Bias Against Caregiving
Faculty members rarely take advantage of family-friendly workplace policies. What are we so afraid of?
By Robert Drago, Carol Colbeck, Kai Dawn Stauffer, Amy Pirretti, Kurt Burkum, Jennifer Fazioli, Gabriela Lazarro, and Tara Habasevich
Few eligible faculty members take formal leaves for childbearing or caregiving. The Faculty and Families Project at our university, for example, found that between 1992 and 1999, only four of 257 tenure-track faculty parents at Pennsylvania State University took any formal family leave.1 Perhaps faculty parents were unaware of leave policies, or maybe department heads discouraged them from taking leave. We suspect, however, that biases against caregiving in the academy caused them to avoid taking time off.
As a result of such biases, faculty members suffer career penalties for using policies designed to help them balance work and family commitments. To escape these penalties, faculty members rarely use the policies, engaging in what we label “bias avoidance” strategies. Because the biases are often hidden, faculty who even inquire about relevant policies risk damaging their academic reputation. Many choose to avoid doing so, fearing that if they so much as ask about the rules, they will not be considered serious players in the academic game.
In 2002, we set out to study the extent and nature of bias avoidance in the academy, and how it might be alleviated, with funding from the Alfred P. Sloan Foundation.2 The research included a national survey administered to 4,188 faculty members in English and chemistry at 507 colleges and universities in the United States. We also prepared case studies of eleven institutions and documented the experiences of thirteen faculty parents in English and chemistry for three days each.
Bias Avoidance
Some bias avoidance strategies can improve work performance. Minimizing or reducing family commitments, for example, can free up time and energy for paid work. University of California, Berkeley, researchers Mary Ann Mason and Marc Goulden reported in the November–December 2004 issue of Academe that women who take fast-track academic jobs are more likely to remain single or childless or get divorced than either their male peers or their female counterparts in second-tier jobs. Major life decisions, such as remaining childless, may reflect what we call productive bias avoidance.
When careers are structured so that only workers with few family responsibilities can succeed, there is indeed a bias against caregiving. That bias is gendered in that women face more demands than men on the home front when a spouse or children are present. Combining any family commitments with the extreme demands of a fast-track career makes simultaneous career and family success difficult to achieve.
In our national survey of faculty, 10.2 percent of men, but 16 percent of women, reported that they remained single because they did not have time for a family and a successful academic career. Among parents, 9.1 percent of men and 17.2 percent of women “had one child, but delayed considering another until after tenure.” In total, 12.6 percent of men and 25.6 percent of women had fewer children than they wanted to have “to achieve academic success.”
These examples reflect productive bias avoidance—minimizing family commitments to meet extreme job demands—and they all affect women more than men. One woman acknowledged the burden of these demands in explaining why she delayed childbearing until after tenure: “I could not have done it while the tenure clock was ticking . . . [it] would have just sent me over the edge.”
Although the reasons for productive bias avoidance may be comprehensible, the phenomenon contributes to continued and largely hidden gender inequality in the academy. It is also inefficient to the extent that talented individuals choose not to enter the academy to escape biases against caregiving.
In her 2000 work Unbending Gender, law professor Joan Williams suggests that norms about ideal workers and motherhood shape women’s choices, not simply their work and family responsibilities. Ideal workers—typically men but increasingly professional women as well—should be fully committed to their careers, willing to work long hours each day for years, and even decades, at a stretch. The traditional role of mother casts mothers as selfless and endlessly giving to their children or others in need of care.
To the extent that an academic workplace accepts the norm of the ideal worker and mother, some faculty members will behave in ways that hinder their work and their family performance. We call this mode of behavior unproductive bias avoidance. Faculty parents in two departments, for example, may face similar dilemmas if their children’s day care center closes early on the day of an important faculty meeting. If one parent perceives that the norms for ideal workers and mothers prevail in her department, she might miss the meeting entirely and lie about the reason for doing so, or go and rely on less satisfactory care for her child. If the other parent fears no bias against his caregiving, he might bring the child to the meeting or set up a teleconference from home.
The gendered character of the norm for motherhood means that women who admit to caregiving responsibilities will be penalized more heavily than men. As a focus group participant stated: “I think women have an issue of proving they’re committed. . . . It’s always bizarre to me that I could have gone through four years of college, five years of graduate school, nine years as a postdoctoral fellow . . . and . . . [be] in my sixth year here now working my butt off, and people are wondering about my commitment.”
We expected women to exhibit higher levels of unproductive bias avoidance than men when we initiated our faculty survey. Its results confirmed that faculty members—particularly women—do engage in unproductive bias avoidance. A total of 18.9 percent of men and 32.8 percent of women did not ask for a reduced teaching load when they needed it for family reasons, “because it would lead to adverse career repercussions.” Thirty-three percent of faculty fathers and mothers did not ask for parental leave even though they agreed that it would have been helpful to them, and just under 20 percent of faculty fathers and mothers did not ask to stop the tenure clock for a new child even though they reported that they would have benefited from doing so.
