Wednesday, September 20, 2017

Discussing Diversity: How Emphasizing and Minimizing Intergroup Differences Affect Bias and Empowerment with Ashley Martin

Over the last several years, there has been an increased push to talk about diversity, under the assumption that being blind to difference is counterproductive and that to leverage the benefits of diversity, we have to emphasize it. But is it really a good idea to talk about difference? This week’s WAPPP seminar featured Ashley Martin, PhD Candidate at Columbia Business School, as she presented a series of studies about awareness of difference and effects on inclusion and empowerment.

There is a great deal of literature about awareness of difference with respect to race: talking about difference reduces bias and increases engagement. However, there is almost no research on this awareness strategy for gender. The single study that had been published showed that being aware of or emphasizing gender differences was related to benevolent sexism. As a field, researchers have not systematically explored the types of differences attention to diversity emphasizes, whether they are similar for race and gender, and how this strategy affects outcomes for each group.

Effects of awareness on attributions of difference

In the first study, the researchers asked 143 participants to list ten differences either between men and women or between black and white people. When asked to categorize these differences, it became clear that the listed racial differences were far more in the realm of opportunity and culture, and the gender differences were more related to personality and biology. The results of this initial study demonstrate that the types of differences that people generate when asked to consider race and gender are difference – for race, people focus on external differences related to opportunity, but for gender, people focus on internal differences.

The second study served to measure individuals’ beliefs about how one should approach difference. Study participants were rated on an awareness scale (how often they agreed with statements like “there are differences between groups that should be acknowledged” versus “there is no reason to categorize people based on their membership in a certain group”) and how often they attributed difference to external versus internal factors (opportunity/culture or personality/biology). Study participants who said that we should celebrate racial differences were more likely to attribute difference to opportunity; by contrast, those who said we should celebrate gender differences were more likely to attribute difference to biology. Recognizing external differences with respect to race is a positive development, as recognizing inequality is the first step in disrupting it. However, attributing gender differences to personality and biology has consequences for women.

In particular, the third study revealed that “personality differences” really boil down to stereotypes about gender and leadership; that men are agentic and assertive, whereas women are communal and warm. By telling people to recognize and embrace differences, we may be inadvertently telling them to embrace stereotypes that limit women’s opportunities. In this experiment, subjects were divided into three groups: the first read an article on why it is good to be aware of differences, the second an article on why it is good to be blind to differences, and the third were simply asked to think about diversity.

In the second stage of the study, participants were asked to rate whether differences between groups were due to opportunity or biology, to what extent a list of character traits regarding assertiveness versus communality were associated with each group, and a measure of “denial of inequality” – whether participants thought that discrimination was not a problem, and that on average all groups were treated equally.

On the race side, participants were more likely to attribute difference to opportunity rather than biology. However, those who reflected on the importance of acknowledging difference in the first stage were more likely to attribute differences to opportunity, while those who read the article deemphasizing difference made fewer opportunity attributions. On the gender side, participants endorsed biological reasons for difference in both the control and the “aware” condition. Only the group who read the article about being blind to difference attributed gender differences more to opportunity than to biology. Further analysis of these findings indicated that there was little connection between race and stereotyping. By contrast, all participants associated men with agency much more than women, though in the blind condition, this assessment decreased somewhat.
Overall, this study reveals that being blind to difference is the baseline for racial differences, and that awareness pushes people to consider the antecedents of difference more closely. For gender, awareness of difference doesn’t change results from the status quo, but being blind to difference can improve assessments of women.

Effects on bias

The next study was designed to test the effects of awareness ideology on bias. Two groups of male study participants were asked to read either the “aware of difference” article or the “blind to difference” article, and then were given a woman’s resume and asked to rate her leadership potential and assess whether they would hire her. In the awareness condition, participants rated the applicant lower on leadership potential than in the blindness condition, and were less likely to hire her. The decision to hire is mediated by the individual’s leadership evaluation, which is itself affected by stereotypes regarding agency and what makes a good leader that are emphasized through awareness of gender difference.

Effects on women’s confidence

How do these assessments affect women at work? In the next study, 163 women were asked to make lists of differences or similarities between genders, and were then asked to what extent these qualities undermined their ability to be seen as a leader. When women listed differences, they perceived these differences as having a major negative effect on their leadership potential and listed more words related to agency and assertiveness than when they listed similarities.
In the next phase of the study, 115 women read either the article about the benefits of being aware of difference, the benefits of being blind to difference, or a control article that was completely devoid of any gendered content. Next, the participants took a workplace confidence scale: in the control condition and in the awareness condition, women rated their workplace confidence above the midpoint. However, in the blindness condition, women rated themselves as significantly more competent.

