substantial
portion of the sex gap in research time preferences can be explained by
female faculty members’ lower likelihood of having a doctorate degree
(insofar as those with doctorate degrees prefer to spend a greater percentage
of their workweek on research than do those without doctorates). Rank
explains little of the sex gap in research time preferences because female
faculty members are more likely than male faculty members to be both
assistant professors and instructors or lecturers; the former is associated with
greater research preferences, while the latter is associated with a smaller
preferred percentage time allocation to research. Institutional context also
plays a role—female faculty are less likely to be located at research institutions
or at institutions that value research over teaching, both of which are
associated greater preferences for research.
Table 2 shifts the focus to actual time allocations to teaching (panel A)
and research (panel B). The first model of panel A indicates that female
faculty members spend about four percent more of their workweek on
teaching than male faculty; while female faculty members spend approximately
56 percent of their workweek (or approximately 29 hours per week)
on teaching, male faculty members spend less than 52 percent of their
workweek on instructional activities (or just less than 28 hours per week).
The second model of panel A includes a control for respondents’ preferred
percentage of teaching time, which is positively associated with actual
teaching time. Differences in preferences—which, recall from Table 1, are
themselves constrained—account for approximately 50 percent of the sex
gap in time allocated to teaching. Yet given equal preferences, female faculty
allocate approximately two percent more of their workweek to teaching
than male faculty.
Downloaded from gas.sagepub.com by guest on September 18, 2011
780
TABLE 2: Ordinary Least Squares Regression of Actual Percentage of Time Spent on Teaching and Research
Panel A: Teaching Panel B: Research
Variable Model 1 Model 2 Model 3 Model 1 Model 2 Model 3
Female 4.4215** 2.0850** 1.7642** –5.4248** –1.3166** –0.9439**
Preferred teaching % 0.8930** 0.8002**
Preferred research % 0.7718** 0.7312**
Highest degree 1.6832** –0.2773
Rank
Full — —
Associate 1.6467** –0.3916
Assistant 3.6284** 0.2368
Instructor/lecturer 3.8001** 0.1786
Other rank/No ranking system –3.8409** 3.7739**
Field
Humanities — —
Biological sciences –1.1062 4.0112**
Physical sciences 0.8824 1.2456*
Medicine –5.5509** –0.5731
Nursing 0.3990 0.0794
Other health fields –2.4869** 1.3123*
Architecture/Engineering –0.2538 1.3228*
Business –1.2085 1.1969**
Computer science/Math 0.1015 0.9695*
Social science –1.1101* –0.5088
Education –3.2126** 0.6157
Vocational fields –1.7460 1.2336*
Other –4.0035** 1.5256**
(continued)
Downloaded from gas.sagepub.com by guest on September 18, 2011
781
TABLE 2: (continued)
Panel A: Teaching Panel B: Research
Variable Model 1 Model 2 Model 3 Model 1 Model 2 Model 3
Institution type
Research — —
Doctoral 1.7303** –3.0986**
Comprehensive 3.9832** –4.6844**
Liberal arts 4.8287** –5.4743**
Two year 3.5982** –4.1624**
Other 0.6105 –2.0755**
Research valued over teaching –1.2962** 0.4378
Teaching commitments
No. of classes 0.8832**
Average class size 0.0813**
No. independent students 0.0147*
No. of student committees 0.1553**
Women treated unfairly 0.1901 –0.5936*
Family status
Married with dependents — —
Married without dependents 0.706* –0.3390
Single without dependents 0.5815 0.1548
Single with dependents 0.1974 –0.7604
SOURCE: 1999 National Study of Postsecondary Faculty (National Center for Education Statistics 1999).
*p < .05. **p < .01.
