Intended for healthcare professionals

Research Christmas 2018: Equal to the Task

Physician mothers’ experience of workplace discrimination: a qualitative analysis

BMJ 2018; 363 doi: https://doi.org/10.1136/bmj.k4926 (Published 12 December 2018) Cite this as: BMJ 2018;363:k4926

Linked editorial

Unto them a child is born

  1. Meghan C Halley, research scientist1 2 *,
  2. Alison S Rustagi, physician3 *,
  3. Jeanette S Torres, strategic director1,
  4. Elizabeth Linos, assistant professor of public policy4,
  5. Victoria Plaut, professor of law and social science5,
  6. Christina Mangurian, professor of psychiatry6,
  7. Esther Choo, associate professor of emergency medicine7,
  8. Eleni Linos, associate professor of dermatology1
  1. 1Program for Clinical Research, Department of Dermatology, University of California, San Francisco, San Francisco, California, USA
  2. 2Palo Alto Medical Foundation Research Institute, Palo Alto, California, USA
  3. *Contributed equally
  4. 3Santa Clara Valley Medical Center, San Jose, California, USA
  5. 4Goldman School of Public Policy, University of California, Berkeley, California, USA
  6. 5University of California, Berkeley School of Law, Berkeley, California, USA
  7. 6Department of Psychiatry, University of California, San Francisco, California, USA
  8. 7Center for Policy and Research in Emergency Medicine, Department of Emergency Medicine, Oregon Health & Science University, Portland, OR, USA
  1. Correspondence to: Eleni Linos eleni.linos{at}ucsf.edu
  • Accepted 20 November 2018

Abstract

Objectives To report woman physicians’ experiences, in their own words, of discrimination based on their role as a mother.

Design Qualitative analysis of physician mothers’ free-text responses to the open question: “We want to hear your story and experience. Please share” included in questions about workplace discrimination. Three analysts iteratively formulated a structured codebook, then applied codes after inter-coder reliability scores indicated high concordance. The relationships among themes and sub-themes were organized into a conceptual model illustrated by exemplary quotes.

Participants Respondents to an anonymous, voluntary online survey about the health and wellbeing of physician mothers posted on a Facebook group, the Physician Moms Group, an online community of US physicians who identify as mothers.

Results We analyzed 947 free-text responses. Participants provide diverse and vivid descriptions of experiences of maternal discrimination. Gendered job expectations, financial inequalities (including lower pay than equally qualified colleagues and more unpaid work), limited opportunities for advancement, lack of support during the pregnancy and postpartum period, and challenging work-life balance are some of the key themes identified. In addition, participants’ quotes show several potential structural drivers of maternal discrimination and describe the downstream consequences of maternal discrimination on the physician herself, her career, family, and the healthcare system.

Conclusions These findings provide a view of maternal discrimination directly from the perspective of those who experience it. Women physicians report a range of previously uncharacterized ways in which they experience maternal discrimination. While certain aspects of these experiences are consistent with those reported by women across other professions, there are unique aspects of medical training and the medical profession that perpetuate maternal discrimination.

Introduction

The proportion of medical school graduates who are women has increased from 7% in 1965 to 53% in 2017.123 However, there is ample evidence of persistent gender discrimination against women physicians, from disrespectful treatment to persistent pay discrepancies that are evident among early and mid-career academic physicians as well as those in private practice.45678 When directly asked, 70-77% of women physicians report experiencing gender discrimination,910 and 30% of academic women physicians report having experienced sexual harassment.9 Discrimination may also explain at least part of the gender gap in leadership roles, a phenomenon particularly well documented in academic medicine: while women comprise 46% of US residents (doctors in training), they comprise only 38% of faculty (academic medicine) positions, 21% of full professors, and 15% of department chairs.1112

In addition women physicians, an estimated 80% of whom are or will become mothers, also report experiencing “maternal discrimination” based specifically on their motherhood status.13 However, there are limited data on the experience of maternal discrimination or its sequelae on the careers of women in medicine. The primary goals of this study were to (a) characterize the ways in which physician mothers experience maternal discrimination; (b) examine cultural and structural characteristics of the healthcare system that may facilitate maternal discrimination; (c) describe a range of outcomes of maternal discrimination; and (d) identify organizational targets that have the potential to minimize discrimination against physician mothers.

