 Good morning, everyone, and thank you for being here. It's a pleasure to give you a perspective from the social sciences from somebody trained in the pure scientific tradition. So let me get started with an overview of today's talk. I will start with describing a perspective, my perspective, and two driving questions for today. Then I'll describe two larger frameworks, one about contextualizing measured gaps, like the one this project has sought or is seeking to measure, and one about striving for gender equity in STEM. I will then go into three examples of gaps and equity. The first one is about results of a study that stemmed from the data of this project. The last two are other studies. And they concern single versus double blind reviews. That's the study that I led with an undergraduate tied to data from this project. Then academic service, and I'll define what that is, which is an inverted gender gap in the United States. And then the case of parental leave, why gender neutral policies, and I'll describe what that means, remain gendered in practice. And then I'll close with some conclusions, again bringing this back to the initial questions and my perspective. So my perspective, oops, OK, I need to go slow with this. My perspective, just to give you a little bit of background on myself and where I come from and what informs what I'm going to share today. So my training is in mathematics as a geometry. I work with integrable systems, but I haven't done that for a long time. Gradually, I moved within academic positions to working on research on teaching and learning, measuring of mathematical knowledge, lots of connection to curriculum and pedagogy and assessment. And what I work in now has to do with equity, inclusion, and increasing diversity in universities in the United States, particularly those that enroll a lot of minority students. So in a sense, I did with gaps like the ones where, like the one we're focused in in this project, but they mostly have to do with ethnic and racial minorities at large US universities. So that's the perspective I come from, and I focus a lot on what it means to create institutions that support and enhance the assets that these minorities bring to the sciences in particular. So that's the perspective with which I'll talk today. In my work, transformative change is something we, it's a word that in the US is used a lot. It's a phrase that it's used a lot. And this means change that transforms, much like what I envision the gender gap project seeks to instill and begin. And so transformative change requires not just these things, but some of these things. So it requires first knowledge of culture, context, and realities. Whatever we have with data needs to be framed within the spaces in order to drive transformative change. It requires attainable goals today, a vision of tomorrow, but attainable goals today. It requires strategy, and hopefully that strategy will build on available assets. This is a little bit what I heard Petra said the other day. It requires intentionality. We need to be there. We need to want to do this. It also requires solidarity and empathy. Compassion was the word used a couple of days ago, or maybe that was yesterday. My head is not completely sure. And it requires attention to equity, not just equality. And I will make sure that is something that's a notion that we share, and it's clear in our heads, and I'll describe that in just a second. So the driving questions that I propose from how I'm framing this, and much of this frames my work, is our two. How do we leverage our study, this study of the gender gap in STEM, our study results for transformative change at the local, regional, national, and global level? And each one of those is important, and the idea of contextualizing plays a big role in each one of these. And it's different in different countries and different communities. The second question is, how do we promote equitable human centric engagement in this kind of change? And the sort of framework that I like to think the broader context in which the project sits from the perspective I'm talking from is an intersection of several things, but I'll simplify them to two here. This idea of contextualizing measure gaps, like the gap that the project is seeking to measure, and this striving for gender equity in STEM. So in the middle lies the question of shrinking this gender gap in STEM, which is one of the hopes of this project. So a little bit about equality and equity before we move on. In sciences, a lot of us scientists tend to think that striving for equality is what we need to do, but here's a cartoon that is fairly popular in the US that sort of explains that equity is about equal opportunity and equal access, and that's not the same as equal resources. So in the left, we have equality where everybody gets the same bench to stand on. In the right, we have equity, where actually people get equal access or equal opportunity to do what they're seeking to do. So this is sort of the framework in which I propose we should think in the back of our heads about whatever transformative change connected to projects like the gender gap should be framed in. So I'll start, I said I will talk about three examples. I'll start with the first. So the first deals with single versus double-blind reviews in journals in the sciences, and I want to make a case for more research. The question of single versus double-blind reviews came up the first day