Downloaded from gas.sagepub.com by guest on September 18, 2011
782 GENDER & SOCIETY / December 2010
The inclusion of controls in the third model of panel A reduces the sex
gap in teaching time allocation further. This is driven by female faculty
members’ lower likelihood of having a doctorate degree as well as their
concentration in more teaching-intensive ranks and at more teaching-intensive
institutions. Perceiving that one’s institution values research over teaching
is, as predicted, negatively associated with actual teaching time allocations,
although additional analyses indicate that this effect is concentrated among
male faculty. Those who are married without dependents spend a greater
percentage of their workweek on teaching than do those who are married
with dependents, offering partial support for the effect of fitting the idealworker
norm. Additional analyses indicate that this relationship is concentrated
among male faculty. Controlling for preferences, educational attributes,
rank, and institutionalized constraints, a sex gap of nearly two percent
remains. In the context of weekly work hours, controlling for all measures
in the model, female faculty spend approximately an additional hour more
per week on teaching than do male faculty.
The three models of panel B present similar analyses of the percentage
of time faculty members report actually devoting to research. The first model
indicates that female faculty spend more than five percent less of their workweek
on research than do male faculty. Descriptive analyses indicate that
female faculty members spend approximately 13 percent of their workweek
(approximately seven hours) on research, while male faculty members spend
approximately 18 percent of their workweek (approximately 10 hours)
engaged in research. This weekly three-hour gap accumulates to approximately
45 hours in a 15-week academic semester. The second model indicates
that preferences explain approximately three-quarters of the sex gap in the
percentage of time allocated to research. Yet controlling for their smaller
preferred percentage of research time, which, recall from Table 1, is itself
shaped by a multitude of factors, female faculty members spend about one
percent less of their workweek on research than do male faculty members
(roughly one half hour).
The sex gap in time allocated to research is further reduced by the inclusion
of controls in the third model. This is largely due to female faculty
members’ disproportionate concentration in less research-intensive disciplines
and institution types. Unlike the percentage of time allocated to teaching,
research time allocations are not significantly associated with the
perception that one’s institution values research over teaching, suggesting
that institutional climate discourages teaching time rather than increasing
research time allocations. While teaching time allocations were not affected
by the perceived treatment of women, faculty members who perceive that
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 783
their institutions treat women unfairly spend a smaller percentage of time
on research than do those who perceive that their institution treats women
fairly. Although not testable here, it is possible that a reduced ability to spend
time on research is a manifestation of an unfair climate.
Family status is not significantly associated with research time allocations
in the pooled sample. Additional analyses indicate that for male faculty, those
who are married without dependents spend a smaller percentage of time on
research than those who are married with dependents. For female faculty,
those who are single with dependents spend a smaller percentage of time on
research than those who are married with dependents. Controlling for all
measures in the model, female faculty members spend a smaller percentage
of their time on research than do their male colleagues. More specifically,
there is a sex gap in research time allocations of just less than one percent,
or the equivalent of just less than one half hour per week less on research for
a female faculty member compared to an otherwise identical male.
Table 3 examines time allocation mismatches. For each set of analyses,
I present four models—one for preferences (identical to the model presented
in Table 1), one for actual time allocations (similar to the final model in
Table 2, but without a measure of preferences), and two for the mismatch
between preferred and actual time allocations (one including just sex and
the other including all covariates). I focus on the mismatch model, referring
to the other two models for comparative purposes. Given how the mismatch
measure is calculated, a negative value indicates wishing to spend less time
on teaching or research than one currently does, while a positive value
indicates wishing to spend more time than one currently does.
The first allocation mismatch model in panel A indicates that, compared
to male faculty, female faculty would prefer to spend nearly two percent
less of their workweek on teaching than they currently do. The inclusion
of all other covariates in the second mismatch model has little effect on the
sex gap in teaching time allocation mismatches. Stepwise inclusion of all
measures in the model indicate that while some of female faculty member’s
attributes (namely, their lower likelihood of having a doctorate degree)
exacerbate the relationship between sex and teaching time mismatches, they
have some attributes (e.g., they are more likely to be assistant professors
and instructors) and face some constraints on their time (e.g., they teach
more classes and provide more students with individualized instruction)
that attenuate the relationship between sex and teaching time mismatches.