Awareness of difference not only affects confidence, but also action-taking. The next study featured 126 female managers, who read either the aware or blind to differences article, then completed a self-assessment of their agency, the workplace confidence scale, and then participated in several risky decision tasks. In the awareness condition, the participants identified less with agency, rated themselves lower on workplace confidence, and took less action than in the blindness condition.

Dyadic effects

In the final study, the researchers examined how men’s awareness beliefs affected male-female interactions. Study participants were divided into male-female pairs. In the first part of the study, the men read either the awareness or blindness article. Next, the pairs were asked to complete the bushfire survival task, requiring them to discuss and rank a list of survival items in order of importance. Independent raters were asked to evaluate the amount of time men versus women spent talking, the level of openness in the conversation, and the overall quality of the interaction. In the awareness condition, men spoke more than women and were less open. In the blindness condition, women spoke significantly more than men, men and women were equally open, and the interactions were overall higher in quality. Indeed, it seemed that men were “leaning out” so that women could lean in.

The results of these studies indicate that awareness of difference can be beneficial in attributing race to external differences, but that deemphasizing gender difference leads to better outcomes for women. This research demonstrates that diversity strategies shouldn’t be monolithic, as they may not work for all groups. Perhaps instead of discussing differences between men and women, discussing similarities may be an effective intervention to promote gender equality.

Wednesday, September 13, 2017

Should She Lead? Why Colleagues of Women, and Women Themselves, Often Think the Answer is No with Alison Wood Brooks

Should women lead? We here at WAPPP strongly believe yes, but many people, including women themselves, often think the answer is no. This year’s first WAPPP seminar featured Alison Wood Brooks, Assistant Professor of Business Administration at Harvard Business School. Professor Brooks presented results from two published papers regarding perceptions of women in leadership roles and women’s experiences of those roles, as well as two earlier-stage projects.

We know women are underrepresented in leadership positions: less than 15% of executive officers are women, less than 11% of US firms with venture capital backing have been founded or led by women, and less than 5% of Fortune 500 CEOs are women. To understand why, Professor Brooks examines both demand-side factors (how people perceive and react to women in the workplace) and supply-side factors (how women themselves think and behave in the workplace, including individual differences in personality and preferences).

The Demand Side: How Gender and Attractiveness Affect Entrepreneurial Pitch Success

On the demand side, Professor Brooks presented a study demonstrating that investors prefer entrepreneurial ventures pitched by attractive men. An analysis of the pitches given at the three largest entrepreneurial pitch competitions across the US over three years revealed that male entrepreneurs were 60% more likely than female entrepreneurs to win the competition and earn investor funding. In their study, Professor Brooks and her colleagues asked 521 people to watch two entrepreneurial pitch videos and choose one to invest in, switching whether the video was narrated by a male or a female voice. Sixty-eight percent of participants chose a venture pitched by a male voice, at a very high level of statistical significance, controlling for the pitch itself, the order the pitches were presented, participant gender, and participant age.

However, field data reveals that this effect is not just about gender: attractive male entrepreneurs were statistically significantly more likely to win pitch competitions than less-attractive male entrepreneurs. Interestingly, in the field data, there was no statistically significant impact of attractiveness on female entrepreneurial success. In the next stage of the study, to understand whether attractive people might just be better at pitching, the researchers asked 194 people to watch one pitch video and rate it. The researchers varied whether participants heard a male or female voice, and whether the photo of the narrator they were shown was more or less attractive. Participants were asked how likely they were to invest in the venture and to what extent the pitch was persuasive, fact-based, and logical. Again, attractive male entrepreneurs were significantly more likely to secure investment than less-attractive men, while there was no statistically significant difference for women. However, male-narrated pitches were rated as more persuasive, fact-based, and logical. These findings indicate that male entrepreneurs – especially attractive ones—are more likely to receive investor funding than are female entrepreneurs based on their gender and physical attractiveness rather than the content of their ideas.

Extending the Analysis: Data from Kickstarter

Professor Brooks embarked on a new project to replicate and extend this study using big data from Kickstarter.com, a large crowdfunding platform. Professor Brooks and her colleagues took a sample of 1,249 projects from Kickstarter that included a picture of a single entrepreneur whose gender could easily be identified. Ten independent judges rated the age and attractiveness of the entrepreneurs in every photo. The researchers used the number of investors as a measure of pitch success and included other available information (including number of days left in the campaign) as control variables. Analysis of the Kickstarter data replicates Professor Brooks’ earlier findings. Male entrepreneurs had about twice as many investors compared to female investors, and for male investors (but not female investors) attractiveness garnered more investors.