Although only 14.4 percent of fathers came back to work sooner than they would have liked after having a new child because they “wanted to be taken seriously as an academic,” 51.1 percent of faculty mothers did so. Perhaps most troubling because the time involved is minimal compared with the entire span of an academic career, 37 percent of fathers and 46.2 percent of mothers missed some of their young children’s important events, because they “did not want to appear uncommitted” to their jobs.
In total, 23.9 percent of men and 49.6 percent of women reported at least one productive bias avoidance behavior, while 47.6 percent of men and 53.9 percent of women reported at least one unproductive behavior. We believe these figures understate the true extent of bias avoidance in the academy. Because the behaviors are strategic, they are almost limitless in their potential permutations. As a faculty mother stated in a focus group about her university’s policies on family care: “I mean, I don’t discuss this stuff with anybody.” We did not ask about silence as a form of bias avoidance, but we suspect it is widespread and that we have not even conceived of many alternative forms of bias avoidance.
Not surprisingly, we found fewer reports of bias avoidance in departments in which the chair supports the work and life commitments of faculty members. We also discovered that positive affect among faculty members led to fewer incidents of bias avoidance for about half of the behaviors cited above. The fact that individuals with a more upbeat approach to life exhibit bias avoidance behaviors less frequently than their more pessimistic colleagues implies that these behaviors are related to individual psychological differences.
Although we did not obtain data on the sexual orientation or race or ethnicity of our respondents, we did identify single and adoptive parents. Sixteen percent of the mothers and just over 5 percent of the fathers among our parent respondents were single, and just under 8 percent of mothers and fathers had adopted at least one child. When we looked at just mothers, those who were single were more likely to have stayed single to advance their careers, but they were less likely than married mothers to have limited their number of children or to have delayed a second child until after tenure. Single parenthood sometimes occurs because of divorce, but the evidence also suggests that a small group of women make family commitments without men as partners. These women seem less willing to engage in bias avoidance behaviors. The adoptive mothers looked much like the other mothers except that they were almost three times as likely to have stayed single to advance their careers; staying single and rearing children are far from mutually exclusive for some academic women.
Payoff to Avoiding Bias
Do faculty members behave rationally when they use bias avoidance strategies, or are they merely responding to fears that are fundamentally groundless? To answer this question, we looked at a subsample of our faculty respondents who had been tenured since 1990. We asked whether strategies to avoid bias resulted in tenure at a younger age or reduced the time between receipt of the PhD and the year they were tenured.3
For men, reports of unproductive bias avoidance are associated with a reduction in the years between receipt of the PhD and the achievement of tenure. Men who did not use the strategies took an average of 9.7 years, while those who did took only 8.9 years to attain tenure. On average, productive bias avoidance behaviors among women reduced the time to tenure from 8.6 to 7.6 years.
Age at Tenure
Women who deployed the strategies also earned tenure at an earlier age: 39.1, on average. Those who did not engage in productive bias avoidance were 41, on average, when tenured. Unproductive bias avoidance also helps women, reducing the age at which they attained tenure from about 40.9 to 39.9.
Faculty members, particularly women, who believe their institutions will reward bias avoidance behaviors—whether productive or unproductive—are usually correct. Not discussing “this stuff with anybody” pays off.
We as individuals and institutions can and should reduce bias avoidance in the academy. By doing so we could attract more talent and improve gender equity and faculty performance (by limiting unproductive bias avoidance). Through our research, we identified the following overarching strategies that institutions and individuals now use to counter bias avoidance.
Formal Policies
We matched data from our case studies to policy data reported by Beth Sullivan, Carol Hollenshead, and Gilia Smith in the November–December 2004 issue of Academe. We discovered that faculty members often reported lower levels of bias avoidance at institutions with an array of formal policies, including those that permit caregivers to stop the tenure clock, allow new parents to take on modified duties, offer paid maternity and dependent-care leave, and have professional work-life staff.
For example, two otherwise similar schools—large, state- supported but nonflagship universities—diverged on the number of formal policies they offered and on reports of bias avoidance. One institution had twice as many formal policies on the books; at that school, only a minority of men reported bias avoidance, and the rates of parenting were above average. A mother in a focus group said that her third child was born in October and that she took the rest of the semester off, “so it was about an eight week maternity leave . . . at which point they had hired in some colleague to continue my classes . . . and they were very supportive.”
At the other university with fewer policies, most men reported bias avoidance behavior. When asked about leave for new parents, the chemistry chair responded, “I don’t think they would ask for it. I’m not sure they would want it.” A woman in a focus group at the school noted that she was lauded by her colleagues because both of her children were due the week after finals, thereby ensuring that no parental leave would be taken.
Although policy use often remains low, our research supports efforts to expand and strengthen formal work-family policies as a way to reduce bias avoidance. Sullivan, Hollenshead, and Smith provide guidelines for pursuing this strategy, which also coincide with recommendations from the AAUP’s 2001 Statement of Principles on Family Responsibilities and Academic Work.