A sex gap in teaching time mismatches remains—compared to male faculty
members, female faculty members wish to spend 1.7 percent less of their
workweek (or approximately one hour less per week) on teaching than they
Downloaded from gas.sagepub.com by guest on September 18, 2011
784
TABLE 3: Ordinary Least Squares Regression of Time Allocation Mismatches
Panel A: Teaching Panel B: Research
Allocation Mismatch Allocation Mismatch
Variable Preferred Actual Model 1 Model 2 Preferred Actual Model 1 Model 2
Female 0.1682 1.8988** –1.8052** –1.7306** –0.9144* –1.6125** 0.1017 0.6981**
Highest degree –2.9366** –0.6674 –2.2692** 10.2840** 7.2427** 3.0413**
Rank
Full — — — — — —
Associate 0.9789 2.4300** –1.4511** 0.2151 –0.2343 0.4494
Assistant 0.6855 4.1769** –3.4914** 3.4239** 2.7405** 0.6834
Instructor/lecturer 4.4619** 7.3703** –2.9084** –2.3308** –1.5258** –0.8050*
Other rank/No ranking system –8.9771** –11.0239** 2.0469** 0.6958 4.2826** –3.5869**
Field
Humanities — — — — — —
Biological sciences –2.3279* –2.9689** 0.6410 9.2278** 10.7589** –1.5311**
Physical sciences 2.5225* 2.9007* –0.3783 1.1850 2.1121* –0.9271
Medicine –11.1122** –14.4421** 3.3302** –8.6163** –6.8737** –1.7427**
Nursing 2.5280* 2.4218* 0.1062 –4.5395** –3.2401** –1.2994*
Other health fields –4.5759** –6.1484** 1.5724 –2.9415** –0.8387 –2.1029**
Architecture/Engineering 2.8364** 2.0158 0.8206 1.6857 0.0901 –1.7758**
Business 2.2931* 0.6263 1.6668* –2.4865** –0.6213 –1.8652**
Computer science/Math 5.1208** 4.1989** 0.9219 –1.8334** –0.3711 –1.4623**
Social science –1.2188 –2.0853** 0.8665 –0.5823 –0.9346 0.3523
Education 0.9002 –2.4923* 3.3925** –5.2087** –3.1931** –2.0156**
Vocational fields 3.1189 0.7496 2.3693* –2.3207* –0.4634 –1.8573*
Other –2.7238** –6.1829** 3.4591** –2.9108** –0.6028 –2.3079**
(continued)
Downloaded from gas.sagepub.com by guest on September 18, 2011
785
TABLE 3: (continued)
Panel A: Teaching Panel B: Research
Allocation Mismatch Allocation Mismatch
Variable Preferred Actual Model 1 Model 2 Preferred Actual Model 1 Model 2
Institution type
Research — — — — — —
Doctoral 6.7111** 7.1002** –0.3891 –7.1583 –8.3330** 1.1747*
Comprehensive 12.3944** 13.9006** –1.5062** –13.8566** –14.8169** 0.9603**
Liberal arts 16.7755** 18.2517** –1.4762* –15.9977** –17.1723** 1.1746*
Two year 21.4674** 20.7754** 0.6920 –9.4673** –17.1352** –0.6057
Other 3.3804* 3.3153 0.0651 –8.2736** –8.1255** –0.1481
Research valued over teaching –0.2682 –1.5108** 1.2426** 1.0080 1.1750** –0.1669
Teaching commitments
No. of classes 1.2583** 1.8901** –0.6318**
Average class size 0.1502** 0.2015** –0.0513**
No. independent students 0.0456** 0.0512** –0.0056
No. of student committees 0.2237** 0.3343** –0.1106*
Women treated unfairly –1.1512* –0.7310 –0.4202 0.1236 –0.4666 0.6402*
Family status
Married with dependents — — — — — —
Married without dependents 2.3890** 2.6177** –0.2287 –0.8433* –0.9556* 0.1123
Single without dependents 1.1583* 1.5083* –0.3500 –0.1098 0.0745 –0.1843
Single with dependents 0.5356 1.1905 –0.6549 –0.1992 –0.9061 0.7069
SOURCE: 1999 National Study of Postsecondary Faculty (National Center for Education Statistics 1999).