Methods

Data collection

Details of study methods have been previously published.10 Briefly, data were collected prospectively through a voluntary, anonymous, opt-in, online survey with informed consent. The survey was developed by the research team on secure REDcap survey software. The link to the survey was posted on a Facebook group: the Physician Moms Group, a diverse online community of physicians who identify as mothers. At the time of the survey in June 2016, there were approximately 60 000 members, of whom 34 956 were active members visiting the page at least once per month. In order to participate in the survey, participants had to click on a link that directed them to a cover page with a consent form that explained the goals of the study, and that it was voluntary and anonymous. The survey was posted three times starting June 17, 2016, and the data were collected through the end of August 2016. A total of 5782 physician mothers completed the full survey, with a participation rate of 36.0% based on the number of individuals who saw the post (estimated to be 16 059 unique views), or a participation rate of 16.5% based on the total number of active members (34 956), which is similar to past survey participation rates in this group.14 The vast majority of participants completed the full survey (99.2%).

The full survey included questions about demographics, physical and reproductive health, perceived discrimination, potential workplace changes, and burnout. All questions were optional. Regarding discrimination, participants were asked: “Have you ever experienced any of the following forms of discrimination at your workplace? (Please select all that apply),” followed by the open ended statement: “We want to hear your story and experience. Please share.” Comments were provided by 1009 physicians and varied in length from a short phrase to multiple paragraphs.

Ethical approval for this study was obtained from the University of California, San Francisco’s institutional review board.

Analysis

Participant comments were analyzed using Dedoose, an online program for analyzing qualitative and mixed methods data (www.dedoose.com). A team of three trained qualitative researchers (MCH, ASR, JST) drew on the techniques of grounded theory in analyzing the data. This approach is designed to identify emergent social and structural patterns in human experience and to generate a theory or theories to explain these patterns.1516 First, the team read the data in their entirety and developed a structured codebook that included inductively derived codes that represented common and/or unique ideas that emerged from the data (such as missed opportunities, job changes) as well as deductively selected categories to organize the dataset. The latter included codes to identify any comments that were unrelated to either gender or maternal discrimination (n=62) for exclusion, leaving 947 excerpts in the final analysis.

The analytic team conducted multiple rounds of codebook revision and interrater reliability testing until a minimum pooled kappa value of greater than 0.8 was achieved by all coders to ensure the clarity of coding definitions—including which portions of a comment related directly to maternal discrimination—and their consistent application. The analytic team then reviewed and consolidated these memos into a list of key themes and subthemes designed to characterize the range of experiences of maternal discrimination as well as the processes that led to and resulted from these experiences. Only ideas relevant to maternal based (not only gender based) discrimination were included in the conceptual model. We include instances of discrimination that are tightly related to gender stereotypes (for example, that women should be warm and nurturing), which are based on traditional social roles related to motherhood. We used an iterative process to refine and organize these themes into a conceptual model.16

To evaluate the representativeness of our sample, we calculated median age, race and ethnicity, marital status, number of children, geographic location, practice type, and medical specialty of our study group and compared these demographics with women physicians reported by the 2016 Association of American Medical Colleges (AAMC) survey.17

Results

Demographic data for the full sample have been reported previously.10 The median age was 39 years (range 24-62), median number of children was two (range 0-6, where 0 children included mothers who were currently pregnant or had lost children). Overall, respondents were similar to the general population of US physicians: 74.2% White (US 75%), 12.9% Asian (US 12.8%), 8.4% Hispanic or Latina (US 5.5%), and 5.3% Black (US 6.3%). Those who responded to the discrimination prompt were similar to non-respondents in terms of demographics (age, race/ethnicity, geographic location), and career factors (training stage, medical specialty, years in practice). Most participants were residency-trained practicing physicians (90%) and most worked >40 hours per week (43% worked 41-60 hours per week, 17% worked >61 hours per week).