of the conference, and we have this perspective that in the sciences, almost everything is single-blind reviews, and that is OK. We know each other. We're objective, and I want to show you data that might suggest something different. So can people see, is this big enough for the back? If people can't see, you can move to the front. We're going to have some graphs later on. So this is a study that I did with an undergraduate. I'll describe the study highlights, and I'll describe some results. So the goal was to explore the switch to double-blind reviews in women authorship representation in the American Mathematical Monthly, which is a journal of the Mathematical Association of America. This is a journal that's been around since the late 1800s, so it's fairly old for a US journal, and it's not a research journal. The focus is on expository articles, but it's actually quite quality people tend to present, because it's not their research scientist, but they try to expose for the general public to an extent. The data that we used for this analysis was CB Math data, so that was generously shared with the GenderGar project from this team about two years ago. And that data covered the 1900s to 2014. We also added some recent data that was available online and was not part of the CB Math data the years from 2016 to April 2018. The team was, under my direction, was an undergraduate at the University of Arizona. It's now working in industry and has a master in finance, so that's a picture of Ben, right there. Why this journal? I already did say a little bit about that. The longevity was part of it, and CB Math data was available. It had a recent double-blind switch, we understand, in 2015, and this is something unusual for just about every journal in Math. The history of the journal was quite interesting, and there's some stuff available online, but I maintain communications with email with the current editor, who is the first female editor of the journal in all those years. And that was also part of the reason that helped sort of centered or drive this research. So the driver question was, is there evidence of a significant change in female AMM, that's the name of the journal, authorship, after the 2015 switch to double-blind reviews for contributed article submissions? Ah, okay, so let me explain that. So thank you, please do interrupt me and question. I tried to explain, but we never do that enough. So single-blind, so when we submit an article to a journal that is not invited, so we just contributed, a single-blind review means that you don't know who the reviewer is. I submit an article, I don't know who the reviewer is, but the reviewer knows who I am, knows my name. A double-blind review means that neither person knows. So I don't know who the reviewer is, but the reviewer doesn't know the name of who submits. And there is some evidence, again, double-blind reviews tend to be rare in science in general, but there's some evidence from prior studies in the biological sciences, the ecology in particular, that there in fact changes after switches for generals to double-blind that don't really interfere with the quality of the articles in the journal, but really change female authorship rather significantly and increase it. Okay, so some of the analysis highlights. We excluded, so this journal doesn't just contain contributed articles, it has some invited articles, it has book reviews, and it has other types of articles. So this was the communication with the editor helped a lot because it's hard to find this information online, but we excluded from the analysis, invited articles, book reviews, another non-reviewed pieces that come to the journal. So we only included those that were evaluated. Python code created, that Ben created, was used to identify and count abbreviated unknowns. Elena said a little bit about this, but so when we analyze, genderize data, we use first names. So if first names are abbreviated with initials, we just can't genderize them. So in order to speed up the process, we use methods that are robust but are free, and so Ben wrote a code that separated everything that was abbreviated before he put it through the engine to genderize it. Then he genderized, so assign gender to all of the data that we had, and that was about 12,000 entries, combining Python code and genderize IO, which is a free version of a program that is robust, and many studies have used online. We did some other confidence testings to decide on definite or almost definite, confident gender assignment, and that was based on the work of Topaz and Zen. Okay, chat Topaz was going to be here, but I guess he couldn't be here, but somebody who's done, he is one of the authors of the Editorship Genderization Papers, which is this one that we counted here, and somebody mentioned the other day. So the automated genderization using the programs was supplemented by hand, Ben actually looked at names that for example had special characters like accents and so on, and tried to see if he could decide a gender association. So the overall genderization results, which are not about the double blind single blind yet, we genderize 41% of this almost 12,000 authorship and the abbreviated authorship were 51%. So that's pretty high. The difference was the names that were spelled out as first names, but could not be assigned gender, okay? And out of those 41%, throughout the history of the journal, 3% were female, 38% were male. So here are a couple