*p < .05. **p < .01.
Downloaded from gas.sagepub.com by guest on September 18, 2011
786 GENDER & SOCIETY / December 2010
currently do. This stems largely from the sex gap in actual time allocations.
In other words, while female and male faculty members have similar teaching
time allocation preferences (net of controls), female faculty spend more
time on research than do their male counterparts, thus producing a greater
negative mismatch for female faculty than for male faculty.
The first mismatch model in panel B indicates no significant sex difference
in research time allocation mismatches. The second model of panel B
indicates that, net of controls, female faculty members, more so than males,
would prefer to be allocating more of their time to research than they actually
are. Stepwise inclusion of all measures in the model (analyses not shown),
indicate that the sex gap in research time allocation mismatches stems largely
from differences in education level and rank. Looking across all three models
of the panel, we see that female faculty members have a greater mismatch
because although they prefer to allocate just less than one percent less of
their weekly work time to research, they actually allocate approximately 1.5
percent less of their weekly work time to research than male faculty. In
other words, the magnitude of the sex gap in actual time allocations is larger
than the magnitude of the sex gap in preferences, thus producing a sexdifferentiated
mismatch disadvantaging female faculty.
Also of note, those who perceive that their institutions treat women unfairly
have a larger positive time allocation mismatch than those who perceive
their institutions to be fair (a relationship that does not differ by sex), suggesting
that a climate of unfairness translates into a greater difficulty in
getting one’s preferences met. Additional analyses also indicate that female
faculty members who are single with dependents have a significantly greater
positive research time allocation mismatch than do female faculty members
who are married with dependents, indicating that those who do not fit the
ideal-worker norm—and, arguably, those who deviate the most from it—are
less able to get their preferences for research time met.
CONCLUSION
The results offer support for the contention that gender is embedded in
the structure of academic careers. They also add to our understanding of
gender inequality in academia by identifying and attempting to explain
differential time allocation preferences, behaviors, and mismatches. Several
findings stand out. First, women faculty report preferring to spend a larger
percentage of their workweek on teaching and a smaller percentage on
research, although these preferences are themselves shaped by gendered
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 787
opportunities. Women’s greater preferences for teaching time allocation can
be explained by their lower likelihood of having a doctoral degree—there
is no evidence for gender-differentiated teaching preferences beyond their
dependence on educational credentials. Moreover, a sizeable portion of the
gender gap in research time allocation preferences can be explained by
educational credentials and institutional features. While human capital
theorists would argue that gender-differentiated teaching and research
preferences largely or completely reflect the choice to differentially invest
in one’s credentials, one must consider how the “choice” to invest in one’s
education is itself constrained. Given that the data utilized here represent a
cross-section of American faculty in the late 1990s, the legacy of men’s
historic advantage in doctorate degree receipt is evident. Recent data indicating
that women have now surpassed men in the receipt of doctoral degrees
(Jaschik 2010) suggest that gender gaps in time allocation preferences ought
to be narrower among current and future academic faculty.
The results also indicate that research and teaching time allocation preferences
are shaped by the features of the institutions in which men and women
are located. This suggests that gender-differentiated preferences may in part
reflect the constraints women face in obtaining positions comparable to those
of men. While exploring the sources of gender gaps in education and institutional
location is beyond the scope of this article, research on preferences
for work time demonstrates that individuals accommodate their preferences
to their situations rather than vice versa (Reynolds and Aletraris 2006, 2007).
Women faculty members may prefer to spend a larger percentage of time
on teaching and a smaller percentage of time on research not because this
is what they prefer under ideal circumstances but because, as a result of a
legacy of gender inequality in educational and occupational opportunities,
these are the options they perceive as realistic.