The themes and relationships from the 947 participant responses summarized in the conceptual model (fig 1) are organized into five broad categories related to maternal discrimination: experiences, drivers, interpersonal mediators, impacts, and modifying factors (tables 1-5).

Fig 1
Fig 1

Conceptual model of themes and relationships derived from physician mothers’ reported experiences of discrimination

Experiences of maternal discrimination

This category captured the diverse ways in which physician mothers experience discrimination due to pregnancy, maternity leave, breast feeding, and subsequent parental responsibilities (table 1). Many of these reported experiences of discrimination are consistent with research documenting the application of different standards for pregnant women, and the implications of these perceptions for hiring and promotion-related evaluation.18192021

Table 1

Experiences of maternal discrimination as described by self identifying physician mothers among 947 qualitative excerpts, by theme and subtheme

View this table:

Gendered performance expectations—The first theme described higher standards applied to physician mothers than their colleagues, which physician mothers described as a requirement to prove their commitment and competence as a physician. Conversely, other participants described being held to a lower standard, in that their colleagues assumed that they were not interested or able to take on new tasks or leadership roles due to their maternal responsibilities, and, therefore, they were never offered.

Limited opportunities for advancement—The second theme included participants’ descriptions of being excluded from administrative decision making because of pregnancy or after returning from maternity leave. Participants also described being passed over for leadership roles in favor of colleagues perceived as less qualified or having their contracts grossly modified or terminated in response to announcing a pregnancy or when returning from maternity leave.

Financial inequalities—This was based on participants’ experiences of receiving lower salaries than colleagues with comparable (or significantly less) experience and productivity. In addition, women reported being asked or required to take on more work (such as serve on committees) without additional pay, while male colleagues were more typically compensated for additional work. This also was an issue following maternity leave, after which women reported being required to “make up” the time without pay or benefits, despite the fact that their leave time was already unpaid.

Lack of support during pregnancy and the postpartum period—This theme captured participants’ descriptions of the disparagement of maternity leave and motherhood in their workplaces, including colleagues referring to maternity leave as “vacation,” being pressured to forgo leave to which they were entitled, and/or being subjected to rules and expectations regarding leave that were not applied to male colleagues. Participants described feeling “punished” for their pregnancy, maternity leave, and/or pumping breast milk through personal mistreatment or unreasonable expectations.

Challenging work-life balance—This incorporated comments about challenges finding childcare for unusual or extreme work hours, the lack of alternative work schedules to support work-life balance, and balancing home responsibilities with growing demands on physicians to complete administrative tasks after clinic hours.

Drivers of maternal discrimination

This category was developed to capture the underlying causes of maternal discrimination as perceived by participants in three themes (table 2).

Table 2

Drivers of maternal discrimination as described by self identifying physician mothers among 947 qualitative excerpts, by theme and subtheme

View this table:

Broader cultural norms included participants’ descriptions of how their experiences working in medicine reflected gender norms in society, such as the expectations of women’s interpersonal interactions and social, financial, family, and workplace roles, consistent with well documented gender norms and expectations in feminist sociology, anthropology, and social psychology.22232425

Culture of medicine aggregated participants’ descriptions of the norms, values, and expectations underlying individuals’ social roles and interpersonal interactions specifically within the medical workplace. The culture of medicine was evidenced by several assumptions or expectations: that mothers cannot be successful doctors, that doctors cannot be good mothers, that women physicians should delay childbearing, that childbearing necessarily ruins women physicians’ careers, and that physician mothers do not value their careers.