Professor Brooks also included two follow-up experiments. The first was designed to test the relationship between entrepreneur gender, entrepreneur attractiveness, and investor gender. The results reveal that male investors are more likely to fund good-looking male and female entrepreneurs, but that female investors are more likely to fund attractive male entrepreneurs and less likely to fund attractive female entrepreneurs. The second experiment was to test to what extend these results are about threat. Are women threatened by attractive women? Are men threatened by successful men? The data bears out this initial hypothesis: women are less likely than men to invest in ventures run by attractive women, but are more likely than men to invest in ventures run by successful men.

The Supply Side: How Do Women Themselves View Professional Advancement?

On the supply side, Professor Brooks presented a study indicating that compared to men, women view professional advancement as equally attainable, but less desirable. When asked to list their core goals in life, female participants listed more goals than men overall, but a smaller proportion of women’s goals were related to achieving power at work. These findings hold when people make their own lists of goals, when they select goals from a pre-determined list, and when they reject goals from a pre-determined list. When asked to what extent they expect certain positive or negative outcomes resulting from a promotion to a higher-power position at work, men and women report equivalent levels of positive reactions to the promotion, but women report stronger negative reactions to the promotion. Men and women rate the promotion as equally attainable, but women rate the promotion as less desirable and, compared to men, are less likely to accept the position. These results are consistent across high-level executives, mid-level managers, recent MBA alumni, and undergraduate students.

Importantly, these findings are descriptive, not prescriptive. It may be that men and women are correctly predicting the unique experiences they will each face in high-power positions; alternatively, women may be overestimating the negative consequences associated with power, men may be underestimating the negative consequences associated with power, or both.

New Areas for Research: Gender, Power, Laughter, and Happiness

In a new project, Professor Brooks is working to investigate the relationships between gender, power, laughter, and happiness. Two sources of data provide converging evidence that men and women might use laughter differently: the U.S. Gallup Poll 2013 and speed-dating conversations. In the Gallup poll, women report laughing and smiling more than men in their daily lives, but also report experiencing more negative emotions (worry, sadness, anger, stress). In the speed dating context, women are more likely to laugh during their dates, regardless of how funny they find their partner. However, female laughter is a worse indicator of interest in a second date compared to male laughter. Said differently, women laugh more often, irrespective of humor, but not in a way that is necessarily connected to interest. As there has been little research done on the supply-side dimension of women’s unique experiences, particularly in leadership roles, this is an area ripe for exploration.

For more on Professor Brooks’ research, check out our Gender Action Portal:

http://gap.hks.harvard.edu/investors-prefer-entrepreneurial-ventures-pitched-attractive-men

http://gap.hks.harvard.edu/compared-men-women-view-professional-advancement-equally-attainable-less-desirable

Thursday, May 4, 2017

Self-Affirmation as Strategy for Reducing Gender Effects on Negotiation with Chiara Trombini

Women face serious obstacles on the road to success. Despite reporting the same career aspirations and ambitions, women are underrepresented at the top and tend to be less satisfied with their careers than men. Women make up 44% of S&P 500 companies, but only 25% of corporate boards and only 6% of CEOs. Gender discrimination accounts for a large part of this discrepancy: certain attributes that are considered essential for career advancement are stereotypically male, and professional women face backlash when they display these traits.

How do men react to the prospect of working with agentic women in male-dominated environments, and how can we make men more willing to work with agentic women? The final WAPPP seminar of the year figured Chiara Trombini, AY’17 WAPPP Fellow, as she presented three studies on this line of research.

Why focus on men evaluating agentic women in male-dominated environments? Gender differences are more likely to emerge in male-dominated environments, and women are evaluated more harshly when they occupy male-dominated roles, exhibit stereotypically male attributes, or are evaluated by men. In particular, women face a competence-likability double bind: either they are well-liked but considered incompetent, or are competent but socially unattractive and therefore less worthy of hiring or promotion.

Is there a way to overcome the “threat” posed by agentic women? Chiara argues that self-affirmation is one way to reduce backlash and promote gender equality. Self-affirmation is when individuals reflect on values that are personally relevant to them. Self-affirming individuals are less likely to experience distress, less likely to react defensively, and are more likely to be objective in their decisions and less likely to rely on stereotypes. From a cognitive perspective, self-affirmation is effective at reducing the effects of prejudice and stereotypes, makes individuals more likely to accept threatening information and modify their behavior in response to threat, and increases concession-making and openness to compromise in negotiations.

How can self-affirmation reduce backlash? The mechanism lies in men’s emotional response to agentic women. Men dominate the current gender system and are sensitive and responsive to threats to their masculinity. When gender status is uncertain or challenged, men experience anxiety and react aggressively. There is some evidence that self-affirmation reduces cortisol levels and may be able to reduce stress and anxiety responses in these situations.