Second, women faculty devote a larger percentage of their weekly working
time to teaching and a smaller percentage to research than men. A
sizeable portion of the gap in teaching and research time allocations can be
explained by gender-differentiated (and constrained) preferences, women’s
lower likelihood of having a doctorate degree, their overrepresentation in
teaching-intensive ranks and institution types, and their underrepresentation
in research-intensive ranks and institution types. The perception that one’s
institution values research over teaching is associated with lower teaching
time allocations for men but is not significantly related to research time
allocations, suggesting that such an institutional climate discourages teaching
time rather than increases research time allocations. Family status affects
teaching and research time allocations and in gender-specific ways. For men,
Downloaded from gas.sagepub.com by guest on September 18, 2011
788 GENDER & SOCIETY / December 2010
those who are married without dependents spend a smaller percentage of
time on research than do those who are married with dependents, offering
some support for the claim that those who fit the male ideal-worker norm are
advantaged in academia (insofar as research is a more highly regarded activity).
For women, those who are single with dependents spend a smaller percentage
of time on research than do those who are married with dependents,
offering support for the advantages accruing to those who fit the ideal-worker
norm and suggesting that women in arguably the most demanding family
context—single mothers with children—reduce the time they allocate to what
is likely the most discretionary area of faculty time allocation, research.
Finally, women are more likely than men to be teaching more than they prefer
and researching less, gaps that cannot be fully explained by educational, institutional,
and family status attributes. This result offers support for the contention
that higher-status groups, in this case, men, are seen by administrators as more
deserving of rewards (Krefting 2003), indexed here by the ability to get one’s
preferences met. Moreover, insofar as power can be defined as the ability to
get one’s preferences met, the gender gap in time allocation mismatches represents
a previously undocumented example of men’s continued greater power
in the academy. It is important to underscore that men have a time allocation
advantage over and above the effect of their having other, measureable attributes
associated with status (e.g., rank, institutional affiliation, etc.).
Taken together, the analyses presented here indicate substantial gaps in
the use of faculty members’ time, gaps that cannot be explained by measureable
attributes included in the models. What might explain these gaps? The
results are consistent with research and theory on the impact of gender status
beliefs in shaping behaviors, expectations, and perceptions of the rewards
that we and others deserve (Ridgeway 1997). This perspective would argue
that despite growing equality in men’s and women’s representation in academia,
the interactional processes by which gender status beliefs produce
inequality result in gender differences in faculty members’ actual and preferred
time allocations and time allocation mismatches.
Research indicates that preferences (including career aspirations) are
shaped by individuals’ assessments of their likely success in a given activity;
these attributions for success are themselves shaped by cultural ideologies
about gender (Correll 2004). It may be that men prefer to spend a larger
percentage of their time on research because gender status beliefs dictating
their greater competence (particularly at “masculine” tasks) shape their
individual time allocation preferences. The results are also consistent with
the argument that gender status beliefs influence behavior over and above
their effect on preferences because individuals adjust their activities in light
of the gendered expectations others have of them (Correll 2001). Women’s
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 789
larger actual time allocations to teaching, for example, suggest that they may
be influenced by students’ expectations of a more time-intensive teaching
approach from their women professors (Sprague and Massoni 2005). At the
same time, the results are consistent with research indicating that competencerelated
gender status expectations raise one’s definition of the work required
to be successful at tasks deemed appropriate for one’s gender (Eccles 1994).
Specifically, gender-differentiated time allocations may be attributed to a
higher internalized teaching bar for women and a higher internalized research
bar for men.
Although these results offer compelling evidence for persistent gender
inequality in academia, they are limited in at least two ways. First, the data
are cross-sectional, so we cannot know whether individuals accommodate
their preferences to their behavior over time, as some research has suggested
(Reynolds and Aletraris 2007). In addition, the data were gathered more than
a decade ago, before institutions of higher education were profoundly affected
by the recent economic downturn. At many institutions, this has meant
increased student and/or course loads, which may have increased faculty
members’ actual time allocations to teaching and, given evidence that the two
obligations are competing (Fox 1992), decreased time on research. If this has
happened with little change in preferences, time allocation mismatches have
likely grown as well.