of graphs, and the top graphs shows by decade in the horizontal axis since the 1900s. So if you can't see these numbers, which you probably can't, on the left we have 1900, on the far right on the top graph we have 2010. The bottom graph is by years, and it starts in 2003, so the last 10 years since 2018, it goes from 2003 to 2018, and the blue vertical line marks the year 2015, which is where that double blind switched happened. There's, so if you're color blind, maybe you don't see these two colors, but if you're not color blind, there are two main colors. There's red tones on the top and green tones on the bottom. The red tones all bunched together if you don't distinguish light and dark are described as ungenderized authorships. So the ones that were not assigned gender assignment, and in the top graph you can see that that's most of them at the beginning and then as the decades progress that shrinks. The green represents those that were genderized, okay? And in both graphs is the same. Now, and we can see that in the bottom graph, roughly in the last decade, the stuff that the entries, the authorships that were not genderized shrunk quite a bit. The other interesting thing, the different tones in this graph describe, within the red tones, so within the authorships that were not genderized, the abbreviated unknowns, I said there were the light reds, they're the dark red, they're not the light red, okay? So that's, there's a mistake here, those are switched. The abbreviated unknowns are the majority. So the dark red is the abbreviated unknowns, and we see that at the beginning in the 1900s, this seems to say most things were abbreviated and that seems to shrink, same consistent in both graphs. And in the green graphs, the dark green were the female authorships, okay? Out of the genderized, well, within the genderized, the female authorships. So one of the things that we see here, let me just move to the next graph. And the next slide has the same graph we had at the bottom exactly where it was, okay? So it's the decade graphs, I'm just now focusing on the decades, and has a different graph at the top, where proportions are counted a little bit different. But let's just focus on the bottom graph first. And so what we can see there is that female authorships sort of stabilizes after the year 2000, and is about 7% on average, and abbreviated authorships are about 10%. This is in absolute over the total number of authorships. Then we also see that a female authorship, we look after 2015 seems to consistently increase after the double blind switch. We don't have too much data after the double blind switch because it's just a couple of years. So this is something that we need to investigate probably some more with data from the rest of 2018 and 2019. Yes, online means going to the articles. It's possible if the CV math data was reliable, and we did check for that because we compare in years what we had both kinds of entries that the kind of data was the same. Then I don't see that that could have been a big issue. The online thing really didn't mean that having some other source of database was directly going to the journal and working article by article and deciding what kind of article it is and pulling it up, sort of building the equivalent of the CV math database manually. So it wasn't looking at a third party data. But that's a good question. So, and also the other interesting, what top? Just a second, the top graph. In the bottom graph, the red is everything. I haven't gotten there, just one second. So in the bottom graph, the other thing I want us to notice back here is that the percentage of abbreviated and no strings to zero, there's just nothing. So there's a switch apparently with, well, or an increase of female authorships and a drop to zero of abbreviated unknowns, which is also interesting. The top graph sort of shows that switch more dramatically here, okay? But the top graph has different percentages because the red is abbreviated unknowns only out of those entries that were genderized, not the total. And the green is females only out of those who were genderized, not the total. So both percentages are higher in the top graph, okay? And so that's what we have. And some takeaways from what we have so far, these preliminary results, is that a switch to double blind reviews may in fact increase female authorship and may enable more legitimized female authorship. By that I mean to draw attention to the switch the drop in abbreviated unknowns coinciding seemingly with the increase authorship by females. Most journals in math do single blind reviews, but all major journals of the Mathematical Association of America, MAA, may have recently switched to double blind reviews to, and I quote, the editor of the AMM, minimized by as maximized inclusiveness of the sets of authors and their papers. And I say may only because she was almost sure this had happened, and she actually commented that to her knowledge, the particular journal we looked at here was one of the last ones to make the switch, okay? But this is not necessarily so evident from the information online. So some recommendations for increasing or legitimizing female authorship simply to undertake further research of the kind we've done for the American Mathematical Monthly, particular for other MAA journals and other journals that may be incurring this switch. So this is worth more looking at data. Again, I wanna draw us back to the perspective