Chiara hypothesized that self-affirmation would make individuals less likely to rely on stereotypes and prejudices, so men who practiced self-affirmation would be more willing to work with agentic women and would feel less anxious at the prospect of doing so. She presented three studies testing these hypotheses on self-affirmation and gendered backlash.

Study 1

In the first study, evaluators were asked to rank 11 values (sense of humor, politics, religion, creativity, etc.) in order of personal importance. In the self-affirmation condition, evaluators were asked to write why their most important value was important to them. In the control condition, evaluators were asked to write why their least important value could be important to someone else. Then, evaluators watched a job interview and evaluated an internal candidate for job placement. The measured variable was willingness to work with the candidate. As hypothesized, self-affirmation increased male evaluators’ willingness to work with agentic women.

Study 2

Study 2 followed the same design as Study 1, but included measures for negative trait perception (whether the evaluator found the candidate to be arrogant, greedy, etc.), level of anxiety, and neutral feelings. The findings from Study 2 replicated Study 1: evaluators were less willing to work with women than men overall, but self-affirmation increased general willingness to work with candidates and particularly increased men’s willingness to work with women. In addition, men in the self-affirmation condition had lower rates of negative trait perception and lower anxiety.

Study 3

Study 3 set out to evaluate whether anxiety decreases willingness to work with a job candidate. Participants were asked to evaluate a job candidate based on their behavior in a job interview after being exposed to a music clip designed to induce feelings of anxiety or feelings of calm – in this case, either the theme from Psycho or “Weightless” by Maroni Union. Evaluators then rated their current mood in terms of anxiety or calmness. In general, male evaluators were more willing to work with male candidates, and female evaluators were more willing to work with female candidates. Participants in the high anxiety condition were less willing to work with candidate across the board. In addition, male evaluators in the high anxiety condition were less willing to work with agentic women.

These studies lend a cognitive and affective perspective to examining the roots of gender discrimination in hiring. It appears that perception of negative traits and feelings of anxiety are wrapped up in hiring discrimination. Self-affirmation is a low-cost, powerful affirmation that can increase men’s willingness to work with women in male-dominated environments, which could ultimately reduce gender gaps in professional settings. Organizations could institute self-affirmation practices before hiring, promotion, and performance review decisions in order to alleviate gender discrimination and ensure that they are benefitting from the entirety of the hiring pool.

Tuesday, April 25, 2017

Fathers and Work Family Balance: Mix Methods for Understanding Fatherhood Involvement and Enrichment Experiences with Marc Grau-Grau

Fatherhood is in transition: the traditional “distant breadwinner” model of fatherhood has eroded to a more nurturing role for fathers. Fathers are doing more care work, and work-family balance is increasingly a priority for men. At the same time, work-life conflict is increasing, leading to high levels of stress, absenteeism, and burnout, along with a lower fertility rate. In Europe, 21% of workers report that they “always” or “almost always” feel too tired when they get home from work to spend time with their families. However, few studies focus on understanding men and work-family balance.

Much of the research on work-family conflict has focused on mothers. We know that involved fathers are good for kids and for gender equality, but less about how these dual roles affect fathers and their work and family lives. This week’s WAPPP seminar featured Marc Grau-Grau, WAPPP Fellow 2016-2017. Marc presented three studies from his thesis on Catalonian fathers to better understand how patterns of paternal involvement may give some clues on how to promote gender equality in parenting.

Study 1: Understanding the predictors of fatherhood involvement 

Fatherhood involvement is positive for children and gender equality, but also for fathers and their jobs. The aim of the first study was to explore the care work contributions of Catalan working fathers with children under 10 to understand how patterns of fatherhood involvement differ based on metrics like education level, income, age, number of paid working hours, occupation, and partner’s occupation. The motivating research question is: Why are some working fathers more involved than others?

Marc hypothesized that well-educated fathers are more likely to devote more time to their children, especially in “developmental care." As the number of paid working hours increase, fatherhood involvement is expected to decrease. Finally, men in managerial occupations are more likely to devote less time to their children.

A sample of 471 fathers and their partners were asked to complete time-use diaries recording what primary and secondary activities they were engaged in throughout the day in ten-minute intervals. Childcare was operationalized into basic care (feeding, dressing, bathing), developmental care (teaching, helping, playing, reading), and secondary childcare. On average, fathers did 59 minutes of basic care, compared to 126 for mothers. Fathers also completed 33 minutes of developmental care, compared to 37 minutes for mothers.

As predicted, fathers with higher levels of education spend more time with their children. There is no clear trend between fathers’ income and childcare. Fathers in elementary occupations and in high-management occupations spend the highest proportion of quality time with their children. Younger fathers spend most time with children, but this effect could be because their children are younger or because these fathers might be unemployed. As expected, more hours spent at paid work means less time spent on care work. Finally, having an adult dependent is associated with a significant decrease in fatherhood involvement – by 29 minutes – all else being equal.