Structure of medicine was based on participants’ perceptions of ways in which the administrative, procedural, and physical structures of the medical workplace perpetuated maternal discrimination, including policies and procedures that limit maternity leave, the lack of coverage and flexibility in physician schedules, the lack of physical space and time to breastfeed or pump milk, the long (and often overnight) work hours generally required of physicians, and incentive structures that financially penalize mothers for taking maternity leave or pumping breaks.

Interpersonal mediators of maternal discrimination

This category captures the avenues by which the drivers of discrimination manifest in interactions in the workplace (table 3). This is consistent with longstanding sociological theory that interpersonal interactions mediate the relationship between individual experience and culture norms and expectations of gender, and with more recent research on the role of implicit bias in perpetuating discrimination.4 This category is organized by the core groups that physician mothers interact with in the workplace, including physician colleagues, administrators, support staff (such as nurses), and patients and their families. These examples demonstrate the personal and sometimes blatant ways in which discrimination is experienced by physician mothers.

Table 3

Interpersonal mediators of maternal discrimination as described by self identifying physician mothers among 947 qualitative excerpts, by group of individuals that physicians interact with in the workplace

View this table:

Impacts of maternal discrimination

This category includes themes focused on the downstream effects of maternal discrimination on the women, their families, and the healthcare system (table 4).

Table 4

Impacts of maternal discrimination as described by self identifying physician mothers among 947 qualitative excerpts, by theme and subtheme

View this table:

Psychological, captures participants’ descriptions of the psychological consequences of maternal discrimination, including extreme stress due to work and family demands, guilt from feeling unable to meet those demands, and fear of reporting discrimination due to possible retaliation.

Career focuses on the negative impacts of maternal discrimination on careers. A common example cited by participants was switching from full time to part time work, a “choice” that was not always described as voluntary. Participants also described leaving academic or clinical medicine, leaving group practice for solo private practice, or leaving medicine altogether in response to an unfriendly work environment characterized by maternal discrimination. Many women’s career changes also came with immediate financial consequences, including lower salaries and fewer benefits. Additional financial consequences included being denied earned salary increases or bonuses due to maternity leave (despite reaching or exceeding productivity goals), being denied higher paid positions due to motherhood, paying full time overhead costs despite holding a part time position, and forgoing pay to recover from birth or take pumping breaks.

Family outlines the impact of maternal discrimination on the families and children of physician mothers. The negative financial consequences described above also undermined participants’ family’s financial stability, particularly for mothers who were the primary breadwinner in their family. In addition, women were expected to perform tasks while pregnant that presented risks to the fetus (such as exposure to communicable diseases known to cause fetal defects, continuing work while in preterm labor or while having a miscarriage) and many described prematurely curtailing breastfeeding or pumping due to lack of time, space, and support. Participants also described generally feeling that they were expected to ignore the advice regarding child care that they were giving to their own patients due to inflexibility in their workplace.

Healthcare system—Under this theme, participants described negative impacts on the healthcare system and their patients due to maternal and gender discrimination. This included potential impacts on healthcare quality due to their orders being ignored (whereas those of male colleagues were not), putting patient safety at risk. In addition, participants described how their part time status, which many found to be the only option for work-life balance, meant they were automatically excluded from leadership and other opportunities to be involved in decision making. Participants also described being passed over for these leadership positions in favor of less qualified colleagues.

Modifying factors

Finally, table 5 captures the modifying factors that participants’ pointed to as either perpetuating or mitigating maternal discrimination.

Table 5

Modifying factors regarding maternal discrimination as described by self identifying physician mothers among 947 qualitative excerpts, by theme and subtheme

View this table:

Discussion

Our findings suggest insidious, persistent, and sometimes blatant discrimination experienced by physicians, based on their status as mothers. While these experiences may be mitigated or exacerbated by stage of training, institution, and choice of specialty, participants’ reports suggest these experiences are pervasive. Our conceptual model illustrates the role of broader cultural norms and expectations of mothers, as well as both the culture and structure of the medical workplace in perpetuating maternal discrimination. In addition, our data suggest multiple ways in which maternal discrimination has short and long term consequences, not only on the individual physician, but also on her family and the larger healthcare system.