that we had before about equity and about transformative change and things that might be a bit uncomfortable but are helpful to drive that change we hope to see. And again, the context, the place in which we can carry out and the meaning of these things changes with structure and context. Okay, the second example, so that was the first example of perspectives on gaps and equity. The second example is about academic service and I'll explain what that is or what that entails but it's an inverted gender gap in the United States and I'll explain how we mean that as well. So first of all, what is academic service in the United States? So types of examples of things that fit under academic service are, this is within the context of universities or higher education, participating in student recruitment activities or faculty recruitment activities. Serving on committees, committees at the level of a department, college or at the university level. Serving on an institutional board or an editorial board. Sometimes some of the service work can be extramural, happens, is tied to our being at the university but it happens beyond the walls of the university. Mentoring students and mentoring faculty, writing letters of recommendations for both students and faculty. So the inverted gender gap in service, what I mean by that is that in academia in the US we have essentially three blocks or types of workload, research, teaching and this service hours. So the research hours is the graph that sort of pulls from institutional, national data in the US for associate professors. So this is people that have tenure but are not full professors. And we have a breakdown on research, teaching and service hours each of the blocks. The females are in blue and the males are in red, okay? So we can see that males spend more time on recent hours and teaching hours, sorry. And females spend more time in teaching hours and females spend more time in service hours. And in fact, national and institutional faculty data show that even when we control for rank, race and discipline, women perform significantly more service than men. I want to draw attention that this is not particular to STEM, this is just more generic. But again, this shouldn't be discouraging but actually have the opposite effect. This is work to do in particular in STEM and in the different context, we might gather different information that confirms this or not. There is some sub data on STEM that doesn't change the picture quite so much. In fact, it tends to make it worse depending on where it is, okay? So where does this data come from? Where are this type of studies that pull this and what are some larger questions that frame this type of information? So the types of study that help draw these kinds of conclusions or do this analysis are of several kinds and this is not necessarily my field so I have some examples here. Some of them deal with organizational data from institutions, okay? Some use time diary approaches. So these are researchers asking to faculty document workload and activities in real time rather than recall from the very far past which sometimes is not so accurate. And some of them are interview based, okay? Examples of the research but also framing questions for that we can ask ourselves when we try to dig into this type of data are, I have six there and I'll go through them but each connects to either policy or institutional levels, cultural levels or individual levels and these are not so separate from each other. There's a lot of overlap between these types of questions. So one question that is really a framing question but I want to, I think this is a word that is important to know and it's not so evident what it means, what does it mean to say United States universities are gendered institutions? So that's the word right there, okay? And they are gendered institutions. The research also says that this essentially means that there is, they're kind of, they're created or organized to bias or benefit one gender, they're gendered masculine so to speak and this is part of the information that helps understand and frame the data we analyze. Does actual academic workload vary with gender controlling for academic rank and other variables? We saw in the previous slide that it does, okay? And that's pulling from a variety of data and not just the data that I looked at for the presentation today. If so, how? So if it makes a difference, how exactly does it make a difference? At the cultural level, so these are things, questions of the policy or institutional level, at the cultural level, what counts, we can ask what counts as documentable service? So if we are documenting our work or our activities or our time in our CVs or in a report for our universities, what counts as things we can document? What counts as a valuable service that will count for rank promotion, seniority and tenure at universities? And at the individual level, but again, not separate completely from the culture and policy, do women engage in different type of service activities than men? And do women say yes more or volunteer more for service activities? So I'm not gonna necessarily answer all of those questions, but those are framing and the research will answer some and then we have motivation for answering some more. So the research shows, and there's three main things and I'm gonna go one by one in different slides. So I wanna first focus on this one in connection to service load, so