What this means for gender inequality is an empirical question. On one
hand, if increased teaching demands have been relatively equally distributed,
overall time allocations may have shifted, but any gender gaps ought to have
remained consistent. On the other hand, gender gaps may have grown even
if increased teaching obligations are distributed equally insofar as men’s
general reluctance in moving into female-dominated activities (England
2010) suggests they may be slow to increase their teaching time significantly
(even when faced with growing teaching obligations). Furthermore, if
increased teaching demands are disproportionately imposed on those already
concentrating more on teaching, the current economic downturn may further
disadvantage women faculty members. Although recent data (National Science
Foundation, Division of Science Resources Statistics 2009) indicate
few differences in men’s and women’s teaching and research time, suggesting
that time allocation gaps have narrowed, the analyses focused on only
handful of fields at research institutions. Examining the contours of faculty
time use—and how they may be gendered—during periods of institutional
transformation is a ripe area for future research.
How and why do these time allocation differences matter? Previous
research indicates that those with heavier teaching obligations, particularly
women, report more stress (Hart and Cress 2008), while those who spend
Downloaded from gas.sagepub.com by guest on September 18, 2011
790 GENDER & SOCIETY / December 2010
a greater percentage of their time on research report higher workload satisfaction
(Jacobs and Winslow 2004). Taken together, these findings suggest
that women faculty may be more stressed and less satisfied than men. The
time allocation differences presented here also have implications for research
productivity and career success. The percentage of time one spends on
research is positively associated with research productivity, while the percentage
of time spent on teaching is negatively related to research productivity
(Jacobs and Winslow 2004). Since research time is linked to research
productivity (Jacobs and Winslow 2004) and research productivity remains
a key component in employment reviews (Lewis 2004), gender differences
in research time allocations may contribute to gender-differentiated patterns
of tenure and promotion.
At the same time, women may be more likely than men to exit academia.
Reynolds and Aletraris (2006) find that individuals are more likely to resolve
work hour mismatches by changing employers than by making adjustments
while maintaining employment continuity. Moreover, Boheim and Taylor
(2004) find that women are more likely than men to exit an employment
situation when their preferences and actual behavior do not match—a position
in which women are disproportionately likely to find themselves. If they
choose to leave any given position, the data presented here, in combination
with previous research, suggest that women faculty may be less marketable
than their male colleagues. Since women bear a disproportionate responsibility
for labor that is institution specific (e.g., institutional housekeeping,
mentoring individual students), their investments are less likely to be portable
across institutions. This stands in stark contrast to men, whose investments
in research make them more highly desirable candidates should they choose
to leave their own institutions. Finally, the results have implications for the
ability of American educational institutions to attract young, newly minted
women PhDs to academic employment. Faced with institutionalized expectations
for their behavior and the prospect of disproportionately constrained
time allocations, these women may pursue employment outside academia
(Mason, Goulden, and Frasch 2009).
NOTES
1. Because of restricted-access data limitations, the sample size reported here
is rounded to the nearest 10.
2. National Center for Education Statistics (1999).
3. A limitation of the National Study of Postsecondary Faculty for faculty timeuse
studies is the manner in which it categorizes service and administration. For
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 791
example, administrative activities are described as including “departmental or
institution-wide meetings or committee work,” which many might well consider
to be service. This may account for the lack of difference between men’s and
women’s time spent in service activities.
4. Analyses not shown indicate that 43.9 percent of faculty members (41.2 percent
of males and 48.5 percent of females) spend a greater percentage of their time on
teaching than they would prefer, 32.5 percent (34.7 percent of males and 28.8 percent
of females) allocate exactly the percentage of time to teaching that they prefer, while
23.6 percent (24.1 percent of males and 22.7 percent of females) teach less than they
would prefer. More than half (54.5 percent; 54.2 percent of males and 54.9 percent
of females) spend a smaller percentage of their time than they would prefer on
research, 35.8 percent (36.2 percent of males and 35.0 percent of females) allocate
their research time as they prefer, and 9.8 percent (9.6 percent of males and 10.1 percent
of females) spend a greater percentage of time on research than they would prefer.