These findings extend research on gender discrimination in the medical workplace by focusing on motherhood and, notably, provide a view of maternal discrimination directly from the perspective of those who experience it. These reported experiences are consistent with a wide range of sociological and psychological studies documenting the stereotyping of mothers and the implications for workplace outcomes.2025 These include gendered expectations of women’s interpersonal interactions and gendered assumptions about workplace ambition and professional capabilities.

Our findings also contribute to a broader debate on the causal explanations of the “motherhood penalty”—the empirical finding that mothers face an additional wage penalty (over and above any gender wage penalty) even as men enjoy a “fatherhood bonus.”2627 Although prior research has shown that the motherhood penalty is a problem among women from low socioeconomic status, our data suggest that this problem also applies to women in higher paying fields such as medicine.28

Other aspects of physician mothers’ experiences are specific to the environment of healthcare and institutional policies and norms around physician training, suggesting ways in which maternal discrimination may manifest for physicians in particular. The intensive and extended time course of medical training provides a particular conundrum for women physicians, as their training is rarely complete until they are already nearing the end of their childbearing years.

Some evidence suggests that gender-specific “work preferences” and “life values” may explain some of the gender differences found among parents in the workplace.29 Indeed, when looking at national data, many speculate that women choosing “motherhood-friendly” occupations may account for the persistent wage gap, and find a much stronger motherhood penalty among low income and low wage workers.28 These data underscore that “work preferences” and “life values” may themselves be shaped by the limitations of the specific work environment mothers face, including physician mothers, and what seems to be a career “choice” may actually stem from the internalized preclusion of other options and interests.192930 In fact, exposure to gender discrimination and sexual harassment may influence medical students' choice of specialty.31 Unfortunately, when motherhood itself is viewed as a choice, mothers’ incomes are penalized more.32

Certain aspects of the structure of medicine and medical training are unique, deeply ingrained, and may inadvertently discriminate against women who become mothers. The limitations of this work environment perpetuates a cycle in which women, and mothers in particular, are excluded from leadership positions in which they might otherwise be able to drive incremental change by advocating for more family-friendly policies in their institutions. Identification of specific structural biases that contribute to maternal discrimination—such as institutional policies that exclude part time physicians from leadership roles—is essential to identifying avenues for intervention. In addition, our findings indicate that maternal discrimination varies by institution, suggesting that solutions exist. Further research examining “positive deviants”—that is, institutions where physician mothers report better work-life balance and support—could provide models for combating discrimination elsewhere. Further, we see men as allies who can help make long lasting change by advocating for family-friendly policies.

It is noteworthy that some of the downstream effects of maternal discrimination might be mitigated by policies and procedures: flexible schedules including shift work, on-site high quality child care, longer paid parental leave, etc. A recent analysis of parental leave policies of top US medical schools showed high variability in the duration of paid leave, and that duration of leave was often negotiable and at the discretion of supervisors.33

However, there is promising evidence that efforts to increase awareness of family accommodation policies (including presentations, workshops, and web or social media presence) improves uptake of those policies.34 The pervasive misconceptions and negative cultural norms around maternity leave documented in this study strengthen the argument that decisions about maternity leave cannot be left to the discretion of individual employers if they are to be implemented consistently. Because generous family-friendly policies may unintentionally exacerbate discrimination against women in the workplace, research is needed to understand how to most effectively create structural changes to the work environment given the underlying cultural biases against mothers.3536