quantity and frequency of service. Research shows that women receive more frequent requests for service than men and are asked to perform more service. Again, this is the context of the US nationally and mostly at research universities, large research universities, okay? In fact, the research says that women do not volunteer more and do not say yes significantly more than men, they're simply asked more. And of note is that women are asked to serve more not by men mostly, but by other women, okay? So this is also interesting because it drives back that in a way we need to be our own advocates and have awareness of our own tendencies. When we are and we live inside gender institutions, a lot of the time we assume, especially if we are successful, masculine gendered roles, okay? And we forget a little bit the empathy and the advocacy and the bigger picture of who we represent or who we might be intentional in representing. Another piece of the research is research says in connection to service being visible and valued that women's service roles are often more time consuming and less prestigious, okay? So this is an interesting slide because some of the research identified two types of service, relational service and task-oriented service. So relational service has to do with building connections, community is quite essential for universities because it helps retain students, recruit students and keep them healthy within the institutional environment. But it's sort of a bigger scope than task-oriented when we are circumscribed to a particular goal. So relational service includes things like mentoring, writing letters of recommendation, attending recruitment events, participating in retention. Task-oriented service entails things such as serving on promotion and tenure committees, serving on editorial boards and other boards. So women do more relational service and women do less task-oriented service. Generally speaking, that's what the research says. And the interesting thing, so again, this goes back to why university are gendered institutions. Part of relational service is genderized as feminine. And task-oriented service is genderized as masculine. The differences between these two are noteworthy in the categories that seem to matter. So relational service is critical yet typically devalued as opposed to task-oriented service, which is prestigious. And not every committee's serving is prestigious and that's something to be aware of. Some committees are not as prestigious as others. Relational service also tends to stay largely undocumented. I don't know about people, but generally in the academic spheres that I move in, we don't write how many letters of recommendation for students or faculty in our CV. That's not something we document or count or how many recruitment events we may attend. So again, it's things that speak to the health of the academic institutions but are not documentable. So they remain invisible relative to the highly documentable and prestigious services. They're also emotionally taxing. They're effortful versus more objective for the task-oriented ones. And they're time-consuming versus more circumscribed. Some committees go on forever. I'm not trying to say that's not the case, but relational service is sort of long-term and more time-consuming. Lastly, the last aspect that I want to talk about today in connection to service is that inequitable service loads may occur even among women. Faculty of color do a disproportionate share of service and this may be especially true of women of color. And I think this came up at some of the presentations earlier today. Women of color are what we call double minorities, at least in the US. So they're female and they're ethnically minoritized. And they're sought after increasingly more and more by students who do not see themselves represented anywhere else, faculty that look like the students. And this is a big problem in the US because we have a demographic that is becoming a majority, especially in the Southwest, where I come from, Hispanics and other racial, traditionally historically minoritized populations, but they're not represented in leadership or in faculty in academia. Okay, so some takeaways. I have sort of a breakdown at this associate professorship level, 75% of females engage in major service roles versus 50% of males. An extra service hinders women's academic careers that it is less valuable in tenure and promotion or research universities and in effect takes away from research time, which is the kind of thing that leads to the publications and the gender gaps we've been seeing as documented by the projects. So some recommendations for increasing equity in service and some of these come from the research, go beyond training faculty to say no, there's a lot of research dedicated to tell, to train women as to how to say no. So as to build up women research assets and empowerment to become champions for women's work. The little sons there focus on something I said in the beginning of my presentation about the importance of focusing on an asset perspective rather than a deficit perspective. So women saying no is sort of deficit perspective, is trying to change women and trying to tell them to not do something to protect themselves. Building to the assets that women have in research and empowering men to become champions is building on both the assets of men and women. So those are particularly valuable. There's more recommendations in