5. The slight difference between descriptive and multivariate analyses is attributable
to the different estimation techniques employed.
REFERENCES
Acker, Joan. 1990. Hierarchies, jobs, bodies: A theory of gendered organizations.
Gender & Society 4 (2): 139-58.
Bailey, Jeffrey G. 1999. Academics’ motivation and self-efficacy for teaching and
research. Higher Education Research and Development 18 (3): 343-59.
Bird, Sharon, Jacquelyn Litt, and Wang Yong. 2004. Creating status of women
reports: Institutional housekeeping as “women’s work.” NWSA Journal 16 (1):
194-206.
Boheim, Rene, and Mark P. Taylor. 2004. Actual and preferred working hours.
British Journal of Industrial Relations 42 (1): 149-66.
Clarkberg, Marin, and Phyllis Moen. 2001. Understanding the time-squeeze:
Married couples’ preferred and actual work-hour strategies. American Behavioral
Scientist 44 (7): 1115-36.
Correll, Shelley J. 2001. Gender and the career choice process: The role of biased
self-assessments. American Journal of Sociology 106 (6): 1691-1730.
Correll, Shelley J. 2004. Constraints into preferences: Gender, status, and emerging
career aspirations. American Sociological Review 69 (1): 93-113.
Cress, Christine M., and Jeni Hart. 2009. Playing soccer on the football field: The
persistence of gender inequities for women faculty. Equity and Excellence in
Education 42 (4): 473-88.
Eccles, Jacquelynne S. 1994. Understanding women’s educational and occupational
choices. Psychology of Women Quarterly 18 (4): 585-609.
England, Paula. 2010. The gender revolution: Uneven and stalled. Gender & Society
24 (2): 149-66.
Fox, Mary Frank. 1992. Research, teaching, and publication productivity: Mutuality
versus competition in academia. Sociology of Education 65 (4): 293-305.
Downloaded from gas.sagepub.com by guest on September 18, 2011
792 GENDER & SOCIETY / December 2010
Fox, Mary Frank. 2001. Women, science, and academia: Graduate education and
careers. Gender & Society 15 (5): 654-66.
Fox, Mary Frank. 2010. Women and men faculty in academic science and engineering:
Social-organizational indicators and implications. American Behavioral
Scientist 53 (7): 997-1012.
Hakim, Catherine. 2002. Lifestyle preferences as determinants of women’s differential
labor market careers. Work and Occupations 29 (4): 428-59.
Hart, Jennifer L., and Christine M. Cress. 2008. Are women faculty just “worrywarts?”
Accounting for gender differences in self-reported stress. Journal of
Human Behavior in the Social Environment 17 (1/2): 175-93.
Jacobs, Jerry A., and Kathleen Gerson. 2004. The time divide. Cambridge, MA:
Harvard University Press.
Jacobs, Jerry A., and Sarah Winslow. 2004. Overworked faculty: Job stresses and
family demands. Annals of the American Academy of Political and Social
Science 596 (November): 104-29.
Jaschik, Scott. 2010. Women lead in doctorates. Inside Higher Ed. http://www
.insidehighered.com/news/2010/09/14/doctorates (accessed September 15, 2010).
Kan, Man Yee. 2007. Work orientation and wives’ employment careers: An evaluation
of Hakim’s preference theory. Work and Occupations 34 (4): 430-62.
Kaufman, Gayle, and Peter Uhlenberg. 2000. The influence of parenthood on the
work effort of married men and women. Social Forces 78 (3): 931-49.
Krefting, Linda A. 2003. Intertwined discourses of merit and gender: Evidence
from academic employment in the USA. Gender, Work and Organization 10(2):
260-78.
Link, Albert N., Christopher A. Swann, and Barry Bozeman. 2008. A time allocation
study of university faculty. Economics of Education Review 27 (4): 363-74.