Study 2: Understanding the positive side (work-family enrichment)

Most research on work and family is based on a conflict perspective, under the assumption that time, energy, and attention are finite and that work-family conflict is a zero-sum game. However, an expansion approach may be more appropriate: high performance skills in one role may spill over into one’s other roles. This expansion of skills into other roles is known as work-family enrichment. Do the rewards and benefits perceived by working fathers in one role spill into the other role

In this study, Marc interviewed father and their partners. Seven of 20 fathers chose to do their interviews in the workplace, including 5 of 6 managers. The fathers described skills that they perceived that they learn at work that help them at home, including organization and time management, technical skills, people management, new perspectives, and cultural capital. In the opposite direction, fathers described skills that they perceived that they learn at home that help them at work, including sensitivity, patience, responsibility, people management, support, values, and long-term project management.

Fathers in high-level occupations described their work-family enrichment in terms of people management and cultural capital; fathers in lower-level occupations reported enrichment in organization and technical skills. This study reveals that enrichment is not the opposite of conflict and that soft skills are mainly developed at home.

Study 3: Understanding potential barriers to fatherhood involvement

What happens with these involved fathers? Having a child has an impact on fathers’ professional careers, but not the same effect as for women: previous research has established a motherhood penalty as compared to a fatherhood premium for having children. That said, fathers can receive backlash in terms of being a “poor worker” (that is, not committed 24/7 to their jobs) or a femininity stigma.

This study employed qualitative interviews on a different sample of fathers to examine barriers to fatherhood involvement. Marc identified five key barriers in particular:

  • Barrier 1: Poor organizational support. Though organizational policies for fathers exist, there is little institutional support for them, and fathers are rarely explicitly informed of potential options like flextime.
  • Barrier 2: Anticipation of negative career consequences due to transgressing the “ideal worker” norm.
  • Barrier 3: Personal obstacles to fatherhood involvement based on gender norms and self-image; many fathers reported barriers to fatherhood involvement linked to their masculine identity.
  • Barrier 4: Many fathers perceive that being a working father is not a legitimate reason to ask for more flexibility; they may ask for accommodations when completing an MBA or starting a new business, but not for spending time with their children.
  • Barrier 5: Money – Fathers perceive that using flex options is associated with wage reductions in a sort of “fatherhood penalty.”
Future potential work

There are a great number of possible research projects within fatherhood involvement. There is certainly more to do on the “sociology of thriving” and how fatherhood involvement can benefit organizations. On the flipside, more evidence on stigmas, barriers, and costs of fatherhood involvement could illuminate obstacles to gender equality in parenting. We look forward to hearing more!

Monday, April 17, 2017

Intersectionality and Women’s Health: Sexual Orientation, Race/Ethnicity, and Cervical Cancer Screening with Madina Agénor

Over the last few years, there has been a significant increase in public health literature that employs an intersectional approach. Intersectionality is an analytical tool that allows us to more accurately describe how human experience is shaped by multiple forms of social inequality that act in diverse and mutually reinforcing ways. This week’s WAPPP seminar featured Madina Agénor, Assistant Professor of Social and Behavioral Sciences at the Harvard T.H. Chan School of Public Health. Professor Agénor presented her research on how sexual orientation and race/ethnicity simultaneously affect disparities in cervical cancer screening among U.S. women.

Professor Agénor presented the results of three studies to demonstrate how intersectionality matters for population health and health equity. The first study examines how sexual orientation affects cervical cancer screenings without using an intersectional approach, and the second does the same for race/ethnicity. The third study employs an intersectional approach to illuminate how sexual orientation and race/ethnicity interact to produce discrepancies in cervical cancer screening.

Sexual orientation and cervical cancer screenings 
Lesbian and bisexual women (collectively referred to here as “sexual minority women”) are less likely to receive cervical cancer screenings and routine gynecological care than heterosexual women. Women with female sexual partners may also be more susceptible to cervical cancer because of higher rates of smoking and HPV. In addition, sexual minority women face a number of barriers to healthcare, including lower average income, lack of access to regular care and health insurance, and discrimination in society and in the healthcare system.

However, sexual minority women’s health is an understudied area, and there are significant gaps in research, including overreliance on convenience samples of mostly white, college-educated women; problems with measurement of sexual orientation (some studies conflate sexual orientation identity with sexual behavior); lack of appropriate comparison groups; and limited attention to potential drivers of sexual orientation disparities in pap testing.

Professor Agénor’s research uses a national probability sample of U.S. women, operationalizes sexual attraction, sexual orientation identity, and sexual behavior as different components of sexual orientation, uses appropriate comparison groups for each component of sexual orientation, and assesses whether healthcare factors (including access to insurance, receiving contraception, and STI services use) contribute to sexual orientation disparities in pap testing.