this brochure which I'm happy to make available. This was created by the University of Arizona Commission for the Status of Women and was presented by the University at large by leadership. Okay, so that's also another good practice. Okay, the last example I wanna end with is the example of parental leave and the question there is why gender neutral policies and we'll explain what that means, remain gendered in practice. We already know what gender means. It means no matter how things are construed there's a particular gender than their bias towards. Okay, so let's start with a big, with a photo. So the question is, is paid leave? And when I say leave here, I mean family leave or parental leave, I'm generally referring to that and I'm not gonna distinguish, but generally sometimes in these examples is parental leave, sometimes is family leave. Is paid leave available for both parents of infants? So this is the map of the world. It's a very recent study, 2018, and you have color codes by countries. So the US notably is one of the countries where there's no national policy, there's no parental leave that is paid. Of course states have their own laws and there is parental leave but it's state based, it's not national. There's no countries in which parental leave is only available for fathers, not surprisingly. There's some countries where it's mothers only and there's some countries where it's mothers specific as an entitlement but it can be transferred to the father, not very many of those. I think Spain used to have that before. Yes, yes, yes. And then there's the entitlement for both parents. And the entitlement for both parents really means that there's maternal leave, there's paternal leave, or there's what we call gender neutral leave. So that means that it's utilizable by all genders, okay, that's gender neutral. Now gender means it's actually utilized by a particular gender regardless of how the policy makes it available as gender neutral, for example. Okay, so the picture here says that policies vary a lot and again the attention to context for understanding and driving change. The types of studies that exist here and a subset of which I pulled from, then again this is not my area directly. Our organizational institutional data studies as well. They're also experimental studies. These are studies where participants respond to reality-based scenarios which are well-construed. These are professional researchers and interview-based studies. So workers or workers' partners respond to questions. Again the examples of the types of questions we can ask, span the gamut of policy, institutional, cultural, individual, and they are such as how pervasive or how common are gender neutral family leave policies, okay? When they are in place, to what extent do men utilize family leave policies? How are men who utilize such policies rewarded and penalized? What factors influence men's utilizations of parental leave and related policies? How are men utilizing, that utilize these policies perceived by their peers, the employers, and the institutions? And how do men who utilize these policies see themselves, okay? So this is again drawing the attention to the fact that it's not just policy that drives what happens here, but there's cultural and individual factors that also do that. And now I'm zooming out so I imagine this isn't my field but this quite complex to find information that really tracks down what happens in different countries and how things varies. But this is a study that's actually not published yet. It's coming out in 2020 but has data from 2012, especially for these reasons that compares Korea, Spain, and the US. And I want us to focus, there's a lot of stuff here on the bottom part of that graph. First, if we start on the last column, we see three zeros and a 12, which seems simpler, so let's start there. Basically the last column is the US and we see that there's no paid parental leave at the national level. So the first one is maternity leave, zero weeks, paternity leave, zero weeks, and parental leave, zero weeks while it's paid. If it's unpaid, then the national policy in the US in 2012, and I'm assuming now but I don't know for sure, is 12 weeks, okay? And that means gender neutral if it's parental. Then Spain and Korea have paid leave. Korea in particular has less than Spain for fathers, right here, much less, this is measured in weeks. And Korea in particular has 52 weeks of gender neutral parental leave, but it's not necessarily fully paid or paid at full time. Again, these are the nuances that are hard to pin down unless one looks more carefully. And it has no unpaid leave. Spain sort of leads the world, I think, in these three years of unpaid parental leave available to either parent. Now if we look at the top part of this graph, we see other nuances to these policies. So for example, we see that in the US and Korea, not very many males agree, or the percent of males that consider the best childcare option to have both parents is not very high, especially compared to Spain. And we also see here that in the United States and in Spain, parents tend to agree, I'm sort of using the counter of the statement that it's not necessarily true that males should have more right to a job when jobs are scarce than women. That's different in Korea. We also have the gender gap index standing of these countries. These are leaning towards gender parity. These ones leaning towards not. So that just contextualizes