Cultures of Inclusion
We were surprised to discover among our case studies an institution with low levels of bias avoidance and few work-family policies. This public university has a reputation as an alternative school. It attracts faculty, staff, and students largely because it promotes intellectual and personal diversity within a context of democratic decision making. Diversity is not only welcomed and expected, but also fostered by expectations that the voices of all will be heard and accounted for. We believe the institution’s culture of inclusion explains why only a minority of men and women reported using bias avoidance strategies.
The faculty reported difficulties with high workloads at the institution, as would be expected at many other public universities today. Still, a male focus group participant noted, “When I’m . . . running out of here at 5:40 to go pick up my kid everyday . . . I know these people aren’t looking . . . at me for the fact that I’m running out.” A woman participant said that when her child was an infant, older students who encountered child care problems would bring their children to her classes. They asked why faculty could not do the same. “That . . . gave me some flexibility,” she said. “So I had a portabed in my office, and . . . we were actually together a lot in the first year.”
The logic for this model is intuitively powerful: where students, faculty, and staff with diverse ethnic backgrounds, lifestyle choices, and intellectual paths are made to feel welcome, it makes sense that the institution would also be friendly to faculty members with family commitments.
Bias Resistance
In analyzing our findings, we encountered yet another model for creating family-friendly workplaces: bias resistance strategies. These behaviors explicitly acknowledge one’s own family commitments or provide support for others’ personal lives. Bias resistance was reflected in requests for midday department meetings, inclusion of family members in work-related social functions, and the development of informal support networks to cover work responsibilities when colleagues experienced family crises. Lone individuals working in departments where the ideal worker norm prevailed were particularly courageous. In some departments, however, individual bias resistance inspired others and helped create a climate supportive of faculty commitments to family.
The career choices and subsequent actions of two faculty members from different departments in our study exemplified resistance to the general bias against caregiving in the academy. One woman stated of her prospective departmental colleagues regarding her decision to join their department: “They know they have . . . famil[ies] and they spend time with family, and I appreciated seeing that, because I want to have that kind of balance. I don’t want to neglect my family for five or six years until I get tenure.” She continues to bring her preschooler to receptions for job candidates in the department.
A male chemistry professor told prospective departmental colleagues up front that his family was a priority. “When I was re-cruited,” he said, “it was me and my family. It’s illegal to ask somebody if they’re married and about their family, but I volunteered that information because I wanted . . . them to know who[m] they were hiring. They’re not just hiring a teacher and a researcher; they’re hiring a dad and husband, too. And so, . . . that wasn’t a bone of contention. If anything, that was something that was terrific.”
Evidence from our case studies suggests that bias resistance may be more common among men. This result is somewhat surprising, but makes sense given that men are less likely to be pegged as “just a mom” and therefore penalized for engaging in bias resistance strategies.
Although we discovered the models described above at different institutions, they are not necessarily incompatible. Theoretically, the implementation of formal work-family policies may fit hand in glove with improvements in organizational culture affecting family issues and other forms of diversity. Similarly, individuals pushing the envelope of permissible and expected behavior may help foster policy and culture change, partly by using existing work-family policies. Better policies and institutional respect for family commitments, in tandem with individual resistance, can help to reduce the extent of bias avoidance behaviors.
Today, bias avoidance remains part of everyday life for many faculty members. Where the behavior is productive, it may enhance an individual’s work performance, but it nonetheless contributes to gender inequality in our colleges and universities. Bias avoidance behaviors that are unproductive reduce work performance while sustaining gender inequality. In no case are they fair to faculty members, and they should and can be curtailed, whether through formal work-family policies, cultures of inclusion, or explicit resistance to institutional biases against caregiving.
Notes
1. For details, see Robert Drago, Anne C. Crouter, Mark Wardell, and Billie S. Willits, Final Report: The Faculty and Families Project, 2001. http://lsir.la.psu.edu/workfam/facultyfamilies.htm. Back to text.
2. For most of the results reported here, see Robert Drago and Carol Colbeck, Final Report: The Mapping Project, 2003. http://lsir.la.psu.edu/ workfam/mappingproject.htm. Back to text.
3. The figures reported are averages from the subsample. The sub- sample excludes faculty who had delayed their academic careers for family purposes. Differences discussed are for cases in which the productive or unproductive bias avoidance coefficient was significant at the 5 percent level in linear regressions predicting age tenured or time to tenure. Control variables included the discipline of chemistry, parental status, immigrant status, and the Carnegie classification of the institution. Back to text.
Robert Drago is professor of labor studies and women’s studies at Pennsylvania State University and professorial fellow at the University of Melbourne. Carol Colbeck is associate professor of education and director of the Center for the Study of Higher Education at Pennsylvania State University. The remaining authors were graduate students at Penn State who worked on the study described in this article.
|