Lesbian-identified women exhibited significantly lower pap test use compared to heterosexual women (43% versus 69%), as did women with only female sexual partners compared to women with only male sexual partners (46% versus 71%). These disparities persisted even after controlling for age, household income, and other similar factors. Interestingly, differences in healthcare factors – access to health insurance, contraception, and STI services use – completely attenuated this disparity. According to the results of this study, focusing on only sexual orientation without an intersectional lens, healthcare factors may explain the discrepancy in cervical cancer screening among sexual minority women.

Race/Ethnicity and cervical cancer
In 2012, incidence of cervical cancer was highest among black and Latinx women, and cervical cancer mortality was three times higher among black women than white women. However, there has only been one subnational population-based study on sexual orientation, race/ethnicity, and pap test use, and this study also lacked appropriate comparison groups and didn’t seek out the drivers of disparities. Again using a national probability sample, Professor Agénor examined sexuality, race/ethnicity, and healthcare factors across racial and ethnic groups within sexual minority groups.

Professor Agénor found a lower prevalence of pap test use among women with only female sexual partners for black and white women, with the greatest disparity in pap test use between white women with only female partners and white women with only male partners. There was no difference in pap test use among Latinx women, regardless of whether they had male or female sexual partners. Importantly, the results of this study indicate that healthcare factors completely attenuated the disparity between women with female sexual partners and women with male sexual partners for white women only.

The magnitude and mechanisms of sexual orientation disparities in pap testing vary by race, which we wouldn’t have known without employing an intersectional approach. The crucial implications of this study are that addressing healthcare access barriers may mitigate disparities for some but not all women, and that we should be wary of “one-size-fits-all” interventions.

Qualitative Intersectional Study 
In order to better understand the mechanism of these effects on black sexual minority women, Professor Agénor held focus group discussions that centered around four key themes: healthcare provider communication style, heteronormative healthcare provider assumptions, heterosexism, racism, and classism, and healthcare provider background.

  • Participants preferred healthcare providers who took time to build relationships, were knowledgeable about sexual minority women’s health, and provided them with relevant sexual health information. 
  • Many participants reported that their healthcare providers assumed heterosexuality, and therefore provided patients with limited relevant sexual health information. Similarly, many participants reported both a fear of disclosing their sexual orientation to healthcare providers and negative experiences when they did so. 
  • Many participants reported that their healthcare providers made social class assumptions based on their race/ethnicity. Some healthcare providers made further assumptions based on perceived social class about their patients’ ability to understand health information, which affected both how they communicated with them and how they involved them in decision making. 
  • Participants reported negative experiences with OB/GYNs and expressed a preference for nurses and physician assistants, who provided more individualized attention. Finally, respondents reported a preference for healthcare providers with similar lived experiences, particularly for black sexual minority women healthcare providers. 

The results of this study indicate that patient-provider communication may be an important contributor to sexual orientation disparities in pap test use among black women. Sexual orientation and race influence black sexual minority women’s pap testing experiences and outcomes by shaping their exposure to multiple forms of discrimination and their access to and rapport with healthcare providers. Interventions designed for black sexual minority women should address multiple forms of discrimination and will have to look different than interventions designed for white sexual minority women.

Whether or not we employ an intersectional approach can be critical for designing appropriate interventions to promote health. Intersectionality matters for population health and health equity.

Tuesday, April 11, 2017

Quotas Matter: The Effect of Gender Quota Laws on Work-Family Policies with Ana Catalano Weeks

Gender quota systems, though controversial in the US, have proliferated in over 50 countries. In parliamentary elections, quota laws require a certain number of women to be on the party lists for all political parties. Quotas can have a significant effect on female political representation – in one election cycle in Portugal, female representation increased from 21% to 27%. But are these increases indicative of real change, or just “window dressing”? What effect do quota laws have on policy, and under what conditions? This week’s WAPPP seminar featured Ana Catalano Weeks, WAPPP Fellow and College Fellow in the Department of Government at Harvard University, as she discussed her research on the effects of gender quota laws on work-family policies.

In order to assess these questions, Ana examined data from Italy, Belgium, France, Spain, and Portugal. She limited her research to advanced democracies, as her theory is dependent on men and women having different policy preferences – i.e., women prefer spending more on social policy programs compared to men – which is less true in developing countries. She found that quotas are especially likely to lead to policy changes in areas in which there is a gender gap in preferences, and particularly in areas that don’t align with the main left-right dimension in politics (that is, issues that aren’t well-described by economic or class-based cleavages in politics).