that it's not just about policy. Okay, so the other question is, does policy existence imply usage? And the short answer is no. Economic, cultural, and individual factors influence this usage. And again, the data becomes scarce if I want to look at usage, but I'll just have three examples from one of these papers. And just to remind us, the types of family leave that we're considering here are gender neutral, mothers only, fathers only, paid, and unpaid. Okay, and there's combinations of these. So here are some examples that come from one of the recent papers. So US has nothing as far as paid leave nationwide, but California has state laws apparently that allow gender neutral paid leave and there's 26% male usage. California is a fairly liberal state in the good sense of that perspective. In France, apparently there is gender neutral leave, but 3% males use it. In Sweden, fathers used to use the gender neutral leave about 4% of the time, apparently until Sweden switched to fathers only leave for some of that time that would get lost if the fathers didn't use it. And so the reported percentage of male usage now is 80%. So what can we learn about why do men not use family leave policies more if we take that question as a driving question? In connection to cultural and individual factors, there's the standard things we tend to think about. There's strong masculinity and workplace norm that drive the choice. This is the case in Korea where we saw those disagreements with when jobs ask scars, who has more right to them. For example, individual choices in countries like the US also drive this quite a bit. Men tend to see parental leave as a choice or a privilege rather than a norm which is completely opposite for women. So this is genderized towards females. But that's not the whole story. In Spain, for example, job security and gender gap in salaries between females and males only determine that only one parent can work because the leave is unpaid, then the male working brings is more profitable. There's also other interesting things in the US and Australia, this is where most of the studies come from, in connection to stigmas that male users of parental leave undergo. So the poor workers stigma and the femininity stigma are related but different. The poor workers stigma basically means that male leave users are seen as prioritizing family and outside responsibilities over work. This is completely counter the masculine norms. And the femininity stigma is sort of a derivative of that that says that in fact male users are seen as acting like women, weaker and communal. These are the adjectives that are associated and family leave is the norm for women and not for men and those are the sort of adjectives that are attributed to us. So both of these are problematic. The first one, the research shows decreases work-based reward recommendations for men. And the second one, the research shows undermines in addition to undermining personal identities increases work-based penalties. Of note for both of these, the poor working stigma, women appear to stigmatize male users more than men. That's what some of the research says that has studied this. And of note but not new in connection to femininity stigma, women feminize men as much as other. Okay, and just to close, almost to close, some takeaways, paid parental leave policies do not guarantee usage by males as we've seen. Economic, cultural and individual factors influence how these are used. And male users of parental leave policies are at risk at least in some countries of poor workers and femininity stigma as well as organizational penalties. Some recommendations, some of these come from the research and they're not all of them. And again, I'm asking a fairly narrow question here for changing the trends. We could do, or it's possible, we can not only do it. And again, the cultural context might repeat this completely. Increase or scale up fathers only paid leave like Sweden did. But in the side of asset perspectives incentivize the development of emergent new identities. And this is research, this is not just my idea for male parental leave users. So there's some emergent identities where males that use these leaves are seeing themselves as resilient and strong rather than weak and communal which is what the femininity stigma is. And so that is positive. Just to close, again bringing us back to the beginning, circle back to those driving questions as closing or as the gender gap project comes to partial end at least. How do we leverage this study for transformative change at the local, regional, national and global level? And each one of those is important. The group here is focused at the global level but really driving change happens locally and a lot of these things that I say might apply to some places and might not but I hope they can give us a broader perspective. And how do we promote, again, equitable and human centric engagement in this transformative change? And again, my thoughts are that we need intersectional knowledge within each of us. I'm a person that's developed a little bit of that myself but we also need broader groups of people where we can bring different intersectional knowledge that's well integrated and also inform empathy. Empathy is not alone without knowledge, knowledge is not useful without empathy and we need the combination of these two is more and more important if we seek to drive change that is equitable and counts. So thank you.