Although our participants were mostly from the United States, the problems highlighted are not unique to the US: a cross national comparison of 22 countries found that wage penalties associated with motherhood were smaller in countries with public policies that maintain mothers’ labor market attachment (for example, moderate length job-protected leaves, public funded childcare, tax rate policies that tax second-earners at a lower rate, and paternity leave). The highest motherhood penalties were found in countries with short leave policies, but the penalty was also high in countries with long leave policies (perhaps because they reinforce maternal stereotypes and weaken labor force attachment).37

Strengths and limitations

This study is the largest qualitative study to examine physicians’ experiences of maternal discrimination to date, and one of the largest qualitative studies on maternal discrimination in any profession, providing an extremely rich dataset. However, the findings of this exploratory analysis should be viewed as preliminary given the limitations of the methodology. The dataset is limited to mothers, and we therefore did not have the opportunity to study women physicians who are not mothers, or fathers, and penalties they face. Entirely disentangling the effects of gender versus motherhood on discrimination was not possible with these data. We also did not have the perspective of employers or cultural context for these motivations. For instance, administrators may have to make difficult choices regarding individual employees based on broader organizational constraints. Additionally, the relatively low participation rate could bias responses, and we recognize that this is highly selected voluntary sample of participants. Women who experienced discrimination may have been more likely to respond, overestimating prevalence of discrimination. Conversely, the prevalence of discrimination could have been underestimated if women who left the profession due to the negative impacts of discrimination were not included in our sample of mostly practicing physicians. We believe response bias is likely not a major concern because our survey was not limited to questions about discrimination, and was a broader survey about health and wellbeing of participants. In addition, research suggests few differences between responders and non-responders in physician surveys.38 Another limitation is that the relative lack of racial diversity in the sample did not permit us to examine how maternal discrimination and racial discrimination may intersect.

Conclusion and next steps

These preliminary findings suggest insidious, persistent, and sometimes blatant manifestations of discrimination experienced by physician mothers. The cyclical nature of maternal discrimination in medicine is concerning because it reinforces and perpetuates key drivers of maternal discrimination in medical training and practice. As we strive to build more equitable workplaces, our findings suggest that challenging norms around motherhood in the medical workplace, as well as structural changes that address pregnancy, parental leave, and childcare, are needed in order to mitigate the impacts of maternal discrimination in medicine.

What is already known on this topic

  • Over two thirds of women physicians report experiencing gender discrimination and a third of academic women physicians report having experienced sexual harassment

  • Discrimination may also explain part of the gender gap in leadership roles, and gender pay discrepancy in medicine

What this study adds

  • This large qualitative study of discrimination among physician mothers reveals frequent, persistent, and sometimes blatant discrimination experienced by physician mothers based on their status as mothers

  • Some of the experiences of maternal discrimination are consistent with those reported by women across professions, but there are unique aspects of medical training and the medical profession that exacerbate maternal discrimination

  • Structural changes that address pregnancy, parental leave, and childcare are needed to mitigate the impacts of maternal discrimination in medicine

Acknowledgments

MCH and ASR contributed equally to this work.

We thank Lily Morrison for technical support with this manuscript, and Hala Sabry and Dina Seif for their help with recruitment.

Footnotes

  • Contributors: MCH and Eleni L conceived of the study and analysis plan. MCH, JST, and ASR analyzed the data. MCH drafted the conceptual model. MCH and ASR wrote the first draft of the manuscript. All authors contributed to interpretation of the data, edited the manuscript, and approved of the final manuscript. Eleni L had full access to all the data in the study and had final responsibility for the decision to submit for publication.

  • Funding: Eleni L is supported by the National Cancer Institute (grant No R21CA212201, the National Institute of Aging (grant K76AGO54631), and the National Institute of Health (grant DP2CA225433); VP is supported by NSF #1535435; and CM is supported by grants K23MH093689, R01MH112420, and R03DK101857.

  • Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf and declare: no support from any organization for the submitted work.

  • Ethical approval: Obtained from University of California, San Francisco’s institutional review board.

  • Data sharing: No additional data are available.

  • The lead author affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

References

View Abstract