Lewis, Philip. 2004. The publishing crisis and tenure criteria: An issue for research
universities? Profession 2004: 14-24.
Mason, Mary Ann, and Marc Goulden. 2002. Do babies matter? Academe 88 (6):
21-27.
Mason, Mary Ann, Marc Goulden, and Karie Frasch. 2009. Why graduate students
reject the fast track. Academe 95(1): 11-16.
Massachusetts Institute of Technology. 1991. A study on the status of women
faculty in science at MIT. http://web.mit.edu/fnl/women/women.pdf (accessed
May 5, 2010).
Massy, William F., and Robert Zemsky. 1994. Faculty discretionary time: Departments
and the “academic ratchet.” Journal of Higher Education 65 (1): 1-22.
McRae, Susan. 2003. Constraints and choices in mothers’ employment careers:
A consideration of Hakim’s preference theory. British Journal of Sociology 54 (3):
317-38.
Milem, Jeffrey F., Joseph B. Berger, and Eric L. Dey. 2000. Faculty time allocation:
A study of change over twenty years. Journal of Higher Education 71 (4):
454-75.
Downloaded from gas.sagepub.com by guest on September 18, 2011
Winslow / TIME ALLOCATIONS AMONG ACADEMIC FACULTY 793
National Center for Education Statistics. 1999. 1999 National Study of Postsecondary
Faculty: Faculty questionnaire. U.S. Department of Education Office
of Educational Research and Improvement, Washington, DC.
National Center for Education Statistics. 2009. Digest of education statistics: 2009.
http://nces.ed.gov/programs/digest/d09/ (accessed May 5, 2010).
National Science Foundation, Division of Science Resources Statistics. 2009. Doctorate
recipients from U.S. universities: Summary Report 2007-08. Special Report
NSF 10-309. http://www.nsf.gov/statistics/nsf10309/ (accessed August 27, 2010).
Reynolds, Jeremy. 2003. You can’t always get the hours you want: Mismatches
between actual and preferred work hours in the U.S. Social Forces 81 (4):
1171-99.
Reynolds, Jeremy, and Lydia Aletraris. 2006. Pursuing preferences: The creation and
resolution of work hour mismatches. American Sociological Review 71:618-38.
Reynolds, Jeremy, and Lydia Aletraris. 2007. For love or money? Extrinsic rewards,
intrinsic rewards, work-life issues, and hour mismatches. Research in the Sociology
of Work 17:285-311.
Ridgeway, Cecilia L. 1997. Interaction and the conservation of gender inequality:
Considering employment. American Sociological Review 62 (2): 218-35.
Risman, Barbara J. 2004. Gender as a social structure: Theory wrestling with
activism. Gender & Society 18 (4): 429-50.
Risman, Barbara J., Maxine P. Atkinson, and Stephen P. Blackwelder. 1999. Understanding
the juggling act: Gendered preferences and social structural constraints.
Sociological Forum 14 (2): 319-44.
Schuster, Jack H., and Martin J. Finkelstein. 2006. The American faculty: The
restructuring of academic work and careers. Baltimore: Johns Hopkins
University Press.
Sprague, Joey, and Kelley Massoni. 2005. Student evaluations and gendered expectations:
What we can’t count can hurt us. Sex Roles 53 (11/12): 779-93.
Survey of Earned Doctorates. 2007. Doctorate recipients from United States universities:
Selected tables 2007. http://www.norc.org/NR/rdonlyres/2D5FD7C8-
4AE0-4932-B777-0BC8EA7965EF/0/2007_selectedtabs.pdf (accessed March 6,
2009).
Williams, Joan. 2000. Unbending gender: Why family and work conflict and what
to do about it. New York: Oxford University Press.
Sarah Winslow is an assistant professor in the Department of Sociology and
Anthropology at Clemson University. Her research focuses on gender, paid
labor, and family and has been published in Journal of Marriage and Family;
Journal of Family Issue; Community, Work, and Family; and Annals of the
American Academy of Political and Social Science.
Downloaded from gas.sagepub.com by guest on September 18, 2011