Previous literature indicates that quotas are effective at increasing the share of women in office. While there is a great deal of work on the preliminary stage of quotas, getting women elected, there is less data on how gender matters for legislative behavior. Evidence that greater female representation translates into specific policy outcomes is rare. One notable exception is a study of political reservations in India, which found that female leaders were more likely to invest in resources that women favor, such as water and roads. However, it is not clear that the same outcomes will translate to advanced democracies, particularly because of the party system, and the mechanism between female political representation and policy outcomes remains unclear.

The key argument in this case is that after implementation of a quota law, policies are likely to shift in the direction of women’s preferences, especially for issues that are orthogonal to left-right dimensions in politics. Women face high barriers to entry into politics, be it because of discrimination or an “ambition gap.” Quota laws that increase the number of women in office should also increase the representation of their political priorities. Ana argues that absent women’s presence, political parties have little incentive to address orthogonal issues: such issues are cross-cutting, which may split the party’s traditional constituencies and cause conflict, detract from the party’s signature issues, and are not recognized by primarily male party elites to offer electrical opportunity. Quotas solve this problem by requiring parties to include women, who are more likely to care about these issues and be able to point out the electoral opportunities they present.

Ana posits two possible mechanisms for this effect. It may be that as quotas increase the number of women in parliament, these women act like any faction within the party. More people within the faction means more leverage, which leads to policy change. However, it may also be that the quota law and the debate around it raises the salience of these issues in public discourse, and this increased attention alone can lead to policy change.

The data confirm what we already know about gender gaps in preferences (women tend to prefer more social spending and government intervention than men). The biggest gender gap in social policy has to do with working mothers. When asked whether “preschool children are likely to suffer when the mother works,” across countries and over time 30% of men disagreed, compared to 39% of women. This gender gap is increasing over time: greater proportions of both men and women are disagreeing, but women are disagreeing “faster.” These gender gaps increase with education and persist across political parties. An exploratory factor analysis confirmed that work-family policies are actually orthogonal to the main left-right political dimension. While the latent underlying ideas around redistribution and work-family policies were correlated, they were not a result of the same underlying factor.

This preliminary work generated three key hypotheses. First, that quota laws will lead to an increase in spending on child care and parental leave. Second, quota laws will lead to a decrease in spending on family allowances, sometimes called child benefits, that are typically motivated by concern for fertility rates and that are designed to make it easier for women to stay at home. Finally, we should expect to see a larger change in spending in areas that exhibit a large gender gap in policy preferences.

Ana examined OECD social expenditures data, which allowed her to operationalize spending on child care, parental leave, and family allowances. She included controls for the number of women already in parliament, women in the labor force, GDP, party quotas, union density, fertility rate, EU membership, and overall social spending. Her methodology included two-way fixed effects models comparing changes within quota countries, along with a series of placebo regressions that set the date of the quota law’s implementation back to see if another factor in the law’s effective date was responsible for the result.

The results of the study confirm the first two hypotheses: gender quota laws were associated with increased spending on child care and decreased spending on family allowances. Given that overall spending remains about the same, there is a major shift in funding from family allowances to child care. These changes in spending are sizable: data from France showed a $500 increase in spending per child on child care, compared with $300 less per child on family allowances. In a smaller state like Portugal, there was a small increase per child in child care, around $40, and also small decrease in allowances per child, approximately $125. These effects are conditioned by the gender gap in each country. A larger gender gap in preferences leads to a bigger change in the amount of money spent on child care versus family allowances.

Robustness checks demonstrated that the results are actually an effect of the quota law rather than an artifact of the model or due to some other contemporaneous factor. Quotas did not affect issues not characterized by a gender gap, nor was there an effect between quotas and spending on issues well situated on the main left-right dimension.

Qualitatively, Ana compared the case of Portugal, which had a quota law at the time, with Italy, which had a quota law up for debate and chose not to pass it. Though both countries were in a period of major economic depression and austerity, they handled family policy budgeting very differently. Italy increased family allowances and new birth grants while decreasing funding for child care. By contrast, Portugal started a program to build 400 daycare centers, passed a law creating a right to preschool for children over the age of four, and eliminated family allowances for the top two income groups.

What was the mechanism at work in these changes? Interviews with policymakers in each country indicated that while there was little evidence that women acted differently in parliament, quotas did give women more leverage to push party leaders on their different priorities. However, there was also support for the second hypothesized mechanism. Party leaders began to champion gender-related issues, partially as a way of credit-claiming and partially because many leaders feared looking outdated or backwards. Interestingly, these changes persisted even when right-wing parties took over the government. While there is evidence for both mechanisms, it appears that they are not as distinct as initially theorized – when there are more women in office who have greater leverage, the salience of their policy priorities is also greater.

In conclusion, the results of quotas are not just window dressing: quotas increase the substantive representation of women’s interest. Quotas shift legislative priorities toward policies that help parents combine work and family. Importantly, this indicates that identity matters even in parliamentary democracies. However, there are many more questions left to answer, including how budgetary changes affect real people’s lives. We look forward to hearing more from this exciting line of research!

Sunday, April 2, 2017

Inclusive Talent Management: How Business Can Thrive in an Age of Diversity with Stephen Frost

Many organizations believe that they are truly meritocratic. This belief can make C-suite executives react defensively when faced with evidence of bias. How can diversity and inclusion advocates best communicate to such an audience in ways that leverage existing research and, most importantly, in ways that will be accepted and incorporated by those in leadership roles? This week’s WAPPP seminar featured Stephen Frost, Founder and Principal at Frost Included and WAPPP AY14 Fellow, whose latest book, Inclusive Talent Management: How Business Can Thrive in the Age of Diversity, demonstrates how to align talent management and diversity and inclusion agendas.

Steve began with a video that encapsulates the notion that “it’s easy to miss something you’re not looking for.”


When viewers are concentrating on counting the number of passes one team makes, it’s easy to miss the moonwalking bear! A similar logic applies to talent management concerns – when managers protest that they “can’t find qualified women” for a leadership role, it behooves us to ask how we can restructure talent management processes with an eye toward diversity and inclusion to make sure that such qualified candidates – who certainly exist! – are properly identified and given due consideration.

One reason it’s so easy to “not look for” diverse candidates is because we anchor on our in-groups. Think of your five closest friends, your five closest colleagues, your partner(s), and where you live.  How diverse is this group? Often, our in-groups are very similar to ourselves. Our in-groups also serve as our anchors for perceiving traits like intelligence and trustworthiness in others. Without a diverse in-group that would counteract this implicit bias, we create schema of what a “leader” looks like, for example, based on the people we know who are similar to us. This bias can also manifest in corporate decisions, so that we tend to hire, retain, and promote people like ourselves.

The challenge, Steve says, is that in an age of diversity, we are still addicted to likeness. The resolution is how to change the system and our own behaviors. The key to creating change is to employ both unconscious nudges – changing the way we structure our environment to incentivize less-biased decisions – as well as conscious leadership efforts.

Steve and his colleagues interviewed 66 organizations, asking whether they believed themselves to be meritocracies and asking about their talent management in terms of recruitment, retention, and promotion. The current state is homogenous talent management, infused with large amounts of bias at each stage. However, Steve also found three models of diversity and inclusion programs. Diversity 101 is “diversity for diversity’s sake,” characterized by a compliance-based approach – “we do it because we have to.” Diversity 2.0 is “diversity for social responsibility,” which makes a marketing approach to diversity designed to draw out the benefits of difference. Diversity 3.0, by contrast, treats diversity as a boardroom-level issue and employs a bottom-up, integrated approach to systems designed to embed the benefits of difference.

One such example, as Steve described, was a decision to interview candidates in groups of ten in order to save time and money. In these larger settings, interviewers tended to see skill sets as more salient, rather than identity, and were primed to think about team dynamics rather than the individual. In these group interview settings, recruiters hired more diverse candidates. This change was to save resources, not because of the research on nudges toward diversity and inclusion, but had the effect of creating a more diverse work environment.

Steve also discussed an intervention in talent management specifically related to gender at KPMG. The organization’s talent management approach involved all of the (predominantly straight white male) partners gathering in one room, talking about who they want to promote to the partnership, putting candidates on table, and agreeing on 10-20. Unsurprisingly, given what we know about in-groups and unconscious bias, the candidates up for promotion tended to look like the existing partners. The question is how to make this group less biased and more meritocratic without making them defensive.

The simple intervention that Steve proposed was to whiteboard the pipeline of potential incoming partners, separating the pool that were considered ready for partnership from those who were one to two years out, but to write the names of male candidates in red and female candidates in green. As soon as the names were written out, the discrepancy was clear – women were considered one to two years out for consideration for partnership, while men were disproportionately considered ready for promotion. On top of this, Steve attached a quantitative measure of performance for each name. Then, it became clear that several names in the “ready for partnership” group, predominantly men, were less well-performing than women in the “one to two years out” group. This intervention employs simple strategies to engender a rational, rigorous, calm, and systematic approach to checking in-group bias in talent management.

One practical challenge with diversity and inclusion interventions is that everyone is stressed and has little time and attention, to say nothing of resources, to put toward this critical work. Fortunately, even simple interventions can decrease unconscious bias, and when paired with conscious leadership efforts, these interventions can be very powerful. The simplest approach is simply to ask people to remember to ask “WADI”: What About Diversity and Inclusion.